2020 In Sight - Innovation Pulls the Bits Together - Ed Maguire

2020 in sight Innovation pulls the bits together

Special report

June 2016

2020 innovation

2 [email protected] 17 June 2016

Contents

Executive summary .......................................................................... 3

Add the bits together and stir ........................................................... 4

The Unicorn Era winds down ............................................................34

Bits flow into currents .....................................................................53

10 themes from digital to physical ...................................................63

All prices quoted herein are as at close of business 14 June 2016, unless otherwise stated

Follow our 2020 innovation series

Ed Maguire [email protected] +1 212 549 8200

Reagan Tangney +1 212 549 5028

Find CLSA research on Bloomberg, Thomson Reuters, FactSet and CapitalIQ - and profit from our evalu@tor proprietary database at clsa.com

Produced by CLSA Americas LLC. For important disclosures please refer to page 136.

Visit our 2020 mini-site at clsa.com

http://www.clsa.com/

https://www.clsa.com/member/report/487625560

https://www.clsa.com/member/analysts/index.cfm?alogin=MAGUIREE&pagename=bios


http://www.clsa.com/member/sector/index.cfm?scode=tech


Executive summary 2020 innovation

17 June 2016 [email protected] 3

2020 in sight Technology-driven “creative destruction” is impacting many different industries as information technology matures into the backplane of the global economy and society as a whole, and businesses undergo digital transformation. Following a frothy period, a funding slowdown for startups will provide established firms with attractive acquisition opportunities. Change is already accelerating within information-intensive industries like tech, media, retail, communications and financial services. Next up will be transport, manufacturing, agriculture, energy, materials and life sciences. We explore 10 innovation themes with disruptive potential and highlight Microsoft, Red Hat, Salesforce and Splunk as our top software picks.

The information and communication technology (ICT) era is 40 years old. Cloud computing has industrialized IT, deflating costs and squeezing sector margins to the benefit of users and businesses pursuing digital strategies. Value has migrated from infrastructure to platforms, applications and applied analytics. The stage is set for a coming era of combinatorial innovation. Artificial intelligence, robotics, mobility and cloud will power new transformations at the intersection of digital, physical and biological domains. IT-driven deflation disrupts wages and economic “rents” from providers, as benefits accrue to consumers, innovators and transformational organizations.

After years of growth in venture-capital (VC) funding and investment, sentiment is turning more cautious, as the era of Unicorns (privately funded companies valued over US$1bn) fades with “down-rounds” and IPO investors increasingly leery. The pace of new-business formation has rebounded however. Expect belt-tightening among startups to lead to favorable tech M&A opportunities for buyers from different industries. Structural headwinds in the US include slippage in science, technology, engineering and mathematics (STEM) education, broken immigration policies, costly patent litigation and over-regulation.

We update our technology “meta-themes” - transparent IT, intelligent systems and convergence - highlighting the importance of platforms for innovation. Software remains at the top of the tech value chain as the driving force in innovation, gaining strategic importance for non-technology companies as well.

A new generation of high-growth businesses builds on connectivity, massive data-processing power, near-ubiquitous reach and powerful analytics. We explore 10 themes with disruptive potential that bridge the digital and physical worlds: artificial intelligence, augmented/virtual reality, Blockchain, open source, security, clean disruption of energy and transportation, autonomous vehicles, robotics, 3D printing and the Internet of Things.

Recommended innovation stocks

Ticker Rating Target price Last close Currency Akamai AKAM-US BUY 69.00 52.66 US$ Alphabet GOOGL-US BUY 970.00 733.25 US$ Amazon AMZN-US O-PF 770.00 719.30 US$ Apple AAPL-US BUY 115.00 97.46 US$ Facebook FB-US BUY 161.00 114.94 US$ Microsoft MSFT-US O-PF 60.00 49.83 US$ Red Hat RHT-US BUY 94.00 77.17 US$ Salesforce.com CRM-US BUY 101.00 81.09 US$ Samsung Electronics¹ 005930-KR BUY 1,600,000.00 1,380,000.00 won Splunk SPLK-US BUY 86.00 56.64 US$ ¹ Covered by CLSA; all others by CLSA Americas. Source: CLSA

The Unicorn Era winds down

10 themes frame innovations in the digital

sphere and physical world

Add the bits together and stir

Bits flow into currents

Change is accelerating in information-intensive

industries

Our top software picks are Microsoft, Red Hat, Salesforce and Splunk


Section 1: Add the bits together and stir 2020 innovation


Add the bits together and stir Since we published our 2020 forces converging: Crossing creative disruptions report in March 2015, US stock indexes have touched all-time highs with robust capital markets and M&A activity. In 2016, markets have shaken off early jitters, but there are signs that the sentiment that propelled the boom in privately funded tech firms is turning more cautious. As the “Bubble 2.0” cycle winds down, we see promising consolidation and expansion opportunities ahead for cash-rich firms. As software continues to “eat the world” we expect innovative cross-pollination across domains to accelerate, with the value-creation criteria inherent to technologies - platforms, applications and applied analytics - increasingly relevant across industries.

Our subtitle for this report, Innovation pulls the bits together, alludes to a coming “Cambrian explosion” of innovation across disciplines. Information technology has become industrialized, and is becoming a substrate of the broader economy. We see less incremental value creation in technology itself than the increasing potential to apply technologies to an increasingly digitized physical world. The next era of innovations will be powered by cognitive software, robotics and ubiquitous computing applied in combinatorial ways to transform industries.

The era of Unicorns (privately funded companies valued over US$1bn) is fading with “down-rounds” and IPO investors growing leery. Expect belt-tightening among startups to lead to favorable tech M&A opportunities for buyers from different industries. Innovation holds the key to sustainable value creation, with R&D estimated to drive 1.4% of GDP growth. Businesses continue to face headwinds including lagging science, technology, engineering and mathematics (STEM) education, broken immigration policies, costly patent litigation and over-regulation. With Moore’s Law and corollaries as a backdrop, software becomes the defining vector as businesses undertake digital transformation. Tech mastery will differentiate winners in every sector.

Digitization of business creates deflation, disruption and dislocation. Cloud computing industrializes IT and accelerates deflation, impairing infrastructure hardware and software providers’ ability to charge economic “rents”. This causes the cost to field test a new idea or start a business to plummet. Startups can scale at unprecedented pace with a fraction of the resources previously required, increasingly disrupting powerful incumbents in media, transportation, hospitality and other industries.

There are downsides too. Digitally enabled automation depresses wages for less skilled workers, exacerbating unemployment and social costs from income inequality. Economic dissatisfaction in the middle class is fueling a volatile political climate in 2016, though for the most part, proposed policy prescriptions do not address technology’s role in the underlying dynamics.

We remain optimistic about potential for value creation from combinatorial innovation. Sustained progress in computing, storage and connectivity powers increasingly sophisticated ideation, design, prototyping, research, product development and business creation. Value creation migrates “up the stack” to platforms, applications and applied analytics. Red Hat, Salesforce, Microsoft and Splunk are our top thematic software picks, as they exhibit key value creating characteristics of platforms, applications and applied analytics

A coming convergence between the digital and

physical

We highlight Salesforce, Red Hat, Microsoft and

Splunk as our top thematic software picks

There’s growing evidence that technology may be indirectly exacerbating

economic anxiety

Digitization of business allows rapid scale, but it

creates deflation, disruption and dislocation

The era of Unicorns (privately funded

companies valued over US$1bn) is fading


https://www.clsa.com/member/reports/557704675.pdf



Signs of a new industrial revolution emerge In 2016, employment and a range of industries are navigating widespread disruption from an industrial revolution driven by ICT. It’s not exactly “bloodless”. Leaders have been dethroned, the status quo is fluid, massive wealth is being created with unprecedented velocity and there is bubbling discontent in the broader US economy as machines displace jobs and a disproportionate share of income accrues to an elite few.

For investors, there are attractive opportunities and potential minefields as pace of innovations accelerates. Innovation (inextricably conjoined with technology) is key to sustainable value creation. With Moore’s Law and its corollaries as backdrop, software becomes the defining vector separating winners and losers. In this report, we characterize 2016 in a big-picture context, pulse-check the current state of funding and entrepreneurship, examine drivers and developments within frameworks and explore 10 critical innovative themes.

Why is innovation so important? Innovation is critical to value creation in an economy increasingly shaped by advanced technology. The most innovative industries typically contribute outsized gains to GDP, create businesses that generate attractive returns for investors, and in the case of information technologies, enable productivity gains across a continuum of industries. R&D contributes directly to gains in productivity. Macroeconomist Charles Jones estimates that R&D accounts for around 1.38% of annual economic growth. If the US economy had invested only the same amount on R&D as a share of GDP today as in 1950, productivity would be 17 to 32% lower. The Brookings Institute’s 2015 study, America’s Advanced Industries, found that advanced industries (those with the top 20% of R&D spending per worker and a higher proportion of STEM workers) employed just 9% of the active workforce, but generate 17% of total US GDP. These advanced industries employ 80% of the country’s engineers, generate roughly 85% of US patents and account for 60% of US exports.

Bits come together for a Cambrian innovation explosion We are on the brink of a veritable Cambrian explosion of combinatorial innovation, catalyzed by the confluence of cloud computing, mobile connectivity, automation and artificial intelligence. What has happened to the technology industry over the past decade is harbinger for disruption in other industries - with rapid, highly concentrated value creation on one hand, and deflationary forces that erode growth and employment on the other. Andrew McAfee and Erik Brynjolffsen’s 2014 book The Second Machine Age posits that the world is at an inflection point where digital technologies will manifest in full force through automation, resulting in the creation of an unprecedented range of “things”.

Technology provides leverage for value creation, accelerating the ability for platforms to dominate and exacerbating the division between winners and losers in the economy. Starting a company is cheaper by a factor of 1,000x versus a decade ago. Just US$5,000 in funding can be enough for a new business to disrupt large incumbents.

It’s not one single factor driving acceleration of innovative activity; it’s the combination of enabling technologies that creates powerful leverage to the economy. The convergence of Moore’s Law-enabled cloud and mobile computing, advanced analytics (artificial intelligence/machine learning/cognitive computing),

Software becomes the defining vector that

creates winners and losers

2016 sees continuing disruption from the digital

industrial revolution

R&D accounts for around 1.38% of annual economic growth

Digital technologies will manifest in full force through automation

Starting a company is cheaper by a factor of

1,000x versus a decade ago

It’s the combination of enabling technologies that creates powerful

leverage




robotics and automation is extraordinarily powerful. Gartner refers to a “Nexus of Forces”, IDC defines a “3rd Platform” comprised of cloud, mobile, big data analytics and social technologies. Boston Consulting Group describes a “software-driven digital metasystem” - a mobile-first environment of near-infinite compute power and connectivity with the capacity to scale massively sophisticated analytics in real time. However you define it, barriers to ideation, experimentation and true innovation are falling away, promising to empower billions of new minds coming online in the near future.

Technology has become an embedded, foundational strata across the economy and society, representing a core infrastructure analogous to energy in the Industrial Age. Market boundaries are also shifting because of sharing and platform economy dynamics. Businesses that were previously based on the sale of products or assets are now being turned into renting, sharing or subscription businesses.

We share an optimistic though tempered view. Cloud computing, mobile internet, non-traditional user interfaces, advances in programming science and artificial intelligence, and falling costs of compute and bandwidth place unprecedented power in the hands of everyone, from a child with a cellphone, to entrepreneurs, to researchers seeking to solve the challenges of medicine. For investors, it’s critical to time investments in disruptive technologies appropriately, as there are risks being too early or too late.

Grownups don’t believe in unicorns Market sentiment in mid-2016 has apparently shaken off jitters over weakness in China and Brazil, but there are growing concerns of macro headwinds with US GDP in 1Q16 trudging along at a 0.5% rate.

Figure 1

Official employment trends upward while stocks struggle for further gains

Source: Research.stlouisfed.org, FactSet

Sentiment is growing more cautious, judging by the dropoff in VC funding and decline of late-stage investments. GDP in the USA is plodding along at barely 1% growth despite the official unemployment rate in the 5% range. There are concerns that the post-recession recovery may be losing steam. There’s an ongoing debate regarding the future of US growth between the techno-optimists and those that view technology as ultimately contributing to slowing GDP growth.

4

5

6

7

8

9

10

11

12

6,000

8,000

10,000

12,000

14,000

16,000

18,000

20,000

May

06

Sep

06

Jan

07M

ay 0

7Sep

07

Jan

08M

ay 0

8Sep

08

Jan

09M

ay 0

9Sep

09

Jan

10M

ay 1

0Sep

10

Jan

11M

ay 1

1Sep

11

Jan

12M

ay 1

2Sep

12

Jan

13M

ay 1

3Sep

13

Jan

14M

ay 1

4Sep

14

Jan

15M

ay 1

5Sep

15

Jan

16

DJIA (LHS) US unemployment rate(%)

The challenge always remains a combination of

timing and careful selection

Technology has become an embedded,

foundational strata across the economy and society

Markets and employment recovered from the

financial crisis

Signs of trouble or a wall of worry ahead?




The volatility in the markets in early 2016 marked a shift in investors’ tolerance for high-growth software and internet companies with high perceived risks. “Internet Bubble 2.0” appears on its way to deflating not with a bang but with a whimper. Increasingly disciplined investors are focused on profitability and consistent execution. Top-line growth accompanied by losses is not commanding the valuation premiums previously. Skeptical sentiment among public company investors is impacting VC funding and M&A sentiment. Startups need to conserve cash and build sustainable, profitable businesses.

Investors are increasingly cautious over venture-backed IPOs. On the IPO demand front, most recent tech IPOs are trading below issue price. When newly public companies have hiccups, investors are punishing the stocks. After a mis-step, there needs to be a sustained period of rebuilding a business before investors will award a premium valuation.

According to an analysis by CB Insights, the average return of 90 US VC-backed tech IPOs since Facebook‘s May 2012 IPO through October 2015 was just 7.05%. For the same period, the S&P 500 return was 60.5% (over 8x higher than if an investor had purchased every VC-backed tech IPO during this period). For comparison, the Dow Jones Industrial Average gained 42.8% (6x higher) and equally weighted S&P 500 return (a better proxy for theoretical tech IPO investments) would have gained 67% (9x higher than IPOs). SecureWorks (SCWX) has been the only tech IPO in 2016, pricing lower than initially expected.

According to the industry analyst firm 451 Group’s 1Q16 survey of tech bankers and corporate development execs, 58% expect the number of new offerings to decline, versus 43% expecting this in 3Q15 and 27% in 1Q15. Sentiment has also turned more bearish regarding M&A potential over the next year, although the flurry of software acquisition activity in late May and early June may help. Volatility in the capital markets is a contributing factor: in the first six weeks, stocks including Tableau (DATA), Hortonworks (HDP) and LinkedIn (LNKD) saw valuations cut in half. Public market sentiment impacts willingness to do deals. With economic uncertainty from Europe, the impact of lower energy prices and equity price volatility, participants are finding it more difficult to get deals done.

Valuations in the M&A markets are polarized as well, with high multiple and low multiple deals and little in between - not a healthy indicator for M&A prospects. In 2015, investors were also rewarding acquirers, but this is not happening this year. Private equity is the only hope for many aspiring tech Unicorns, but the debt markets are also impacted by stock volatility, macroeconomic and some political uncertainty. There’s not an obvious investment cycle playing out in tech like the PC or internet wave either, so M&A focus is moving to life sciences and biotech.

By 1Q16, the late-stage VC financing environment has become more challenging and more discerning. There has been an increase in failures among high-profile startups including Quirky, Secret, Fab.com and others. Mutual funds have also begun to mark down valuations from private investment rounds, and the pace of private investments by large mutual fund firms is slowing as firms find it more difficult to agree on acceptable valuations.

“Internet Bubble 2.0” appears on its way to

deflating not with a bang but with a whimper

Investors are increasingly cautious over venture-

backed IPOs

M&A sentiment has turned negative and

valuations are highly polarized

Late-stage VC financing has become more

challenging and more discerning

Average return of 90 US VC-backed tech IPOs from

May 2012 to October 2015 was just 7.05%

Valuations in the M&A markets are polarized




While we do expect some pain for underfunded startup companies and overly optimistic venture and private investors, we do think these dynamics set up favorably for established companies with plenty of cash on hand. The deflation of the Unicorn Era bubble will set up unique opportunities for cash-rich companies.

Innovations shifts from atoms to bits . . . to atoms One of more prominent questions is whether technology driven innovation has run its course as IT has become industrialized. We’ve seen prior computing architecture cycles - mainframe, PC and first-generation web technology cycles have run their course. Now the mobile/cloud architecture is maturing and consolidating. For investors, there’s a change in opportunities to invest in technologies as market value creation shifts away from providers of the technology to applications of the technology.

Taking a broader view of technology systems (beyond computing) Robert D. Atkinson of the Information Technology & Innovation Foundation lays out how major tech system adoption waves layer upon one another in the economy. In Think Like an Enterprise: Why Nations Need Comprehensive Productivity Strategies, Atkinson posits that one can model the evolution of technology in the US economy through S-curves. The post-World War II growth wave was powered by electro-mechanical technology systems (televisions, electric appliances, etc). This drove growth until the mid-1970s when improvements in performance and reductions in cost diminished. It was not until the next wave of digital technologies based on computing and the internet that growth resumed in the 1990s. The chart below outlines how the cycles of the adoption S-curves for electro-mechanical tech systems, digital-electronic tech system and the artificial-intelligence-robotics tech systems.

Figure 2

Evolution of technology systems

Source: Robert D. Atkinson, Think Like an Enterprise: Why Nations Need Comprehensive Productivity Strategies - Information Technology & Innovation Foundation 2016, licensed under Creative Commons

In Atkinson’s view, there are good arguments that we may be closer to the end than the middle of the current digital-electronic technology S-curve, as the incremental improvements in broadband speed, storage capacity and computing speed have far less incremental impact than they did in the 1990s, when the introduction new microprocessors drove steady PC upgrade cycles

Technology adoption cycles occur in nested

“S-curves”

Value creation shifts away from providers of

technology to applications of technology

We may be closer to the end of the current digital-

electronic technology S-curve

M&A dynamics are favorable for established companies with plenty of

cash on hand

Tech system adoption waves layer upon one

another in the economy




among businesses and consumers, along with significant productivity gains from the adoption of business applications. In his view, advances such as cloud computing are more stepwise in nature compared to the launch of the Netscape browser in 1995 and client server computing. It’s clear to us that cloud computing has accelerated the deflationary characteristics of information technology, making it more difficult for providers of infrastructure hardware and software to charge economic “rents”. As such, value creation is migrating “up the stack” to platforms and applications. Newer technologies such as artificial intelligence (AI), autonomous vehicles, drones, robots and genomics are still in their nascent stages.

Atkinson thinks that it could be a decade or more before there is a new growth wave powered by technologies like robotics and machine learning - but we think this may be conservative. However, we think this analysis overlooks the nature of accelerating change. A period of slow deceptive growth is characteristic of new technologies. The early deceptive phase occurs before growth goes exponential then matures - and the pace of new technology phases in a “Nested S-Curve” sequence may be accelerating.

The coming “Fourth Industrial Revolution” as growth catalyst Dr Klaus Schwab’s book, The Fourth Industrial Revolution, posits a new wave of innovation that goes beyond smart and connected machines and systems. This analysis looks back further: the First Industrial Revolution began in the late 18th century with development of machinery and automation. The Second Industrial Revolution began around 1870 using electrical power to create mass production systems with division of labor. The Third Industrial Revolution began around 1969 using electronics and information technologies to automate production.

The coming Fourth Industrial Revolution will be powered by exponential growth of information technologies, catalyzing broad disruption across nearly every industry. Building on information technology advances of the Third Industrial Revolution, the combination of artificial intelligence, robotics, the Internet of Things, autonomous vehicles, 3D printing, nanotechnology, materials science, energy storage and other technologies will to be catalysts for exponential innovation.

The Fourth Industrial Revolution will be characterized by the fusion of technologies across physical, digital and biological spheres. The biological component (reflected in advances in nanotechnology and gene sequencing) makes it fundamentally different from prior revolutions. The Fourth Industrial Revolution is expected to drive massive supply-side gains of efficiency and productivity. The decline of costs and frictions involved with delivering, creating, and consuming products and services will be so significant that demand will increase, opening up new markets and driving economic growth. This view downplays concerns about technological unemployment and inequality resulting from labor market disruption.

Both the view of an AI-robotics technology system S-curve and the Fourth Industrial Revolution anticipate value coming increasingly from applications of technology to the physical world - Internet of Things, automation, materials science, manufacturing, bioengineering, precision agriculture, distributed energy production. While the exact timing is under debate, we think the signs are clearly pointing to more innovation coming from applying digital technologies to the physical (and biological world) - and this is where the most promising investment opportunities like ahead.

A period of slow deceptive growth is characteristic of

new technologies

Fourth Industrial Revolution is about the fusion of the physical,

digital and biological

Expect more innovations from applying digital

technologies to the physical (and biological

world)

Exponential growth of information technologies

will power the Fourth Industrial Revolution

This biological component will be fundamentally

different from prior revolutions




Creating value: Platforms, applications, applied analytics If one subscribes to the premise that “Software is eating the world” in Marc Andreessen’s words, that every company ultimately becomes a software company, then it follows that many of the operative investing principles for software companies are relevant more broadly. We’ve previously outlined our views that value in technology migrates upward, from infrastructure to platforms to applications. The adoption of cloud computing represents the industrialization of information technology, paralleling the adoption of electrification of industry beginning in the late 19th century.

There are three characteristics of software and technology based businesses that are capable of generating sustainable value creation in our view: platforms, applications and applied analytics. We prefer businesses with dominant concentration in one or more of these areas because of the ability to sustain differentiated value and competitive advantage, which leads to higher margins.

Platform businesses harness innovation from an ecosystem of partners, which results in far greater creation of value than a single firm could create. (We cover platforms in greater detail in Section 2 of this report).

Application businesses (more broadly referring the application of various technologies - software, hardware, and services) create value by addressing general or specific needs to businesses and users.

Applied analytics businesses make better decisions, reduce risks, are more effective at sales and marketing and realize more operational efficiencies.

Platform companies create value by engaging multiple stakeholders, creating value across multiple dimensions. They are able to benefit from third-party innovations and compete more effectively in industries outside their core domains. Amazon’s competition is not just e-commerce and retail, but publishing, grocery, media, IT and other firms. Apple’s competition is not just PCs and phones, but watches, music and movie streaming services. Google and Nest compete against companies like Honeywell, Phillips and Toshiba. Platform firms innovate faster because they harness the capabilities of partners. Leveraging technology, they are also are able to generate more value from fewer resources. Uber is the biggest taxi company, but the company owns no taxis; AirBnB is the biggest accommodations marketplace but it owns no inventory; Alibaba is the biggest merchant but it owns no inventory.

Platform companies also generate value at a greater proportion than product companies. According to MIT Professor Marshall Van Alstyne, platforms represent over two-thirds of privately held companies valued over US$1bn; three of the top five companies in the US by market cap - Apple, Google and Microsoft - were platform companies. Platform firms are also becoming more important in the economy, with the platform companies among the top 20 firms by market cap representing a growing proportion of total market value: from 10% in 2001 to over 25% in 2014. Platform companies dominate the technology industry, and leading platform companies in other industries also tend to be technology intensive. Leading platform companies include Apple, Facebook, Google, Microsoft, Samsung Electronics, IBM, Intel, Cisco, Oracle, Amazon, American Express, SAP, eBay and Alibaba.

Platforms, applications and applied analytics are

relevant across industries.

Many of the operative investing principles for

software companies are relevant more broadly

Platform companies create value by engaging

multiple stakeholders

Platform companies generate value at a

greater proportion than product companies




Figure 3

Platform companies have outperformed the S&P 500 since 2006

Source: CLSA, Factset - Index components: Apple, Google, IBM, Microsoft, Samsung Electronics, Intel, Cisco, Oracle, Amazon, American Express, SAP, eBay

In our software coverage, we favor Salesforce (CRM), Red Hat (RHT), Splunk (SPLK) and Microsoft (MSFT) as top platform picks. All four companies foster vibrant ecosystems of application developers and partners that create powerful network effects through engagement and innovation.

Narrowly defined, applications are software programs that address a business or consumer need. We prefer an expanded definition: the “application” of technology, or technologies and services that provide distinct utility of value to users. Applications simply consume technology infrastructure in order to deliver value to the users. They are the raison d’etre for technology itself. Successive generations of IT architecture have decoupled compute, storage and networking infrastructure from the application logic itself, which is where business value resides.

Enterprise applications have predominantly moved to a software as a service (SaaS) model over the past decade. We increasingly see applications that run on infrastructure as a service (IaaS) from Amazon Web Services, Microsoft Azure or Force.com. Because application value is measured by the perceived utility to the user, application vendors benefit from infrastructure technology cost deflation which is absorbed by hardware, infrastructure software or IaaS. As a result, application vendors can better sustain higher gross margins. Applications are not immune to deflation from feature expansion, bundling, competition, and the transition to cheaper subscription-based models, but the inherent value is not impacted directly by the dynamics of infrastructure commoditization.

Applications can be horizontal in nature (customer relationship management/CRM, human capital management/HCM, supply chain management/SCM, financial, procurement, marketing etc) or specific to a particular industry (retail, financial services, health care, ecommerce etc). Applications can also be monetized in ways beyond licenses and subscriptions. Large internet and e-commerce companies are application

0

50

100

150

200

250

300

350

2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016

Platforms S&P 500(2006=100)

Leading platform companies gained 173%

vs the S&P 500’s 59% gain since 2006

Applications are software programs that address a

business or consumer need

Applications can be horizontal, vertical and monetized in different

ways

Applications are less exposed to deflationary

characteristic of technology

We favor Salesforce, Red Hat, Splunk and Microsoft

as top platform picks




businesses: eBay is an application that makes money from transaction fees; Facebook and Google make money from user engagement by selling advertising. Although not typically regarded as such, IT security products are applications as well.

There are dozens of pure-play publicly traded application companies. Leading companies include SAP, Oracle, Microsoft, Salesforce, Workday, NetSuite, ServiceNow, Guidewire Software, Autodesk, ANSYS, Cadence, PTC, Adobe, Intuit, Blackbaud, Ultimate Software Group, Cornerstone On Demand, Fleetmatics, Marketo, Real Page, and many others. We’d include security leaders like Palo Alto Networks, FireEye, CheckPoint, Imperva, Fortinet, Qualys, Trend Micro and others as well.

Figure 4

Application companies have outperformed the S&P since 2006

Source: CLSA, Factset - Index components: SAP, Salesforce, Workday, NetSuite, ServiceNow, Guidewire Software, Autodesk, ANSYS, Adobe, Intuit, Blackbaud, Ultimate Software Group

From our software coverage, we favor Salesforce (CRM), Akamai (AKAM) and Microsoft (MSFT) as top application picks. Salesforce has leveraged its leadership in Sales Force Automation to expand into service, marketing and commerce. Akamai’s performance and security businesses leverage its extensive content delivery network infrastructure to solve technically complex business problems. Microsoft Office dominates personal productivity software, and its Dynamics applications are strong in the mid-market.

Applied analytics businesses use data and sophisticated mathematical, statistical and cognitive techniques to reduce risk, increase sales and optimize efficiencies. We distinguish applied analytics from traditional data warehousing and business intelligence tools, which are general purpose technologies used to build applications. Companies that use analytics strategically can realize significant financial and competitive advantages. The retail industry provides a great example: in the 1990s and 2000s Wal-Mart was able to capture enormous share gains by applying advanced analytics to its pricing and supply chain. In the 2010s we’ve seen Amazon gain share of e-commerce by applying analytics to recommendation engines, dynamic pricing and its own supply chain, arguably to Wal-Mart’s disadvantage.

0

50

100

150

200

250

300

2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016

Applications S&P500(2006=100)

Leading platform companies gained 138%

vs the S&P 500’s gains of 59% since 2006

Applied analytics techniques reduce risk,

increase sales and optimize efficiencies

We favor Salesforce, Akamai and Microsoft as

top application picks




The field of data analytics itself is evolving beyond the data warehousing, statistical analysis, multi-dimensional analysis and simple reporting that originated in the 1980s to embrace artificial intelligence (also known as machine learning or cognitive computing). We cover AI/machine learning in greater detail in Section 4 of this report. The use of AI techniques by large internet companies like Google and Facebook is a significant contributing factor to their financial success.

Applied analytics also provide the means for companies in maturing or declining industries to protect their competitive advantages and margins. Information-intensive industries have long used predictive analytics: in financial services to reduce risk through credit scores, in retail to improve yield form marketing campaigns, in mobile telecommunications to reduce churn, in logistics to optimize efficiencies. It’s not always obvious which companies are best at using applied analytics for competitive advantage, but the success of the Oakland A’s baseball team chronicled in Michael Lewis’ book Moneyball has charted course for companies in many industries.

Figure 5

Applied analytics vs S&P 500

Source: CLSA, Factset - Index components: Amazon, Google, Facebook, FICO, Netflix, Microsoft, Verisk, IBM, Visa, American Express

Companies that are leaders in applied analytics include Amazon, Google, Facebook, Fair Isaac/FICO, Netflix, Microsoft, Verisk, IBM, Visa and American Express. Firms that are leading investment into AI/machine learning include Google, Facebook, LinkedIn, Microsoft, Baidu, IBM and others. There are also companies outside of technology that use analytics to drive new business opportunities, for instance GE for connected products and IoT, and Tesla for tracking user experience to drive product improvement.

The market for pure-play AI is nascent so there are few if any ways for investors to gain direct exposure to the components - rather those firms with the foresight and expertise to apply analytics most effectively are those that will continues to generate sustainable value over time. In our coverage universe, we favor Akamai (AKAM), which employs sophisticated algorithms

0

50

100

150

200

250

300

350

400

2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016

S&P 500 Applied Analytics(2006=100)

Leading applied analytics companies gained 254% vs. the S&P 500 gains of

59% since 2006

Applied analytics help companies in mature

industries protect competitive advantages

We favor Akamai and Microsoft as top picks for

Applied Analytics




to solve complex optimization problems, and Microsoft (MSFT), which has been aggressively investing in AI to embed predictive capabilities into its platforms and applications. These two are our top software picks for applied analytics.

Playing secular trends can pay off, but watch the hype Cloud computing, the mobile internet, non-traditional user interfaces, advances in programming science and artificial intelligence, and falling costs of compute and bandwidth place unprecedented power in the hands of everyone, from a child with a cellphone to entrepreneurs to researchers seeking to solve the challenges of medicine. For investors, it’s critical to time investments in disruptive technologies appropriately, as there are risks being too early or too late.

Cloud computing has proven a sustainable growth trend for the leading providers, particularly for applications. Newer generation SaaS and high-growth software stocks have rewarded investors over the past five years. There has been pronounced and sustained contrast between the high-growth SaaS and cloud names and on-premise “Big Tech” stocks. SaaS consistently outperformed the S&P500 over the past five years while big (on-premise) tech has underperformed since 2013.

Figure 6

High-growth software versus big tech stock performance

High growth index components include Salesforce.com, Workday, ServiceNow, NetSuite, FireEye, Palo Alto, Red Hat, Splunk, Tableau, Veeva. Big tech index components include Microsoft, IBM, Oracle, SAP, Cisco, Hewlett-Packard. Source: FactSet, CLSA

Some secular themes provide a bit more consistency but which need constant vigilance. The IT security industry benefits from the need for constant innovation to protect data, systems and reputations from myriad evolving threats. IT security is highly dynamic with accelerated adoption (growth) and maturity cycles, making for periodic disconnections between investor sentiment and fundamentals. Over the past five years, the sector has both outperformed and underperformed the market at different times. In 2016, the group has struggled as companies from different generations like Symantec, FireEye and Imperva have disappointed investors.

50

100

150

200

250

300

May 11 May 12 May 13 May 14 May 15 May 16

High growth tech S&P 500 Top big tech companies

107%

(May 2011=100)

54%

36%

IT security benefits from constant innovation

cycles, but investors need to monitor constantly

High-growth software is more volatile, but has outperformed big tach

and the S&P 500

Traditional big tech and SaaS/high-growth software performance

have bifurcated

The challenge always remains a combination of

timing and careful selection




Figure 7

IT security - strong fundamentals but a crowded market

IT Security index components: AVG, Check Point Software, Barracuda Networks, Cyber-Ark, FireEye, Fortinet, Imperva, Guidance Software, NICE Systems, Palo Alto Networks, Proofpoint, Symantec, Qualys, Varonis, Qihoo360, Finjan Software. Source: FactSet, CLSA

We’ve highlighted the 3D printing sector as a key innovation trend since 2011, and since that time the sector has captured popular imagination and investor attention. Interest in and awareness of the potential for 3D printing, a technology that incorporates a range of techniques including additive manufacturing, peaked in 2014, and investors aggressively sought out the limited vehicles to gain exposure to the trend. This pushed valuations to a peak, followed by a crash, corporate restructurings and protracted period of underperformance. Despite being a transformative technology for manufacturing, investor expectations around the consumer opportunity may have been misplaced, and the expiration of key patents in 2013-2014 has led to a flood of new market entrants. Declining costs benefit end users, but put pressure on vendors of the technology.

Figure 8

3D printing stock performance

Index components include 3D Systems, Stratasys, ExOne, Proto Labs, VoxelJet, Materialise. Source: FactSet, CLSA

90

100

110

120

130

140

150

160

170

180

May 13 Nov 13 May 14 Nov 14 May 15 Nov 15 May 16

Nasdaq Composite IT Security(May 2013=100)

39%

21%

0

50

100

150

200

250

300

350

400

450

500


3D printing stocks Nasdaq Compsoite

-5%

67.3%

(May 2011=100)

The 3D printing sector has ridden a cycle of hype

and disillusionment

IT security exhibits greater volatility than the

market overall

We’ve been highlighting the 3D printing sector as

a key innovation trend since 2011




Wearable technology was at the peak of hype in July 2015 when FitBit went public, but the shares have disappointed investors since the initial enthusiasm. In our view, wearable tech has been a classic hype cycle story, but it’s yet uncertain whether the sector overall will be profitable enough to support further IPOs.

Figure 9

Overly optimistic views on wearable growth caught up with Fitbit stock

Source: FactSet, CLSA

The declining cost of solar energy promises to disrupt the traditional energy industry for the next two decades as the cost curve undercuts utilities’ own generation costs. This chart from author and futurist Ramez Naam shows solar cost undercutting natural gas-generated electricity in the early 2020s.

Figure 10

Solar cost projected to undercut natural gas by the early 2020s

Source: Ramez Naam

20

40

60

80

100

120

140

160

180

Jun 15 Aug 15 Oct 15 Dec 15 Feb 16 Apr 16

Nasdaq Composite Fitbit(Jun 2015=100)

-57%

-7%

FitBit was the first wearable tech pure-play

IPO

Wearable technology was at the peak of hype in July

2015 when FitBit came public




Of course, innovation does not insulate investors from the risks of globalization and competition. Declining cost curves do stimulate adoption and disrupt incumbents, but they also can prove challenging for direct participants as well. Despite being one of the most disruptive technologies longer term, solar stocks are still a difficult investment and have been an ongoing disappointment to investors over the past five years.

Figure 11

Solar technology is real, but solar stocks have been a losing proposition

Source: FactSet, CLSA

If one assesses stocks according to the Gartner “Hype Cycle”, the 3D sector may be nearing the “Trough of Disillusionment” before reaching the “Plateau of Productivity”, while solar remains mired in the trough for some time.

A brief survey of the rich landscape of innovation We explore 10 key innovation themes worthy of investors’ attention in Section 4. A key thread across each category is the essential role that software and data analytics plays enabling accelerating innovation. In our view, these are longer-term themes that could profoundly reshape markets, the economy and society at large. As is typical of long-term technology evolution, early stage activities are small, but with exponential progress, inflection points are likely to surprise.

Innovations in the digital sphere Artificial intelligence (AI), machine learning and cognitive computing are different terms that all refer a similar range of technologies. AI powers everything from speech recognition to search software, airplane navigation and auto-pilot systems, video image recognition systems and intelligent assistants for smartphones. A new generation of self-learning computing algorithms will be integrated into applications of all kinds. Integration with advanced robotics will power a new generation of autonomous and semi-autonomous machines.

Augmented-reality and virtual-reality technologies are coming to mass markets in 2016. The technologies have captured public imagination and the first generation of applications and viewing devices have made their debut. Virtual-reality is making its entrance first with Facebook’s Oculus VR, the Sony PlayStation VR and HTC Vive. Next up is augmented reality, with Microsoft HoloLens and offerings from Magic Leap and Meta paving the way for a new generation of applications and entertainment.

0

20

40

60

80

100

120

140

160

180

200


Nasdaq Composite Guggenheim Solar ETF(May 2011=100)

-73%

67%

A key thread across each category is the essential

role of software and analytics

Artificial intelligence is seeing another

resurgence - this time as foundational technology

Augmented reality and virtual reality devices are coming to mass markets

in 2016

Solar technology stocks have been a risky

investment

Innovation does not insulate investors from

the risks of globalization and competition




Blockchain, the distributed ledger that powers Bitcoin and other cryptocurrencies, is emerging as a transformative technology for the financial services sector in particular. The combination of advanced mathematics, access to massive computing power through peer-sharing, the open-source ethos and powerful new software enables a new universe of applications for managing transfer of value. There is an explosion of new blockchain startups targeting healthcare, media, finance, insurance and other industries.

Open-source principles inherently enable innovation, not just in software, hardware and services, but through derivative benefits to technology users in any endeavor. The open-source model has transformed software development and is increasingly being applied in hardware, networking, crowdsourcing, media and new business models.

Trust is the basis for essential functions of commerce and society. With the explosive growth of connections, applications, communications, information and systems, threats become more pervasive, driven by technological advances and growing involvement of organized crime and governments. As such, there’s growing need for security to facilitate e-commerce, electronic money transfers and modern conveniences such as ATMs. The IT security market is a dynamic market, conducive to startups offering fertile ground for innovators and investors.

Figure 12

Innovations in the digital sphere

Innovation What it means Who could benefit Potentially at risk Related companies

Artificial intelligence/ cognitive computing

Artificial intelligence governs everything from speech recognition to search, airplane navigation and auto-pilot systems, motion-detection systems and intelligent assistants for smartphones

Advertisers, businesses, consumers, government, society at large

Jobs across a wide range of capacities from blue-collar drivers, security guards and others to knowledge workers like translators, paralegals, medical professionals, investment analysts

Google (GOOG), Microsoft (MSFT), IBM (IBM), Baidu (BIDU), Facebook (FB), Amazon (AMZN), LinkedIn (LNKD),Salesforce (CRM), many startups

Virtual reality/ Augmented reality

Gaming, entertainment, commerce, travel

Consumers, game developers, content creators

na Facebook (FB), Microsoft (MSFT), Samsung, HTC, Sony (SNY), Google (GOOG), private firms Meta, Magic Leap

Open-source everything

The open-source model is transforming software development, crowdsourcing, prototyping, datacenters and the replacement of proprietary systems

Entrepreneurs, operators of cloud datacenters, corporations and service providers, SaaS independent software vendors (ISVs), consumers, industrial designers, military, consultants

Traditional proprietary hardware and software vendors including HP, Dell, Oracle, IBM, Microsoft, VMware, Cisco, EMC, Juniper, etcc

Red Hat (RHT), Hortonworks (HDP), Microsoft (MSFT), Facebook (FB), Google (GOOG), Intel (INTC), AMD (AMD), many private firms including DataStax, MongoDB, Acquia

Blockchain and cryptocurrencies

Blockchain technologies are distributed ledgers that enable non-repudiable exchanges of value between unrelated parties without an intermediary

Startup businesses, low-income workers, citizens in unstable countries, investors

Banks and other financial services firms, credit card and money transfer firms

Microsoft (MSFT), IBM (IBM), many private companies - Ripple, Ethereum, CoinDesk, Coinbase, BitPay, many others

Security Trust is paramount in a connected world. Rising levels of increasingly complex IT security threats compel increasingly innovative defenses

Consumers, businesses, government, society at large

Everyone and everything connected to the internet, including consumers, businesses, utilities, governments

AVG (AVG), Barracuda Networks (CUDA), Check Point (CHKP), CyberArk (CYBR), FireEye (FEYE), Fortinet (FTNT), Imperva (IMPV), Imprivata (IMPR), MobileIron (MOBL), NQ Mobile (NQ), Palo Alto Networks (PANW), Qihoo360 (QIHU), Rapid7 (RPD), SecureWorks (SCWX), Symantec (SYMC), Qualys (QLYS), Proofpoint (PFPT), Cisco (CSCO), IBM (IBM), CA (CA), EMC (EMC) and many others

Source: CLSA


Blockchain technologies have the potential to

upend the financial services sector

Security is an ongoing “arms race” between bad

actors and security professionals




Innovations in the physical world The energy and transportation sectors are poised to undergo massive disruption as the combination of increasingly cheap solar, advances in energy storage and adoption of electric vehicles will disrupt the traditional oil, gas, utilities and automotive industries - without regulatory intervention. There are increasing signs that solar energy is on the verge of becoming the cheapest source of power, without any subsidies. Innovations and economies of scale in battery and other energy storage technologies pave the way for large-scale adoption of utility-scale solar and electric vehicles.

The fundamental nature of transportation is changing as autonomous vehicles prove technological viability and sensor-based vehicle communications systems promise to ease traffic jams and improve safety. The first generation of self-driving prototypes is being tested on public roads, with all of the major automakers and several technology companies investing in research. The technology is also being extended to trucks, other commercial vehicles and unmanned aerial vehicles (UAVs). The bigger challenges to adoption lie ahead with laws, lawmakers and insurance companies.

Advances in robotics are having a transformative effect on manufacturing and industry as a new wave of personal and collaborative robotics comes to market. The top three drivers of the market are increased processing, reduced cost and size of sensors and programming languages and interfaces. Beyond nanorobotics, drones and autonomous vehicles (covered robots are transforming industries like manufacturing, warehousing and distribution, healthcare, retail and other areas.

Interest and hype around 3D printing peaked in 2014, leading to disillusionment among investors as stocks continued to come under pressure in 2015. The consumer market appears saturated with low-cost completion. Meanwhile, advances in industrial additive manufacturing continue to progress with adoption among manufacturers doubling between 2013 and 2015 and moving beyond prototyping toward full production.

Hype around the Internet of Things reached fever pitch in the mainstream media in 2015, as technology and industrial raced to articulate their out IoT strategies. Our 2014 report Deep Field: Discovering the Internet of Things focused on growing relevance and opportunities across consumers and businesses in a full spectrum of industries. Interest in wearable computing and smart home also reached fever pitch in 2015 as a flood of new market entrants resulted in a highly fragmented market. In 2016, we are seeing industrial proof of concept projects chart paths to sustainable ROI, but some businesses are taking a wait-and-see approach. While the major inflection point is not expected until 2017-20, seeds for immense transformations are already being sowed.

Autonomous vehicles and robotics will remake

transportation and manufacturing

A world connected - the Internet of Everything

3D printing - in a hangover from peak of hype, but technology is

real

The energy and transportation sectors are poised for disruption form

solar and energy storage

Advances in robotics are having a transformative effect on manufacturing

and industry





Figure 13

Innovations in the physical world


Clean disruption of energy and transportation

The declining cost of solar, advances in energy storage lower the cost of energy and make electric vehicles the default choice by 2030

Consumers, businesses, entrepreneurs, cities, society at large.

Carbon-based fuel businesses (oil, coal, gas), utilities, some automakers

Tesla (TSLA-US); ADAS: Mobileye (MBLY-US); Battery: Panasonic (6752-JP); Sony (6758-JP)

Smarter moving machines (autonomous vehicles, drones)

Self-driving cars, trucks, buses, drone aircraft

Consumers, businesses, automobile manufacturers, auto supply chain, military

Transportation based employment (taxi, truck drivers, logistics)

Google (GOOG), Toyota Motor (TM), Ford Motor (F), General Motors (GM),Mobileye (MOBL), Raytheon (RTN), AeroVironment (AVAV), Boeing (BA), Northrop Grumman (NOC), Textron (TXT), BAE Systems (BAESY), Adept Technology (ADEP), Amazon (AMZN), Lockheed Martin (LMT), AeroVironment (AVAV), General Dynamics (GD), SAIC (SAIC), GoPro (GPRO), Ambarella (AMBA), IXYS Corp, (IXYS), InvenSense (INVS) and others

Robotics Automated manufacturing, surgical robots, trainable robotic assistants, domestic robots

Manufacturers, healthcare, consumers, military

Labor, especially employees doing repetitive tasks in manufacturing, service, etc

Amazon (AMZN), iRobot (IRBT), Google (GOOG), Raytheon (RTN), Moog (MOG), Intuitive Surgical (ISRG), Cognex (CGNX), Accuray (ARAY), AeroVironment (AVAV), Northrop Grumman (NOC), Rockwell Automation (ROK), General Dynamics (GD), Boeing (BA), Teledyne (TDY), Textron (TXT)

3D printing Custom fabrication, prototyping, spare parts

Consumers, designers, industrial designers, manufacturers, service providers, materials producers

Spare parts, machine tooling, mass manufacturing

3D Systems (DDD), Stratasys (SSYS), ExOne (EXONE), Proto Labs (PRLB), VoxelJet (VJET), Arcam (Sweden), envisionTEC, EOS (Germany), Renishaw (UK), Organovo (ONVO), Autodesk (ADSK), Staples (SPLS), Adobe (ADBE), Microsoft (MSFT)

Connected everything (IoT, eHealth, sharing economy)

Myriad implications for both industrial and consumer

Consumers, businesses, manufacturers, logistics, military, public safety, wireless sensor network providers, analytic software vendors

Companies with high reliance on manual processes

IBM (IBM), Cisco (CSCO), GE (GE), PTC (PTC), National Instruments (NI),Google (GOOG), Intel (INTC), AMD (AMD), Siemens (SI), Oracle (ORCL), Salesforce (CRM), Amazon (AMZN), Teradata (TDC), SAP (SAP), Splunk (SPLK), Broadcom (BCOM), Qualcomm (QCOM); Apple (AAPL), Samsung, Sony (SNY), Nike (NKE), Intel (INTC), Qualcomm (QCOM), Microsoft (MSFT), GoPro (GPRO), FitBit (FIT) , Buddy Platform (BUD - ASX) wireless network, sensor and analytics vendors

Source: CLSA

Biology, healthcare and new worlds We don’t cover advancements in biotech, genetics and healthcare in-depth in this report, but want to highlight the potential value creation opportunities at the intersection of information technology, biology and medicine. There has also been significant progress in the space race, with SpaceX achieving several key milestones toward viable reusable boosters in 2015 and 2016.

The declining cost of computing, more powerful systems and the capacity to store and process massive quantities of data create conditions are conducive to accelerating innovation in the life sciences. With Illumina’s latest machines lowering the cost of a sequenced human genome below US$1,000, genomics is actively decoding elusive mysteries of DNA, the “source code” for the human body, with promise of proactive avoidance and better treatment for cancer, Alzheimer’s, multiple sclerosis and other chronic diseases.

Computational genomics are decoding the software

of nature




One of the most significant advances is a new technology, CRISPR, which enables scientists to edit genomes with unprecedented precision, efficiency, and flexibility. CRISPR (Clustered regularly interspaced short palindromic repeats) are segments of prokaryotic DNA containing short repetitions of base sequences followed by short segments of “spacer DNA” from previous exposures to a bacterial virus or plasmid. CRISPR interference technique has enormous potential applications, including altering the germline of humans, animals, and other organisms and modifying genes of food crops. While there is a lot of potential, there are also ethical considerations: Chinese scientists have applied the technique to nonviable human embryos, which hints at CRISPR’s potential to cure genetic diseases but also raises the prospect of genetically designed babies.

Mobile healthcare technology is seeing robust innovations among consumers and professionals, though adoption is concentrated at the ends of the spectrum among the very healthy and very sick. There has been an explosion of FDA-approved apps for diagnostics and treatment as well as monitoring. Despite robust VC investment and high-profile media coverage, the market remains nascent and highly fragmented. We expect Apple, Google, IBM, Microsoft and Qualcomm to foster a robust ecosystem of startups and partnerships with care providers and pharmaceutical companies.

A new generation of privately funded companies is pursuing a range of ventures including commercial space-cargo flights, low-Earth-orbit space tourism, asteroid exploration for resource extraction and longer-term plans for manned space ventures to the Moon and Mars. SpaceX has successfully landed its reusable booster rocket, potentially saving tens of millions of dollars per launch and opening up an opportunity for greater activity. Major aerospace firms Boeing, Lockheed Martin and Northrop Grumman are actively engaged, alongside leading private companies SpaceX, Virgin Galactic and Planetary Resources.

Techno-optimists versus secular growth skeptics There’s an optimistic school of thought that convergence of information technologies, clean energy and connectivity bringing billions of new minds online will catalyze enormous gains in wealth, health and quality of life. The most passionate advocates like Ray Kurzweil and Peter Diamandis of Singularity University see the current technological disruptions to industry and employment as temporary, paralleling the dislocations seen in prior industrial revolutions. People have dealt with change in the past, ergo this time it’s just another cycle set to play out along historical lines.

Technologists and investors tend to project the future in stepwise terms. Human beings naturally have linear intuitions about the future because linear thinking progresses logically from experience. However, innovations and paradigm shifts occur at an accelerating, often exponential pace. This creates a disconnect. The way the exponential progress in technology accelerates change can be illustrated by comparing mass adoption of inventions over the past 150 years. One only has to look at the rapid growth of Facebook and the explosive growth in tablet computing that the iPad catalyzed to appreciate accelerating paradigms.

Tapping into accelerating change

Techno-optimists see current disruptions to

industry and employment as temporary

The new space race

Connected health poised for steady adoption

CRISPR is a technology that allows unprecedented

ability to edit genomes




Figure 14

Adoption paradigms are accelerating

Source: Ray Kurzweil, KurzweilAI.net

Working our way out of the trough of a “super cycle” Klaus Schwab’s view of a “Fourth Industrial Revolution” differs from economist Carlota Perez who views us in the middle of a multi-decade ICT supercycle. Perez draws from economists Nikolay Kondratieff and Joseph Schumpeter in framing multi-decade waves of technological change that occur in surges about once every 50 years. The current ICT revolution is the fifth upheaval experienced by the capitalist system since the first industrial revolution in the late 18th Century. Since we first encountered this thesis in 2011, we grow more convinced of its applicability to current conditions.

Carlota Perez defines the “turning point” as the current phase of the grand ICT supercycle that commenced with the crash of the internet bubble in the early 2000s and persists today. Perez’ view is that this turning-point period may last a few more years, with structural changes to the economy working themselves out until we emerge into a new economic “golden age” in the decades ahead.

Figure 15

The historical record: Bubble prosperities, recessions and golden ages

Source: Carlota Perez

Telephone Radio Televison

PC

MobilePhone

TheWeb Facebook

?

0.1

1

10

1001860 1880 1900 1920 1940 1960 1980 2000 2020

(Years)

Maturity

1771The IndustrialRevolutionBritain

1829Age of Steamand RailwaysBritain

1875Age of Steel andheavy EngineeringBritain/USA/Germany

1908Age of Oil, Autosand Mass ProductionUSA

1971The ICTRevolutionUSA

Internet mania, Telecoms, emerging markets,

Financial casino & housing

Sustainable global "golden age"?5th

London funded global market infrastructure build-

up (Argentina, Australia, USA)

Belle Epoque (Europe) "Progressive Era" (USA)

3rd

The Roaring Twenties Autos, housing radio,

aviation electricity

Post-war Golden Age4th

1st Canal mania Great British leap

2nd Railway mania The Victorian Boom

INSTALLATION PERIOD Bubble collapse DEPLOYMENT PERIODNo., date, revolution, core country

TURNINGBubble prosperity POINT Golden Age prosperity

1793-97

1848-50

1890-95

Europe1929-33

USA1929-43

2007/08-????

The turning-point period may last a few more years

The rate at which new inventions reach

widespread adoption is accelerating

Carlota Perez frames current problems as a

consistent pattern seen in prior technology shifts




An installation period typically begins within a mature economy, with a great deal of experimentation in the free markets. What typically follows is a frenzy of investing, inflated asset prices and speculation followed by a crash (or crashes). This happened in the 1920s followed by the Great Depression, and again in the 2000s with the crashes of the first internet bubble and global financial crisis and Great Recession. This decade we are experiencing a protracted transition as the economy readjusts and the ICT revolution propagates through society and the economy.

The coming “deployment” period (which Perez has estimated may commence before the end of the decade) will see expansion of both new and rejuvenated sectors, as the potential of new technologies comes to fruition. This is a period of creative construction, where benefits of wealth are spread more broadly and capital and finance decisions are directed towards production rather than speculation.

Technology exacerbates economic imbalances Despite the promises of innovation, economist Robert Gordon and others are quantifying painful economic imbalances that are becoming more pronounced. The modern digital revolution has yet to deliver on the promise of better jobs and higher productivity. Instead, the new economy is creating immense wealth with far fewer workers, with income growth stagnating for the vast majority of employed. The stratification of income growth over the past decade has concentrated the majority of income gains in the top percentile while the median income in 2011 is 8% lower than in 2007, having peaked in 1998. These figures show earnings from labor are declining as a share of total economic output, but the share gains from highly skilled sectors have been growing.

Figure 16

Median income and employment have diverged from GDP growth

Source: Research.stlouisfed.org, FactSet

Since the 1990s there has been a transition from an industrial-based to an information-based economy where big investments are being made in technology and technology industries and although innovations in technology are awe-inspiring some (like economist Tyler Cowen) believe they are having less transformative effects as inventions once did. Robert Gordon, professor of economics at Northwestern University, believes that after an initial boost in

50

100

150

200

250

300

350

400

450

500

1967

1969

1971

1973

1975

1977

1979

1981

1983

1985

1987

1989

1991

1993

1995

1997

1999

2001

2003

2005

2007

2009

2011

2013

2015

Private employment Labor productivityReal GDP Median household income

(1967=100)

Despite the promises of innovation, there are

increasingly painful imbalances

The “Great Decoupling” of income and employment

from GDP growth accelerated in the 1980s

Every major technology development surge has seen the same

pattern play out

The installation phase is a period where new firms are formed and

older firms fail

Some economists believe growth is over in the US




productivity and growth from 1994-2004, the potential impact from further technology innovations will be minimal. He predicts that for the next 25-40 years, real per-capita disposable income of the bottom 99% of the US income distribution will grow at an average annual rate of 0.2%, a tenth of the pace of the 2% per year for the century before 2007.

Author Nicholas Carr writes in The Arc of Innovation “There has been no decline in innovation; there has just been a shift in focus. We’re as creative as ever, but we’ve funneled out creativity into areas that produce smaller-scale, less far-reaching, less visible breakthroughs”. From 1876-1886, we saw the development of the internal combustion engine, light bulb, steam turbine, railroad, car, phone, movie camera, and toilet among others. From this time until the 1950s life in America changed drastically. These were large-scale inventions that had a big impact on the way we live. Carr (as well as others like Tyler Cowen) believes that innovation is fundamentally less impactful today. There are other possible causes for declining growth such as a decline in education, lack of R&D investment, or short-sighted investors that pressure companies away from long-term investments.

Cutting a broad swath through unskilled jobs The phenomenon of technology replacing jobs is not at all new. What’s different now is that it’s software and artificial intelligence replacing tasks and jobs at a dramatic pace. The downside for employment is that fewer people are required in many industries (eg, airlines, manufacturing and the supply chain). As service-sector tasks are increasingly automated, this reduces the need for cashiers, toll takers and potentially taxi drivers as we’ve previously seen technology replace stenographers, typists or bookkeeping people.

Pessimists foresee a jobless future There’s been a lot of discussion of the impact of technological unemployment. Martin Ford’s Rise of the Robots - Technology and the Threat of a Jobless Future paints a dire scenario arguing that technological unemployment is inevitable, it will be pervasive, and societies will need to implement a universal basic income (UBI) to address citizen’s needs. A number of prominent economists including Lawrence Summers, Nouriel Roubini and Paul Krugman have publicly expressed concerns that successes in technology are eliminating jobs. Robert Reich has said that robots will “take away good jobs that are already dwindling. They will in short supplant the middle class.”

Concerns over technological unemployment are not new, stretching back most prominently to the Luddites in the early 19th century. In the US, Congress has commissioned studies at the end of the 1897 recession and in the late 1930s to ascertain the impact of labor-saving devices on human labor, and in the 1961 recession, President John F. Kennedy created an Office of Automation and Manpower in the Department of Labor on the premise that the major challenge of the Sixties was “to maintain full employment at a time when automation, of course, is replacing men.” Robert Atkinson of the Information Technology & Innovation Institute argues that if technology-led productivity growth really has been the culprit behind America’s anemic job growth since 2009, one would expect that America’s productivity growth rate would be higher than normal. In fact, US productivity growth has tracked at about half the previous rate since the end of the Great Recession.

One of most widely quoted studies on technological unemployment (by Osborne and Frey) estimates that 47% of US jobs could be eliminated by technology over the next 20 years. However, the McKinsey Global Institute

Software and artificial intelligence are replacing

tasks and jobs at a dramatic pace

Concerns over technological

unemployment are not new

Some believe tech unemployment will lead

to need for universal basic income (UBI)

Digital innovation is not as significant to society as

late 19th century inventions




sees more job redefinition instead of unemployment, foreseeing that very few occupations will be automated in their entirety in the near or medium term. A recent OECD report sees less risk of unemployment, rather certain activities will be automated, business processes will transformed, and jobs redefined.

There is a “productivity paradox”, which refers to an observation made in that as more investments are made in information technology, worker productivity declines instead of increases. Part of the challenge is measuring value from ICT is the typical delay between the installation of new systems and realization of productivity benefits. MIT professor Erik Brynjolffson found a five- to seven-year lag between deployment of an enterprise-resource-planning (ERP) system and subsequent benefits. In a review of over 50 ICT and productivity studies between 1987 and 2002, Jason Dedrick, Vijay Gurbaxani and Kenneth L. Kraemer of the University of California, Irvine concluded that the productivity paradox had been effectively refuted, with greater investment in ICT associated with greater productivity growth.

Living in a world of digital deflation Deflation is an inherent characteristic of information technologies, and nearly every business where technology dominates must reckon with deflation. Moore’s Law has continued unabated since 1958, and there are corollaries across storage and connectivity. Digitization and hyperconnectivity have served to eliminate frictions that allowed value to accrue to intermediaries, particularly in media businesses. So many businesses are built on a model of scarcity.

Digitization typically brings significant deflationary pressure. This impacted newspapers and music, TV is coming next. The print newspaper industry increased from US$20bn in 1950 to US$60bn in 2000, then saw revenues drop in half by 2010 because of web advertising. The music industry saw revenues decline from US$16bn in 1998 to around US$6bn in 2008 as listeners access music through YouTube, Spotify, Pandora, iTunes and other sources.

What’s increasingly clear is that there is less opportunity to extract economic rents from IT itself with adoption of cloud computing. Amazon’s founder Jeff Bezos charted the course for Amazon Web Services by targeting highly profitable business models of HP, Oracle, IBM and others: “Your margin is my opportunity”. With over 50 price cuts since 2006, Amazon Web Services has accelerated the deflationary trajectory characterized by Moore’s Law with economies of scale and highly aggressive pricing, leaving traditional enterprise server and storage hardware vendors hurting. Apple (and for a time Samsung) were the only vendors of smartphones able to sustain profits, while early industry leaders like Nokia, Ericsson and Blackberry have failed or struggled. Open source software has also eroded both revenue growth and profitability for infrastructure software providers as well.

Highly automated companies need fewer people to generate more value. While tech employees (especially with specialized skills) tend to be well paid, by nature they are deflationary to employment; there are a lot fewer jobs required at tech firms to generate net profits than in other sector. A vivid example of the leverage that technology provides is illustrated by the difference in net income per employee across different industries. Tech leaders generated an average of US$250,029 of net income per employee, banking generated US$84,238, autos generated US$38,533, retail generated US$11,485, and fast food generated US$6,787.

Technology will mean more job redefinition

instead of unemployment

Benefits from ICT investment could play out

on a larger scale over time

That there is less of an opportunity to extract

economic rents from infrastructure IT

The music industry saw revenue decline from

US$16bn in 1998 to around US$6bn in 2008

Highly automated companies need fewer

people to generate more value

Nearly every business where technology

dominates must reckon with deflation




Figure 17

Tech leaders’ net income per employee vastly outpaces other industries

Source: CLSA, company filings, Brett King - Tech components: Apple, Microsoft, Google, Facebook, IBM and Oracle; Banking components: Bank of America, Wells Fargo, Citibank and JP Morgan Chase; Autos components: Ford and GM; Retail components: Walmart, The Home Depot, Target and Costco; Fast food components: McDonalds and Yum!

Adoption of technology creates value by automating tasks and improving productivity, but also this has deflationary impact on wages. It’s our view that technological advances are also a key factor behind the growing divergence between rising corporate profits and labor’s declining fortunes. When technologies arise that reduce the need for human workers at lower costs, the cost of technologies also tends to decline over time.

This allows productivity gains without impacting corporate profits, while reducing labor’s share of the pie. Andrew McAfee of MIT compared labor’s share of corporate expenses versus corporate profits as a share of GDP. 2015 data show ever-widening divergence since the last recession. Corporate profits, meanwhile, have never been higher in absolute terms or as a percentage of GDP.

Figure 18

Labor’s share of corporate profits declining since 2002

Source: Research.stlouisfed.org, Andrew McAfee (MIT) Corporate profits calculated by profit per unit of real gross value added of nonfinancial corporate business.

0

50,000

100,000

150,000

200,000

250,000

300,000

Fast Food Retail Autos Banking Tech

(US$)

84

86

88

90

92

94

96

98

100

102

104

6080

100120140160180200220240260280

1947

1949

1951

1953

1955

1957

1959

1961

1963

1965

1967

1969

1971

1973

1975

1977

1979

1981

1983

1985

1987

1989

1991

1993

1995

1997

1999

2001

2003

2005

2007

2009

2011

2013

2015

Corporate profits (LHS) Labor (RHS)(1947=100) (1947=100)

Technology firms generate far greater profits per employee than other industries

Increasing evidence that technology drives

value creation but is deflationary to wages

Technology reduces frictions that support

wages




Authors John Seely Brown and John Hagel of the Deloitte Center for the Edge identified a fundamental paradox facing business that they dub “The Big Shift”. On the one hand, new technologies create vast possibilities for doing things better, faster, cheaper, more convenient and more personalized. Per-capita labor productivity has been steadily improving. However, returns on assets for businesses have declined over the past 50 years and businesses are not effectively capturing value from these new possibilities. Return on assets for US companies has fallen to almost a quarter of 1965 levels.

Digitization and dematerialization turn everything into software The twin forces of digitization (of information and media) and dematerialization (reducing goods and services to software) have profound implications for the economy and for society on a global basis. Digitization has transformative and disruptive implications. Dematerialization (of personal electronics, recorded music, books, video, etc) alters the economics of many businesses based on physical goods. As more computers become connected, a fundamental shift occurs as business processes that were previously handled by human beings are now being executed electronically. Because these processes usually involve computers talking to one another, they are unseen.

Rendering physical goods into digital bits and bytes impacts manufacturers, the supply chain, distribution, retailers and end users. A prominent example of the impact of digitization is tax-preparation software. Software that costs less than US$100 can replace many of the functions previously performed by a certified public accountant (CPA) or other tax-preparation professionals. Dematerialization refers to the transformation of physical products into software. Apps on a US$600 iPhone can replace pocket calculators, cameras, mobile telephone, video camera, clock radio, portable CD player, video cassette player and other items that historian Steve Cichon calculated would have cost over US$3,000 at Radio Shack in 1991. This is a 10:1 compression of value.

The on-demand economy turns products into services On-demand or sharing business is a large and growing component of the economy. Online platforms where sellers can offer goods or services to customers are attracting more attention and spending. Data from the 2015 National Technology Readiness survey estimates that over 22 million consumers spend nearly US$58bn annually in the on-demand economy. According to Crowd Companies, which tracks on demand platform businesses, there are over 280 companies that provide on-demand goods and services in 16 industries. This is an increase from just 76 companies and six industries two years ago.

The on-demand economy both creates and destroys jobs and businesses. In the case of Uber, there were investors that passed on investing in early rounds because they though the market was too small. Before Uber, the size of San Francisco’s taxi market was US$150m per year. In early 2015, Uber CEO Travis Kalanick disclosed that it had increased to US$650m with Uber accounting for US$500m. Uber expanded the market by delivering an experience to customers. However, this still came at the expense of incumbents - San Francisco’s Yellow Cab Co-Op filed for bankruptcy in January 2016.

Digitization turns services into software;

dematerialization turns gadgets into bits

There are over 280 companies that provide

on-demand goods and services in 16 industries

A fundamental shift as human business processes become

digital communications

The on-demand economy both creates and destroys

jobs and businesses




One of the challenges with the advent of digitization is that measuring digital goods and services as a proportion of GDP is difficult, if not impossible. Hours spent on the internet continue to climb (doubling from 2001 to 2011) and consumers access ever more free goods in the form of Facebook postings, blogs, online videos, games and other pursuits. As the volume of digital goods increases, this renders the traditional GDP measure less useful. Erik Brynjolfsson, professor of management at MIT, and Joo Hee Oh, assistant professor of management at the Erasmus University Rotterdam School of Management analyzed how much time people spent on the internet, and using that method they valued free internet services at about US$106bn a year. However, economist Tyler Cowen points that that this accounts for less than 1% of GDP.

The nature of organizations changes with digitization There is a lot of change happening in the nature of organizations, as businesses no longer profit from the frictions of scarcity but find ways to tap into abundance. Examples include Zappos and Valve Software, which have no organizational hierarchies - teams self-direct their work. Haier makes 55 million refrigerators a year with 80,000 employees, and the company’s CEO created 2,000 independent teams that elect their own leaders and vote on features, with the resulting successes resulting in tripling the company’s market cap. For big business, transformation is difficult but disruptive changes are often best developed at the edge of the organizations. Apple is master of this approach: their real innovation is organizational - creating small teams that work in stealth to go after new industries.

Singularity University’s Salim Ismail’s 2014 book Exponential Organizations highlights organizations with common characteristics - leveraging information technologies, massive transformational purpose and innovative organizations. Information-based industries are the best at leveraging technology to create value. Singularity University, in conjunction with the Hult School of Business, developed an ExO Score, and identified a correlation between stock market performance and the ExO score. The top five exponential organizations are GitHub, AirBnB, Uber, IndieGoGo and Google.

Figure 19

ExO market-cap improvement

Age (years) 2011 valuation 2014 valuation Increase (x) Haier 30 US$19bn US$60bn 3 Valve 18 US$1.5bn US$4.5bn 3 Google 17 US$150bn US$400bn 2.5 Uber 7 US$2bn US$17bn 8.5 Airbnb 6 US$2bn US$10bn 5 Github 6 US$500m (est.) US$7bn 14 Waze 6 US$25m US$1bn (in 2013) 50 Quirky 5 US$50m US$2bn 40 Snapchat 3 0 US$10bn 10,000+

Source: Salim Ismail, Yuri Van Geest, Michel S. Malone

In the past only large organizations could drop their transaction costs, but now small firms can take advantage of forces of technology and innovation. Financial services firms lack scalability largely because of regulation, which perversely protects against startups. Software and information-based industries are the best adoptors of technology. There are no digitally scalable industries in healthcare - yet. The measure of performance improvement from using Exponential Organization principles is powerful:

Despite disruption, traditional GDP measures

may not capture change

There is a lot of change happening in the nature

of organizations

Exponential organizations leverage IT, and

innovative organizations to create massive value

The Fortune 100 is disruptable with software

for the first time




Figure 20

ExO performance improvement

Company Business Performance improvement Airbnb Hotels 90x more listings per employee GitHub Software 109x more repositories per employee Local Motors Automotive 1000x cheaper to produce new car model, 5-22x faster

process for a car to produce (depending on vehicle) Quirky Consumer

Goods 10x faster product development (29 days vs 300 days)

Google Ventures Investments 2.5x more investments in early stage startups, 10x faster through design process

Valve Gaming 30x more market cap per employee Tesla Automotive 30x more market cap per employee Tangerine (formerly ING Direct Canada)

Banking 7x more customers per employee, 4x more deposits per customer

Source: Salim Ismail, Yuri Van Geest, Michel S. Malone

Because of technology, winners are winning bigger and faster Technology is accelerating value creation for new companies. Play Bigger Advisors’ study, Time to Market Cap: The New Metric that Matters, found that “category kings” (Facebook, Twitter and Uber, for example) that dominate their markets command over 70% of the total market value in their category, leaving everyone else to share the remaining 30%. While this dynamic is far from new, what’s different now is that winners are winning faster than ever. Companies headed for a US$1bn valuation will do so in a third of the time it took at the end of the first internet bubble. Companies founded in 2009-13 took 2.9 years on average to reach US$1bn valuation versus 8.5 years for those founded in 2000-03.

Digitization exacerbates divide between “haves” and “have mores” The dynamics of the shift to a technology-driven economy favors developed markets. Emerging markets tend to have under-skilled workers and less flexibility to improve their skills. Developed economies are early adopters of new technologies and with them, new business models such as on-demand and platform businesses. There are also differences between industries. The McKinsey Global Institute’s study Digital America: A tale of the haves and have-mores finds that the US economy is digitizing unevenly. ICT is the most digitized sector, followed by media, professional services and financial services. The least digitized sectors include healthcare, hospitality, construction and agriculture.

The McKinsey study finds that as the engine of digitization for the broader economy, the ICT sector accounts for about 5% of 2014 US GDP. Unlike prices for goods and services in most other industries, ICT prices have declined 63% from 1983 and 2010. Accounting for this price decline and impact on other sectors, McKinsey estimates the ICT sector represented roughly 10% of the 2014 US GDP. McKinsey devised an index that quantifies a growing gap between the most digitized sectors, “the have-mores” and the rest - “the haves”. The study found the gap between digital haves and have-mores is growing as the most advanced users pull away from everyone else. The less digitized sectors’ index accounted for just 14% of the most digitized sectors. McKinsey estimates that the US economy as a whole is realizing only 18% of its digital potential; digitization could increase 2015 GDP by over US$2tn based on its impact on labor markets, capital efficiency and multi factor productivity.

The shift to a technology-driven economy favors

developed markets

McKinsey estimates the US economy is realizing

only 18% of its digital potential

Exponential Organization realizes enormous

performance gains from tech-led innovation

What’s different now is that winners are winning

bigger and faster than ever




Software - Value migrates upward as it “eats the world” We’ve long expounded the view that value migrates upward in the technology “stack” from hardware to infrastructure software to software applications. The tech-sector business model is in transition from products to products plus services to services. The declining cost of compute and storage allows system architecture to evolve from optimized, tightly coupled systems (mainframe) to computationally “wasteful”, loosely coupled cloud-computing services. Hardware and services vendors are increasingly moving into software for technology and business reasons. Technology vendors that have a high concentration of revenues coming from infrastructure like IBM and Oracle are increasingly investing in and emphasizing SaaS platforms and applications.

Manufacturers and retailers increasingly attach software and cloud services to their physical products. Non-technology companies are compelled to invest in applications and services to differentiate their products in a global market. It’s no longer tech and internet companies vying for the most promising startups in Silicon Valley; companies in other sectors are seeking out investment and acquisition opportunities. Ford announced in May 2016 it would invest US$182m in Pivotal, the software development company spun off from EMC.

Companies in retail, agriculture, industrial equipment, automotive, consumer-packaged goods, energy, utilities, telecommunications, media and nearly every sector are compelled to embrace digitization, evolve or face disruption. There are numerous recent examples of non-technology companies making investments or acquisitions in software:

GE continues to invest aggressively in GE Digital, a separate business division to advance its vision for “software-defined machines”.

Under Armour has spent US$710m to acquire MyFitnessPal, MapMyFitness and Endomondo.

A German automotive consortium of Audi, BMW acquired Nokia’s HERE digital mapping business for US$2.7bn.

Conde Nast parent company Advance/Newhouse’s acquired big data analytics company 1010data for US$500m.

Defense contractor Raytheon acquired cyber-security firm Websense for US$1.9bn.

Audio and infotainment manufacturer Harmon International acquired software engineering and integration firm Symphony Teleca for US$548m and connected device software management provider Red Bend Software for US$170m.

Boeing acquired Peters Software, a provider of aviation training content for commercial and private pilots, as well as 2d3 Sensing, an imagery software company that processes intelligence and surveillance data.

Companies in many industries are seeking to replicate the Silicon Valley innovation model by establishing their own startup incubators and venture funds. Hundreds of companies have established innovation outposts in Silicon Valley and other hotspots like Herzliya Israel. According to Global Corporate Venturing there are over 1,500 corporate venture units responsible for nearly 1,800 corporate venturing deals in 2015, double the amount of 2014 deal activity, with transactions worth US$75.4bn in 2015, five times the amount in 2012. If software is in fact transforming every industry, investors need to evaluate non-tech companies based on their command and embrace of these new technologies.

There’s a transition from an economy of products

to an economy of products plus services

Non-tech corporations are investing to compete in a

digitized, connected world

Non-tech companies are establishing startup

incubators and venture funds

Non-technology companies are compelled

to invest in applications and services




What the future economy will look like There’s always risk in trying to predict what the future economy will look like, but there are threads that will characterize the coming decade. Many of the changes will be generational, as the digitally native millennials increasingly exert impact on commerce, culture and innovation. Author Don Tapscott in Growing Up Digital identified 88 million offspring of 85 million Baby Boomers; this is a generation larger and more connected than their parents. This next generation has internalized their digital behaviors with implications for business models and industries across the board.

The digital economy will continue to be defined by technology-driven business transformation. Over the past several years the rise of new sharing and public transportation models - exemplified by companies like Uber and Lyft has been accompanied by a declining rate of teenagers getting their first driver's licenses. Much of the value in the trade of traditional goods and services is the friction involved with connecting the asset with demand at the time and place of need. For companies that seek to compete more effectively, every company becomes a software company.

With a world of users connected to the internet, this gives rise to services that intermediate and provide a trusted framework. We believe companies like Netflix, Uber and AirBnB offer a blueprint for what the next generation of high-growth businesses will look like. The nature of new businesses is less asset intensive with more sharing of resources.

Products become experiences - with more subscriptions. This also touches on a potentially more profound shift away from a consumerist culture based on ownership of physical products. Sharing-economy businesses like Uber, AirBnB and Lyft address needs through a shared-asset model. Futurist Paul Saffo sees the economy moving to subscription models, predicting that users will subscribe to - and not own - robotic cars among other goods.

Connected products transform development. In “How Smart, Connected Products Are Transforming Companies,” an October 2015 Harvard Business Review cover story, co-authors Michael Porter of Harvard Business School and Jim Heppelman, CEO of PTC, outline how an emerging generation of cloud-service-enhanced products will change how companies will increasingly integrate data from customers and the distribution chain to drive truly interdisciplinary product development.

Design drives differentiation. Companies like Apple, Chrysler, Oxo and others have embraced the premise that smartly designed computers, cars and kitchen tools can stand out in established categories. While engineering remains critically important, creative design is paramount for developed-market firms to prosper in an environment where production is global, choices are myriad and competition is omnidirectional.

More value from fewer resources. Businesses are finding ways to produce more with less energy and materials. Additive manufacturing (3D printing) reduces manufacturing waste; smart grids and intelligent thermostats optimize energy consumption; and cloud computing allows startups and enterprises to innovate faster with lower costs and overhead. The ability for startups like Instagram and WhatsApp to scale to hundreds of millions of users with only a few dozen employees is a harbinger of hyper-efficient organizations.

There will be growing emphasis on the total

experience rather than discrete products

Creative design is paramount for developed-

market firms to prosper

Businesses are finding ways to produce more

with less energy and materials

There are threads that will characterize the

coming decade

The digital economy is defined by technology-

driven business transformation




A view of winners and losers It’s always challenging to identify the winners over the long term when the races have barely begun. What we’ve tried to do is provide a framework for investors to survey the pace of key innovative trends where we see value creation potential. One of the more insightful prognostications comes from author and entrepreneur Brett King in his recent book Augmented: Life in the Smart Lane (with Alex Lightman, Alex Lark and JP Rangaswani) lays out some distinct categories of long-term winners and losers once the Cambrian explosion of digital, physical and biological innovation comes to fruition. Some of the winners might include:

Big tech companies. Market leaders like Apple, Google, Amazon and Facebook have the ability to fund innovation and growth. Scale and a strategic focus have allowed Microsoft to reverse course. For other tech firms with high exposure to hardware and on-premise infrastructure like IBM, HP and Oracle there are still challenges ahead.

Artificial intelligence players. We are still at the beginning of the market but there are a number of public companies that have made big investments including Google DeepMind, Facebook’s Wit.ai, Salesforce with MetaMind, Microsoft, Baidu, and IBM’s Watson. There are a number of startups including Sentient Technologies, The Grid, Enlitic and x.ai among others. The relevance of AI in autonomous vehicles, healthcare, finance and other industries is broad and the industry has barely begun. ]

Next generation/Smart infrastructure. Brett King sees autonomous and electric vehicle manufacturers, smart grid operators, consumer renewables and household battery deployment, nanotech-based water treatment and desalination, robot and drone delivery networks and general smart city infrastructure as booming businesses over the next two decades. He estimates that 30 million people will be involved deploying solar and renewables by 2030.

Internet of Things. Physical things will increasingly get smarter, starting with appliances, smart cars, smart homes, and smart glasses. This will extend to sensors, screens and algorithms embedded in the world around us. Over time all sorts of day-to-day applications will be connected machine-to-machine. This will drive adoption of sensor nets as small computing devices will be ubiquitous. Every device that can be turned on or off, or that needs to be monitored will be linked to the cloud. Wearables, ingestibles and sensors will monitor health and well-being in real time.

Connecting the developing world. Emerging regions will see 2-2.5 billion people connected with access to the web predominantly via mobile devices. With cheap smartphone will emerge internet access followed by commerce. Facebook, Google, OneWeb and SpaceX have all committed to provide free internet access. African mobile growth is expected to result in over 1 billion connections by the end of 2016. The expectation is that by 2020, 66% of the world’s population will be connected to the internet, versus 23% in 2010.

HealthTech and FinTech. Both healthcare and financial services offer massive opportunity for new technology expertise. Startups will displace inefficient and ossified incumbents and partner for core capabilities.

One of the more insightful prognostications comes

from author and entrepreneur Brett King

Scale and a strategic focus have allowed

Microsoft to reverse course

30 million people will be involved deploying solar and renewables by 2030

Every device that can be turned on or off, or that

needs to be monitored will be linked to the cloud

By 2020, 66% of the world’s population will be connected to the internet,

versus 23% in 2010




The augmented and virtual reality ecosystem. VR and AR will give rise to a new generation of entertainment and applications, Content creators will develop new hybrids of games, live experiences and film. There will be increasing integration of social and collaborative capabilities, especially in business.

“Metamaterials” and 3D Printing. Development of new composites and hybrid materials combined with additive manufacturing technologies will give rise to essentially smart manufacturing and construction. Techniques like nanotechnology will enable development of transparent aluminium, carbon-fibre nanotubes, smart clothing, reactive and responsive materials that morph or alter their properties, smart polymers, piezoelectric, thermoelectric and photomechanical materials. Over time, materials will be able to be engineered at the atomic level. We will see 3D printers that can fabricate machines, homes, and appliances.

Losers could include:

Big energy. King sees four key forces that will challenge fossil fuel producers and incumbent energy systems: ultra-cheap renewables, smart grids, electric vehicles and energy storage systems (batteries like the Tesla Powerwall and fuel cells, etc). Gas and service stations and other participants in the ecosystem will go out of business. Mining will shift from extracting fuels to extracting resources to power the new applications.

Big healthcare and pharma. The new paradigm for healthcare businesses will be data, engineering and models rather than chemistry, patented medicine and surgery time. There will still be hospitals and doctors, but diagnosis will be derived from sensors. Medical emergencies will decline as self-learning, healthcare-dedicated healthcare application will prove better than humans at diagnosing and offering treatment suggestions.

Banking, insurance, regulators and finance in general. The adoption of technologies like blockchain will eliminate the frictions to transactions that banks have capitalized on. AI-based applications will replace many jobs performed by humans, whether a bank branch teller, financial adviser, accountant, loan officer or others focused on advising on bank products or financial rules.

Small- to mid-sized colleges and universities. The US$1.3bn in student debt and perceived lower value of graduate studies will lead to consolidation in tertiary education. We are already seeing declines in law school applications as a harbinger of this trend.

Over time, materials will be able to be engineered

at the atomic level

Gas and service stations and other participants in

the ecosystem will go out of business

The new paradigm for healthcare businesses will

be data, engineering and models

The adoption of blockchain will eliminate

transactions frictions that banks have capitalized on


Section 2: The Unicorn Era winds down 2020 innovation


The Unicorn Era winds down Funding and entrepreneurship remain critical to realize value from innovation. In 2016, funding is getting tougher for late-stage private companies and IPOs are harder to come by. M&A remains active as an exit strategy for private firms however, and as valuation expectations ratchet down (along with mutual fund markdowns) we anticipate healthy levels of acquisition activity. The pace of startup creation, while steadily in decline in recent decades, rebounded in 2015 from a more recent five-year slump. Whether this can be sustained is yet to be seen, but the cost and availability of enabling technologies continues to be conducive to startups.

“Unicorn” was a term that entered the broader lexicon in 2015, referring to privately held companies valued at over US$1bn. As a symbol of robust technology-led innovation propelling a new boom, and also as a symbol of VC excess, hype around Unicorns peaked in late-2015 and there’s been a return to more measured assessments of valuation and profitability. Unlike the first internet bubble in the late-1990s, valuations for Unicorns are not being fueled by a frenzy of IPOs accompanied by lofty valuations in the tech sector. In contrast, the pace of IPOs has been more measured, and despite some volatility in the public markets early in 2016, tone of business among public companies does not appear unsettled. For VC-funded private companies, it’s a different story.

2015 ended up being a year of adjustment for the venture-capital world as some of the largest venture-backed businesses struggled to meet their valuation expectations. As a result, we have seen a weaker IPO and M&A market. Investors have begun to pull back from overvalued startups, particularly in the later-stages as they become more discriminate about which of these companies can succeed in the volatile public markets

In January 2015, Fortune magazine heralded the “The Age of the Unicorns”, noting that there were over 80 startups valued by venture capitalists at US$1bn or more. By January of 2016, there were at least 229. In early May 2016, Fortune listed 174 companies with valuations over US$1bn, following a wave of markdowns. Research firm CB Insights identified 588 US tech companies backed by venture capital or private equity with valuations over US$100m in the IPO pipeline in 2015. Of these, 39% raised additional financing or exited via IPO or M&A, with the remaining 61% still active. There’s a robust roster of aspiring public companies, with over 500 US tech companies estimated to be in the IPO pipeline (internet companies account for nearly two-thirds of these).

Figure 21

Top 10 companies valued at US$1bn or more by VC firms Company Latest valuation

(US$bn) Total equity

funding (US$bn) Last

valuation Uber 51.0 7.4 Aug 15 Xiami 46.0 1.4 Dec 14 Airbnb 25.5 2.3 Jun 15 Palantir 20.0 1.9 Oct 15 Meituan-Dianping 18.3 3.3 Jan 16 Didi Kuaidi 16.0 4 Sep 15 Snapchat 16.0 1.3 Feb 16 Flipkart 15.0 3 Apr 15 Space X 12.0 1.1 Jan 15 Pinterest 11.0 1.3 Feb 15 Source: Wall Street Journal

2016 is shaping up to be a challenging year for

startups seeking funding

In 2016, funding is getting tougher for late-stage private companies

“Unicorn” was a term that entered the broader

lexicon in 2015

2015 ended up being a year of adjustment for the

venture-capital world




The bubble begins to deflate Late 2015 saw the beginning of the end of the Unicorn Era. A couple of Wall Street Journal investigative articles into Theranos and Zenefits shone a spotlight on what could go wrong at highly valued private companies with lofty growth expectations. There has also been an increase in failures among startups including Quirky, Secret, Fab.com and others. By 1Q16, the late-stage financing environment has become more challenging for startups.

Mutual funds also began to mark down valuations from private investment rounds, and the pace of private investments by large mutual fund firms is slowing as firms found it more difficult to agree on acceptable valuations. In March 2016, an analysis by the Wall Street Journal found that Blackrock, T. Rowe Price, Fidelity Investments and Wellington Management valued investments in 13 of 40 closely held startups valued at US$1bn at an average of 28% below the purchase price.

2016 could be the year a number of these so-called “Unicorns” find themselves unable to continue to operate with losses as they continue to burn cash and access to funding dries up. Despite the pull-back in late-stage investment, there has been a recent uptick in early-stage investments as venture capitalists move away from mature startups with absurd valuations and look to auspicious companies in their early stages.

As the bar to the capital markets grew higher in the wake of Sarbanes-Oxley legislation, the ability to secure returns through IPOs has been increasingly difficult since the number of annual offerings peaked at 650 in 1996. For venture-backed firms, the past decade has been challenging, as the number of offerings dropped from 92 in 2007 to just seven in 2008. There is a difference in 2016 from the first internet bubble in 1999-2000: the number of IPOs rebounded to 117 in 2014 and 77 in 2015, but ground to a halt, with no venture-backed IPOs in 1Q16.

Figure 22

Annual US venture-capital exit activity

Year Total M&A deals (No.)

M&A deals with disclosed

values (No.)

Total disclosed M&A value

(US$m)

Average M&A deal size (US$m)

No. of

IPOs

Total offer amount (US$m)

Average IPO offer amount

(US$m) 2007 488 200 30,745.5 153.7 92 12,365.5 134.4 2008 423 134 16,236.9 121.2 7 765.0 109.3 2009 361 109 12,364.9 113.4 13 1,979.8 152.3 2010 545 151 17,713.5 117.3 66 7,409.5 112.3 2011 500 169 24,093.2 142.6 50 10,441.1 208.8 2012 492 132 22,694.2 171.9 48 21,354.1 444.9 2013 394 95 16,909.8 178 80 10,998.2 137.5 2014 485 140 48,139.9 343.9 117 15,512.1 132.6 2015 372 84 16,291.2 193.9 77 9,378.9 121.8 Source: Thomson Reuters, National Venture Capital Association

In 2015, venture-backed IPOs raised US$9.4bn from 77 listings compared to a record year in 2014 where venture-backed IPOs raised US$15.3bn from 115 listings. The decline in venture-backed IPOs was in line with the overall decline in filings of public exchanges with the 77 venture-backed IPOs accounting for 42% of total IPOs last year. Increased investment activity by non-traditional firms has allowed companies to stay private longer - which is

Venture-backed IPOs were down in 2015

For startups, access to venture capital and angel

investors is particularly important

Late 2015 saw the beginning of the end of

the Unicorn Era

Mutual funds also began to mark down valuations from private investment

rounds




one reason IPO activity has declined However, there are currently 44 venture-backed companies that have filed with the SEC for an IPO, but it is uncertain whether there will be a rebound.

Acquisitions prove the exit of choice According to the 451 Group there was US$594bn in total M&A transactions in 2015, with almost 4,000 deals in telecom media and technology. Overall spending was equivalent to all of 2013-14 combined. There were nearly 20% more deals in 2015 than in 2014. Big names stepped back into M&A like HP which bought Aruba Networks. Microsoft completed over 20 transactions. IBM completed 15 deals in 2015, 3x the company’s yearly average. Divestitures also accounted for 156 deals worth US$60bn, the highest ever in the US. In 2014 there were 151 for US$43bn.

Software accounted for almost half of the number of M&A transactions reported. The average M&A deal size declined to US$194.8m in 2015 from US$343.9m in 2014, representing a YoY decline of 43%. Despite the decline from last year, total deal value in 2015 was only down 4% compared to 2013 and up 60% from 2012. A key reason for the decline in M&A is that valuations had been too high and companies are less willing to pay premiums without significant revenue generation.

Figure 23

US venture-backed company M&A deal activity

Source: Thomson Reuters, National Venture Capital Association

The activity of VC-backed exits in 2016 has dropped, as well with only two so far over US$1bn and not a single deal so far from US$500m-US$1bn this year. By comparison the previous five years saw between seven and nine exits between US$500m-US$1bn. There were six exits over US$1bn in 2014 declining to one in 2015.

In terms of sentiment the private market tends to lag the public market. Two-thirds of respondents in 451 Group’s 1Q16 survey see private company valuations declining for M&A. Valuations in the M&A market are polarized as well with high multiple and low multiple deals and little in between. Ingram Micro bid US$6bn for a company with US$40bn in sales. Polycom and Lexmark were valued at 1x trailing 12 month sales. At the high end there are

7.6

4.1

8.6

27.8

2.23.7

6.9

3.6

116

120136

113

95

77

109

91

0

5

10

15

20

25

30

35

40

0

20

40

60

80

100

120

140

160

180

200

1Q14 2Q14 3Q14 4Q14 1Q15 2Q15 3Q15 4Q15

Total deal value (RHS)Total M&A dealsDeals with disclosed values

(No. of deals) (US$bn)

Last year was stronger for venture-backed M&A

Overall 2015 M&A spending was equivalent

to all of 2013-14 combined

In terms of sentiment the private market tends to

lag the public market




similarly high multiples. Cisco paid US$1.4bn for Jasper’s US$100m in bookings, Cvent was taken private at 8x sales and Textura was acquired by Oracle for 7x revenue. A schism like this in the M&A markets is not a healthy sign. According to 451 Group, deals would normally be 3.5-4x sales. When pricing is polarized, it’s feast or famine in M&A.

“Down-round” exits are coming According to a 1Q16 survey by the 451 Group, tech bankers expect valuations to decline in 2016 - 64% of those surveyed, with only 14% expecting valuations to increase. About 80% expect valuations for private companies will decline versus 2% expecting valuations to increase. Some 57% of bankers think VC funding will get less available versus only 7% that expect more available funding. The only time this sentiment was more bearish was in 2009. So far in 2016 there have been 13 companies acquired with eight sold for less than their trading price in 2015. Companies are getting discounted in the public markets, and this is freezing the market in terms of venture-backed exits.

Venture capital raising less money and closing fewer funds 2015 saw 235 US venture funds (13% fewer YoY) raise a total of US$28.2bn, down 9% YoY in dollar terms. Despite being down, overall it appears that the fundraising environment is healthy with a diverse mix of fund sizes and almost US$60bn raised for distribution in just two years.

Figure 24

Annual US venture-capital fundraising activity

Year Number of funds Venture capital (US$m) 2008 214 25,054.9 2009 162 16,103.8 2010 176 13,283.6 2011 192 19,080.5 2012 218 19,904.9 2013 209 17,753.4 2014 271 31,094.4 2015 235 28,151.7 Source: Thomson Reuters, National Venture Capital Association

The Center for Venture Research found that US angel investors funded 73,400 entrepreneurial ventures worth some US$24.1bn in 2014, down 2.8% from US$24.8bn in 2013. The number of active investors increased 5.9% from 2013 to 316,600, but data showed that the change in both total dollars and the number of investments brought about a smaller deal size by 6.4% from 2013, indicating that angel investors were active investors but at decreased valuations than from prior years pointing towards a market correction in valuation. Angels invested in a variety of sectors, with software leading at 27%, followed by healthcare (16%), IT (10%), retail (9%), finance (8%), and industrial (5%).

Putting money to work in 2015 According to the MoneyTree Report by PwC and the NVCA, startup investments in 2015 were US$59.1bn, up 22% from 2014, making this year the second highest in 20 years. The total number of deals declined 1.3% to 4,380. Although IPO and M&A activity have been down this year as investors

Funding decreased 9% YoY for VC firms in 2015

but . . .

. . . VC investments increased 22% YoY in

2015

Tech bankers expect valuations to decline in

2016




look to make smarter investments, solid VC capital deployment is indicative of a healthy ecosystem that is committed to supporting entrepreneurs. The amalgamation of technology across sectors is playing an important role in companies’ business models and we are seeing how new emerging tech-enabled segments are disrupting traditional industries. Over half the deals in 2015 went to seed and early stage companies with more than 1,400 companies raising VC money for the first time.

The software industry continues to receive the highest level of funding with a total of US$23.6bn in investments from 1,763 deals down 5% from 2014. Internet-specific company investments increased 35% compared to 2014 as the number of deals remained flat. Of importance to note is that despite a strong year for VC investment in 2015, the fourth quarter saw the lowest number of deals worldwide in nearly three years. This drop in deal flow was most prominent in later stages of venture investing. Expansion stage of investment was down 53% in dollars and 10% in deals from 2014 and later investments saw a decline of 33%. The reason for this slowdown - there isn’t much of an appetite to invest in companies with ambitious growth plans that are burning cash especially during these economically volatile times. However, this slowdown in funding is not necessarily a bad thing as Silicon Valley experiences its own soft landing in the venture business. Valuations have become exceptionally inflated and it is our belief that this normalization is healthy and also warranted.

Startup activity bounces after steady decline Entrepreneurial activity has long been a core engine of growth in the US, but the pace of new business creation has been on the wane. Business startups exceeded business failures by about 100,000 per year until 2008 according to Gallup. However, that year the number of failures begin to exceed the number of startups by 70,000 for the first time since records were kept in the late 1970s. The US startup rate has been falling, a concern for those that view entrepreneurship as a cornerstone of a healthy economy.

Declines in both job creation and job losses provide evidence that there is less risk-taking in the economy. As of 1Q15 the number of jobs created by new businesses fell 7% from 4Q14 on a seasonally adjusted basis, an 18% decline from a decade ago. The number of job losses due to business closures is also down 21% from 2005. The share of firms less than one year old (a key measure of business vibrancy) has declined from over 15% in 1977 to 8% in 2013 according to data from the US Commerce Department and analysis by the Wall Street Journal. According to data from the Labor Department, newly formed businesses are adding jobs at a much slower pace than a decade ago.

The pace of new business formation has rebounded after a five-year decline. By one measure, startup activity in the US remains healthy, with VCs investing US$59.1bn in 2015. The Kauffman Index of Entrepreneurial Activity is the leading tracker of startup activity in the USA and a solid indicator of future innovation. The 2015 report points to the largest YoY increase in startup growth in 20 years, reversing a five-year downward trend. This rise in activity is a good sign for job creation, innovation, and economic advancement as it is an early indicator of new business creation. Notably, Baby Boomers and immigrants are among the most active starting new companies.

Startup activity in the US remains healthy

Declines in job creation and job losses provide

evidence that there is less risk-taking

The software industry continues to receive the highest level of funding




Figure 25

Kauffman Index of Entrepreneurial Activity

Source: Kauffman Foundation

A couple of interesting trends to highlight from the Kauffman report indicate that the biggest startup gains have been among men, who make up 63% of new entrepreneurs. 52.4% of entrepreneurs are over the age of 45 deflating the hype that most are millennials just out of college - the fact is that younger people have a harder time securing resources and are burdened with high student loan payments.

One recent study concludes that there’s no shortage of entrepreneurs with good ideas, but they are struggling to transform vision into companies with a broader reach. A New View of the Skew: A Quantitative Assessment of the Quality of American Entrepreneurship, a research and a policy brief from Massachusetts Institute of Technology’s Catherine Fazio, Jorge Guzman, Fiona Murray and Scott Stern sought to explain why only a very small fraction of all new businesses experience the explosive growth in terms of jobs, revenue, or valuation that propel the economy. They found a divergence between quantity and outcome based measures of startups.

Quantity-based measures indicate a three-decade-long decline in the US rate of entrepreneurship and business dynamism (the pace at which the economy reallocates economic activity). In contrast, outcome-based measures based on early-stage angel and VC financing for new ventures show the rate of entrepreneurship on a significant upswing in recent years. The authors apply different criteria, including average growth potential for a group of new firms, probability of success for regional cohorts and regional acceleration potential - to derive a qualitative measure of US entrepreneurial activity. They found that the expected number of successful startups growth outcomes has followed a cyclical pattern sensitive to capital markets and overall macro conditions. In their findings it’s not the number of startups that translate to greater successful outcomes, but that regional differences may be far more impactful.

The increasing capital efficiency of software and internet firms (with the proliferation of open-source software and public cloud computing) reduces initial capital needs for many startups. This, in turn, lowers barriers to entry for startups, while altering the value equation. Because the amount of capital

0.30

0.28

0.30 0.30

0.32

0.34 0.34

0.32

0.30

0.28

0.31

0.26

0.28

0.30

0.32

0.34

0.36

2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014

(%)

Entrepreneurial Index ticks up after five down

years

Lightweight innovation has reduced capital needs

for software and internet startups

There’s no shortage of entrepreneurs with good

ideas, but they struggle to grow businesses

Quantity and outcome based measures tell a

different story on entrepreneurship




needed is less, a VC investment in itself is no longer a de facto competitive advantage for startups competing for the same markets to the extent that capital access itself is strategic. The value that venture investors provide is management insight, access to executives and customers, mentoring and other intangibles. It’s a combination of external factors, such as the quality of board, advisors, and regional culture steeped in entrepreneurial success that determines whether a startup can create scalable value.

R&D spending increases, but US pace is lagging other nations R&D investment is still a key indicator of innovative activity. According to the 2016 Global R&D Funding Forecast, sponsored by the Industrial Research Institute (IRI), global R&D investments are expected to increase by 3.5% in 2016 to US$1.95tn in purchasing power parity (PPP) values for the over 110 countries with R&D investments more than US$100m. R&D spending is highly correlated with sustained revenue growth. We compared leading technology firms’ 10-year Cagr for revenues and R&D spend, and what we found was an R2 of 0.9366. It’s not necessarily a leading indicator but undoubtedly a coincident indicator, reinforcing that tech companies that deliver sustained revenue growth maintain the pace of investment.

Figure 26

Growth in R&D investment is highly correlated with revenue growth

Source: Factset

IRI/R&D Magazine forecasts 3.4% growth in US R&D spending to US$514bn. After accounting for 2016’s expected inflation rate of 1.5%, R&D growth in real terms will be closer to 1.9%. 2016 was a difficult year to plan for as enterprises, government and academia struggled to get their heads around the macro uncertainties of a volatile global economy. Although spending is forecasted to increase, there are some concerns as a result of China’s slowdown, recent US stock market volatility, restrictions on federal spending due to earlier sequestration programs, and limpid growth in the US in addition to a strengthening dollar. Despite these drags, the US still accounts for the largest single country’s R&D investment, representing more than a quarter of all global R&D spending.

Apple

GOOGL

NFLX

AMZN

CRM

MSFT

ORCL

IBM

RHTAKAM

CTXS

SAP

CHKP

ADSK

ADBE

CA

CSCO

INTC

R² = 0.9366

(5)

0

5

10

15

20

25

30

35

40

(5) 0 5 10 15 20 25 30 35 40 45 50

Sales growth (%)

R&D growth(%)

Global R&D investments expected to increase by

3.5% in 2016 to US$1.95tn

The US still accounts for the largest single

country’s R&D investment

Companies with sustained growth in R&D also have

sustained sales growth




Figure 27

Change in business expenditures on R&D, 2009-13

Source: OECD, Duke University’s FUQUA School of Business

There is a correlation between a country’s economic growth and innovation investment as measured by GDP and its R&D growth. China has shown a strong willingness to enhance its research and development despite its recent slowdown. The IMF forecasts GDP growth for China at 6.3% and a 2.8% increase for the US with smaller increases across many European countries. For 2016, R&D Magazine forecasts China R&D to increase 6.3% to US$396bn. In the US, investors weary of high R&D spending without payback have become activists. In 2015, investor Nelson Peltz tried to wrest control of DuPont from then-CEO Ellen Kullman arguing that the company was not moving in the right direction and that cuts to the R&D budget needed to be made. There are two schools of thought about the value of R&D investment but the data favors the benefits of stronger R&D spend promoting growth and value creation.

While R&D spend does not predictably translate into robust earnings, the greatest companies invest in innovation as it is a primary source of competitive advantage, efficiencies and productivity. Based upon our experience, we find there is a positive connection between R&D spending, sales growth and stock returns. Essentially, R&D is the key to staying ahead of competition in a quickly evolving landscape.

Software and internet investment outpaces other industries According to PwC’s annual study of the 1,000 largest globally publicly traded firms based on R&D investment, spending rose 5.1% to US$680bn in 2015 versus 1.4% in 2014, marking the strongest increase in the last three years. The top three industries by R&D spending continue to be computing & electronics, healthcare and auto. Software and internet companies increased R&D spending by 27.4% from 2014 to 2015 which was the largest increase among all industries. This reflects the highly dynamic nature of the competitive landscape and market dynamics for software and internet companies, compelling aggressive investment in innovation. R&D spending in China increased 150% to US$55bn between 2007 and 2015. The US remains the top R&D spender at US$145bn, remaining the number one location for

(20)

(10)

0

10

20

30

40

50

60

70

80

90

China Korea Germany Japan US UK Canada

(%)

Calculated in constant dollars and adjusted for purchasing power parity

Software and internet were the biggest movers

in R&D spend between 2014 and 2015

China and Korea lead in growth of R&D

investment

There is a correlation between a country’s

economic growth and innovation investment




innovation according to PwC. However, China is increasing the scope of its R&D investments. In 2007, R&D spend in China accounted for 23% of the US total, today it is 38%.

PwC found distinct benefits to a company with a global R&D footprint. Companies that distribute spending outside their headquarters country tend to outperform companies that are less globalized. Companies that spent 60% of R&D budgets overseas saw operating margins and return on assets that on average 30% higher than those companies that kept R&D locally concentrated. The reason for the outperformance is credited to the payoff companies receive from the placement of competencies and knowledge on a global scale as well greater success in understanding local market needs.

Figure 28

Global innovation spending by industry

Source: PwC

Innovation benefits from technology enablers We identify several technological accelerators of innovation. The convergence of enabling factors that reduce innovation cost - open source, cloud computing, mobility and the commoditization of hardware/software and bandwidth - provides a catalyst to create value and accelerate innovation, all facilitated by software. Key enabling trends include:

Exponential declines in the cost of computing, bandwidth and storage, which lower barriers to ideation and enable unparalleled scale for new ventures.

Cloud computing, which enables users to dynamically procure IT resources and content on an as-needed basis.

Open-source software, which empowers users with flexibility and increasingly robust functionality at little or no cost.

The mobile internet, which connects users with information and applications regardless of device, connection or location.

Growing penetration of internet users and social networking, which results in near-ubiquitous reach of services.

Computing and electronics

25.9%

Healthcare21.1%

Auto16.2%

Industrials10.7%

Software and internet9.2%

Chemicals and energy6.5%

Aerospace and defense3.3%

Consumer3.3%

Telecom2.1%

Other1.7%

Over the next decade, computing will become increasingly embedded

in daily life

Computing and electronics invest the most of any

industry in R&D

There are benefits to a company with a global

R&D footprint




Computing overall is becoming more intuitive and pervasive with the evolution of more powerful software, rapid growth of endpoint devices, availability of instant-on connectivity and declining costs of hardware, bandwidth and storage. We see the continuous elevation of simplicity of experience to the user as logic controls the underlying systems, processes and infrastructure with increasing power. These concurrent trends have a transformative impact, accelerating innovation across the broader economy in traditionally non-tech as well as technology industries.

Commoditization of computing, bandwidth and storage The commoditization of computing, bandwidth and storage has proven to be a continuous dynamic. The principle of Moore’s Law, which holds that processor performance can double every 18 months, has held fast since the 1970s, while price performance of Dram continues to improve along a similar dynamic.

In fact, the dynamic of exponential cost and performance improvement is occurring across a broad range of technologies. While improvement occurs at different rates, the consistent historical trend remains a common dynamic across different hardware technologies.

Figure 29

Moore’s Law

Source: Wikipedia (Wgsimon)

Butters’ Law (named after Gerald Butters, former head of Lucent's Optical Networking Group at Bell Labs) posits that the amount of data coming out of an optical fiber doubles every nine months, essentially cutting the cost of data transmission in half over that period.

Cost of computing, bandwidth and storage

continues to decline

Moore’s Law has held fast since the 1970s

Technology has a transformative impact, accelerating innovation

across the economy

Exponential cost and performance gains occur across a broad range of

technologies




Figure 30

Time to double (or half) Dynamic Ram memory “half pitch” feature size 5.4 years Dynamic Ram memory (bits per dollar) 1.5 years Average transistor price 1.6 years Microprocessor cost per transistor cycle 1.1 years Total bits shipped 1.1 years Processor performance in MIPS 1.8 years Transistors in Intel microprocessors 2.0 years Microprocessor clock speed 2.7 years MIPS = Millions of instructions per second, a measure of processing capacity. Source: Ray Kurzweil, KurzweilAI.net

Beyond Moore’s Law In 2Q15, Intel CEO Brian Krzanich announced that Moore’s Law - the principle that the number of transistors incorporated in a chip will double every two years - is slowing to a pace of every two and a half years. At the current pace of innovation, there’s a widely held view that Moore’s Law will end with the 7nm in 2020 or 5nm in 2022 and then beyond that, physics will simply not make it practical to shrink the transistors further. In July 2014, IBM announced plans to invest US$3bn in R&D for two research programs that will extend computing capabilities beyond Moore’s Law. Futurist Ray Kurzweil believes that photolithography will run its course by 2019, but quantum or optical computing will continue the exponential growth in computing power. Some of the most promising research involves graphene, organic transistors, 3D stacking, neurosynaptic chips and quantum computing.

Moore’s Law has held up for over 50 years and many businesses have used predictable advances to time PC purchases, databases and servers. If chips are doubling in power every two years then software was being written in tandem in anticipation of the increase. This law was always considered more of an economic rule of thumb that helped companies schedule manufacturing targets. With the move to more centralized cloud computing, chip power is still important for the capabilities of a cloud data center but the speed of the processor in client devices is no longer as relevant, as new applications increasingly rely on compute power delivered as cloud services. Businesses however are no longer purchasing hardware in regular upgrade cycles as they have done in the past. Companies like Apple are seeing that sales are declining as customers begin to hold on to older models satisfied with the performance improvement they receive from the cloud. Over time, business models predicated on hardware upgrades eventually give way to commoditization or reliance on higher-margin services and software.

Cloud computing accelerates incubation of new ideas Cloud computing has lowered barriers for technology adoption, reduced time to market and placed unprecedented computing power in the hands of startups, departmental IT staff and others heretofore lacking access to scalable systems. Importantly, the availability of cloud-based resources is lowering barriers for startups and fuels accelerating innovation. The capital efficiency that can be applied to developing, introducing and then iterating on these new applications is declining. The capital efficiency for a venture firm trying to launch new companies is becoming much better. The Law of Accelerating Returns on these technologies is at play.

Pricing declines continue to benefit users of cloud computing. According to a report by Tariff Consultancy (TCL), average entry-level cloud computing prices declined by 66% over a two-year period ending in November 2015. In March

Availability of cloud-based resources is

lowering barriers for startups

Exponential cost and performance

improvement occurs across technologies

There’s a widely held view that Moore’s Law

will end with the 7nm in 2020 or 5nm in 2022

Pricing declines continue to benefit users of cloud

computing

Moore’s Law has held up for over 50 years




2016 Amazon Web Services (AWS) reduced prices for the 51st time since 2006, with Microsoft following close after. Google announced at the beginning of 2016 that it’s pricing for compute prices would be 15-41% cheaper than AWS. While the pace of price cuts has slowed over the past couple of years, the market has consolidated around AWS and Microsoft, with Google running fast to catch up, as hyper-scale public cloud service providers.

Cloud computing facilitates a range of new use cases around social computing, mobile services and high-performance analytics, all of which have the potential to create new economic value. Cloud computing reflects the industrialization of information technology. While this has disruptive effects on certain providers of infrastructure components and services, there are many beneficiaries of broad-based adoption of cloud computing.

Businesses benefit from lower costs for technology, IT staff and overhead, while gaining benefits of agility. Lowering the marginal cost of failure accelerates innovation and allows for rapid value creation. The extensible aspects of cloud computing allow businesses to scale online operations far more quickly than possible on their own, for less capital, with less risk.

Consumers benefit directly and indirectly from the availability of cloud-enabled applications. Cloud-enabled search, entertainment, information services, location-based services and applications democratize access to culture, knowledge and commerce. Dematerialization of physical goods (for instance, the digitization of books, music and video) into cloud-delivered services reduces friction around information flow.

It’s our view that cloud computing and related technologies such as open-source software and higher-level programming languages are analogous to James Watt’s steam engine in the first industrial revolution, with compute power substituting for coal. As such, disruption within the IT infrastructure market is likely to continue along a path of creative destruction as proprietary infrastructure hardware and software providers see their ability to charge premium “rents” and enjoy attractive profit margins steadily erode.

A wholesale economic model disrupts a bespoke industry. We point to the 2006 introduction of Amazon’s EC2 as the turning point where scalable compute and storage became available on a self-serve, pay-per-use model. Amazon Web Services’ annual run rate now exceeds US$10bn, with hundreds of thousands of startups using the service.

Native cloud architectures commoditize infrastructure hardware and software. A mature Platform-as-a-Service (PaaS) layer decouples the application logic from the underlying infrastructure and makes infrastructure components increasingly interchangeable and subject to the forces of commoditization.

Open source - Freeing innovation and enabling cloud computing Open-source principles inherently enable innovation, not just in software, hardware and services, but through the derivative benefits to technology users in any endeavor. There are over one billion lines of freely available open-source code that developers and business users can access to create applications and new businesses. Free open-source software reduces costs for startups as well as new projects within IT organizations. Even with paid technical expertise and support, the ROI tends to be overwhelmingly favorable for open source. Developers looking to for code can choose among millions of free projects uploaded to GitHub, download what they need and use it within minutes.

Open source has been critical to enable

innovation, not just in software

Cloud computing enables social computing, mobile

services and high-performance analytics

Open-source software and higher-level programming languages parallel James

Watt’s steam engine




Open-source alternatives to proprietary technologies gain ground The shift to cloud computing accelerates as open source is increasingly viable for the enterprise. There are increasingly viable open-source alternatives to proprietary software, all the way up the stack from OS (Linux, Android) to database (MySQL, MongoDB, Cassandra), Big Data/data-warehousing management (Hadoop, Talend), content management (Drupal), application server/ middleware, (JBoss, MuleSoft), analytic tools (R language, Pentaho, Jaspersoft), customer relationship management (CRM) (SugarCRM) and many more. It is not unusual for startups, particularly in the internet or e-commerce arena, to avoid use of proprietary software entirely. Large technology vendors including IBM and even Microsoft have embraced open source as a development model, and Microsoft has open-sourced a number of key offerings including the .NET application framework and Xamarin mobile application development tools.

The mobile internet: “Any device, always on, anywhere” The explosive adoption of internet-enabled smartphones and the growing availability of wireless internet provide key innovation vectors for new services and applications, including micropayments, content streaming, multiplayer gaming, location-based services and a plethora of targeted applications. In the coming decade, the deployment of high-speed 5G and beyond mobile networks will help advance the vision of always-on high-speed internet connections and adoption of context-aware, data-based applications.

Figure 31

Global IPv6-capable devices will reach 7.7bn by 2020

Note: Internet Protocol version 6 (IPv6). Source: Cisco 2016

The explosion of internet-enabled smart devices has created a need for each device to have its own unique address that can communicate with other devices and the internet. IPv4 addresses are the current protocol that devices use to communicate on the internet. However, these addresses have been exhausted. Transitioning to the better IPv6 protocol will offer significant advantages giving every device a globally routable public IP address on the internet. Cisco estimates that by 2020, 92% of smartphones and tablets will be IPv6 capable.

Advanced mobile networks will support a new generation of innovative applications, including virtual reality content streaming, shopping, 4K (ultra-high definition) television, collaboration, social networking, video conferencing, robust gaming, additional personalized offerings and Internet-

2.67

3.53

4.44

5.38

6.47

7.65

0

1

2

3

4

5

6

7

8

9

2015 2016 2017 2018 2019 2020

Number of devices (bn)

Adoption of smartphones and the wireless internet helps realize the vision of

pervasive computing

The leading cloud platforms, including

Google and Amazon EC2, are built on open source

Smartphone penetration will connect billions to the internet for the first time

More advanced mobile networks will support a new range of innovative

applications




of-Things applications. This in turn expands the innovation of connected devices beyond tablets, smartphones and laptops. Increasingly, there will be wireless-connected audio, video, sensors, industrial equipment, automobiles and appliances. The availability of greater bandwidth will facilitate adoption of advanced applications, which should drive further growth of data traffic.

New types of mobile applications expected to see healthy growth include money transfer by chat-based bots, short message service (SMS), mobile search and browsing, location-based services, mobile music and video services, near-field communications services, mobile health monitoring and many others.

Mobile explosion a hotbed for innovation According to the International Telecommunications Union, there are approximately 7.1bn mobile telephone subscribers, representing 96.8% of the world’s population - up from just 738mn in 2000. The penetration rate for fixed-line telephone service in 2015 was just under 11% of the world’s population.

Mobile adoption is still growing but at a slower pace than in the early 2000; developed markets are growing more slowly as they begin to approach saturation levels. In 2020, GSMA estimates average annual subscriber growth rate will be 3.9% compared to 7.7% over the last five years with a penetration rate of 88% of the population. In developing countries, the current 59% penetration rate suggests room for growth. In 2015, 2.5bn users throughout the developing world were able to get internet access with a mobile device. GSMA estimates that 90% of the incremental 1 billion new mobile subscribers forecasted by 2020 will come from developing markets.

In 2015, the mobile ecosystem accounted for 4.2% of global GDP, representing US$3.1tn of economic value added. As the mobile sector’s global footprint continues to grow, it provided employment for approximately 17 million people across the world and contributed to US$430bn in taxation.

A major advance for mobile in developing countries is the role in payments and banking. Mobile money services are currently available to 1.9bn people globally, bringing financial inclusion to those in developed countries that were once unable to access banking services. Currently there are 270 live services in 90 countries and over 100 planned new service launches.

Figure 32 Figure 33

Global smartphone connections Global average mobile network connection speeds

Source: GSMA Intelligence Source: Cisco VNI Mobile 2016

0

1,000

2,000

3,000

4,000

5,000

6,000

2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020

(m)

0

1

2

3

4

5

6

7

2015 2016 2017 2018 2019 2020

(mbps)

Innovations and connectivity enable a new

class of mobile applications to emerge

The mobile ecosystem accounted for 4.2% of

global GDP

The 7.1bn mobile telephone subscribers,

represent 96.8% of the world’s population




By 2020, GSMA predicts there will be 5.8bn smartphones. Data traffic is expected to achieve a Cagr of 49% from 2015 to 2020. Mobile 3G and 4G broadband connections accounted for almost 50% of the total connections in 2015. The GSM Association estimates that this could increase to more than 70% by 2020. Globally, the average fixed broadband connected speed is estimated to grow 2.2-fold from 20.3Mbps in 2014 to 42.5Mbps by 2019. In the first half of 2015, 47.4% of American homes were considered “wireless only” compared to 44% one year ago according to the National Center for Health Statistics.

Figure 34

US mobility statistics Dec 98 Dec 02 Dec 03 Dec 08 Dec 12 Dec 13 Dec 14 Wireless subscribers (m) 69.2 140.8 158.7 270.3 326.5 335.6 355.4 Wireless penetration % of total US population 24.6 48.0 53.6 87.2 102.2 104.3 110.0 Wireless-only households % of US households na na 4.2 20.2 38.2 39.4 47.0 Annualized total wireless revenue (US$bn) 33.1 76.5 87.6 148.1 185.0 189.2 187.8 Annual voice minutes of use (bn) 89.0 619.8 829.9 2,200.0 2,300.0 2,618.0 2,455.0 Annual wireless data usage (MB) na na na na 1,468.0 3,230.0 4,060.0 Source: CTIA

Cisco’s Visual Networking Index (VNI) estimates that global mobile data traffic will increase nearly eightfold between 2015 and 2020 to 30.6 exabytes per month (a 53% Cagr), with mobile video increasing 11-fold to 75% of the global mobile data traffic in 2020. Global mobile data traffic grew 74% in 2015 and although 4G only represented 14% of total mobile connections, it accounted for 47% of all mobile data traffic. VNI estimates that by 2020, 4G will represent 40.5% of all connections and 72% of total traffic. In 2015, 4G exceeded 3G traffic for the first time as 4G connections are targeted for higher-end device users where higher speeds encourage the use of high-bandwidth applications.

Smartphone apps are the nexus of innovation The broad adoption of smartphones paves the way for rapid innovation of new internet-based applications and services. Apple’s AppStore has more than 1.5m applications with over 100bn downloads reported as of June 2015. The evolving ecosystem around smartphone platforms has given rise to many successful new businesses - not just games, but thousands of applications for productivity, news, lifestyle and other uses. Statista predicts annual mobile application downloads to reach 269bn in 2017, with mobile applications revenue of US$41.1bn

Figure 35

Free versus paid mobile app downloads worldwide

Source: Statista

0

50

100

150

200

250

300

2011 2012 2013 2014 2015 2016 2017

App (bn)Free downloads Paid-for downloads

More devices, faster connections: more bits

Adoption of smartphones paves the way for rapid

innovation of new applications and services

Statista forecasts an increasing ratio of free

application downloads for every paid app download

Global mobile data traffic will increase nearly

eightfold between 2015 and 2020




Internet adoption expands potential end markets for innovators According to Cisco’s estimates, the total number of internet users worldwide was 2.8bn and expected to grow to 3.9bn globally by 2019, representing more than half of the world’s population. Cisco estimates that in 2014, only 40% of total IP traffic originated with non-PC devices, but by 2019 the non-PC share of total IP traffic will grow to 67%.

Figure 36

Global internet users by country, 2013

Source: Wikimedia Commons Author Jeff Ogden, (W163)

The International Telecommonications Union estimates that over 43% of the global population was connected to the internet in 2015, with the most dramatic growth occurring in Africa, the Arab States and the former (Russian) Commonwealth of Independent States.

Figure 37

Worldwide internet usage Region 2005 2015 2005-2015

Cagr (%) Penetration of total

population, 2015 (%) Africa 17 193 27.7 19.0 Arab States 26 141 18.3 40.6 Asia & Pacific 344 1,506 15.9 32.4 Commonwealth of Independent States

29 170 19.5 55.9

Europe 277 487 5.8 74.8 The Americas 316 651 7.5 65.5 World total 1,009 3,147 12.0 43.2 Source: ITU World Telecommunication/ICT Indicators database

Never a straight line - The headwinds to innovation The US has been struggling to sustain its competitive edge, which has slipped over the past decade but is showing some signs of improvement. The Global Innovation Index, sponsored by the World Intellectual Property Organization, INSEAD and Cornell’s Johnson School, has the United States up one spot from sixth in 2014 to fifth overall in 2015. Switzerland, the UK, Sweden, the Netherlands were ranked higher.

Cisco estimates nearly 4bn internet users

worldwide by 2019

Number of internet users worldwide exceeded

3.1bn in 2015, c.43% of worldwide population

America’s remains ahead in innovative capability

Over 43% of the global population was connected

to the internet in 2015




Among high-income economies, the US ranks No.1 in innovation quality, benefiting from its second place slot in top university rankings. The World Economic Forum’s Global Competitiveness Report for 2015 has the United States ranked third, retaining its spot from last year. The US had been No.1 in 2008 and 2009, slipping to No.2 in 2010 and gradually declining since then. The WEF sees a number of potential risks to US competitiveness such as the slowdown of accommodating monetary policy, stagnant growth, and a weak education system at the primary level but believes that the US can recover if it can improve government efficiency, steady the macroeconomic environment (including addressing high health and social security costs), and keep its financial markets sound.

However, US spends more on education while performance lags In the most recent findings published by the OECD, the 2012 Program for International Student Assessment (PISA) measured students’ scores around the globe. Among the 34 OECD countries, the United States performed below average in mathematics in 2012 and is ranked 27th. The US ranks 17 in reading and 20 in science. Mathematics scores for the top performer, Shanghai-China, indicate performance equivalent of over two years more formal schooling than students observed in the strong-performing state of Massachusetts. The US spends more than most countries on education but unfortunately this has not rendered better performance. In the report, the OECD claims that the Slovak Republic spends US$53,000 per student as opposed to the US which spends US$115,000 while both sets of students test at the same level.

Concerns over the higher-education bubble One of the greatest challenges for students - and the economy as a whole - is the untenable state of student debt, which now exceeds US$1.3tn in the aggregate in the US. The average class of 2016 graduate has US$37,172 in student loan debt, up 6% from last year. Debt service limits borrowing capacity for buying a car, home and starting a family, and ultimately could cause real-estate-asset deflation for retiring baby boomers who fail to find buyers for trade-up homes. In our view, higher education is the next bubble to deflate. There are signs of price declines at lower-tier colleges and declining grad-school enrollment in certain fields, particularly at law schools.

An educated workforce is critical for innovation A highly educated workforce with significant participation in science and technology is key to drive innovative effectiveness. The USA is one of the world’s most educated countries, ranking fifth among OECD countries in terms of tertiary level educational attainment for those aged 25-64, with 44% holding a degree from an institution of higher learning compared to the OECD average of 33%. According to Pearson, the US has a “cognitive skills and educational attainment” rank of 14th out of 40 countries.

However, graduates are not choosing the right majors. The USA severely lags in science, technology, engineering and math (STEM) achievement, which is critical for an R&D-intensive workforce. According to the Department of Education only 16% of American high school seniors are proficient in math and interested in a STEM career. The US is falling behind internationally, ranking 29th in math and 22nd in science among industrialized nations. The USA ranks 23rd among developed nations in terms of annual STEM graduates per person aged 20-34. It ranks 32nd in the percentage of its graduates majoring in STEM fields - just 13% of graduates majored in science, computer science, or engineering (compared to 27% in South Korea and Germany)

One of the greatest challenges is the

untenable state of student debt

The USA severely lags in STEM achievement

The USA ranks No.5 in advanced degrees

The US ranks No.1 in innovation quality




Immigration policy challenges entrepreneurs Immigration and innovation are inextricably linked, as many of the most creative entrepreneurs, scientists and investors have been attracted to the US environment and culture. Today, the USA places onerous restrictions on entry into the country, no matter what the reason. Annual caps and large fees on visas for highly skilled potential employees (H-1B) limit the number of jobs available to foreign graduates of America’s university system. Immigration policy is at the center of the 2016 Presidential race, with presumptive Republican nominee Donald Trump advocating more aggressive border controls and deportations to combat illegal immigration. Meanwhile, legal immigration processes remain inefficient and frustrating for prospective immigrants and employers alike.

According to the Institute of International Education, a record high of 974,976 international students were enrolled in US higher education institutions in 2014/2015 representing 4.8% of total enrollment. This represented an increase of 12.8% YoY, acceleration from the 8.1% increase the year before. According to the National Science Foundation, foreign students at US universities earned 36% of science and engineering doctoral degrees in 2012.

Immigrants play a vital role in driving American innovativeness and growth. Immigrant inventors contribute to more than a quarter of US global patent applications. A study by the National Foundation for American Policy found that immigrants started more than half (44 of 87) of America’s startup companies valued at US$1bn, and are key members of management or product development teams at over 70% of these companies. Among billion dollar startups, immigrant founders created an average of 760 jobs per company in the US. As many companies are dissuaded by the process of sponsoring a highly skilled foreign employee (which can be timely, costly and aggravating), many engineers, scientists and entrepreneurs are choosing to pack up and return home - often India or China - to work for competitors to US firms or to start businesses that create jobs abroad.

Patently difficult to fight the trolls Patent activity remains a key measure of innovative activity. According to research by Gridlogics, the number of patent applications filed in 2013-2014 globally totaled 4.4 million, a 6.2% rise over the previous period 2011-2012. The US remains the world’s largest patent filer, with over 57,000 patent applications in 2015 representing 26.3% of global filings. This was a decline of 6.7% from 2014, a year with an exceptionally high number of filings.

There are ongoing concerns that America’s patent system squelches competition, slows innovation and enables egregious predation through the legal system. The Wall Street Journal last year reported that the pace of patent growth may be decelerating. In 2014, patents did grow 3.3% to 2.15m but that is compared to 17.7% in 2013 and 20% in 2012. In fact, the growth of patents is at its lowest point since the financial crisis. US patents are still growing at an 8% Cagr since 2013. During the recession patents grew at an abysmal 1% in the US. Japan, usually a leader in patent issuance saw its patents decline over 4% in 2013 as companies in Japan found holding on to patents were becoming too costly.

“Patent trolls” are also known as patent assertion entities (PAEs) and often come in the form of non-practicing entities (NPEs), which are commonly shell companies with no operations, just a portfolio of patents used as the basis for litigation. Most PAEs do not use patents to create products or build

Since 9/11, the USA places onerous

restrictions on entry into the country

International students earn 36% of science and

engineering doctoral degrees in the USA

Immigrant inventors contribute to more than a

quarter of US global patent applications

Non-practicing entities - “Patent trolls” file more

than half of US patent lawsuits

The USA continues to file the greatest share of

global patents




businesses; the great majority exist only as a means to drive revenue through litigation. According to RPX Corporation, PAEs initiated 62% of all patent litigation in 2012. In a modestly encouraging development, the number of patent lawsuits filed in US courts hit a peak in 2013 at 6,030 and declined to 5,002 in 2014.

A May 2014 study by Catherine Tucker, professor of marketing at MIT Sloan School of Business, finds that VC investment would have likely been nearly US$22bn higher but for litigation brought by frequent litigators (defined as companies that file 20 or more patent lawsuits). The study covered the period from 1995 to 2012 and estimated with a 95% confidence interval that the amount of investment was between US$8.1-41.8bn relative to a baseline of roughly US$131bn of investment that actually occurred during that period. A 2012 study by James Bessen and Michael J Meurer of Boston University estimated the annual cost of NPE patent assertions at around US$29bn for defendants’ direct costs as a result of litigation as of 2011, up to US$80bn if indirect costs are included.

Tech companies were far and away the top targets of litigation initiated by NPEs, with HP, Apple, AT&T, Sony and Microsoft among the hardest hit. There were 171 lawsuits brought against Apple alone between2009-2013. 84% of all high tech-tech patent lawsuits filed every year are filed by patent trolls according to K. Jakel at the IP Counsel Café Conference in 2015. However, the big firms are not the only ones hit. Companies with less than US$10m in revenue comprise 55% of unique defendants to PAE suits. Of the sectors with the most litigated patents, PatentFreedom (an online community of companies that share information about NPEs) placed the semiconductor industry first, followed by software applications, financial services, communications equipment and system-infrastructure software.

Efforts to reform the patent system were stalled in the 113th Congress (2013-2015) largely due to two decisions from the US Supreme Court that made it easier for district court judges to award attorney’s fees in appropriate circumstances. In 2016, there are four serious proposals for patent reform in various stages of consideration in the 114th Congress - The Innovation Act; The Targeting Rogue and Opaque Letters (TROL) Act; the Support Technology and Research for Our Nations Growth (STRONG) Patents Act; and the Protecting American Talent and Entrepreneurship (PATENT) Act. This include efforts to force patent trolls to pay defendants’ legal fees if they lose a frivolous lawsuit, limit scope of discovery so targets are not overwhelmed with costly requests; and require infringement charges to be more specific.

Companies with less than US$10m in revenue

comprise 55% of unique defendants

The annual direct cost of NPE patent assertions is

estimated at around US$29bn

Bipartisan legislation underway to limit the

damage of patent trolls

PAEs initiated 62% of all patent litigation in 2012


Section 3: Bits flow into currents 2020 innovation


Bits flow into currents We reiterate the importance of our technology “meta-themes” - transparent IT, intelligent systems and convergence - to highlight the importance of platforms for innovation. “Transparent IT” describes how technology becomes increasingly ambient, interactions more natural and tools more intuitive and powerful. The combination of ubiquitous connectivity and big data analytics linking the physical world with information technologies enables powerful, interactive “intelligent systems”. Integration and coupling erase categorical distinctions between hardware, software, services and content in a process we refer to as “convergence”. Software remains at the top of the tech value chain as the driving force in innovation, gaining strategic importance for non-technology companies as well.

Transparent IT. Complexity gives way to simplicity. Powerful capabilities become accessible and pervasive, while advanced technologies become subsumed into systems and the environment.

Intelligent systems. Software and technology systems increasingly drive intelligent automation, decision enhancement, process optimization and risk management in self-directed, recursive systems. Advancements in analytics and artificial intelligence create leverage from the exponential growth of data from users, devices, sensors, applications and systems.

Convergence. Software, hardware, services, content and business processes straddle previously discrete definitional and categorical barriers. We expect vertical and horizontal integration, cross-disciplinary development and M&A, and organic evolution.

Transparent IT - computing melds into the walls Innovation over the next decade will be characterized by the trend towards transparent IT, technology that appears so simple that the underlying complexity becomes invisible to the user. The early Internet-of-Things vision advanced the idea of “ambient computing,” which in many respects is synonymous with transparent IT. We can see this trend play out in automobiles, which are increasingly coming to resemble “iPads on wheels”.

Computing is increasingly embedded in the daily lives of consumers, businesses and other organizations. It becomes more intuitive and pervasive with adoption of natural user interfaces, more powerful software, proliferation of devices, ubiquity of instant-on connectivity and declining costs of hardware, bandwidth and storage. We expect the continuous elevation of simplicity of experience to the user as logic controls the underlying systems, processes and infrastructure with increasing power.

We identify four vectors that drive accelerating innovation towards transparent IT: cloud computing, mobile internet connectivity, natural user interfaces and more powerful development tools.

User interfaces - “See me, feel me, touch me - think me?” Natural interfaces expand the experience of computing beyond the traditional keyboard/mouse, touch and speech recognition currently available. Touch and haptic interfaces enable new types of applications: gaming, empowering the disabled, medical procedures, industrial processes, training, simulation and therapy. Haptic interfaces have applications in virtual and augmented reality (by enabling real touch to operate in artificial environments) and through tele-operation (using real touch to operate in real environments via computer). Motion detection has gone

Simplicity wins

Things get smarter

Solutions are neatly wrapped and delivered

Complexity gets pushed down, logic moves

to higher levels

Technology is dissolving into the walls

Updating our technology “meta-themes”




mainstream in gaming, with adoption in retail, healthcare and other scenarios. Brain computer interfaces (BCIs) are also making strides, notably to assist the physically disabled. When coupled with self-learning capabilities and advances in natural language understanding, users will increasingly see computing as a natural extension of personal intent. A new generation of augmented reality (AR), virtual reality (VR) and “holographic” computing technologies will lead to more immersive experiences. We cover this topic in more depth in Section 4 of this report.

Cloud computing - A paradigm shift in computing. Cloud computing represents the transition of information technology to a utility model - the “industrialization of computing”. The emerging cloud software architecture causes a shift away from the enterprise IT economic model built around tightly coupled systems, proprietary software & hardware and extensive IT implementation, development, monitoring and support services towards a more open model built for massive scale. It emphasizes commodity hardware, open-source software and more automated processes.

The mobile internet - “Any device, always on, anywhere”. The rapid adoption of smartphones and the growing availability of wireless internet are key vectors for realizing the vision of pervasive computing and a wealth of related applications, including micropayments, content streaming, multiplayer gaming, location-based services, enterprise applications and Internet-of-Things use cases. High-speed 4G mobile networks encourage adoption of data-based applications. We expect the move to 5G wireless to make significant progress towards always-on high-speed internet connection. Additionally the trajectory of cost declines in bandwidth follows a corollary of Moore’s Law known as Butters’ Law, with potentially even more profound ramifications for adoption.

Development tools & standards - Simpler, more powerful. Higher-level software programming languages bring development closer to the business process and put increasing power in the hands of business users. There are over 31 million open-source software projects available on GitHub, freely available that developers and business users apply to create applications and new businesses. Standards such as HTML 5 enable a new class of rich, interactive mobile applications. Broad adoption of representational state transfer (REST) APIs enables web services to be combined into Web 2.0-style mashups, giving rise to new business models in “The API economy”.

Intelligent systems lead to the cognitive era Analytics and self-learning are increasingly embedded into systems. Their power accrues from the growing predictive strength of software, standards-enabled integration, always-on connectivity and the proliferation of sensors and remote-controlled devices. The concepts behind intelligent systems have been around for decades, but constraints around limitations of connectivity, throughput and computational power are falling away. The emergence of big-data technologies like Hadoop enables a vastly expanded scope of unstructured data analytics.

The surge of data generated by applications offers unprecedented visibility into operations and businesses. Smaller and more powerful sensors generate huge amounts of data, which can be used to improve energy efficiency and optimize broad functional aspects of the supply chain. Data warehousing, business intelligence and predictive analytics have matured and are easier and faster than ever to deploy. Systems will leverage the power of predictive

Adoption of smartphones and the wireless internet help realize the vision of

pervasive computing

Standards such as HTML 5 promise to enable a new class of rich, interactive and mobile applications

Moving towards a more open model built for

massive scale

Real-time intelligence is increasingly embedded

into systems

More data are available and analytic technologies

have matured




analytics and advanced techniques to optimize business processes, improve collaboration, target information flow and reduce risk. New types of devices that generate data for analysis and receive instructions remotely give rise to new solutions. The adoption of IPv6 will allow for billions more devices to be connected to the internet, each with a unique IP address. The Internet of Things will enable far more pervasive analytics.

Breakthroughs in artificial intelligence/machine learning will make it cheaper and easier for predictive capabilities to be embedded in ever smaller devices and systems. IDC predicts that within two years, 50% of application developers will incorporate some form of AI/machine learning capabilities into new application.

Prediction and optimization embedded in the walls Predictive analytics grow in value and effectiveness with increases in computational power. If traditional analytic apps and business intelligence sought to understand the past, predictive analytics looks to the future, to calibrate marketing and operational efficiencies and reduce risk of all types. Predictive analytics (which are the output of artificial intelligence/machine learning processes) are relevant in many situations including:

Online search and advertising, to improve query results and better target ads and recommendations. Much of the leading intellectual horsepower in the analytics community has been focused on web-based analytics. Google, Facebook, Amazon, Microsoft, Baidu, LinkedIn and other firms are leaders in this regard.

All types of marketing, to improve effectiveness and revenue “lift” from campaigns. IBM, Oracle, Salesforce.com and many other firms are focused on analytic marketing.

Manufacturing and operations, to anticipate potential safety, maintenance and quality-assurance issues. GE, Rockwell, Siemens and other industrial conglomerates leverage new sensor technologies as data inputs to drive improvements.

Financial-risk management, to help mitigate portfolio or credit risk (a prominent example is credit scores). IBM, Verisk, FICO and many other firms address a multitude of different aspects of risk management.

We highlight several manifestations of intelligent systems:

Marketing optimization. The use of analytics is a longstanding part of marketing and campaign management, where vendors, such as SAS and IBM, have provided the analytic horsepower to fuel effective campaigns. Vendors such as Salesforce, Oracle and IBM continue to drive incremental value through the use of scoring and predictive analysis.

Risk management, churn analysis and fraud prevention. For financial-services firms, solving the risk-management problem requires pulling together multiple sources of data and applying risk analytics to detect potentially consequential changes. Vendors such as RiskMetrics, Verisk, NICE, IBM, SAS and many others are focused on evolving challenges associated with detecting electronic fraud.

Smart grid, smart buildings, smart datacenters. The “smart grid” employs two-way technologies to control appliances and heating, ventilation and air conditioning (HVAC) systems at a home or building, allowing customers and utilities to keep costs under control. Smart home

Risk management continues to gain

importance in financial markets

Smart-grid technologies employ predictive

techniques to conserve energy

The value of predictive analytics continues to

grow with increases in computational power

Half of all application developers will

incorporate some form of AI/machine learning




thermostats from companies like Nest learn the behavior of occupants and use this information to optimize energy usage and comfort. Companies focused on this problem include Google, Honeywell, iTron, EnerNoc, Silver Spring Networks and privately held GridPoint, among others.

Convergence - Products become solutions Users and consumers of IT are concerned foremost with the experience and utility, not technology for its own sake. Increasingly, vendors combine hardware, software, services, connectivity and content into packaged solutions. Advantages include accelerated time to deployment, reduced integration and implementation service costs and increased efficacy of the overall solution. This also blurs the lines of product categories.

Apps are top of the tech food chain As we first discussed in our February 2010 Compubiquity report, it’s our consistent view that applications are where ultimate value in IT resides. Automating business processes, facilitating communications and collaboration, entertaining and informing users and enabling business models - the value to the users is defined by their utility. It’s a broad generalization, but the higher the vendor operates up the stack, the better cushioned from forces of commoditization. Applications can be deployed both as SaaS and on-premise offerings, and broadly include enterprise and consumer applications as well as the proprietary code powering e-commerce and internet sites. Latency and data gravity considerations inform the choice to deploy on premise or in the cloud.

Applications are no longer just software Along these lines, the distinction between discrete software, hardware, services and content blends into solutions. Vendors are at once more diversified and vertically integrated. Business models blur categories as value creation accrues in different points in the delivery chain. Apple, Microsoft, Amazon and Google all offer a combination of hardware, software, services and content. We expect increasing convergence over the next decade across technology and media sectors as software, hardware and content providers seek to improve customer stickiness, drive incremental revenue and erect higher competitive barriers.

Figure 38

Value moves up in stacks

Source: Jonathan Murray

Application

Platform

Infrastructure

Val

ue

Software offerings increasingly combine hardware, software,

connectivity and content

Business models will evolve as value creation

accrues in different points in the delivery chain

Value migrates from infrastructure to platform

to application





The “SMAC” stack + “things” = new innovations There is increasing focus on the characteristics of the next generation of applications. In fact, a new acronym has emerged to describe the new generation of apps: SMAC, which refers to social, mobile, analytics and cloud.

Social - Socially enabled applications need global scale, ability to perform multiple real-time operations, relatively low latency and a distributed architecture.

Mobile - With mobile applications, more of the application logic resides in centralized datacenters, delivered as a service to mobile endpoint devices. There’s increasing importance of location-based services. Mobile smartphones turn people into sensors, extending the edge of the network to individuals.

Analytics - With the immense proliferation of data, the opportunities to deliver context-aware, real-time analytics demand scalable data-management systems with intelligent filtering capabilities. There are many different types of analytics - “in-flight” including operational analytics (trends, statistical variations and calculated metrics), operational insights, predictive analytics and anomaly detection.

Cloud - Public and private cloud services increasingly draw from mega datacenters that provide massive scalability. The flexible architecture accelerates the forces that commoditize underlying infrastructure components. A key consequence of adoption of the new cloud computing architectures is that we will see more organizations elevate the importance of the software development function. We are seeing large investments pour into the development platforms for next-generation architecture. GE, for example, has invested US$100m in the EMC/VMware Pivotal spinout.

Adding the “T” to SMAC - When social, mobile, analytics and cloud technologies connect to physical “things”, devices become smarter. This is where the magic happens. Combining location-based services with social technologies along with mobility and analytics creates new opportunities for products. We believe there’s enormous potential for smart, context-aware objects and devices enabled by apps built on the SMAC stack.

Figure 39

The SMAC stack will enable new applications that connect to “things”

Source: Cognizant Technologies

Next-generation applications embrace

social, mobile, analytics and cloud

More organizations elevate the importance of the software development

function

When SMAC technologies connect to physical

“things”, devices become smarter

The SMAC stack represents a new wave

of application architecture




Convergence will reshape landscapes of industry and technology We expect convergence to play out along several vectors over the next decade:

Everything as a Service. Acceptance of Software as a Service (SaaS) paves the way for services to reach lower down the stack. Platform as a Service (PaaS) and Infrastructure as a Service (IaaS) hide the underlying complexity of computing and storage infrastructure and allow users, service providers and application developers to access resources in a holistic fashion. This is a profound shift (and challenge) for incumbent technology providers, financially, organizationally and culturally.

Commoditization of IT infrastructure. A corollary to the trend towards Everything as a Service and transparent IT is the continual commoditization of compute, memory, bandwidth and storage. The next generation of cloud architecture is heavily based on open-source software - and with the rise of open-source hardware and software defined networking, this trend extends into networking and datacenters.

Content integrates into the IT ecosystem. This is a longer-term trend, as proprietary content and information services become increasingly integrated into solutions. Proprietary content includes not just industry data, but also data collected from social applications, IT systems and internet traffic. It’s possible proprietary or exclusive content could be used as leverage for the “platform” players, though content providers are likely to prefer to retain a certain amount of leverage.

Physical and logical control systems converge. Rising interest in the industrial Internet of Things shifts focus on the need for integration between physical and IT systems. A notable area has been energy management in the datacenter around IT power management. Internet-of-Things uses are extending this to public infrastructure, automotive and industrial uses.

Figure 40

Converging IT ecosystem

Source: CLSA

Software Content

Hardware

Connectivity

Solutions

Physical facilities and logical systems controls

are converging

Expect the continual commoditization of compute, memory,

bandwidth and storage

For those that consume these services, the

experience is what counts

Software, hardware, connectivity and content

will converge into holistic solutions




The directions of convergence We summarize the broad directional convergence trends as follows:

Software vendors move into hardware and content.

Hardware vendors move into software and services.

Services vendors move into software and content.

Content vendors move into hardware and software.

Software vendors move into hardware and content Technology vendors increasingly deliver hardware and software bundled with services, both at the consumer and enterprise. Apple is a prime example of vertically integrated solutions for the consumer. Google offers Nexus phones and tablets, Microsoft offers Lumia phones, Xbox, the Surface tablet and laptop hybrids and Amazon offers the Kindle reader, Echo, Fire phone and TV. Enterprise-software vendors (particularly in infrastructure, security and networking) deliver integrated appliances. Oracle’s Exa-systems, IBM’s Pure Systems and Check Point’s security appliances reflect increasingly sophisticated hardware/software systems. There are divergent trends, however. Microsoft has written down much of the Nokia phone business, Google’s divested Motorola Mobility and IBM divested its x86 server business, both to Lenovo.

Software vendors also seek to diversify their revenue streams from products to product and service models. Microsoft is making the transition with Xbox Live, Office365 and Azure. Oracle is increasingly offering its portfolio with on-premise and cloud options that incorporate its engineered systems. Many if not most of the enterprise software perpetual license vendors are moving towards additional subscription-based businesses and/or hardware via acquisitions and partnerships to reinforce the value of an integrated portfolio and reduce reliance on a single-business model.

Hardware vendors move into higher-margin software and services For hardware vendors, the appeal of software’s margins and recurring revenue streams is compelling. We expect continuing development and M&A. This trend has been so consistent we would expect hardware vendors to continue to drive consolidation in the software sector. The past decade has seen a growing number of software companies acquired by firms for which hardware had been a predominant focus.

IBM: SoftLayer, Cleversafe, Kenexa, Weather Channel Data, SPSS

HP: Eucalyptus, Vertica, Voltage, ArcSight, EDS (services)

Cisco: Jasper, Sourcefire, Composite Software, JouleX, WebEx

Dell: Quest Software, Boomi, Perot Systems (services), EMC

EMC (pre-Dell): Virtustream, Pivotal Labs, Greenplum, RSA (software)

Intel: Mashery (SaaS), McAfee, Wind River (software)

Services vendors move into higher value software and content IT services vendors such as Accenture, CSC, Cognizant, Infosys and others specialize in customized solutions that incorporate proprietary IP, packaged software and hardware. With adoption of cloud computing and SaaS, the diminished need for technology-implementation services pressures many traditional types of engagements. As a result, we are seeing value-add resellers (VARs) and consulting firms move into managed services. Over

We expect application appliances to become

more common along with hybrid cloud offerings

Expect vendors to blend software, hardware, content and services

Pressures from cloud computing and SaaS force

IT service providers towards higher value




time, the large business process outsourcing (BPO) and IT outsourcing vendors need to provide increasingly differentiated, higher-value solutions. Infosys’s 2015 acquisition of SaaS ERP vendor Panaya will be followed by more in our view. We expect more IT services vendors to move into SaaS and other offerings.

Content vendors move into hardware and software Content companies are increasingly looking for new ways to monetize their assets, in particular publishers and entertainment companies whose models have been undermined by digitization of books, periodicals and music. The two most common approaches currently include writing custom apps for smartphones (all of the major media companies have done this) or alternatively opening up access to content through web-service application programming interfaces (APIs). Content retailers, including Amazon and Barnes & Noble, use devices (Kindle and Nook) to promote proprietary ecosystems enhanced with digital delivery services. Apple’s iPad has boosted magazines and other periodicals in digital form. Even the DVD kiosk rental service RedBox has moved into online video streaming offerings that compete with Netflix.

Publishers of textbooks and scholarly journals are under significant pressures from digitization and easy availability of used books. Providers of proprietary data, such as Bloomberg, Acxiom, Dun & Bradstreet, Experian and others, already sell raw data, provide benchmarking services and offer analysis for structured data. Increasingly, these providers of proprietary content are using APIs to turn their content into services that can be delivered over the internet.

Value grows as businesses evolve from products to platforms We’ve discussed in the past how the convergence of enabling forces gives rise to a new type of economy, and convergence is occurring because vendors seek to become platforms rather than components. Marshall van Alstyne of the MIT Center for Digital Business has produced a body of insightful work focused on how and why platforms are so valuable. It’s our view that the trend towards connecting “things” or products to services provides promising opportunities for existing businesses to evolve.

Platform strategies drive convergence of technologies Platforms have been enormously important in the IT industry since the introduction of IBM's System360 mainframes in 1964. The ecosystem of complementary hardware, software and services made System360 the leading enterprise computing platform until the 1980s. The Microsoft and Intel duopoly around the PC attracted an enormous ecosystem of hardware and software developers. With the growth of the internet and world wide web in the 1990s, internet-based platforms could connect larger numbers of PC users.

The advantage of platforms is that they can harness innovation from ecosystems of partners, creating aggregate value far greater than could be developed by a single company itself. Apple’s success against Nokia, Sony and Microsoft in phones and music players can be attributed to the breadth of reach. Instead of a linear value chain where users purchase a product in a “razors and razor blades” model through a single provider, Apple’s support of music, video, books and applications created a virtuous cycle where the more devices it sold, the more developers were attracted to write apps for the platform, attracting more consumer demand.

Watch more content companies build business

models from web APIs

From publishers to information brokers

Connecting products to services provides

promising opportunities for existing businesses

Platforms can harness innovation from

ecosystems of partners




Figure 41

Platform strategies create additional value by leveraging third parties

Source: Marshall van Alstyne, MIT Center for Digital Business

A platform can be defined as a system that must provide a useful function or service and allow third-party access, according to van Alstyne, Geoffrey Parker and Sangeet Paul Choudary in their book Platform Strategy. MIT Professor Michael Cusimano defines a platform this way: “A platform or complement strategy differs from a product strategy in that it requires an external ecosystem to generate complementary product or service. Innovations build positive feedback between the complements and that platform. The effect is much greater potential for innovation and growth than a single product-oriented firm can generate alone.” Venture capitalist Marc Andreessen says, ‘A platform is a system that can be . . . adapted to countless needs and niches that the platform’s original developers could not possibly have contemplated . . . ”

The dynamics of platforms are similar to those of network effects: the more products or services offered the more users it'll attract - the greater the scale, the more users. This attracts more complementary offerings and in turn, more users - creating a virtuous cycle of value creation.

With billions of mobile internet users and the advent of cloud computing, platforms have gained even greater scope and value creation. Platform companies are major drivers in innovation with the top companies like Google, Amazon, Facebook and Apple blazing trails for digital transformation. A recent report, The Rise of the Platform Enterprise: A Global Survey by Peter Evans and Anabelle Carr from the Center for Global Enterprise analyzed 176 platform companies with valuations over US$1bn. Their work identified four major types of platforms:

1) Innovation platforms provide a foundation for developers to create complementary products and services. They can attract a large community of external innovators to create an innovation ecosystem. Examples include IBM System360, Wintel and more recently Apple's IOS and Google Android which have established large innovation ecosystems of developers.

The Internet of Things creates opportunity for

more businesses to pursue platform strategy

A platform is a system that must provide a useful

function or service and allow third-party access

The dynamics of platforms are similar to

those of network effects

Innovation platforms help developers create

complementary products and services




2) Transaction platforms connect individuals and institutions. The first generation of e-commerce platforms in the 1990s with companies like Amazon, eBay, Ticketmaster and others have given way to a new generation of on-demand platforms like Uber, Lyft, Airbnb and others.

3) Integration platforms combine characteristics of both transaction and innovation platforms. Both Apple and Google have innovation platforms for developers which can then be made available in their transaction platforms, the App Store and Google Play.

4) Investment platforms are for investors managing a portfolio of platform companies like the Priceline group, which includes Priceline, kayak and OpenTable.

Brand consultancy Interbrand highlights the top global brands in an annual survey. In 2013, three of the top five brands (Apple, Google and Microsoft) were platforms. According to van Alstyne’s analysis, platform companies’ share of market cap out of the top 20 firms has increased from 10% in 2001 to over 25% in 2013.

Figure 42

Twelve of Interbrand’s top 30 brands are platforms

Source: Geoffrey Parker, Marshall van Alstyne and Sanjeet Paul Choudary

Three of the top five brands (Apple, Google

and Microsoft) are platforms

The top platform brands all have active strategies for the Internet of Things


Section 4: 10 themes from digital to physical 2020 innovation


10 themes from digital to physical We profile 10 innovation themes with potential to create new value, disruption and investment opportunities. A key thread across each category is the essential role that software and notably analytics plays in enabling innovation. These are longer-term themes that could profoundly reshape markets, the economy and society at large. As is typical of long-term technology evolution, activities in the early stages of markets represent little revenue, but with exponential progress, inflection points are likely to surprise. We segment these innovations into the digital sphere and the physical world.

The digital sphere Artificial intelligence comes of age Artificial intelligence (AI), machine learning and cognitive computing are different terms that all refer a similar range of technologies. AI powers everything from speech recognition to search software, airplane navigation and auto-pilot systems, video image recognition systems and intelligent assistants for smartphones. A new generation of self-learning computing algorithms will be integrated into applications of all kinds and integration with advanced robotics will power a new generation of autonomous and semi-autonomous machines.

Figure 43


Artificial intelligence/ cognitive computing

Artificial intelligence governs everything from speech recognition to search, airplane navigation and auto-pilot systems, motion-detection systems and intelligent assistants for smartphones

Advertisers, businesses, consumers, government, society at large

Jobs across a wide range of capacities from blue-collar drivers, security guards and others to knowledge workers like translators, paralegals, medical professionals, investment analysts

Google (GOOG), Microsoft (MSFT), IBM (IBM), Baidu (BIDU), Facebook (FB), Amazon (AMZN), LinkedIn (LNKD),Salesforce (CRM), many startups

Source: CLSA

2015 saw growing interest and investment in machine learning/artificial-intelligence technologies, even as luminaries such as Elon Musk and Stephen Hawking sounded alarms about the potential risks from uncontrolled AI. The combination of increased computing power, massive influx of data and a new generation of research drives innovations and new applications of the technology. In the most significant test of machine intelligence since IBM’s Deep Blue defeated Garry Kasparov in chess in 1997, Google’s AlphaGo won two of the first three games against grandmaster Lee Sedol in a Go tournament in March 2016.

The terms “artificial intelligence” or “cognitive” software describe a set of technologies that automate the processes of prediction. Artificial intelligence technologies can be known by several monikers, including cognitive agents, cognitive computing, cognitive reasoning, deep learning, intelligent fabric, machine intelligence, neural automation, neuralytics and other terms. IBM has popularized the term cognitive computing, which has been adopted by industry analyst firm IDC as well.

Cognitive/AI software can support decision making more rapidly, with higher confidence based on applying deeper analytics to more data. Characteristics of cognitive software include the ability to perform natural language

A key thread across each category is the essential

role of software and analytics

Artificial intelligence technologies can be

known by several monikers

Cognitive/AI software can support decision making more rapidly

Artificial intelligence is seeing another

resurgence - this time as foundational technology

2015 saw growing interest and investment in

artificial-intelligence technologies




processing for input and information ingestion. These systems perform learning in real-time as new data arrives into the system, applying past experiences and data to generate and predict possible outcomes. Potential outcomes can be scored for human analysis.

While AI brings to mind dystopian science-fiction movies like Terminator and HAL 9000 from 2001: A Space Odyssey, the current realities are more mundane - for now. AI’s value up until now has been less as horizontal technology or tool set, but in its application to specific problems.

AI software is omnipresent - it governs everything from speech recognition to search, airplane navigation and auto-pilot systems, motion-detection systems and intelligent assistants for smartphones like Apple’s Siri, Google Now and Microsoft’s Cortana. A new generation of self-learning computing promises to instrument the physical world and integration with advanced robotics will power a new generation of autonomous and semi-autonomous machines. AI technologies are a core competency for companies including: Google, Baidu, Facebook and Microsoft, which have helped boost interest and startup activity in the field.

Figure 44

Advances in AI are fueled by declining cost of compute and storage

Source: Tej3478 (Wikimedia Commons)

What is cognitive software (or AI)? According to CB Insights, AI startups have received a total of US$867m in funding since 2010. Funding for AI startups has increased sevenfold from US$45m for six deals in 2010 to US$310m for 54 deals in 2015. 2014 was the high water-mark for deals and investments in AI, with US$394m invested across 60 deals.

AI startups have received a total of US$867m in

funding since 2010

Cognitive systems “learn” and interact with humans

using real language

Deep learning has emerged as the hottest

area of AI

A new generation of self-learning computing

promises to instrument the physical world




By several estimates, over 200 AI startups have received venture funding. Vicarious Systems has raised US$67m, after demonstrating that it can solve Captchas, the visual puzzles that are used by websites to distinguish humans from computers. Sentient Technologies has raised total funding of US$144m with investors including Access Industries and Tata Communications. Sentient develops technology to distribute artificial-intelligence software to millions of graphics and computer processors around the world. Ayasdi raised US$98m to further its platform for automated insight discovery. Kensho has raised US$15m to train computers to replace expensive white-collar workers, such as financial analysts. Context Relevant has raised US$44m for its advanced hosted and on-premise analytics software.

The pace of investment remains active Both large and small companies are active in the market. Acquisitions of AI startups included Whetlab by Twitter, Vocal IQ by Apple, Wit.ai by Facebook and Coldlight by PTC. In October 2015, Intel announced it had acquired cognitive-computing platform provider Saffron Technology for an undisclosed price. Saffron’s cognitive systems technology uses statistics-based continuous learning that generates recommendations that can be used to anticipate market trends, optimize processes, mitigate risk and personalize customer experiences. According to CB Insights, over 45 corporations and corporate venture capital firms have invested in artificial intelligence startups since 2010.

The market for predictive analytics is a subset of the broader data analytics market, but is likely to outpace growth for other categories of analytic software, which is generally expected to grow around 10% YoY in the aggregate. Research firm MarketsandMarkets forecasts the predictive analytics market to reach a Cagr of 27.4%, from US$2.7bn in 2015 to US$9.2bn in 2020. Transparency Market Research forecasts predictive analytics software (including customer intelligence, decision support systems, data mining and management, performance management, security intelligence, risk management and financial intelligence) to see a 17.8% Cagr, from US$2bn in 2012 to US$6.6bn in 2019. IDC forecasts that the market for cognitive software platforms will grow at a 35% Cagr to $3.7bn by 2019.

Figure 45

Worldwide cognitive software platform forecast

Source: IDC

8271,075

1,419

1,916

2,644

3,683

0

5

10

15

20

25

30

35

40

45

0

500

1,000

1,500

2,000

2,500

3,000

3,500

4,000

4,500

5,000

2014 2015 2016 2017 2018 2019

Revenue Growth (RHS)(US$m) (%)

Cognitive Software platforms are forecast to

grow at a 35% Cagr through 2019

Over 200 AI startups have received venture funding

Growth forecasts for AI-related technologies are

in the 20-25% range through 2020




A robust cognitive software market The challenge for companies, large and small, is capturing the talent from a pool of experts estimated to be only a few dozen top experts worldwide. Google has been particularly aggressive in acquiring and hiring technologies with deep-learning expertise. There’s growing interest in extending access to AI for developers, given the relative dearth of available experts.

Microsoft announced it will be offering AI-based vision, speech and language API’s from its Project Oxford. The tools include spell check, video analytics for tracking faces, detecting motion and stabilizing imaging, speaker recognition, customized speech recognition and facial detection. Microsoft recently unveiled a new platform, dubbed AIX, which is specifically designed to harness the power of Minecraft to accelerate the development of AI applications. AIX will enable scientists to use Minecraft to test AI technologies.

Figure 46

Microsoft’s Project Oxford recognizes emotions

Source: Microsoft

IBM has launched its Watson partner program to work with developers, service providers and ISVs to make its capabilities accessible to cloud-based apps. There is a growing cadre of cognitive software platform vendors including IBM, Palantir, Digital Reasoning, IPSoft, Numenta, Cognitive Scale, Intel/Saffron, Tata Consulting Services, Cycorp / Lucid AI, Loop AI, MindMeld, Nervana Systems and others. Other companies working on cognitive systems/AI/deep-learning include Lockheed, SAIC, Google, Facebook, Fujitsu, Baidu, Microsoft, Apple, Amazon, Walmart, Yahoo, Hitachi and others.

From data to insight to prediction to action Analytics and “Big Data” trends have been top of mind for businesses and investors for many years. After the business intelligence (BI) market underwent consolidation of leading players, Business Objects (by SAP), Cognos (by IBM) and Hyperion (by Oracle) in 2007 with all three vendors approaching or exceeding the US$1bn annual revenue mark, a new generation of vendors has emerged including QlikTech, Tableau and a large number of startups focused on “Big Data” analytics using the open-source Hadoop data management technology.

More data, powerful computing and new

algorithms drive a resurgence of AI

The challenge for companies is capturing the talent from a small

pool of experts

IBM has launched its Watson partner program




While BI is typically historical facing, predictive analytics is forward looking and powered by statistics and mathematics. The largest predictive analytics vendor in privately-held SAS with annual revenues in excess of US$3bn. IBM acquired the only publicly traded predictive analytics pure-play SPSS in 2009. Artificial intelligence is an extension of predictive analytics, applied to specific tasks or problems.

Since it was initially conceived at a conference in 1956, AI has experienced waves of excitement and periods of disfavor. Interest has been gradually picking up in the past several years and in 2014 momentum accelerated. AI research saw initial commercial success in the early 1980s in the form of expert systems, which simulated the skills of human experts. With the collapse of the Lisp Machine market in the 1980s, AI fell out of favor. The late 1990s and early 2000s saw a resurgence of AI for logistics, data mining and other areas in technology. In 1997, IBM’s Deep Blue became the first computer chess-playing system to defeat world chess champion Gary Kasparov. In 2011, IBM’s Watson DeepQA system defeated the two all-time champion contestants of the Jeopardy quiz show game. Google’s AlphaGo won two of the first three games against grandmaster Lee Sedol in a Go tournament in March 2016.

Figure 47

It’s tough to beat the machines

Source: CLSA

Use cases for AI continue to expand With the increasingly powerful and cheaper computing and storage resources, it’s getting easier for AI/cognitive systems to be applied to new scenarios. AI powered computers are getting better than humans at many tasks. Some of the most common applications of AI include speech and image recognition. The annual ImageNet competition measures who automatically classify 100,000 photo databases within 1,000 object categories at the lowest error rate. Microsoft won in 2015 with an error rate of 3.5% and a localization error of 9%. In comparison, humans achieved an estimated error rate of 5.1% on the ImageNet dataset.

Artificial intelligence is an extension of predictive

analytics

AI has experienced waves of excitement and periods

of disfavor

IBM’s Watson has been spun into an independent

business unit at IBM

AI powered computers are getting better than humans at many tasks




A growing range of use cases for AI/cognitive computing includes autonomous vehicles, health care (including diagnosing illness, researching cancer and pharmaceuticals), managing power grids, protecting against cyber threats, monitoring credit card systems for fraud, high-frequency stock trading, logistics optimization, real-time translation, facial recognition and other uses.

Deep learning emerges at the cutting edge of AI Deep learning became one of the hottest disciplines in computing over the past decade, largely due to the work of Geoff Hinton, now at Google. The principle is that if one feeds a computer a lot of images of cats, for instance, the computer will eventually learn to recognize cats from video and pictures. Extrapolating this further is the view that machines will be able to understand language and images, with the hope that these techniques can be applied to helping spot tumors, avert financial risks and gauge the mood of a restaurant, for instance.

The term “deep learning” refers to an approach to building and training neural networks, an essential component of artificial intelligence. Neural networks are essentially algorithms that take different inputs such as words, pixels or audio waveforms, run a series of tests, then generate output in the form of predictions (such as the identification of a word or type of image). What’s key is that the networks are trainable. They improve their accuracy the more data is fed through the function. Certain problems involve complex algorithms and involve large amounts of data - with millions of numbers in the case of image recognition - so the increasing speed and power of computation is a key enabling factor. Speech recognition is a real-world example of how multiple iterations and more powerful processing can improve accuracy over time.

In theory, neural networks should be teachable, but scaling problem solving requires the calibration of an enormous number of inputs, weightings and iterations. In 2012, there was a breakthrough as researchers Alex Krizhevsky, Ilya Sutskever and Geoff Hinton showed in their ImageNet paper (ImageNet is an ongoing research effort to provide researchers with easily accessible image database with 15m images currently) that in a few weeks they could train a very complex network to a level that outperformed conventional approaches to computer vision, similar to what had been accomplished with natural language processing and speech recognition.

Teaching computers to solve problems on their own Deep learning takes neural networks to a new level of automation, using the training process to discover useful patterns. Automatic feature discovery enables the ImageNet network to recognize classes of objects on which it was never trained. This approach is relevant to many kinds of data including audio, seismic data or language. Daniel Nadler of Kensho described deep learning as ‘a paradigm shift from putting commands into a box to a time when computers watch you and learn.’

Google’s DeepMind has expertise in reinforcement learning, which involves getting computers to learn about the world even with limited feedback. DeepMind recently published a paper showing that its software could learn to play 1980s-era video games using only the information visible on a video screen as inputs, such as the score. The software gets an instruction such as “maximize the score” then learns the steps on how to get the highest score. The Google DeepMind AlphaGo Project defeated Korean Go champion Lee Sedol for the first two of three games. DeepMind analyzes human moves and gets smarter each time it plays.

Deep learning involves an enormous number of

inputs, weightings and iterations

Deep learning takes neural networks to a new

level of automation

Google’s DeepMind has expertise in

reinforcement learning

Deep learning refers to an approach to building and training neural networks

Deep learning became one of the hottest

disciplines in computing over the past decade




Ethical questions enter the picture During 2015, concerns over the potential dangers of AI were thrust into public discourse with high-profile calls to alarm from the likes of entrepreneur Elon Musk and physicist Steven Hawking, who called AI “our biggest existential threat”. Specifically, these concerns centered on the risks if AI is applied to warfare. It’s not clear who would be responsible for autonomous systems that are capable of selecting targets and operating autonomously without human control. Elon Musk has donated US$10m to the Future of Life Institute, which will run a research program aiming to keep AI “beneficial to humanity”. In July 2015, over 1,000 AI experts and leading researchers signed a letter calling for a ban on “offensive autonomous weapons” in order to avert the dangers of a “military artificial intelligence arms race.”

For now, investments in AI are concentrated in smaller private companies or as part of larger, diversified tech firms. Google, Facebook, Microsoft and IBM have made arguably the most extensive investments in AI to date, but much of the technology is being applied within different offerings rather than sold as standalone components or services (IBM’s Watson is the exception).

Figure 48

Artificial intelligence/cognitive computing related stocks Company Ticker Rating Currency Last

close EPS (US$) PE (x) Market cap

(US$m) FY15CL FY16CL FY15CL FY16CL Google GOOGL-US BUY US$ 732.77 29.58 34.47 24.77 21.26 462,020 Microsoft MSFT-US O-PF US$ 49.81 2.63 2.70 18.94 18.45 391,687 IBM IBM-US O-PF US$ 151.02 14.92 13.44 10.12 11.24 145,021 Baidu BIDU-US BUY US$ 160.82 15.26 5.09 10.54 31.62 43,539 Facebook FB-US BUY US$ 114.90 2.28 3.71 50.39 30.97 265,734 Amazon AMZN-US O-PF US$ 719.15 1.25 6.39 575.32 112.48 339,386 LinkedIn LNKD-US BUY US$ 191.65 2.84 3.78 67.48 50.70 22,611 Salesforce.com CRM-US BUY US$ 81.05 0.75 0.95 108.07 85.32 54,938 Source: CLSA, FactSet

Virtually real - the next computing platform Augmented-reality and virtual-reality technologies are coming to mass markets in 2016. The technologies have captured public imagination and the first generation of applications and viewing devices have made their debut. Virtual-reality is making its entrance first with Facebook’s OculusVR, the Sony PlayStation VR and HTC Vive. Next up is augmented reality, with Microsoft HoloLens and offerings from Magic Leap and Meta paving the way for a new generation of applications and entertainment.

Figure 49

Innovation What it means Who could benefit Potentially at risk Related companies Virtual reality/augmented reality

Gaming, entertainment, commerce, travel

Consumers, game developers, content creators

na Facebook (FB), Microsoft (MSFT), Samsung (5930), HTC (2498), Sony (6758), Google (GOOG), private firms Meta, Magic Leap

Source: CLSA

In 2016, interest in virtual reality surged as users begin to get the first tastes of the potential. With the Oculus headset slated to ship to consumers this year, virtual reality is poised to enter the mainstream, with offerings from Sony and the partnership of HTC and Valve Software also expected to come to market.

In 2016, interest in virtual reality surged

Augmented reality and virtual reality devices are coming to mass markets

in 2016




Figure 50

Worldwide augmented and virtual reality hardware forecast, 2016-2020

Source: IDC, April 2006

There have been a number of divergent estimates about the potential size of the virtual and augmented reality markets. Deloitte Global estimates VR will have its first billion-dollar year in 2016, with roughly US$700m in hardware sales and US$300m from content, with 2.5m VR headsets and 10m game copies sold. Industry analyst firm Statista forecasts a market of US$5.2bn in 2018. Research firm Digi-Capital’s forecasts are far more optimistic, estimating AR/VR could reach US$150bn in revenue by 2020, with AR accounting for around US$120bn and VR US$30bn. Market research firm KZero estimates the total revenue from virtual-reality hardware and software is at US$2.3bn in 2015 increasing to US$5.2bn in 2018 (with US$2.3bn from device sales). Considering the early stage of the technology, it's tough to handicap the accuracy of medium- to long-term projections.

IDC forecasts total annual hardware unit shipments for both tethered and untethered VR and AR growing from roughly 600,000 units in 2015 to 100.4 million units by 2020, a Cagr of 170%. IDC estimates 2-3m head-mounted devices (HMDs) will be sold this year. IDC forecasts standalone HMDs will ship around 45m units in 2021. And tethered VR will probably top out around 20m a year. Unit sales will not approach anything near smartphone volumes. The yield on some of the hardware needed is particularly low at this early stage of the market. There will be verticals that get disrupted if they do not embrace VR and AR.

While discrete HMDs such as the Oculus Rift and HTC Vive (devices that need an attached power source such as a PC to operate) will be used in settings where the user’s experience are limited to a room, mobile headsets (such as the Samsung GearVR) are likely to become a bigger part of the VR market due to usage flexibility. A portable unit enables people to share experiences on the go, which could stimulate industry adoption.

Augmented and virtual reality hardware units are forecast to grow at 186%

through 2020

Mobile headsets are likely to become a bigger part

of the VR market

Industry forecasts vary widely, but no one really

knows how big the market can be . . . yet




VC investment continues at a rapid pace Investors and corporations are seeing the potential of VR and AR. In March 2014, Facebook purchased Oculus VR, a creator of the Oculus Rift headset for US$2bn. The size and scope of the deal created huge waves and have catapulted a new generation of VR technologies to the forefront. Since then Oculus, Google, Apple and Microsoft have each acquired small AR/VR startups. 3D Systems (DDD) acquired Simbionix, a 3D virtual-reality surgical simulation and training company for US$120m. PTC in October 2015 acquired the Vuforia augmented-reality assets from Qualcomm for $65mn.

Venture investment continued to flow into VR/AR startups in 2015 with US$658m in equity funding across 126 deals, according to CB Insights. Most prominently, Florida-based augmented reality startup Magic Leap raised US$794m bringing its total funding to US$1.4bn. Mark Linao of Technicolor Ventures has tracked roughly 290 companies that have taken in a total of ~US$2.3bn in disclosed funding to date across six categories: commercial and industrial applications, content (including education and gaming), discovery and distribution, social, hardware, and infrastructure and tools.

Segment differentiation There are three primary markets for applications and hardware, all depending on the user: consumer, recreational and simulation.

Consumer: This market includes technology for the consumer using either a discrete HMD at home (eg, Oculus Rift, Sony PlayStation VR) or a mobile HMD on the go (eg, Google Cardboard, Samsung GearVR).

Recreational: This market includes experiences that exist in a social outlet (theme parks, malls, sporting events and expos) where systems may consist of either consumer hardware or custom hardware. Consumers do not need to own the hardware and the experience will often be controlled by the exhibitor to ensure quality.

Simulation: This market includes custom systems (including software and hardware) for enterprise clientele to solve specific problems, increase productivity and potentially lessen cost.

Virtual reality is about the immersive experience Virtual reality describes a computer-simulated environment with similar physical presence in real or imaginary worlds. While the vision and concepts have been around for decades, virtual reality technology has become viable only in recent years as the Moore’s Law advances in processing have powered the technology necessary for tracking along six degrees of freedom, an expanded field of view, instantaneous image refresh rates and haptics.

The basic idea of virtual reality has been around since the 19th Century stereoscope that combined two images taken from slightly different angles to render a three-dimensional illusion. Virtual-reality environments have been displayed on a computer screen or specialized monitor (such as a headset). Virtual reality is the immersive cousin of augmented reality, which refers to displays that overlay information and images over the physical world.

There is a new generation of headsets that includes the Oculus Rift, HTC’s Vive, Samsung’s Gear VR (a headset that uses the Samsung Note 4 as a display) and the Sony PlayStation VR. The new generation of devices benefits from declining cost of processing, memory and advances in 3D design and rendering software. What’s critical is to offer an experience with minimal image latency so that viewers see different images when looking up, down or around a virtual environment.

The basic idea of virtual reality has been around

since the 19th Century

Investments are pouring into virtual-reality

technologies

Venture investment continued to flow into

VR/AR startups in 2015





The experience really does live up to the hype Facebook CEO Zuckerberg described his first time using the VR headset as revelatory. Following the announcement of the Oculus deal, he described the technology as one of the few candidates to be the next major computing platform. This analyst shares the sentiment that virtual reality is one technology that is not overhyped.

Figure 51

This is your brain on Oculus

Source: Marco Verch (Wikimedia Commons)

Over time, there will be increasing integration with gesture recognition, haptic interfaces and enhanced sound (Oculus touts its 3D-sound technology as essential to the immersive experience). It’s important to understand a few basic technological concepts that help in understanding the platform:

Tracking: One of the most crucial aspects of VR is the ability to track movement in space with six degrees of freedom (6DoF). This enables the user to look and move around in a virtual world and interact with objects with their hands just as they would in the real world. Two of the more common types of tracking are inertial tracking and optical tracking.

Figure 52

Visual tracking needs six degrees of freedom

Source: GregorDS (Wikimedia Commons)

The VR experience is completely immersive

Using the VR headset is “revelatory”

Tracking allows the user to look and move around

in a virtual world

Two of the more common types of tracking are inertial tracking and

optical tracking




Inertial tracking uses a combination of components (including a gyroscope for calculating orientation and rotation, an accelerometer for measuring acceleration and a magnetometer for absolute direction) usually found in an inertial measurement unit (IMU). Because IMUs are increasingly commonplace in smartphones, prices have declined and quality has improved to a level that’s made VR technology possible.

Optical tracking uses markers (sometimes LED or IR lights) that can be tracked by cameras to locate objects in space. If the shape of the object being tracked is already known, it’s possible to do marker-less tracking by having the camera pick up the shape of the object.

Field of view (FOV). This is simply the angle of degrees available to a user within a headset. The greater the field of view, the more the user can see around in the virtual world - and subsequently the more immersive it feels.

Figure 53

Visualizing field of view

Source: Oculus (Wikimedia Commons)

Refresh rate. This is the rate at which images change on a screen, which is important because a low refresh rate may result in blurriness that could lead to motion sickness. This is also why OLED panels are ideal for screens.

Haptics. Simply put, haptic technologies provide the ability to receive tactile feedback, whether it’s touching an option on a screen or feeling tension. Haptics are an additional dimension of the immersion experience puzzle that truly helps a user feel like they’re in a virtual world.

The key components for capturing video in VR are the cameras and rig as well as the stitching algorithms that assemble disparate images into one cohesive experience. Jaunt (which has raised over US$100m so far, including an investment from Disney) has led the race in the market with its own custom cameras, but has now started to focus more on content.

Meanwhile, companies such as GoPro have entered the market to build rigs and supply cameras. Others like Lytro are investing in light field cameras that, instead of taking traditional video, capture all the data available that allows significantly more flexibility during post-processing.

Inertial tracking uses a combination

of components

Haptic technologies provide the ability to

receive tactile feedback

The greater the field of view, the more the user

can see around in the virtual world

A low refresh rate may result in blurriness




Arguably the most interesting company in the video market is NextVR, a company whose sole goal is to make users feel like they are attending live events, whether a Coldplay concert or the Super Bowl. The idea of sitting at center court and almost feeling basketball player Stephen Curry’s sweat is an idea that can scale on broadcast rights, product placement and the ability to sell the same seat over and over again, a million times.

Seeing the world in new layers Augmented reality describes a direct or indirect view of a physical, real-world environment where elements are augmented by computer-generated input, such as sound, video, graphics or GPS data. The concept of AR has existed for decades, notably in cinema where visuals are augmented with vibrations, smell and sound. AR applications have been used in the aerospace industry and in military, helping workers assemble aircraft and assisting pilots with navigation. Augmented reality differs from virtual reality, which replaces the real world with a simulated experience.

Wearable AR applications for military and emergency services can provide information such as instructions, location of enemy fire and maps. Many examples of this type of AR application have been represented in movies. The Terminator and Robocop films employ augmented-reality systems to “see”.

Figure 54

Entrepreneur Loic LeMeur on Google Glass

Source: Loïc Le Meur, by Rijans007 (Wikimedia Commons)

The Google Glass Explorer program charged users US$1,500 for the device (with a bill of materials that ran around US$300), and after enormous interest around the device, the lack of applications and unfavorable public reaction caused enthusiasm to wane. Google has folded the program into its Google X initiative, and there is likely to be another iteration over time.

There will be a high initial cost for AR but the business ROI will be compelling. Here is an increasingly broad range of applications that employs smartphone camera and GPS functions to overlay rich data on the user’s view - with

The concept of augmented reality has

existed for decades

The Terminator and RoboCop films employ

augmented-reality systems to “see”

We believe the future of Google Glass will be more

commercial applications

NextVR wants to make users feel like they are

attending live events




relevant, topical or even promotional information. Virtual fitting-room apps that enable shoppers to try on different styles of clothing are increasingly common, with several tied to Microsoft’s Kinect interface.

AR offers significant promise across industries In healthcare, AR will impact everything from patient care, to surgery and education. Doctors will be able to assess patient vitals without picking up a book. Medical instruction will also be transformed, and can help replace cadavers. Hololens will enable people to look at walls that haven’t been built yet then help architects figure how to design better. Education offers enormous opportunity enabling students to see examples of places and art. Logistics design can benefit from the use of AR to improve process flows. Field service will benefit from AR as an early use case. AR goggles will be able to walk mechanics through the process of servicing cars. A mechanical engineer can supervise multiple production lines at the same time.

One of the most promising use cases for AR is in field service. PTC has demonstrated integration of its Vuforia augmented reality in a joint connected field service offering with privately held ServiceMax. AR technology can train and guide remote service people and enable customer self-service, improving safety, reducing service time and manual involvement, reducing costs and optimizing results across the board. PTC’s AR technologies enable technicians to access digital schematics by scanning a “VuMark” encoded graphic. AR applications can visually walk technicians through complex service/repair workflows with digital images projected onto views of the physical world. When combined with predictive maintenance functions, the AR technology can shrink time and labor involved to diagnose and service products in the field.

Microsoft’s holographic computing platform comes to market Microsoft’s HoloLens is a headset with dark lenses that allows the users to see (and most importantly) interact with holograms overlaid across the physical environment. Hololens hardware incorporates a CPU, a GPU and holographic processing unit (HPU).

Figure 55

A look at HoloLens

Source: CLSA

Smartphone applications such as Yelp Monocle

overlay information on the user’s camera view

One of the notable use cases for AR is

in field service




HoloLens is not the all-immersive virtual reality of Oculus Rift, nor the mild augmented reality of Google Glass, but an entirely different approach closer to Minority Report than either. The company demonstrated how users could interact with holograms in scenarios such as an architect walking around a building design for his clients watching the holograms. HoloLens projects game images onto the physical world, enabling a gamer to play Minecraft in the living room in 3D, for example. Microsoft management highlights “holographic computing” as a fundamental capability of Windows 10, which should appeal to developers from both the Xbox and Windows platforms. The production version is slated to launch sometime this year.

Figure 56

The Microsoft Hololens augmented reality experience

Source: Used with permission from Microsoft

Successive AR technologies will ride down the cost curve In an earlier generation of the technology, Canon’s MReal AR headset is tethered to a PC used mostly for manufacturing and design. It costs $150k but it replaces multi-million dollar digital caves. This has been shipping for a couple of years. The Microsoft Hololens is looking to displace the MReal - the $3k SDK seems like a lot but there’s a lot of technology and Hololens will be able to integrate with Windows 10. Other AR startups include Meta and Magic Leap. Meta is tethered to a PC and the new unit looks a lot like Hololens. Meta is targeting business not consumers. Another company, DAQRI offers a smart helmet for construction workers. This is a wearable human/machine interface. Magic Leap has not disclosed much about its products other than a few videos, but an April 2016 story by author Kevin Kelly in Wired highlighted its tremendous potential.

As AR/VR software is still in nascent stages of the market, the financial impact on Facebook, Sony and Microsoft is small given their relative size and scope. CLSA Analyst Skye Chen recommends HTC based on the potential for the Vive, and this may represent the most direct way for investors to play the theme in the near term.

Hololens projects digital images over the physical world

Microsoft highlights “holographic computing”

as a fundamental capability of Windows 10





Figure 57

AR/VR related stocks Company Ticker Rating Currency Last

close EPS (lc) PE (x) Market cap

(US$m) FY15CL FY16CL FY15CL FY16CL Facebook FB-US BUY US$ 114.90 2.28 3.71 50.39 30.97 265,734 Microsoft MSFT-US O-PF US$ 49.81 2.63 2.70 18.94 18.45 391,687 Samsung 005930-KR BUY won 1,380,000 121,491 152,165 11.36 9.07 188,491 Sony 6758-JP BUY ¥ 2,941 119 91 24.63 32.46 35,055 Google GOOGL-US BUY US$ 732.77 29.58 34.47 24.77 21.26 462,020 HTC 2498-TW BUY NT$ 93.50 (18.71) (10.63) na na 2,387 Source: CLSA, FactSet

Blockchain - the real revolution of Bitcoin Blockchain, the distributed ledger that powers Bitcoin and other cryptocurrencies, is emerging as a transformative technology for the financial services sector in particular. The combination of advanced mathematics, access to massive computing power through peer-sharing, the open-source ethos and powerful new software enables a new universe of applications for managing transfer of value. There is an explosion of new blockchain startups targeting health care, media, finance, insurance and other industries.

Figure 58

Innovation What it means Who could benefit Potentially at risk Related companies Blockchain and cryptocurrencies

Blockchain technologies provide a cheap, distributed ledger. Open-source currencies provide alternative payment systems not tied to governments

Startup businesses, low-income workers, citizens in unstable countries, investors

Banks, credit card and money transfer firms

Microsoft (MSFT), IBM (IBM), many private companies - Ripple, Ethereum, CoinDesk, Coinbase, BitPay, many others

Source: CLSA

2014 was the year that media and speculative frenzy around Bitcoin reached its peak, but 2015 saw broad recognition of the disruptive power of blockchain, the underlying technology. The technology underlying Bitcoin has promise for a range of different types of applications, and the acceptance of Bitcoin by traditional merchants is growing steadily. In late 2015, The Economist ran an article on blockchain saying it could potentially have the same impact on business as two other profound inventions from the early days of capitalism, double-entry accounting and the joint stock corporation.

Interest in blockchain as a foundational technology has increased even as the price of cryptocurrencies has been volatile. Coinmarketcap.com tracks the market value of 100 top currencies representing a total market cap of US$8.2bn as of 19 April 2016, almost double the US$4.2bn as of 23 February 2014. Bitcoin’s total currency value represents the predominant share at US$6.7bn, far greater than Ethereum at US696m, Ripple at US$258m and Litecoin at US$148m.

Meanwhile, acceptance of Bitcoin continues to grow: at the end of 2015, over 100,000 merchants accepted Bitcoin as a form of payment, including Microsoft, Overstock.com, Dell, Dish Network, Expedia, Intuit and others. Daily transactions occurring with Bitcoins amount to about US$289m per day. VC investments in Bitcoin-related companies continued to grow in 2015, with over US$496m, up from US$326m in 2014, resulting in US$927m total VC investment in cryptocurrency startups to date, according to Crunchbase.

2015 saw growing recognition of the

value of blockchain

Interest in blockchain as a foundational technology

increased as the price declines

At the end of 2015, over 100,000 merchants

accepted Bitcoin

Blockchain technologies have the potential to

upend the financial services sector




What is Bitcoin and blockchain? The creation of blockchain technologies, Bitcoin and other modern cryptocurrencies is made possible by the combination of advanced mathematics, connectivity and the availability of massive (often distributed) computing power used to run cryptographic algorithms.

Bitcoin, Ethereum and other similar systems are composed of several core elements:

The blockchain. This is a database that functions as a distributed ledger, with copies maintained by all the participants in the network.

A cryptographic token. This token represents a store of value in cryptocurrencies like Bitcoin and Litecoin and in systems like Ethereum. The token is a string of numbers and letters with cryptographic properties that are essentially an address that enables people to store and transmit value.

A peer-to-peer networking system. Peer-to-peer networks are a mesh of users and processing nodes that are decentralized - everyone is a consumer and server of resources on that network.

A consensus-formation algorithm. This is a mathematical algorithm that validates a transaction is acceptable by having a majority of participants agree.

Validation. In the Bitcoin system, there is computational work to come to consensus approximately every 10 minutes regarding storage and transmission of value in the form of a Bitcoin token.

Unlike a traditional banking network where transactions are maintained, with Bitcoin/blockchain applications every node sees everyone else’s transaction. Transactions are conducted between two strangers. Holders use digital signatures to “unlock” their Bitcoins. The system is designed so no trust is needed. The integrity of the Bitcoin protocol prevents any duplication of Bitcoins, so theoretically there’s no opportunity for fraud, as is the case with credit cards.

To arrive at a balance requires analyzing all transactions ever made and tallying up the unspent inputs. No records of account balances are kept; instead balances are aggregated by tracking prior transactions to ensure there are enough inputs to cover an output. For each input, nodes check every other transaction in what is call the “blockchain”. To speed things up, Bitcoin nodes keep an indexed list of transaction. Sources and inputs are validated by the owner using a digital signature to assert ownership of the key.

Bitcoin is a peer-to-peer payment system that was introduced as an open-source system by an unknown developer(s) using the pseudonym Satoshi Nakamoto. Bitcoins are essentially long digital addresses and balances that are stored in a shared ledger called the blockchain. The software for Bitcoin is designed to run across a large number of machines, dubbed Bitcoin miners. Bitcoin miners track all transactions and add them to the blockchain ledger.

Bitcoin has captured an enormous amount of publicity and press interest, wand after a crash in 2014, the alternative currency has seen gains in 2016.

Cryptocurrencies are digitally generated

alternatives to traditional currencies

Bitcoins are essentially long digital addresses

and balances

Holders use digital signatures to “unlock”

their Bitcoins

To arrive at a balance requires analyzing all

transactions ever made

Bitcoin has a reputation as being a “rebel”

currency





Figure 59

Bitcoin price - Bubble, crash, recovery

Source: Investing.com

Despite the controversy, Bitcoin is gaining usage by the day, and there are over 100,000 merchants that accept payments in Bitcoin, including PayPal, Microsoft, Time Inc, Dell, Overstock.com, TigerDirect, NewEgg, Virgin Galactic, Dish Network and Zynga. In August 2015 Barclays announced it would become the first UK bank to start accepting Bitcoin, with plans to enable users to make charitable donations using the currency.

Figure 60

Bitcoin transaction visual

Source: Graingert (Wikimedia Commons)

Owning a Bitcoin equates to owning a private cryptography key associated with an internet address that contains a balance in the public ledger. This address and the private key are what enable holders to conduct transactions. For a user to send Bitcoins to someone else, they need the addresses for the sender and receiver as well as the private cryptography key used to authorize a payment. Based on these keys and addresses, the Bitcoin miners (the peer-to-peer computer network running the system’s software) check every transaction that happens on the network. If the math is not correct, the transaction is rejected.

0

200

400

600

800

1,000

1,200

May 13 Nov 13 May 14 Nov 14 May 15 Nov 15 May 16

(US$)

Bitcoin employs cryptographic “hash” functions to validate

transactions

Owning a Bitcoin equates to owning a private

cryptography key

More organizations now accept payments

in Bitcoin

Bitcoin as an investing vehicle remains highly

volatile




Managing these addresses and keys is complicated. A wallet is one of a number of software programs that handles and manages these transactions. These can be web, desktop, hardware or smartphone apps. There are a number of services that enable people to buy and sell Bitcoins. Coinbase is a website that links to a customer’s bank account and charges a 1% fee to buy and sell Bitcoins on an exchange. There are now over 571 Bitcoin ATMs around the world, predominantly located in restaurants, retailers and coffee shops.

There are risks to Bitcoin users that are not characteristic of other currencies and payment systems. Theft is one - if someone gets access to a Bitcoin wallet or private key, the Bitcoins can be stolen for good. There have been successful hacking attempts to services that store Bitcoins.

Figure 61

Bitcoin ATM on display in Toronto

Source: Wikimedia Commons (KryptoNatasha)

The Blockchain: creating an internet of value Blockchain takes the business and transaction logic from large, slow-moving institutions and moves it to the network level. Blockchain is a distributed, secure platform - essentially an immutable database running on billions of devices - purpose built for money and value and also for storing data and building applications. It’s a large distributed and transparent ledger - just like the internet is usable by everyone and owned by nobody, the blockchain is the same way.

There were over 571 Bitcoin ATMs globally as

of March 2015

Managing addresses and keys is complicated and

can be managed by wallet software

There are risks to Bitcoin users that are not

characteristic of other currencies




The more promising uses for blockchain technology occur in the area of contracts and settlement (where the Ripple protocol is emerging as a potentially disruptive threat to the role of foreign-exchange correspondent banks). For financial institutions there are a number of ways where blockchain technology creates new business opportunities. These include:

Serving the unbanked. Identity verification is a big stumbling block for financial institutions. With a public blockchain, two parties can enter into a transaction of any kind without necessarily knowing the other's identity, because the blockchain enables a trustless transaction. There are 2.5 billion people in the world without access to any kind of formal banking or financial institution. One of the reasons banks are reluctant to service those customers is because they can't attest to their identity.

Back-office cost savings. The blockchain network both clears and settles peer-to-peer transactions continuously so that the ledger is always up-to-date. By essentially automating processes normally managed by intermediaries, blockchain offers big cost-saving opportunities.

Faster transactions. In the case of about trade settlements, in can take up to four days (T plus three) for equity trades, seven days for remittances sometimes and correspondent bank settlement times can take weeks. The Bitcoin network takes an average of 10 minutes to clear a transaction and other blockchains that can do it even faster.

Improved risk management. The speed and efficiency of blockchain reduces settlement risk (the risk that a trade will be bounced back), counterparty risk (that a counterparty will default before settling a trade) and systemic risk, which is the sum of all the counterparty risk.

Open source innovation. The open source nature of these technologies engages more developers than any single company could, and this advances capabilities quickly and effectively.

A broad array of uses for blockchain technology The most apparent uses for blockchain are in financial services and there’s been a rush to embrace the technology. The R3 banking consortium is a group dedicated to advancing the use of blockchain. It was founded in September 2015 with initial members including Barclays, BBVA, Commonwealth Bank of Australia, Credit Suisse, Goldman Sachs, JP Morgan, Royal Bank of Scotland, State Street and UBS. It is now composed of over 40 banks and financial institutions that compete both directly and indirectly with each other.

There are many potential uses enabled by this lower-cost infrastructure. Micro-payments on the internet are an example. Blockchain-based micro payments can allow publishers to charge users or readers per article or per minute of YouTube viewing. Stem is a startup that helps music and video content creators to determine the economic split among participants.

The international remittance market is one of the areas targeted for disruption: this is a US$50bn-a-year market dominated by legacy players like MoneyGram and Western Union. Blockchain technology promises to send a vast amount of money in a short period of time for very low fees from any mobile phone in one jurisdiction to a mobile phone in a jurisdiction halfway across the globe.

Promising uses for Bitcoin’s blockchain

technology are in contracts and settlement

The R3 banking consortium is a group

dedicated to advancing the use of blockchain

There are 2.5 billion people without access to

formal banking or financial institution

The Bitcoin network takes an average of 10 minutes

to clear a transaction




Blockchain can be used in healthcare to track information in clinical trials so that doctors can see certain information without seeing the overall efficacy rates of a given drug in a Phase Three clinical trial. The technology can also be used in next generation energy grids, tracking the exchange of credits and debits among networks of solar and battery enabled participants that both produce and consume energy. Blockchain also holds promise as a foundational architecture for next-generation Internet-of-Things applications.

Increased interest and investments from big players Microsoft has announced Ethereum-based blockchain services on its Azure public cloud targeted to the financial services industry. IBM is conducting trials of blockchain based supply chain tracking applications with its partners and has been active in advancing research on the topic. Red Hat has announced that financial startups can experiment with blockchain technology on its OpenShift platform as a service. BNP Paribas has announced it will launch blockchain crowdfunding tools in 2016. Hyperledger is an open source project to accelerate blockchain technology, attracting major players including IBM, Intel, JP Morgan, CME Group, and DTCC into the consortium. For now, revenues from blockchain related technologies are de minimus, but what investors should monitor closely is the potential disruptive effect to established public financial services firms. CLSA analyst Mike Mayo continues to track the implications of blockchain technologies across the financial services sector.

Figure 62

Blockchain related stocks Company Ticker Rating Currency Last


(US$m) FY15CL FY16CL FY15CL FY16CL Microsoft MSFT-US O-PF US$ 49.81 2.63 2.70 18.94 18.45 391,687 IBM IBM-US O-PF US$ 151.02 14.92 13.44 10.12 11.24 145,021 Source: CLSA, Factset

Open-source is eating software is eating the world Open-source principles inherently enable innovation, not just in software, hardware and services, but through derivative benefits to technology users in any endeavor. The open-source model has transformed software development and is increasingly being applied in hardware, networking, crowdsourcing, media and new business models.

Figure 63

Innovation What it means Who could benefit Potentially at risk Related companies Open-source everything The open-source model

is transforming software development, crowdsourcing, prototyping, datacenters and the replacement of proprietary systems

Entrepreneurs, operators of cloud datacenters, corporations and service providers, SaaS independent software vendors (ISVs), consumers, industrial designers, military, consultants

Traditional proprietary hardware and software vendors including HP, Dell, Oracle, IBM, Microsoft, VMware, Cisco, EMC, Juniper, etc

Red Hat (RHT), Hortonworks (HDP), Microsoft (MSFT), Facebook (FB), Google (GOOG), Intel (INTC), AMD (AMD), many private firms including DataStax, MongoDB, Acquia

Source: CLSA

Open-source software is a key disruption vector particularly for cloud computing - as much of the infrastructure in the new cloud-based architecture is based on open-source components. Open-source principles inherently enable innovation, not just in software, hardware and services, but through the derivative benefits to technology users in any endeavor - and we are seeing the ethos and methodologies spread beyond technology.

Blockchain also holds promise as foundational

architecture for Internet-of-Things applications

Major vendors are exploring blockchain

based offerings

Open source has been critical to enable

innovation, not just in software





Open source is a double-edged sword for software vendors Open source commonly refers to intellectual property, most commonly software that can be freely shared and modified by users and programmers. The open-source ethos is fundamentally collaborative, with modifications and contributions from developers occurring concurrently, contributed back to the project. The open-source model allows users to reduce overall IT costs, accelerate collaboration and development across the organization, ultimately lowering barriers to innovation and facilitating value creation.

The role of the user/developer community is self-reinforcing as enhancements shared with the community accelerate the pace of technological innovation. Under the GNU Public License (GPL), the most common open-source license, software can be copied or given away without requiring a fee; the source code (essentially the operating instructions for the software) can be modified by anyone and any derivative works must be subject to the same license.

Flexibility of open source offers significant value to users One of the primary advantages of the open-source model is the ability for users to make changes, modifications and additions to the code at will - and these changes often find themselves incorporated as enhancements in newer versions of the software. Customers of traditional closed-source vendors typically must wait for upgrades from the vendor and are limited in the ability to modify, integrate and extend the software according to how open the vendor decides to be.

Open source has proven a highly disruptive force in technology, not just in software but increasingly in hardware and other businesses (like manufacturing, medicine, digital content, robotics and even beer). Proprietary standards allow vendors to focus on tight integration and the close coupling of hardware and software create frictions that allow an enormous amount of value to accrue to vendors. With Google’s support, open source Android has become the leading global operating system for smartphones.

Investors still pouring money into open source Venture investments in open source continue to grow. Investors are staking a lot on open source with VCs investing US$3.3bn in 188 open source deals from 2011-14. According to Accel Partners, 110 private companies have received a cumulative total of US$7bn in venture funding as of 2015, double the funding total in 2013. Just nine open source companies had risen venture funding in 2004. Many of the most prominent open source companies are focused on big data, with database vendors DataStax, Cloudera and MongoDB commanding valuations over US$1bn. The highest profile deal was in March 2014 when Intel led a US$900m round of investment in Hadoop vendor Cloudera. The December 2014 IPO of Hadoop specialist Hortonworks was the first major open-source software company to go public in the last decade.

The challenge of making money from open-source software The obvious downside to giving away software for free is that it’s hard to build businesses of meaningful scale quickly. Red Hat has built a business model around providing certified versions of its products along with extensive support. After nearly 15 years as a public company, revenue exceeded US$2bn in 2015 and the company consistently posts operating margins in the low- to mid-20s. In contrast, the next largest publicly traded open-source software company Hortonworks reported just US$166m in gross billings for 2015. Stock performance has diverged as well. Since its IPO, Hortonworks

Open source attracts new investment

It’s been tough to invest in open source

Open source is a highly disruptive force not just

to software but hardware and even manufacturing

Reduces overall IT costs, accelerates collaboration

and development

Primary advantage to open source is the flexibility to make

changes and modifications at will




has seen its valuation decline from over US$1bn to US$686m as of 13 April 2016. Red Hat has seen its shares gain 20% in the same period and sports a US$13.5bn valuation.

Figure 64

Red Hat’s scale is an advantage but Hortonworks shares have underperformed

Source: FactSet

Concern over Hortonworks’ valuation is an overhang for investors in open source startups concerned that the open source business model does not work. According to Dow Jones VentureSource, three of MongoDB’s mutual fund investors have written down valuations by 26-58% since their purchases. Cloudera has fared better, seeing its shares increase by 75% over the past two years according to Matt Asay at ReadWrite.

Open source - great for users but don’t expect another Red Hat Peter Levine of VC firm Andreesen Horowitz has advanced the argument that while open source is great for users, the business model breaks down because it takes too long to grow revenue to scale to fund investment in the products. As CEO of XenSource (acquired by Citrix), Levine saw other companies take advantage of the benefits of his company’s code with no revenue benefits coming back to them. Companies like DataStax (Cassandra database), Cloudera (Hadoop), Big Switch (network controller software) and Acquia (Drupal content management) have built their business models around selling proprietary add-on software that extends functionality of the underlying open-source projects.

Proprietary software margins are a target for open source deflation Profit margins in open source companies tend to be narrower, because of the free nature of key components of the technology offering. The main objective of open-source projects is to undercut proprietary technology vendors. The net effect is deflationary to the total market opportunity. Mike Volpi, former Cisco executive with Index Ventures, believes that open-source competition from the likes of Big Switch will diminish the total switching market from US$18bn to US$10-15bn with much narrower margins.

A key example of the dynamics in the market is the relation between market share and revenue share in the x86 server market. Linux servers are outpacing Windows server growth, and could reach parity by 2020.

20

40

60

80

100

120

140

160

Dec 14 Apr 15 Aug 15 Dec 15 Apr 16

RHT HDP Nasdaq(Dec 2014=100)

While open source is great for users, the

business model breaks down

Profit margins in open source tend to

be narrower

Open Source shipments result in far less revenue




Figure 65

Worldwide server shipments by OS, 2014-2019 (000 units) 2014 2015 2016F 2017F 2018F 2019F 2014-2019

Cagr (%) Windows 5,589 5,675 5,681 5,794 5,879 5,970 1.3 Linux 3,525 3,879 4,270 4,591 4,919 5,271 8.4 Unix 104 89 89 82 78 75 (6.3) Source: IDC, 2015

However the revenue contribution from Linux servers is far less than that from Windows. Open source is a highly effective development model for software, but monetization is a different story. Converting free users to paid customers takes a lot of work - Red Hat has been the most effective at this but still most constantly works to convert free JBoss and Linux users.

Figure 66

Revenue share of the server market

Source: IDC, 2014

Enterprises increasingly embrace open source According to the 2016 Black Duck/North Bridge 10th annual future of open source survey of over 1,300 respondents, 65% of respondents reported that usage of open source software increased in their organizations, up from 60% in 2015. It’s not just cost; 90% say that open source improves efficiency, inter-operability and innovation. 78% of 2015 respondents said that open source helps improve margins, with 75% saying it helps grow revenues. More are starting their own projects with 39% planning to start an external project, up 9% from 2014.

Even with the shift to lower-cost open source, companies are choosing to keep IT budgets intact but shift spend away from proprietary software. What’s happening is that investment that previously went to licenses and maintenance goes to development and maintenance of open source. Free open-source software reduces costs for startups as well as new projects within IT organizations. Even with paid technical expertise and support, the ROI tends to be overwhelmingly favorable for open-source users.

There is an enormous amount of open source software available, with over 30 million projects currently available on GitHub. Black Duck Software estimated that there are over 1.8bn lines of freely available open-source code in 2015 that developers and business users can access to create applications and new businesses. Almost two-thirds of companies look for open-source alternatives before they invest in proprietary software. According to one respondent, there’s a generation dynamic as well: “You can’t find a developer under 25 who doesn’t look at open source as the first choice”.

58.9 63.5 64.4 65.6 66.4 64.6 64.0 63.7 62.7 61.9

32.430.7 32.1 31.9 31.8 33.8 34.6 34.9 36.1 36.9

8.7 5.8 3.5 2.5 1.8 1.6 1.5 1.3 1.2 1.2

0

20

40

60

80

100

2009 2010 2011 2012 2013 2014 2015 2016 2017 2018

Windows Server Linux Server Unix Server(%)

Open source has become the go-to approach for

corporate IT

Free (like a new puppy) but taking spend from

traditional software

Windows server represents over 60% of

x86 server market revenues

There are over 1.8bn lines of freely available open-

source code in 2015




Open source software powers the cloud Open source is eating the infrastructure software world. Open-source databases like Hadoop, Cassandra, MongoDB, MariaDB and others provide alternatives to traditional data warehousing and transactional databases. CloudFoundry has displaced many of the proprietary Platform-as-a-Service software offerings, while projects like MuleSoft and Red Hat’s JBoss have displaced traditional application server software. It’s our view that the PaaS layer is the dividing line; infrastructure software and hardware below this layer are being steadily commoditized by open source and “Everything as a Service”. The inherent business logic in applications and analytics provides cushion against commoditization.

Open-source software is also foundational for the current generation of applications built on the de facto standard LAMP stack (Linux, Apache Web server, MySQL database and PHP programming language). Key foundational technologies for cloud computing are open source: the Linux OS, the Xen hypervisor for server virtualization, CloudFoundry PaaS software, the OpenStack cloud computing software platform, and most recently Docker (the application container software) and Kubernetes (container management). It is not unusual for startups, particularly in the internet or e-commerce arena, to avoid the use of proprietary software entirely.

Leading the charge into new technology areas More significantly, open source alternatives are moving from infrastructure into desktop applications like image editing (competing with Adobe Photoshop, for example). SugarCRM is an open source alternative to Salesforce and Microsoft’s Dynamics CRM. Open source analytic tools include the R language (which competes with SPSS and SAS), Pentaho and Jaspersoft. The implications appear far-reaching: every enterprise vendor must co-exist, compete with and integrate with open source. For companies like Oracle, IBM and others with large infrastructure software businesses, we believe open source continues to create a headwind. The leading edge of technology innovation in infrastructure software is happening with open-source projects. OpenStack has been embraced by VMware, IBM, Rackspace, Red Hat and other vendors. Docker is supported by Microsoft, Red Hat, VMware and others as well, and the pace of innovation appears to be accelerating.

Microsoft’s evolution: “Open source is a first-class citizen on Azure” When Linux first emerged in the 1990s, top executives at Microsoft were vehemently opposed to open source software, regarding Red Hat’s ascendance as an existential threat to the business. Microsoft was so intent on hurting Red Hat that it paid US$600m to its rival SuSE Linux so that SuSE could give away free support vouchers to prevent customers from adopting Red Hat subscriptions. This has changed dramatically. Microsoft has embraced support for the open source model, adopting key projects in favor of its own proprietary efforts where user adoption showed a preference. Notably Microsoft abandoned its “Cosmos” project for Big Data storage in favor of Hadoop because of broad adoption.

Microsoft has prominently embraced support for Red Hat Linux and other open source technologies like Cassandra, MongoDB, Hadoop, Redis on the Azure cloud. Microsoft has made its .NET developer framework open source and cross platform. The company established a .NET foundation to curate various open source projects. Recently the company acquired Xamarin (which enables developers to port Windows apps to Android and iOS) then made it free and open source to developers using Microsoft’s Visual Studio development tools.

Open source is eating the infrastructure software

world

Open source is foundational technology

for the cloud

The leading edge of infrastructure software

innovation is open-source

Microsoft’s anti-Linux vitriol is a thing of the

past as the company embraces open source

Microsoft has prominently embraced open source

software




Extending open-source vision into hardware Open-source hardware extends the concept of open source to hardware devices. The open-source principle applies not just to related software, but also to design elements such as mechanical drawings, schematics, printed circuit board (PCB) layout data, hardware description language (HDL) source code and integrated circuit layout. The business case for open-source hardware is in many ways more straightforward than software because vendors make money by directly selling the hardware. Other potential revenue streams include support and maintenance, related services (such as wireless and GPS) and project-based consulting. Open-source hardware initiatives encompass input/output (I/O), circuit boards (Arduino), 3D printers (RepRap), DJ mixers (Aurora 224) and video-game consoles (Uzebox).

Open Compute project promotes the Silicon Valley sharing culture The Open Compute project was founded in 2011 by a team of Facebook engineers who spent two years designing a datacenter from the ground up, with the goal to scale and manage in the most efficient manner possible. The project produced custom-designed servers, power supplies, server racks and battery-backup systems, which were designed with datacenter needs in mind. The team decided to make the specifications available to everyone and launched the Open Compute project.

Open Compute publishes open specifications and CAD drawings for virtual I/O, hardware management, Intel and AMD motherboards and power supplies, datacenter electrical and mechanical designs, a rack standard, battery cabinet and storage. Manufacturers working on Open-Compute-compliant components include Applied Micro, Delta, Emerson and Hyve Solutions, Microsoft, IBM, Yandex, VMware and Box.

The Open Compute consortium as grown to include hundreds of participating companies, with over 2,000 attendees at the recent Open Compute Summit. The designs are tailored for those building hyper-scale data centers but there’s strong adoption in the financial sector. According to a Bloomberg estimate, 80% of new servers at Goldman Sachs and Fidelity follow Open Compute specifications. The Open Compute project has so far focused on servers, storage, cooling and physical design, but the project has moved into networking with the goal of developing along similar principles of open, disaggregated technologies. The aim to produce a switch that could threaten Cisco is coming from Facebook, Intel and Broadcom.

Figure 67 Figure 68

Open compute datacenter design Open CloudServer

Source: OpenCompute.org

The business case for open-source hardware is

more straightforward than software

Google joined the Open Compute consortium in

March 2016

Open Compute originated among a team of

Facebook engineers

Open Compute moving into networking




Open-source business models work outside of pure technology The open-source model extends beyond information technology. Crowdsourcing is a form of open-source product development at companies like Local Motors, where potential customers vote on products and designs to be produced. Tesla’s CEO Elon Musk effectively open sourced the company’s battery patents, promising to share the technology and not seek any compensation for other automakers that seek to incorporate Tesla’s developments into their models.

3D Robotics is an open-source drone company making quadricoptors. The open-source model allows 3D Robotics to compete on quality and price against competition using the designs in China because the global cost of materials is the same. Because of the open-source nature of the designs, 3D Robotics has been able to incorporate contributions from Chinese competitors, who became active participants in the open-source community.

We highlight Red Hat as a top software pick but also as a business that has been able to successfully harness open source as a development model.

Figure 69

Open-source related stocks Company Ticker Rating Currency Last


(US$m) FY15CL FY16CL FY15CL FY16CL Red Hat RHT-US BUY US$ 77.14 1.91 2.25 40.39 34.28 14,001 Microsoft MSFT-US O-PF US$ 49.81 2.63 2.70 18.94 18.45 391,687 Facebook FB-US BUY US$ 114.90 2.28 3.71 50.39 30.97 265,734 Google GOOGL-US BUY US$ 732.77 29.58 34.47 24.77 21.26 462,020 Intel INTC-US N-R US$ 32.14 2.33 2.39 13.79 13.44 151,765 AMD AMD-US N-R US$ 4.39 (0.54) (0.28) na na 3,484 Hortonworks HDP-US N-R US$ 10.31 (3.08) (2.46) na na 597 Source: CLSA, FactSet

Staying secure in a connected world Trust is the basis for essential functions of commerce and society. With the explosive growth of connections, applications, communications, information and systems, threats continue to become more pervasive, driven by technological advances and growing involvement of organized crime and governments. As such there’s growing need for security to facilitate e-commerce, electronic money transfers and modern conveniences such as ATMs. The IT security market is a dynamic market, conducive to startups offering fertile ground for innovators and investors.

Figure 70

Innovation What it means Who could benefit Potentially at risk Related companies Security Trust is paramount in a

connected world. Rising levels of increasingly complex IT security threats compel increasingly innovative defenses

Consumers, businesses, government, society at large

Everyone and everything connected to the internet, including consumers, businesses, utilities, governments

AVG (AVG), Barracuda Networks (CUDA), Check Point (CHKP), CyberArk (CYBR), FireEye (FEYE), Fortinet (FTNT), Imperva (IMPV), Imprivata (IMPR), MobileIron (MOBL), NQ Mobile (NQ), Palo Alto Networks (PANW), Qihoo360 (QIHU), Rapid7 (RPD), SecureWorks (SCWX), Symantec (SYMC), Qualys (QLYS), Proofpoint (PFPT), Cisco (CSCO), IBM (IBM), CA (CA), EMC (EMC) and many others

Source: CLSA

The open-source model proves a good way to

identify global engineering talent






Security is essential to establish and sustain trust in an increasingly hyper-connected world. With the surge in connections, applications, communications and commerce, information and systems remain increasingly vulnerable to threats. With the exponential growth of users, connected devices and applications, securing systems becomes increasingly challenging as architecture becomes more distributed. Threats continue to become more pervasive, driven by technological advances and the growing involvement of organized crime and governments. Given the cybersecurity is one industry whose fundamentals are likely to remain robust for the foreseeable future. Criminals have used advanced technology for years and the future of cybercrime is exponential and automated.

The growing prevalence and complexity of threats to data compel businesses and individuals to defend their devices, networks and information. Attacks are primarily motivated by financial gain as cybercrime has become a multibillion-dollar pursuit. Security has become a sine qua non for anyone seeking to use the internet as vendors engage in an arms race with hackers. Regulatory mandates drive organizations of all sizes to implement policies to protect, back up and archive ever-increasing volumes of data. Proliferation of smartphones, tablets and other connected devices will drive the need for solutions that protect users and data.

Strong VC funding and M&A in 2014-2015 show signs of tightening The security market is enormously broad and complex, with hundreds of public and private companies offering solutions. According to CB Insights, cybersecurity venture funding has increased over the past five years from US$1.1bn across 166 deals in 2011 to US$3.8bn in 332 deals in 2016. There are so many startups that it can be difficult to keep track - various industry analysts that track the market have identified between 1,200-1,400 active companies.

Figure 71

Major security industry segments

Source: IT-Harvest

Security becomes increasingly challenging

with the exponential growth of users

Hundreds of companies, both public and private, offer security solutions

Attacks are primarily motivated by financial gain




M&A continues to fuel investors’ interest in cybersecurity. According to consulting form EY, the value of cybersecurity M&A activity more than doubled from $10.3bn in 2014 to $26.8bn in 2015 with the number of deals increasing 46% YoY to 287. Coming off a period of elevated (and at times seemingly indiscriminate) VC funding and M&A for cybersecurity, there are signs in 1Q16 of tightening in funding for all but the most mature cybersecurity firms with established revenues. IPO activity has slowed a bit, with the SecureWorks IPO from Dell the first IPO of 2016. Prior offerings included Rapid7 in 2015 and Cyberark, MobileIron and Imprivata in 2014.

Secular demand fueled by human-led innovation IT security is different from other technology sectors because demand is driven by external forces - human-led innovation. Whatever metrics one measures threats are getting more fierce, more difficult to detect, more difficult to combat and more numerous. The business of IT security will suffer only when the bad actors decide to go home and the likelihood of this is slim to none. For the foreseeable future, high-quality IT security vendors will continue to see strong sales.

Market forecasts vary across different firms but security spending remains a top priority in corporate IT budgets. Respondents to the 2016 PWC Global State of Information Security Survey increased their IT security budgets by an average of 24% in 2015. Gartner estimates that global spending on IT security grew 4.7% in 2015 to US$75.4bn increasing to US$101bn in 2018. IDC forecasts the global IT Security Products market growing at a 7% Cagr to US$46bn by 2019. According to IDC, the most promising areas for growth are security analytics SIEM, threat intelligence, mobile security and cloud security

Figure 72

Worldwide IT security products forecast, 2014-2019 (US$m) 2014 2015 2016 2017 2018 2019 2014-2019

Cagr (%) Endpoint security 9,032 9,174 9,579 10,012 10,458 10,912 3.9 Identity and access management 5,052 5,470 5,925 6,413 6,907 7,430 8 Security and vulnerability management 4,722 5,392 6,039 6,680 7,320 7,935 10.9 Messaging security 2,071 2,018 2,043 2,078 2,114 2,156 0.8 Network security 9,251 10,445 11,400 12,342 13,276 14,230 9 Web security 1,912 1,992 2,099 2,219 2,345 2,474 5.3 Other security 774 758 776 801 826 833 1.5 Total 32,814 35,248 37,861 40,545 43,245 45,969 7 Growth (%) 7.3 7.4 7.4 7.1 6.7 6.3 Source: IDC

In the view of industry analyst Richard Stiennon, most industry analysts have historically underestimated growth. In 2003 the market was about US$2.5bn and forecasted to grow ~11% over the next decade but actually ended up growing at a 34% Cagr. Steinnon estimates industry growth at 24% from US$68bn in 2013 to US$640bn by 2023 versus mid-teen industry forecasts.

Losses impact consumers and businesses The impact of attacks and theft of data is widely felt. Juniper Research predicts that cybercrime will cost businesses US$2tn by 2019. 80% of hackers work for an organization with an average age of 35.

Symantec reported that 429 million total reported identities were exposed in 2015, a 23% increase from the prior year. According to the PWC Global State of Information Security Survey of 2016, 38% more security incidents were detected in 2015 than in 2014, while intellectual property theft increased 56% YoY.

The value of cybersecurity M&A activity more than doubled from $ 2014 to

2015

IT security demand is driven by external forces

- human-led innovation

Security spending remains a top priority in

corporate IT budgets

Most industry analysts have historically

underestimated growth in the sector

In 2015, 429 million total reported identities were

exposed




According to Symantec there were 54 Zero-Day vulnerabilities discovered in 2015, a 125% increase from the year before. Small businesses are facing growing challenges. The last several years have seen a steady increase in attacks targeting businesses with less than 250 employees. The proportion of “spear-phishing” or highly targeted attacks directed against small businesses increased from 18% in 2011 to 43% in 2015. Globally, one billion data records were compromised in 2014, according to security firm Gemalto.

According to the PWC US State of Cybercrime Survey, 79% of survey respondents said they detected a security incident in the past 12 months, with 76% more concerned about cybersecurity threats this year than in the previous 12 months, up from 59% a year earlier. The 2015 Ponemon Institute Cost of Cybercrime report found that the median annual cost of cybercrime for 252 companies was US$15.4m, an increase from US$14.7m in 2014 and US$11.6m in 2013. The Ponemon Institute also determined that the mean number of days to resolve cyber-attacks is 46 with an average cost of US$21,155 per day (a total cost of nearly US$1m) over the remediation period. In fact, PWC forecasts that revenues in the global cyberinsurance market will increase from US$2.5bn in 2015 to US$7.5bn by 2020.

Figure 73

Internet crime complaint center statistics, 2014

Total complaints received 269,422

Complaints reporting loss 123,684

Total loss (US$m) 800.1

Median dollar loss for those reporting a loss (US$) 530

Average dollar loss overall (US$) 2,971

Average dollar loss for those reporting loss (US$) 6,472 Source: FBI

No “silver bullet” and a robust market for cyber-security companies The nasties grow more sophisticated Cyber threats have become increasingly complex and financially motivated and security is an ongoing “arms race” between bad actors and security professionals. There are growing challenges involved in protecting private data, transactions and intellectual property. Trust becomes essential as more individuals and businesses rely on connected services online.

Online threats continue to proliferate, with increasing frequency and complexity of new attacks forcing innovation.

Increasing number of connections to the internet, including smartphones, tablets, PCs, servers, automobiles, sensors, industrial equipment and other devices create more vectors for attacks.

Threats are financially motivated, as an increased focus on subterfuge and social engineering makes detection increasingly challenging.

A growing tangle of regulations requires organizations of all sizes to take measures to ensure the integrity of systems, processes, networks and data.

The growing number of users and applications creates more data, which need to be protected, tracked, audited and archived according to business needs and regulatory requirements.

There is a chronic shortage of trained cybersecurity professionals

IT security causes significant dollar losses

Median annual cost of cybercrime for 252

companies was US$15.4m






Crime is being committed by software and therefore it scales. The complexity, persistence and damaging impact of malicious attacks are similarly growing more difficult to combat. Over the past 15 years, the nature of the most prominent attacks has evolved from simple pieces of code designed for a single function (to disrupt or steal a password) into sprawling, coordinated attacks known as advanced persistent threats (APTs) that harness multiple vectors to steal personal data, financial information and proprietary intellectual property. According to Chris Young of Intel, 10 years ago security firms faced roughly 25 new threats each day - today it’s more than 500,000.

We have reached the industrial age of cybercrime Technology enables cybercrime to scale - a data theft can encompass 100 million such as the accounts compromised from the Target data breach. Cybercrime is also becoming internationalized: there was a US$1bn online bank heist from 100 banks in 30 countries; teams were targeting banks 24 hours a day. A group of Russian organized criminals removed limits from prepaid debit cards in 27 countries and carried off 36,000 transactions and US$25m. Criminal software is sold in Brazil for aspiring identity thieves - Black Shades is “crime in a box for US$2500” - it enables a whole range of hacks. A hacked Associated Press tweet caused a panic drop of US$160bn from the stock market in minutes and forced the White House to produce the President to prove he was unharmed. Some cryptocurrencies like Dark Coin are designed for criminal activity. The average bank robbery in the US nets US$4,300 and Mt. Gox was a US$460m theft.

One particularly nasty type of attack is “Ransomware”, software that encrypts a user’s data, forcing them to pay ransom to retrieve the “key”. In 2014, 13% of PWC Cybercrime Survey respondents reported they had been victims of ransomware.

Figure 74

Ransomware - Pay up or lose your files forever

Source: Motormille2

Architectural transitions compel new thinking The adoption of mobile and cloud computing creates challenges for security and compliance. Cloud computing impacts multiple areas of concern for businesses: governance (how can organizations ensure that cloud-service providers comply with corporate and regulatory requirements); data (where is

Ransomware is an emerging threat for

consumers and businesses

Mobile and cloud computing create

challenges for security and compliance

Security firms face more than 500,000 new threats

each day

Technology enables cybercrime to scale




the data physically stored, how is it protected); identity (how to federate authentication and access controls across different hosted environments); communications (how to ensure that data in transit is kept secure) and other considerations.

IT security at the enterprise is analogous to physical security, demanding a broad-based approach with multiple layers of protection. The market remains dynamic, with new innovations emerging to address constantly evolving threats as maturing sub-segments are subsumed into broader, converged offerings.

Data breaches target organizations of all sizes Cybercriminals target personal information of customers stored in the systems of websites and business, and regulations typically require that companies disclose incidents that have resulted in the theft of data. This can be a lucrative enterprise for criminals - as the result of a large data breach at Verizon Communications that involved 1.5 million stolen records, the thieves priced the entire package at US$100,000 and selected portions of records for US$10,000 according to data security website KrebsonSecurity. The Identity Theft Resource Center (ITRC) which tracks data breaches, sees the pace of breaches in 2016 tracking 10% ahead of 2015. The ITRC counted 5,810 breaches from 2005 through December 2015, involving nearly 848 million records.

What are the security implications of connected devices? Target was hacked by criminals hacking the air conditioning system. Samsung’s Smart TV used 3rd party voice recognition, so no conversation is private. Cars have over 250 microchips. Airplanes are also hackable - a computer security expert hacked an airplane to make it fly sideways. Cybercrime is going 3D with robots. Drug dealers are flying drones to find other dealers they can steal from. Drones in swarm formation can be a threat if they are armed. Drones can also be hacked as well.

Security is a critical component for IoT applications The Internet of Things (IoT) will see an exponential rise in connections, applications, communications and commerce, information and systems that are increasingly vulnerable to threats. Trust is critical. Securing systems becomes increasingly challenging with the exponential growth of connected devices and applications as architectures become more distributed. Threats continue to become more pervasive, driven by technological advances and the growing involvement of organized crime and governments.

A recent survey by HP’s Fortify division found that 35% of applications scanned exhibit at least one critical or high-severity vulnerability. Additionally Fortify found an average of 25 vulnerabilities per internet-connected device, a sample that included TVs, webcams, thermostats, remote power outlets, sprinklers, door locks, home alarms, scales and garage openers. One of the biggest issues gating adoption of consumer IoT solutions will be the need for users to be comfortable their privacy and personal data are secure. Trust becomes essential as more individuals and businesses rely on connected services online. For organizations including corporations, utilities and the public sector, the risks of network intrusions, denial-of-service attacks, sabotage and theft of intellectual property rise as operational systems increasingly become connected to the internet.

The security market offers rich opportunities for innovators and investors by its dynamic nature. New vectors are constantly emerging for vulnerabilities including Android, industrial control systems and other connected devices. Increasingly, there’s the realization that attacks cannot be completely

Security becomes increasingly challenging

with the exponential growth of connections

The security market offers rich opportunities

for innovators

IT security at the enterprise is analogous to

physical security

Of applications, 35% exhibited at least one

critical or high-severity vulnerability




prevented - rather strategies need to focus on rapid mitigation and risk management instead. There’s a lot of work currently on incorporating machine learning and artificial intelligence to detect and remediate threats.

The market’s dynamism has given rise to several publicly traded companies with over US$1bn in annual revenues including Check Point, Palo Alto Networks, Fortinet, Symantec and Trend Micro. Many of the larger technology vendors have built out and/or acquired security businesses of scale, including Cisco, IBM, Juniper Networks, HPE and Microsoft.

Notable publicly traded security companies include AVG, Barracuda Networks, CyberArk, FireEye, Gemalto, Imperva, Proofpoint, Qualys, LookingGlass, Rapid7, Varonis and others. Notable private companies include Cylance, Carbon Black, Kasperskey, Crowdstrike, ZScaler, Okta, Tanium, Webroot, CyberArk, LogLogic, Ping, SailPoint, Veracode and many others. We would also highlight that while security represents a minority of their businesses, for BUY-rated Akamai and Splunk security is a key growth driver.

Figure 75

Security related stocks Company Ticker Rating Currency Last


(US$m) FY15CL FY16CL FY15CL FY16CL AVG AVG-US N-R US$ 19.49 1.81 1.94 10.77 10.05 990 Barracuda Networks CUDA-US N-R US$ 14.84 0.42 0.47 35.33 31.25 774 CyberArk CYBR-US N-R US$ 45.53 1.00 0.92 45.53 49.69 1,527 FireEye FEYE-US N-R US$ 15.42 (1.61) (1.24) na na 2,567 Fortinet FTNT-US N-R US$ 33.73 0.51 0.70 66.14 47.94 5,794 Imperva IMPV-US N-R US$ 39.08 0.11 0.26 355.27 152.06 1,261 Palo Alto Networks PANW-US N-R US$ 130.24 0.86 1.66 151.44 78.46 11,666 Symantec SYMC-US N-R US$ 18.78 1.03 1.08 18.23 17.37 11,493 Proofpoint PFPT-US N-R US$ 63.37 (0.35) (0.14) na na 2,631 Qualys QLYS-US N-R US$ 32.07 0.70 0.79 45.81 40.83 1,121 Rapid7 RPD-US N-R US$ 14.04 (1.46) (0.91) na na 588 Secureworks SCWX-US N-R US$ 14.96 (0.66) (0.32) na na 159 Trend Micro 4704-JP N-R ¥ 3,585 157.71 141.70 22.73 25.30 4,748 VASCO VDSI-US N-R US$ 17.00 1.06 0.42 16.04 40.96 683 Qihoo QIHU-US N-R US$ 73.66 1.58 3.89 46.62 18.94 6,723 Source: CLSA, Factset

Clean disruption of energy and transportation approaches The energy and transportation sectors are poised to undergo massive disruption as the combination of increasingly cheap solar, advances in energy storage and adoption of electric vehicles will disrupt the traditional oil, gas, utilities and automotive industries - without regulatory intervention. There are increasing signs that solar energy is on the verge of becoming the cheapest source of power, without any subsidies. Innovations and economies of scale in battery and other energy storage technologies pave the way for large scale adoption of utility-scale solar and electric vehicles.

Figure 76

Innovation What it means Who could benefit Potentially at risk Related companies Clean disruption of energy and transportation

The declining cost of solar, advances in energy storage lower the cost of energy and make electric vehicles the default choice by 2030

Consumers, businesses, entrepreneurs, cities, society at large

Carbon-based fuel businesses (oil, coal, gas), utilities, some automakers

Tesla (TSLA-US); ADAS: Mobileye (MBLY-US); Battery: Panasonic (6752-JP); Sony (6758-JP)

Source: CLSA

The energy and transportation sectors are poised for disruption form

solar and energy storage




Information technology is not the only domain where exponential cost curves are having an impact. Energy and transportation as well are poised to see massive disruption from the declining costs of solar panels, energy storage and the consequent declines in cost of electric vehicles (EVs). Author Tony Seba in his 2014 book Clean Disruption has outlined the case that that we are on the cusp of a US$12tn disruption of energy and transportation. His work suggests that the cost of solar will drop so much that by 2030 all new energy will be solar. By 2025 all new vehicles will be electric and all new vehicles will have driverless capabilities. Over time, the energy grid will become an Internet of Energy. Customers will upload and download energy on the grid, with transactions likely tracked on applications running on blockchain architecture.

Industries are headed for disruption The technology driven disruption resulting from a confluence of new technologies and new business models puts at risk conventional electric utilities, the traditional automobile industry and many other peripheral businesses. Solar has zero marginal cost of energy, EVs are basically tablet computers on wheels, technologies are improving exponentially - and there are no breakthroughs required for disruption. Nuclear, coal and gas power will become obsolete as well as diesel, gasoline and the internal combustion engine for automobiles. Meanwhile, there will be investment opportunities in battery materials and technology, energy storage services, software, estate, and a panoply of new business models.

Solar technology costs have fallen sharply over the long term and by 90% just since 2008. Capacity installed is doubling every 2.5 years with 66% of PV capacity installed globally connected in the past two and a half years. The beneficiaries of declining costs are consumer of energy of course, while investors have struggled to make money investing in solar. Solar PV (PhotoVoltaic technology) has increased its price/performance along a 22% learning curve since 1970 and this keeps improving every year. Total PV installed capacity is growing at a 42% Cagr and this has continued since the year 2000. Every source of energy has seen prices increase by a factor of 6-16X since 1970; this includes oil, coal, nuclear and natural gas. Solar PV has improved on a unit basis over 3165X compared to oil, 2600X relative to nuclear, 2965X relative to natural gas.

Figure 77

Forecasting solar domination by 2030

Source: Tony Seba

Solar growth is on pace to provide nearly 100% of

energy needs by 2030

Energy and transportation as well are poised to see

massive disruption

Nuclear, coal, gas, diesel, gasoline and internal

combustion engines will become obsolete

Solar technology costs have fallen by 90% just

since 2008




Solar - great technology, rough for investors Massive increases in production, led by China, have been key to driving continuing price declines. Over the past decade China has displaced Europe as the leading producer of solar PV panels and this has resulted in a decline of both job and clean energy investment in Europe. Bloomberg New Energy Finance (BNEF) estimates that Europe represented 45% of global clean energy investment in 2010, but this has declined to 18% of the global total, or US$58bn, in 2015.

Solar is reaching grid parity today in many markets (unsubsidized) One measure that can mark a tipping point in solar is achievement of “grid parity” which refers to the cost of solar equaling the cost of energy from a utility (of course costs vary widely based on the amount of sun in a given area). Residential electric rates average US$0.12 per kilowatt-hour in the US, while wholesale electric rates (the price utilities pay to power generators) is around US$0.04 per kilowatt-hour on average.

In several countries, solar generated energy is cheaper than energy generated from nonrenewable sources. In Germany, Solar went from 5% of total electricity production to 16% in the eight years ending in 2014. Solar in Chile is already cheaper than wholesale. Today the average cost of energy from solar PV in US is around US$0.12 cents per kWh, about the same as the average retail rate. GTM Research estimates that 20 US states are currently at grid parity, with 42 states expected to reach this by 2020.

Figure 78

First solar desert sunlight solar farm

Source: US Department of the Interior

Regulations are helping advance utilities’ adoption of solar. In the US, the Public Utility Regulatory Policy Act (PURPA) mandates that utilities have to buy power from the lowest cost option. This benefited gas for a while as the marginal cost of natural gas is lower than other carbon-based sources. However, utility scale solar has come down to US$0.06/KWh and no new plant can replicate this. Tony Seba articulates the concept of “God Parity” - this is when the cost of unsubsidized rooftop solar generation is lower than the cost of transmission. He predicts this will happen by 2020 and this will be the tipping point where solar wins.

Utility scale solar farms are on the way to

displacing carbon based power

In several countries, solar generated energy

is cheaper than nonrenewable sources

Regulations are helping advance utilities’ adoption

of solar

Massive increases in production, led by China, have been key to driving continuing price declines




Solar as a service accelerates consumer adoption What’s making solar adoption take off is a business model innovation: Solar as a service. Zero money down solar PPA (Power Purchase Agreements) and leasing started in 2009. In these agreements, consumers or businesses agree to purchase power at an agreed rate from the owner/installer of PV panels, who bears responsibility for installing and maintaining the panels. Third-party financing started with companies like SunEdison and SolarCity.

Batteries the key to the story Battery innovation is reshaping conventional auto and utility industries and the way vehicles and grid systems operate. Electric vehicle battery costs (which currently account for roughly 40% of the cost of an EV) could halve by 2020 due to scale from mega investments, innovation breakthrough and higher energy density. Car makers will launch more competitive mass-market EV models to meet government-mandated carbon-emissions targets.

Battery improvements complement solar advances New innovations in batteries and declines in cost have the potential to disrupt many industries as scale builds. The use of batteries can enable both business and residential consumers to harness and use solar energy in a way that was not possible before. Batteries and energy storage are critical to harnessing the power of solar energy, allowing businesses and consumers the ability to store energy when costs are low (off-peak hours or when the sun is shining) and use energy when needed.

The markets are expanding for batteries as well Lithium ion battery costs are declining at 16% per year and this is accelerating across the industry. There is a race to scale up the energy storage industry to serve autos, utilities and electronics. When this investment occurs this will accelerate the cost curve. Investments in battery technologies are increasing.

Production poised for inflection with GigaFactory The upcoming Tesla GigaFactory in Nevada will give scale to the industry and help continue the cost decline curve. It will be able to supply 50 GWh per year of production - enough for 500,000 when it opens in 2020. This is not included in Tony Seba’s 16% price decline projections. This is a US$5bn investment with 6,500 jobs that will double world battery production and reduce battery pack costs by over 30%. According to Reuters, BYD plans to add six GWh every year. This could ramp up to 34 GWh by 2020 matching Tesla's 35 GWh. Together, Foxconn and LG could add a combined 22 GWh.

Advances in energy storage will help utilities manage peak load There are also different types of storage facilities under development that are used to reduce or eliminate the need for utility “peaker” capacity. Eliminating the need for “expensive capacity” increases the efficiency of existing utility plants and lowers the use of carbon fuels since the plants can now take advantage of solar and wind generation. In Spain there’s a new storage technology that uses potassium nitrate (salt). Molten salt is 60% potassium nitrate and 40% sodium nitrate. It stores heat at about US$50/KWh vs US$500/KWh for lithium ion storage. The Solar Reserve designed Solar Plant in Tonopah NV, has 100MW and 10 hours of storage. There is a 25-year PPA (power purchase agreement) with Nevada Power that will power Las Vegas.

Storage as a service combines software and energy storage There are a number of new companies such as Stem and GreenCharge Networks offering storage as a service to reduce demand charges for businesses. Businesses can charge up industrial scale batteries (or other types of storage) when costs are low or the sun is shining then consume the power when costs are higher or there’s no sun.

Tesla GigaFactory in Nevada will give scale to

the industry

Electric vehicle battery costs (which currently

account for about 40% of the cost) could halve by

2020

Lithium Ion Battery costs are declining at 16% per

year

There are also different types of storage facilities

under development




Stem combines big data, predictive analytics and advanced energy storage to help customers reduce demand charges. The solution is storage plus predictive software, which can help customers save 25% on battery storage charges. Use cases include demand management, load optimization, participation in utility programs, anomaly detection, additional demand response and increased customer engagement. The Stem battery systems are sold on a monthly lease and customers benefit from cost reductions. The financing partner takes 100% of the financing, Stem takes 100% of the SaaS portion of the business. GreenCharge Networks offers customers a percentage of savings model with zero money down.

Figure 79

Stem’s commercial/industrial power storage systems

Source: Stem

“Off the grid” will not mean a log cabin anymore Andrew Birch, CEO of Sungevity predicts that within five years we will be able to buy solar plus storage at lower cost than current retail price for power. As the 30% tax credit steps down, demand may slow down but this is likely to be temporary. By 2020, assuming li-ion batteries are at US$200/kWh in a battery pack, it will be possible for average US home to store electricity for just US$1.20 a day (based on the average American’s energy consumption of 903 kWH per month. This threatens electricity generating companies pricing power. By 2018 Tony Seba forecast all new US Solar installations will include some form of energy storage which will be disruptive. By 2020, storage will be inexpensive enough for millions of people in the US and Australia and other markets to get off the grid. By 2030 all generation and all storage will be distributed.

Vendors like Stem enable businesses to store

energy and save costs

Stem combines big data, predictive analytics and energy storage to help

reduce demand charges

Within five years solar plus storage will be at

lower cost than current retail price for power




Electric Vehicles poised to displace traditional autos There are two major disruptions in transportation technology - electric vehicles and self-driving cars (covered elsewhere in this report). Electric vehicles have a long and mixed history, but with the declining cost of batteries (which account for roughly 40% of the cost of a new care) the cost equation is becoming more attractive by the day. While hybrid cars (internal combustion engine enhanced by electro power) have gained steady acceptance over the past decade with the Toyota Prius the most popular, true EVs have been slower to see adoption. The Nissan Leaf and Chevy Volt have seen underwhelming sales numbers, but it was Tesla’s high-end cachet and high quality vehicle that provided a “cool factor” for EVs. Tesla approach where it started off with a premium product and rode the cost curve down to mass market levels appears to be working with the announcement of the Model 3 expected in late 2017.

Figure 80

EV’s ride a compelling cost curve

Source Tony Seba

There are many reasons why EVs are disruptive:

The electric motor is five times more efficient than internal combustion engines. Internal combustion engine can’t improve on 17-21% efficiency vs electric motors which convert 90-95% of energy.

The EV is ten times cheaper to fuel, as electrons are easier to move than atoms (gas). It costs only US$5 to charge a Tesla to travel 200 miles. Consumer Reports estimated that the cost to fill up a gas Jeep liberty over five years is US$15,000. An electric Jeep Liberty would cost US$1,565 in electricity.

True EVs have been slower to see adoption

Cost curves make EVs increasingly cost

competitive

The electric motor is five times more efficient than

internal combustion engines




Cars can have 2-3,000 moving parts, the EV has roughly 20. This is a 100x difference that translates to lower maintenance costs. Tesla offers an eight-year infinite mile warranty.

Wireless charging is another benefit. Tesla is building a network of hundreds of wireless charging stations around the country.

The modular architecture is a big advantage. Dual motors can be independently adjusted thousands of times per second. MIT has developed a robotic wheel that includes a drive motor, brakes, steering and suspension.

Rapid agile product development enables continuous improvement of the product and maybe even with software downloads.

EVs can provide grid services by both consuming and producing energy. Cars can be a power source for houses.

Figure 81

Tesla supercharging station in Gilroy, CA

Source: Steve Jurvetson (Wikimedia Commons)

How long will it take for EV’s to go mainstream? If we assume a mass market car will need 200 miles of range and a 50 KWh battery, Tony Seba estimates the industry could build a 200 mile range SUV for US$35-40,000 by 2017-2018, and US$31,000 (the average sticker price of a car in the US) by 2020. Tesla’s Model 3 was announced with a $35,000 base price for delivery in late 2017 and the company received 300,000 reservations within the first week. By 2022-2023 the price could be $22,000. New players from the tech industry are targeting the auto market: FoxConn is targeting a US$15,000 car. Xiaomi is targeting a US$6400 super-mini car. The prediction is that mass migration to EVs will occur in 2017-2018 and all new mass-market cars will be EVs by 2030.

Adoption of EV’s is coming out of Silicon

Valley

Cars can have 2-3,000 moving parts, the EV has

roughly 20.

Tony Seba predicts all new mass-market cars

will be EVs by 2030




Figure 82

Signs of inflection at hand

Source: http://newsletter2.statista.com/l/K892xG2YwuMRGWhLxtEEyQoA/aEad0n4wwHSIZZcSa1CaIA/ AYoyV3wC5ssvaqvtST2uAg

For investors, clean disruption offers a broad array of opportunities and risks (see our solar ETF performance chart in section 1 of this report). CLSA analyst Charles Yonts has done extensive work in this sector on solar, battery technology and other clean energy investments.

Figure 83

Clean disruption related stocks

Company Ticker Rating Currency Last close

EPS (lc) PE (x) Market cap (US$m) FY15CL FY16CL FY15CL FY16CL

Tesla TSLA-US U-PF US$ 214.96 (2.30) 1.00 na 214.96 31,359

Panasonic 6752-JP BUY ¥ 894.50 83.40 113.70 10.73 7.87 20,716

Google GOOGL-US BUY US$ 732.77 29.58 34.47 24.77 21.26 462,020

Sony 6758-JP BUY ¥ 2,941 119.40 90.60 24.63 32.46 35,055

First Solar FSLR-US N-R US$ 47.89 5.37 4.33 8.92 11.06 4,897

SunPower SPWR-US N-R US$ 15.13 2.17 1.39 6.97 10.90 2,089 Source: CLSA, FactSet

Vehicles that drive (and fly) themselves The fundamental nature of transportation is changing as autonomous vehicles prove technological viability and sensor-based vehicle communications systems promise to ease traffic jams and improve safety. The first generation of self-driving prototypes is being tested on public roads, with all of the major automakers and several technology companies investing in research. The technology is also being extended to trucks, other commercial vehicles and unmanned aerial vehicles (UAVs). The bigger challenges to adoption lie ahead with laws, lawmakers and insurance companies.

There are now more charging stations than

gas stations in Japan

Autonomous vehicles and robotics will remake

transportation and manufacturing





In 2015 momentum around autonomous vehicles accelerated in a big way. There is a broad range of technologies including artificial intelligence and robotics that go into poweringc self-driving cars, trucks and other vehicles. Investment is pouring into automotive technology. According to CB Insights, investors funded US$374m into 32 deals for the first 10 months of 2015. There has also been a wave of venture investments in self-driving tech startups in 2016 including Comma.ai, nuTonomy, and Nauto.

PWC has estimated that M&A by automotive suppliers in 2015 was nearly $50 billion, three times the levels in 2014. GM spent a reported US$1bn to acquire acquisition of self-driving car startup Cruise Automation. Toyota hired the entire staff of Jaybridge Robotics, an MIT autonomous vehicle spinout. Auto supplier Continental AG acquired a 3D laser-sensor business from Advanced Scientific Concepts. In 2015, auto parts maker Delphi Automotive made two significant investments, acquiring a stake in Quanergy, which is developing a low-cost LiDAR system and Ottomatika, an autonomous-driving software company developed out of Carnegie Mellon University.

Self-driving cars are becoming a reality The market opportunity is expected to be significant. IHS Automotive forecasts annual sales of self-driving or driverless cars reaching 21 million within 20 years. Over the past year, all of the major automakers have made progress toward developing autonomous features. Google has been extensively testing autonomous vehicles since 2009, which have logged more than 1.4 million driverless miles. Tesla has introduced semi-autonomous software updates to its vehicles and the company has stated that technology to support autonomous capabilities will be built into every new vehicle with software updates in the future needed to enable functions.

Other automakers have been getting into the act: Audi, BMW, Cadillac, Ford, GM, Mercedes-Benz, Nissan, Toyota, Volkswagen and Volvo all been testing driverless systems. Google’s prototypes are often seen on the streets around Mountain View, CA, while Audi, BMW, GM, Nissan, Toyota and Volvo have all targeted the introduction of autonomous cars by 2020.

The concept extends beyond passenger vehicles to trucks, busses and other vehicles. As of October 2015. Mining company Rio Tinto is currently operating 69 self-driving ore trucks at mines in Australia, with plans to expand to 150 vehicles within four years. A company called Navia is developing a self-driving electric shuttle. In the US there are several companies developing self-driving shuttles to serve campuses and areas where the speed limit is capped at 30 miles per hour because vehicles that do not exceed these speeds do not fall under the auspices of the Federal Highway Administration.

Figure 84

Innovation What it means Who could benefit Potentially at risk Related companies Smarter moving machines (autonomous vehicles, drones)

Self-driving cars, trucks, buses, drone aircraft

Consumers, businesses, automobile manufacturers, auto supply chain, military

Transportation based employment (taxi, truck drivers, logistic)

Google (GOOG), Toyota Motor (7203), Ford Motor (F), General Motors (GM),Mobileye (MOBL), Raytheon (RTN), AeroVironment (AVAV), Boeing (BA), Northrop Grumman (NOC), Textron (TXT), BAE Systems (BAESY), Adept Technology (ADEP), Amazon (AMZN), Lockheed Martin (LMT), General Dynamics (GD), SAIC (600104), GoPro (GPRO), Ambarella (AMBA), IXYS Corp, (IXYS), InvenSense (INVS) and others

Source: CLSA

Vision becoming reality to improve safety

and access

PWC has estimated that M&A by automotive

suppliers in 2015 was nearly US$50bn

All major automakers are in the process of

developing vehicles with autonomous features

The self-driving concept is being tested in other

capacities such as ore trucks for mining and

shuttle busses for students




Potential safety innovations have the power to be transformative Autonomous vehicles have the potential to solve many of the driving-related problems that occur as a result of human error or negligence. In theory, autonomous cars are safer than driver-operated vehicles because they have a 360-degree range of vision, can react more quickly than humans, cannot be distracted (for instance, by text messaging) and are unable to drive under the influence of alcohol. One of the primary goals is to reduce traffic accidents to zero and Volvo has announced plans for a zero-fatality car by 2020. It’s easy to come up with scenarios where technology would help, in case of the elderly, impaired or inexperienced drivers.

In a world of driverless, connected cars, the road system could be transformed into something akin to a smart grid, where all vehicles are tapped into the system, monitoring traffic patterns and other vehicle movements. This is called “platooning” where trucks follow closely behind each other autonomously on using Wi-Fi connectivity and vehicle-to-vehicle communications. Platooning trucks can reduce fuel consumption by 15% and can also alleviate traffic congestion. In April 2016, a fleet of self-driving trucks built by Daimler and Volvo traveled over 1,000 miles across Europe. Such a system could conceivably obviate the problems of traffic congestion, as well as the accidents that tend to accompany it. Many of the technologies necessary to produce an intelligent highway system are already in place, from traffic monitoring sensors, fiber optic networks, wireless technologies, motion detectors, etc.

A wealth of technologies at work Cost is no longer the issue with autonomous driving as processing power becomes increasingly cheaper. One of the most important technologies is light detection and ranging (LIDAR), an optical remote sensing technology that measures distance and other properties by illuminating an object with light. A LIDAR system cost around US$70,000 in 2012, by end-2013 this fell to US$10,00 and by 2015 the cost fell to around US$1,000. It is led today by private company Quanergy and costs are expected to fall further.

Self-driving cars also benefit from the exponential declines in the cost of computing power. In 2000, the first Teraflop (a measure of computing speed, referring to a trillion floating point operations per second) computer cost US$46m. NVIDIA’s Drive PX GPU (graphics processing unit) built for autonomous vehicles offers 2.3 TerFlops with 15 Watts power consumption for US$59.

New cars increasingly include sensor-based technologies such as automated parking assist and adaptive cruise control. Coming generations of driver-assist systems will provide greater vehicle autonomy. On the horizon are sensor-based solutions that employ stereo cameras, software and complex algorithms to compute the geometry of situations in front of the vehicle. Like all predictive algorithms, the efficacy will improve with more data. There’s increasing research into connected-vehicle systems that use wireless technologies for real-time vehicle-to-vehicle (V2V) and vehicle-to-infrastructure (V2I) communications. V2V and V2I technologies promise to advance crash avoidance and traffic optimization.

The adoption of V2V and V2I communications could significantly reduce traffic congestion. A report from the Eno Transportation Institute estimates that 10% adoption of autonomous vehicles in the USA could create US$38bn in comprehensive cost savings from reduction of crashes, improved fuel efficiency, lower travel time and parking savings.

New cars include automated parking assist

and adaptive cruise control

Ten-percent adoption of autonomous vehicles in

the USA could create US$38bn in cost savings

Self-driving cars also benefit from the

exponential declines in cost of computing power

Autonomous vehicles have the potential to solve many driving-

related problems

Many of the technologies necessary to produce an

intelligent highway system are in place




Current prototype autonomous driving systems employ, for instance, video imaging, GPS, laser scanner, radar and ultrasound technologies that enable the vehicle to recognize everything from surrounding vehicles to traffic light colors and move accordingly. Most systems have built-in redundancies that act as fail-safes and also allow the human driver to override the system at any time.

Automotive active safety suppliers are best positioned for the future In Emmanuel Rosner’s view, automotive suppliers with the best current portfolio of active safety technologies are likely to benefit most from the increased vehicle content in autonomous vehicles. In particular, Mobileye is a pure-play on the active safety and autonomous trends. The automotive industry is migrating to monocular (single) camera based ADAS and Mobileye is the unmatched leader, with 15+ years of validated data and contracts to work with nearly every major automaker. In addition, road experience management (REM), Mobileye’s new real-time mapping technology will eventually be used for autonomous capabilities and creates a moat around the core camera algorithms technology. We expect Mobileye to hold a dominant position, with share rising from 44% today to 64% in 2020 before leveling to 45% in 2025 as competitors join.

Figure 85 Figure 86

Mobileye’s OEM customers Mobileye market share

BMW Jaguar Land Rover SAIC

Fiat Chrysler MAN SE Scania

Ford Mazda Ssangyong

GM Mitsubishi Tesla

Honda Nissan Volvo

Hyundai Kia PSA VW

IVECO Renault Yulon

Source: Company Source: Company, IHS Automotive, CLSA

While not a pure play on the active safety/autonomous trend, Delphi is also well positioned for this high-growth trend. It is partnered with Mobileye on vision based ADAS while also offering collision warnings and avoidance systems that can incorporate radars and other types of sensors. One of its differentiated products in this field is the RACam, which combines radar sensing, vision sensing and data fusion in a single sophisticated module. Adoption of these technologies should see strong growth as automotive manufacturers and regulators advance towards the ultimate goal of zero accidents. Below, we summarize the active safety technologies offered by key suppliers.

Figure 87

Technologies offered by key active safety suppliers Autoliv Bosch Continental Delphi TRW Valeo Forward radar X X X X X Side/360° radar X X X X X X Forward camera X X X X X X Rear/surround camera X X X X Laser X Ultrasonic X X Night vision X X Source: Companies

0

10

20

30

40

50

60

70

2015 16CL 17CL 18CL 19CL 20CL 21CL 22CL 23CL 24CL 25CL

(%)

Automotive suppliers of active safety technologies

are positioned to benefit from autonomous vehicles




Figure 88

Google self-driving car at intersection in Mountain View, CA

Source: Grendelkhan (Wikimedia Commons)

Offering benefits of better space utilization Parking requires large investment of space and money in areas where there are not a lot of mass transportation options. Parking lots cover between 13-14,000 square miles in the US, and accounts for an estimated 20% of the usable land in Los Angeles. One of the potential benefits of a self-driving car is the ability for the car to drop off a passenger then circle around rather than park if there’s a short errand, or pick up another passenger in a sharing scenario. Broad adoption of autonomous vehicles could significantly reduce the need for parking in urban areas.

Bringing on the sharing economy As the world becomes more interconnected through an increasingly mobile internet and cloud-based services, this facilitates a transition away from an ownership to a rental, or sharing economy. The terms “peer economy”, “collaboration economy” or “sharing economy” are gaining currency as virtual communities drive new business models and disrupt incumbent industries. The trend towards sharing and collaboration has been concurrent with the digitization of content.

Companies like Zipcar, Uber, BlaBlaCar in Europe, Lyft and Park 24 (the biggest car park owner in Japan) are moving the market towards car sharing. BMW’s Drive-Now service is an example of an IoT-enabled on-demand mobile service. Launched in several German cities and San Francisco initially, the app allows customers to rent from a fleet of cars by the minute or the hour. The cars can be parked in reserved street spaces and users can monitor the electric charge or gas tank levels, mileage and other diagnostics.

With the addition of autonomous capabilities it’s only a few steps to sharing autonomous car services, with fleets of cars serving users. In extreme scenarios, this could potentially shrink the new car market by 80%, as the concept of car ownership and car insurance fade away while rendering 80% of parking and highway space obsolete.

Prototypes include no pedals or steering wheel

From file sharing to Everything as a Service

BMW’s Drive-Now service is a new example of an

IoT-enabled on-demand mobile service

Adoption of autonomous vehicles could reduce

need for land for parking




A younger generation disconnects from driving The longer-term vision for connected cars will involve rethinking transportation’s role in society. We are already seeing a generational shift related to driving. According to a study by US Public Interest Research Group (US PIRG) Education Fund and the Frontier Group there was a 23% decline in the average number of miles driven by 16- to 24-year-olds between 2001 and 2009. Car sharing is also growing at a rapid rate. According to DOT, there were 1.6 million members in 24 active programs in the US in 2014, over a tenfold increase in seven years.

The future vision of connected cars promises more sharing, as transportation options depend less and less on owning a vehicle. Fleets of shared vehicles, taxis or buses could travel the streets responding to requests from Uber-like services. Campus-based transportation systems could help business, schools and the elderly. Reduction in the amount of parking required could drive redesign of urban spaces for greater efficiency and more open space. Even vehicles themselves could see radical redesign to be adapted where there are no human drivers. Beyond the automakers, potential beneficiaries include Bosch, STMicro, InvenSense, Skyworks Solutions and chipmakers like Nvidia, Qualcomm, Broadcom, Infineon and Texas Instruments.

The biggest challenges are human not technology Regulations and liability remain the bigger challenges ahead. Google has been actively lobbying lawmakers in the US to legalize operation of self-driving vehicles with some success. California and Nevada (AB 511) have legislation allowing autonomous vehicle licensing, while Florida and Washington, DC allow testing. Some of the bigger questions concern liability and the role of human operators in conjunction with the technology. For now the proof of concepts have been a success. The bigger challenges lie ahead with laws, lawmakers and insurance companies. The Vienna Convention on Road Traffic and the Geneva Convention on Road Traffic are two major international treaties that may need to be changed for cars to be used in Europe and the United States, as both require that a driver must be in full control of a vehicle at all times.

Autonomous vehicles in the air: the rise of drones Automobiles and trucks are not the only types of vehicles that can operate autonomously. Most commercial aircraft have auto-pilot capabilities that automate most flying except for takeoff, landing and taxiing. There has been a dramatic increase in awareness around consumer and commercial drones, technically known as unmanned aerial vehicles (UAVs), as growing availability of cheaper and more powerful systems makes the technology ever more accessible. Consumer units have steadily made their way into the marketplace with the cheapest DIY kits starting at just a few hundred dollars.

Promising consumer and commercial market opportunity According to the Consumer Electronics Association, consumer demand is expected to drive the global drone industry to exceed US$1bn with one million units sold within five years. Research firm Radiant Insights estimates that the worldwide unmanned aerial systems (UAS) market will grow from US$609m in 2014 to US$4.8bn by 2021. ABI research forecasts that the small UAS market will reach US$8.4bn by 2018 with the commercial sector dominating the market in 2019 with revenues of US$5.1bn. MarketsandMarkets forecasts the global market for small drones to reach US$1.9 billion by the end of 2020. The Association of Unmanned Vehicle Systems International estimates that than 70,000 jobs will be created in the United States with an economic impact of more than US$13.6bn.

The future vision of connected cars promises

more sharing

Younger people are driving less, as sharing

services gain traction

Most commercial aircraft have auto-pilot

capabilities that automate most flying

Regulations and liability remain the bigger challenges ahead




Figure 89

3D robotics X8-M quadracoptor

Source: 3D Robotics

Commercial UAVs will transform industries Currently, military applications dominate the global UAV market, but commercial applications are poised to see robust adoption over the next decade. There is a broad range of commercial uses currently being explored for drones:

Precision agriculture. Sensors can be attached to ground vehicles and UAVs to scan crops with multi-spectral imaging for health problems, track growth rates and hydration and locate disease outbreaks. Potentially, UAVs can be used for the precision application of pesticides and fertilizer in the appropriate amount and location to save cost and reduce environmental impact.

Product delivery and logistics. Quick turnaround product delivery is a key use case. UPS and Domino’s Pizza are exploring ways that UAVs can be used to deliver products and food to customers. Drones can deliver items such as medications and supplies to remote areas in emerging countries where there may not be road infrastructure for cars.

Oil and gas. Oil companies such as BP are using drones to generate 3Dl maps of roads and pipelines to identify problems and strategize repairs. Drones can help move large equipment in difficult weather by providing real-time surveillance and 3D models of terrain.

Public safety, security and emergency response. Drones can conduct searches to find lost vehicles or identify situations that are too difficult for humans. UAVs can survey locations to help firefighters and emergency responders obtain information to inform tactical and operational decisions that can help ensure the safety of firefighters.

Journalism, filmmaking and photography. Filmmakers are already using drones for aerial shots, particularly for extreme sports content. Reporters are using drones to capture footage for sporting and other media events.

Hi-end consumer drones for photography can run several thousand dollars




Construction, architecture and civil engineering. 3D Robotics’ CEO announced a partnership with Autodesk to perform 3D building scanning that can be integrated into digital workflow and CAD systems. Drones will be able to perform 3D scanning that can be easily deployed or even automated on construction sites and private property.

Figure 90

Autonomous vehicle/drone related stocks


EPS (lc) PE (x) Market cap (US$m) FY15CL FY16CL FY15CL FY16CL

Google GOOGL-US BUY US$ 732.77 29.58 34.47 24.77 21.26 462,020 Toyota Motor 7203-JP BUY ¥ 5,393 741.36 631.00 7.27 8.55 169,957 Ford Motor F-US O-PF US$ 12.83 1.93 2.10 6.65 6.11 50,062 General Motors GM-US O-PF US$ 28.83 5.02 5.70 5.74 5.06 44,764 Mobileye MBLY-US BUY US$ 34.87 0.48 0.69 72.65 50.54 7,644 Amazon AMZN-US O-PF US$ 719.15 1.25 6.39 575.32 112.48 339,386 AeroVironment AVAV-US N-R US$ 31.32 0.10 0.30 308.79 105.91 732 Boeing BA-US N-R US$ 130.50 7.72 8.51 16.90 15.33 83,130 Northop Grumman NOC-US N-R US$ 216.81 10.39 10.68 20.87 20.31 39,124 Textron TXT-US N-R US$ 37.57 2.50 2.72 15.03 13.83 10,100 BAE Systems BA-GB N-R £ 4.75 0.40 0.39 11.85 12.22 21,271 Lockheed Martin LMT-US N-R US$ 239.56 11.46 11.85 20.90 20.22 72,935 GoPro GPRO-US N-R US$ 9.71 0.76 (0.91) 12.78 na 989 Source: CLSA, FactSet

Robotics: replacing and working with humans Advances in robotics are having a transformative effect on manufacturing and industry as a new wave of personal and collaborative robotics comes to market. The top-three drivers of the market are increased processing, reduced cost and size of sensors and programming languages and interfaces. Beyond nanorobotics, drones and autonomous vehicles are transforming industries like manufacturing, warehousing and distribution, healthcare, retail and other areas.

Figure 91

Innovation What it means Who could benefit Potentially at risk Related companies Robotics Automated

manufacturing, surgical robots, trainable robotic assistants, domestic robots

Manufacturers, healthcare, consumers, military

Labor, especially employees doing repetitive tasks in manufacturing, service, etc

Amazon (AMZN), iRobot (IRBT), Google (GOOG), Raytheon (RTN), Moog (MOG), Intuitive Surgical (ISRG), Cognex (CGNX), Accuray (ARAY), AeroVironment (AVAV), Northrop Grumman (NOC), Rockwell Automation (ROK), General Dynamics (GD), Boeing (BA), Teledyne (TDY), Textron (TXT)

Source: CLSA

Robots will be taking over Advances in robotics are having a transformative effect on manufacturing and industry as a new wave of personal and collaborative robotics comes to market. The top three drivers of the market are increased processing, reduced cost and size of sensors and programming languages and interfaces. Beyond nanorobotics, drones and autonomous vehicles (covered robots are transforming industries like manufacturing, warehousing and distribution, healthcare, retail and other areas.


and industry


and industry




Figure 92

Global robot population growth

Source: Stuart Staniford Early Warning Blog 2012

Evolution and advancement of artificial-intelligence software and robotics is having a profound impact on society and the economy at large. Beyond nanorobotics, drones and autonomous vehicles (covered elsewhere in this report), robots are transforming industries like manufacturing, warehousing and distribution, healthcare, retail and other areas. Much of what we see today is an extension of the multi-century trend towards automating human tasks and the acceleration of progress in technology will have outsize impact on the future economy. Robotic technologies and concepts will be increasingly integrated into mechanical and industrial systems, and combined with self-learning software capabilities we expect to see advances in the field accelerate.

The International Federation of Robotics (IFR) estimated that in 2015 over 240,000 industrial robots were sold, representing 8% YoY growth. In its most recent World Robot Statistics report the IFR states that by 2018, 1.3 million industrial robots will be installed in factories around the world with China accounting for more than one-third of the industrial robots installed worldwide. They estimate that the market value for robotic systems hovers around $32 billion. Between 2015 and 2017, IFR forecasts robot sales to grow at a 6% pace in the Americas and Europe, and 16% in Asia and Australia.

China in particular has a booming market expected to reach a Cagr of 25% through 2017. By end-2017, IFR estimates that about two million industrial robots will be installed in factories worldwide. Industry analyst firm Tractict forecasts annual worldwide shipments of consumer robots (which includes vacuums, lawn mowers, pool cleaners and social robots, will grow from 6.6 million units in 2015 to over 31 million units by 2020, a cumulative total of nearly 100 million during the period.

According to CB insights, VC funding for robotics doubled in 2015 with the volume of deals doubling from 45 to 83. Investments grew 115% to US$587m compared to US$273m a year ago, excluding investments in drones. The first three months of 2016 has seen funding rebound but deals to startups have dropped now for three straight quarters. Since March of 2016, 15 companies including Restoration Robotics, Savioke, and 5D Robotics have raised US$89m in funding across 15 deals and since 2011 have raised more than US$1.4bn in funding. Innovation Works was the most active investor in robotics this year with Intel Capital and High-Tech Grunderfonds tied for second.

End-2017, IFR estimates two million industrial

robots will be installed worldwide

Robots to become as plentiful as motors

In 2015 over 240,000 industrial robots were sold, representing 8%

YoY growth

VC funding for robotics doubled in 2015

Welcome our new AI-enhanced robot

overlords




As of the writing of this report, Chinese firm Midea has made a US$5 billion bid to acquire German robotics firm KUKA. If completed, this would be the largest robotics related acquisition ever. The Robot Report, which tracks startup fundings, acquisitions, IPOs and failures most recently reported the General Motors acquisition of Cruise Automation for US$1bn, giving robotics its first startup “unicorn”. Cruise develops autopilot systems for existing cars. Airware, a San Francisco UAS autopilot maker, raised US$30m in a round that included former Cisco CEO John Chambers as an investor and board member. The Robot Report tracked 55 fudings totaling US$1.32bn, 32 acquisitions worth US$2.27bn, and one IPO. Corindus Vascular Robotics raised US$42m in 2015.

New funds target robotics opportunity Dmitry Grishin, CEO of Russia’s largest internet company the Mail.Ru Group, personally launched Grishin Robotics in 2012, a US$25m seed-investment fund. In April 2016 Grishin Robotics launched another fund with US$100m from European and US institutional and individual investors to focus on companies involved in the "hardware revolution." Kazakh petrochemical mogul and entrepreneur Kenges Rakishev founded Genesis Angels, a US$100m fund dedicated to robotics and artificial intelligence and launched a second US$102m fund in January 2015.

Industrial robots account for 86% of the US$30bn market, with service robots (which perform lighter tasks such as picking and packing, cleaning and assisting in surgery) accounting for the remaining 14%. Industrial robots are expected to grow by 5% a year while service robots should grow 25-30% annually over the next few years.

Author of the Robot Report, Frank Tobe, estimates that by 2023 revenue from robots used in services will be greater than those used in industry. According to Tobe, Industrial-related robot revenue Cagr stands at 5.3% through 2023 versus a 27% Cagr for service-related robots. Collaborative robots revenue could be as high as US$1bn by 2020 compared to the current US$95m as these robots find new applications in the 6M SMEs worldwide which represent 70% of the world’s manufacturing.

The US has been investing heavily in its domestic manufacturing sector in order to be more competitive abroad. The 2016, the President’s Economic Report to Congress highlighted the view that increased deployment of robots will be a critical factor in growing production in the US. The automotive sector accounts for 40% of total robot shipments worldwide with consumer electronics the second largest sector at 20%. 55% of the total demand for industrial robots comes from the auto sector currently. Almost one quarter of these robots are at car makers allowing car manufacturers to produce the greatest number of cars after China.

Big tech firms double down on investments in robotics Amazon, Apple and Google have all made significant investments in robots and robotics. Companies like Coca Cola, Walmart, Nike, Ikea, FedEx and UPS are using a variety of automation technologies. Amazon spent US$775m in 2012 for Kiva Systems and renamed it Amazon Robotics. In October 2015 the firm disclosed that it currently uses 30,000 Kiva product-picking robots in its warehouses.

Kiva’s warehouse solutions consist of an army of small forklift robots that lift and move inventory shelves around and bring them to human packers, eliminating the need for human pickers. Meanwhile, a central computer coordinates the robots’ moves and customers’ orders using complex

Service robots are forecast to grow 25-30%

annually over the next few years

Chinese firm Midea has made a US$5 billion bid to

acquire German robotics firm KUKA

By 2023 revenue from robots used in services

will be greater than those used in industry

Kiva’s army of forklift robots

Amazon, Apple and Google have all made

significant investments in robots and robotics




algorithms. Kiva’s robots “know” where they are by scanning barcodes set on the floor using technology from machine-vision leader Cognex (CGNX), which has been providing Kiva with barcode readers since at least 2011.

There is a new generation of ecommerce robotics startups emerging as well. Fetch Robotics is a provider of mobile-picking robots that support freight robots and received US$23m from SoftBank in equity funding to scale up operations. GreyOrange is a Singapore based startup with robots similar to the Kiva systems that recently received a US$30m investment to fund expansion to Japan and China. Harvest Automation is a startup focused on moving potted plants from one location to another in the nursery segment of the agriculture industry.

Google appears to be retreating in part from robotics. According to reports from Bloomberg in March 2016, the company has put Boston Dynamics up for sale after tensions were report hindering the team’s ability to collaborate with Google’s other robot engineers in California and Tokyo and a failure to come up with products that could be released in the near term. Prior interviews with Google’s Andy Rubin suggest that its initial markets would be in manufacturing and logistics, potentially to compete more aggressively against Amazon for home delivery. Apple is in the process of investing US$10.5bn in technologies that include assembly robots, milling machines and equipment to polish the iPhone 5C, according to Bloomberg.

Surgical robots revolutionizing the operating room One area where robots are adding great value is in the operating room where there is low tolerance for errors. According to Alec Ross in The Industries of the Future, 1,300 surgical robots were sold in 2013 for an average of US$1.5 million each, representing 6% of unit value and 41% of sales value for industry robotics. Robotic procedures are growing at around 30% per year; with over 1 million Americans having undergone robotic surgery. The Da Vinci surgical system translates a surgeons hand movements into tiny “micromovements”, assisting in over 200,000 complex surgeries per year such as cardiac valve repair. The Da Vinci systems cost US$1.8m each however, and a newer system designed by the US Army called the Raven cost far less at US$250,000. The systems are not flawless, there have been 174 injuries and 71 deaths related to the da Vinci surgeries according to The Journal for Healthcare Quality.

Personal robotics - Making progress, still don’t do windows . . . yet Housecleaning robots, such as from Roomba or Neato are already several generations into the market, with price points under US$300 making adoption accessible to the middle class. One of the big challenges with robotics is to get people and robots to work together. Safety (and liability) is a major concern where machines are programmed to perform fixed and repetitive tasks, while humans tend to perform tasks with more variation.

There are expanding use cases for robots for telepresence. Systems from companies like VGo and Beam allow users to participate in meetings and even in school lectures, with video and audio streaming. Users can maneuver the wheeled robots from a remote location. In education, a French company Aldebaran has developed the NAO, a two-foot tall humanoid robot that serves as a teaching assistant in math and science classes in 70 countries.

An important development in robotics is the emphasis on open-source software and hardware projects like Arduino. This is driven by dual motives. Free open software encourages adoption and innovation, but also avoids

Robotics emphasizes open-source software

Google established a robotics division after

acquiring eight robotics companies

Housecleaning robots are already several

generations along

One area where robots are adding great value is

in the operating room

There is a new generation of ecommerce robotics

startups emerging




potential issues of liability in case of accidents or property damage. There are two main open-source efforts: a project called Open Source Computer Vision (OpenCV) Library and Robot Operating System (ROS), which is sponsored by the Open Source Robotics Foundation.

There is also a growing category of the personal-assistant robots. In collaboration with Fellow Robots, Lowe's Innovation Labs has placed robots (dubbed the OSHbots) in an Orchard Supply Hardware store in San Jose, California, in order to study the benefits of using robotics to assist customers and employees. The OSHbots can help shoppers navigate the store by providing the in-store location of a product, scan physical objects for identification (useful for screws, connectors and other parts), provide real-time information on product promotions and inventory and communicate via multiple languages.

Where the action is: collaboration, perception and agriculture There are three areas of significant near-term change in the field: collaborative robotics; advances in perception; and robotic products in agriculture. We’ve excerpted from the February 2016 CLSA U Blue Book by CLSA’s Morten Paulson and Editor and Publisher of The Robot Report Frank Tobe: Renaissance bot: Perceptive, Collaborative, out in the fields.

Co-bots coming together Collaborative robots (also called co-bots) are designed to work jointly on a project. To do so, they must be able to communicate and be safe to operate in proximity with humans. There are many reasons for the tremendous amount of buzz around the emergence of collaborative robots, particularly using them to handle ergonomically challenging tasks, using them alongside humans in electronics assembly lines, using them because they are affordable, using them because they are plug and play, using them to offload repetitive and boring jobs, and using them because they are significantly less expensive than their larger robot cousins.

Figure 93 Figure 94

Co-bot machine tending Collaborative robot assembly

Source: CLSA

The robots are breaking out of their cages and new breeds are born

Where are collaborative robots used today





At present, co-bot sales represent 5% of the overall robot market but with strong growth expectations. The collaborative robotics sector stands to increase roughly tenfold between 2015 and 2020, reaching over US$1bn from approximately US$95m in 2014. Insiders suggest more rapid expansion with lightweight co-bots becoming the top seller in the industry in about two years, posting sales in the hundreds of thousands and prices falling to US$15,000-20,000.

The small & medium-sized enterprise (SME) marketplace is huge: six million companies worldwide and almost 70% of global manufacturing. A few lowcost plug-and-play robotics tools can easily fit into the manufacturing process in many of these companies. It is easy to imagine co-bots for SMEs reaching our hundreds-of-thousands unit-sales mark beginning as early as 2018.

As robots move from fixed and caged locations to work alongside us and as the metrics of robot ownership and deployment change and become more affordable, companies of all types and sizes are finding strategic reasons to acquire or invest in robotic ventures to add to their arsenal of products and services. They don’t want to be left behind and they are paying high prices for their acquisitions.

Figure 95

Four collaborating co-robots at IAS in Shanghai 2015

Source: CLSA

Market leader Universal Robots has a big headstart. This year it is likely to maintain its lead but Kuka, ABB, Fanuc and Yaskawa will begin to make inroads and experiment with different prices. By the beginning of 2017, the competition will become even more intense as the number of co-bots sold approaches 15,000 units or US$0.5bn in sales revenue. It’s still too early in the evolution of co-bots for provider consolidation, but some systems are sure to be preferred because of their flexibility, ease of training and support network.

Market poised to grow tenfold

SMEs are the new addressable market

Companies are rushing to get in position

From left to right: Rethink’s Baxter, Kuka’s

LBR iiwa, ABB’s Yumi and Universal’s UR10

Universal Robots is the current market leader




3D vision getting closer Machine vision has been a continuing challenge in robotics. Key advances in sensors, CPU power and software are changing for the better, particularly in the area of 3D perception. Vision systems have become integral in countless applications in factory robotics but also in computer gaming, medical diagnosis and quality control. Emerging today are low-cost vision solutions for logistics and materials handling, for mobile devices of all types as well as for an ever-increasing need in factory robotics. These newer low-cost vision systems offer value in the form of improved product quality, manufacturing flexibility and high throughput.

Research reports covering 2D and 3D vision technologies all forecast a 9-12% Cagr or higher and project the market will reach US$10bn or more by 2020. Low-cost 3D perception is the game changer. There are other methods of machine vision but low-cost 3D solutions are disruptive because of cost, capabilities and software . . . and because we live in a 3D world.

Figure 96 Figure 97

Cognex camera on a UR robot Nachi robot with Keyence laser scanner

Source: CLSA

Capturing and processing camera and sensor data and recognizing various shapes to determine a set of robotic actions is conceptually easy. To emphasize the difficulty, however, Amazon recently challenged the industry to select and pick shelved items robotically. Twenty-eight teams from around the world rose to the competition; none of them could perform as fast as a human.

Agricultural robotics growing With the global population on track to reach nine billion by 2050, agricultural production must double to meet the demand. And because of limited arable land, labor shortages and water considerations, productivity must increase 25% to help meet that goal.

Farmers, ranchers and growers the world over are transitioning to precision agricultural methods, ie, subdividing their acreage into many unique subplots, and in some cases right down to the individual plant, tree or animal, thereby

Vision is a significant bottleneck for robot

deployment

Low-cost 3D vision is a game changer

Analytics are getting a lot better, cheaper and faster

Productivity must increase

Farmers, ranchers and growers transitioning to

precision agricultural methods

Machine vision and laser scanners




enabling increased productivity, traceability and lower overall costs. Unmanned aerial vehicles are integral to the process and are being used to map, observe, sense and spray. Agriculture employs more than 1bn people and generates over US$1.3tn worth of food annually. The subset of precision agriculture’s data capture and analytics market size is estimated at US$25bn per year.

BIS Research, a Minnesota-based company, recently predicted that the precision-farming market will experience a Cagr of 13.09% from 2015 to 2022 and reach US$6.43bn by 2022. Tractica, a Colorado research company, reported that the agricultural robotics market will reach US$3bn by the end of 2015 and US$16.8bn by the end of 2020.

Figure 98

Kubota’s automated CSAS fertilizing system

Source: CLSA

Robotics technology is on the path to impact ag with a multitude of applications which will address farmers’ constant struggle to keep costs down and productivity up. Many factors are precipitating these changes in addition to global population growth and the cost and availability of labor: the diminishing availability and increasing cost of water; political and regulatory procedures (and hold-ups); limited tillable acreage; better, cheaper and faster technological automation products; and climate change, to name just a few.

Digitally controlled farm implements are already in use in developed countries and most Western farmers and ranchers are already high-tech to some extent. Farmers use software systems and aerial or satellite survey maps and data to guide their field operations and are transitioning from satellite and small plane photos to those provided by drones with sophisticated sensors. Farmers are also using optional auto-steer kits in most new tractors that follow RTK/GPS and digital guidance.

Addressing farmers’ constant struggle to

keep costs down and productivity up

Most Western farmers and ranchers are

already high-tech

Kubota presented new automated equipment at

the annual dealers’ conference in Kyoto




Figure 99

Robotics related stocks Company Ticker Rating Currency Last


(US$m) FY15CL FY16CL FY15CL FY16CL Amazon AMZN-US O-PF US$ 719.15 1.25 6.39 575.32 112.48 339,386 Rockwell Automation ROK-US U-PF US$ 116.13 6.40 5.91 18.15 19.65 15,128 Google GOOGL-US BUY US$ 732.77 29.58 34.47 24.77 21.26 462,020 Cognex CGNX-US U-PF US$ 41.90 2.13 1.21 19.67 34.63 3,563 Accuray ARAY-US N-R US$ 5.14 (0.51) (0.27) na na 416 Northrop Grumman NOC-US N-R US$ 216.81 10.39 10.68 20.87 20.31 39,124 Boeing BA-US N-R US$ 130.50 7.72 8.51 16.90 15.33 83,130 iRobot IRBT-US N-R US$ 37.00 1.47 1.31 25.17 28.18 1,016 Raytheon RTN-US N-R US$ 135.80 6.75 7.14 20.12 19.02 40,330 Source: CLSA, FactSet

3D printing - rethinking manufacturing Interest and hype around 3D printing peaked in 2014, leading to disillusionment among investors as stocks continued to come under pressure in 2015. The consumer market appears saturated with low-cost completion. Meanwhile, advances in industrial additive manufacturing continue to progress with adoption among manufacturers doubling between 2013 and 2015 and moving beyond prototyping toward full production.

Figure 100

Innovation What it means Who could benefit Potentially at risk Related companies 3D printing Custom fabrication,

prototyping, spare parts

Consumers, designers, industrial designers, manufacturers, service providers, materials producers

Spare parts, machine tooling, mass manufacturing

3D Systems (DDD), Stratasys (SSYS), ExOne (EXONE), Proto Labs (PRLB), VoxelJet (VJET), Arcam (Sweden), envisionTEC, EOS (Germany), Renishaw (UK), Organovo (ONVO), Autodesk (ADSK), Staples (SPLS), Adobe (ADBE), Microsoft (MSFT)

Source: CLSA

In 2013 and 2014, 3D printing underwent a surge of interest from popular media and from investors as the public awakened to the promise of astonishing low-cost desktop fabrication reviving the moribund US manufacturing sector. Enthusiasm for a burgeoning consumer 3D Printing industry propelling shares of related companies to unsustainable highs. 2013 saw ExOne (XONE) and VoxelJet (VJET) join 3D Systems (DDD) and Stratasys (SSYS) as public companies. 3D Printing software firm Materialise (MTLS) launched an ADR in the US.

Hype peaked with the 2015 Consumer Electronics Show which dedicated a separate section to 3D printing and there were near daily news stories about uses of 3D printing whether for food, hobbyists, aerospace, healthcare and other uses. HP announced that it would enter the 3D printing market and test machines in 2015 before launching in 2016.

In 2015, 3D printing stocks struggled as leaders Stratasys and 3D Systems wrestled with operational challenges arising from rapid expansion. Stratasys and 3D Systems are the leaders in a market that was already 30 years old. 3D Systems is exiting its Cube consumer business. It’s our view that with sentiment around 3D printing is now in the “trough of disillusionment” as Gartner’s Hype Cycle describes it, investors are in a better position to better assess the very real industrial potential for additive manufacturing technologies.

Robust growth forecasts for the sector There’s a range of growth forecasts - all generally bullish for the commercial sector. The Wohlers Report forecasts the worldwide 3D printing industry is now expected to grow from US$3.07bn in revenue in 2013 to exceed US$21bn in worldwide revenue by 2020. Allied Market Research (AMR) projects the 3D

In 2014, the interest around 3D printing

reached the mainstream

Hype peaked in 2015

3D printing - in a hangover from peak of

hype but technology is real

In 2015, 3D Printing stocks struggled

Growth forecasts are generally bullish for the

commercial sector




printing market will grow at a Cagr of 20.6% from US$2.3bn in 2013 to US$8.6bn in 2020. IDC forecasts a 37% Cagr for 3D printers through 2019. There are high volume markets and high value markets. Most of the value comes from high-end machines. There will be a 28% Cagr in the value of the hardware sold. The materials business will exceed services in 2016 - IDC forecasts services and materials growing at a 29% Cagr to US$14bn by 2019. 3D printing for medical applications is forecast to be a US$2.13bn market by 2020, according to market research company MarketsandMarkets.com.

A diverse, fragmented landscape 3D printing has attracted widespread attention from the investment community. Along with the three US publicly-traded pure-play 3D printing firms (3D Systems, Stratasys and ExOne), there are also service bureaus including VoxelJet and Proto Labs and software pure-play Materialise. Dozens of privately held firms are making use of the open-source RepRap technology to reduce the price of 3D printers to make them accessible to average consumers.

The market remains fragmented with dozens of manufacturers selling 3D printers and over 200 startups worldwide developing and selling consumer-oriented 3D printers. Key industrial providers are 3D Systems, Stratasys, ExOne, Arcam (Sweden), envisionTEC, EOS (Germany) and Renishaw (UK). For bioprinting, there is Organovo. In the consumer market, players are 3D Systems, Bits from Bytes, MakerBot (Stratasys), Lulzbot, FormLabs, Sculpteo, Leapfrog and Solidoodle. HP’s inkjet-based technology will be coming to market in the next few quarters. RepRap is an open-source solution. Most software is bundled, though Autodesk is leading the charge to become a standard with its Spark software. Microsoft, Adobe and others have added support for 3D printing to their software. Hundreds of resellers and distributors globally typically resell CAD/CAM software. Service bureaus include Proto Labs, Kraft Wurx, Shapeways, Staples Easyprint and others.

Traditional 2D printing players are looking to expand into this market. HP will release their 3D Printer by the end of 2016. This will incorporate some of the technology they have used in business printers. Canon has also showed some signs around 3D printing. China is focusing a lot on 3D printing. There is a lot of activity in alliances and research institutes and technology parks that are focusing on 3D printing. Retailers and service providers are expanding their offerings with Staples and other service bureaus introducing prototyping businesses. Logistics service providers like FedEx or UPS are also looking into offering 3D printing as a way to provide choices to customers.

Figure 101 Figure 102

Objects printed with Autodesk Spark 3D printed wrench

Source: CLSA Source: Creative Tools from Halmdstad (Wikimedia

Commons)

The price of entry-level 3D printers continues to

drop towards the US$1,000 range

The market is fragmented with over 40 3D printer manufacturers and over

200 startups

Dozens of privately held firms are making use of

the open-source RepRap technology

Traditional 2D printing players are looking to

expand into this market




There are a lot of startups competing in the market and a number of key patents have expired as open source platforms like RepRap have given rise to new entrants particularly in China. Industry analyst firm IDC sees a shakeout coming in the industry. There are hundreds of players in the market from North America, Europe and Asia looking to build share, so margins could be difficult to sustain for hardware-based vendors.

One result of the development of 3D printing is its impact expanding the CAD and PLM markets. Autodesk’s Tinker CAD is a simple, free tool - even though the learning curve is steep. IDC expects expansion of the mid-market CAD market, so there will also be libraries and marketplaces for designs. PLM is much more sophisticated; as designs are increasingly digitized there will be a lot more embedded product intelligence. Autodesk is one of the premier CAD software firms offering design tools for 3D printing. They have open sourced and freely distributed Spark, its 3D printing software platform, and has also launched its Ember line of 3D Printers which retail for US$7,495. The company’s Project Escher networks multiple printers together into a mesh that can produce far larger, more complex objects with faster throughput.

The origins of 3D printing in manufacturing The term 3D printing is interchangeable with additive manufacturing. The technology has its origins in stereolithography, which debuted in the late 1980s. Rapid prototyping persists today, while rapid tooling didn’t catch on in the industry. Rapid manufacturing is at a nascent stage of adoption. 3D printing encompasses these three applications.

Originally used to construct prototypes and models, the technology has now advanced to the point of producing increasingly complex final goods. 3D printing works much in the same way as a traditional printer, except that the “ink” used can be any range of materials, from plastics to powdered metals. Depending on the method, an object can be created, for instance, by depositing molten plastic layer by layer or by blasting powdered metals with lasers that melt and bond the material.

Advantages lie in uniqueness and complexity There are many conversations about the potential for 3D printing among those with a financial interest in the industry: manufacturers of the equipment (OEMs), materials and software, as well as distributors. The whole value chain anticipates when additive manufacturing becomes commonplace as a tool for production because there will be more systems sold and more material consumed compared to the one-off use cases of prototyping. In prototyping, machines will make only a few units versus making tens of thousands or millions of units for production. Equipment initially designed for a prototyping machine does not have the same constraints as a production device in terms of repeatability - ensuring everything comes out exactly the same. Prototyping machines can lack throughput at higher volumes; accuracy can be a concern; and certainly material choices are key.

3D Printing/additive manufacturing - from prototypes to production There is a growing array of practical uses for 3D printing. This year GE expects to print 19 different pieces on their aircraft engines, targeting 100,000 parts by 2020. The automotive industry is ripe for adoption: the average car has 1,800 parts assembled on a production line, with 30,000 total parts. In the future, cars could be made up of only 100 parts. Privately held Local Motors is at the forefront of printing ABS (Acrylonitrile-Butadiene-Styrene plastic) and Carbon Fiber automobiles. Healthcare is using 3D

The term 3D printing is interchangeable with

additive manufacturing

The “ink” used can be any range of materials

from plastics to powdered metals

3D printing is expanding the CAD and PLM markets

There is a growing array of practical

uses for 3D printing




printing for surgical models, implants, and prosthetics. There have been titanium printed rib replacements, as well as 3D printed ears, noses and other body parts. Prosthetics can cost US$40-100,000 so they are expensive. There’s a project sponsored by Rochester Institute of Technology printing prosthetics for US$50. Students at Washington University have 3D printed a robotic arm for about US$200. Traditional robotic limbs can cost between US$50,000 to US$70,000 and they need to be replaced as children grow.

Moving beyond the hype 3D printing drives the evolution in prototyping and manufacturing, but the promise of a revolution for the consumer remains far off. Industrial 3D printing is increasingly mainstream for prototyping, though uses for manufacturing remain limited. Companies are eagerly exploring commercial uses for 3D printing. Several well-known automakers are using 3D printers for around 100,000 prototype parts per year, according to SmarTech analyst Scott Dunham. Boeing uses printers to make 200-part numbers for 10 different types of aircraft. By 2020, GE the company expects 100,000 engine parts to be made using 3D printing processes. PWC estimates that two thirds of manufacturers are already using 3D printing. Of these, 29% are exploring how 3D printing can be best integrated into production processes. 25% are using 3D printing for prototyping.

3D printing has the potential to transform the way we manufacture things. The technology offers a more efficient, less costly means of producing custom items by using significantly less raw material and avoiding the expensive retooling required by mass production lines. In other words, rather than mass producing for customers, it brings custom production to the masses. 3D printing has the potential to be highly disruptive to the manufacturing industry by enabling “mass customization” and just-in-time manufacturing.

Breakthroughs have advanced the relevance of 3D Printing There have been a number of key breakthroughs on 3D Printing technology since 2010 that have changed the market dynamics, according to former CEO of 3D Systems Avi Reichental. It’s now possible to print objects in full color with over 250 different materials and composites. Full color means over a million different color combinations in a single print. The combination of new materials and colors opens up huge possibilities. Medical devices like hearing aids and braces, implants as well as customized surgical tools can be created in bespoke fashion. Metal printing has now been adopted in the aerospace and automotive industries, with the ability to print with more than 20 different alloys. GE and SpaceX are 3D printing parts for jet and rocket engines.

The speed and throughput of 3D Printing has also increased to the point where continuing improvements will make production printing more viable. The tools and technologies have also evolved to make it easier for end-users. CAD (Computer-Aided Design) and 3D modeling tools are getting easier to use, while the emergence of cloud-based marketplaces for models and images helps democratize the technology.

There is a lot of diversity in the classes of materials that can be processed; however, in each class, there are only a few available materials. In the thermoplastic class, there is commercially available polycarbonate, nylon, acrylonitrile butadiene styrene (ABS), and a few specialty materials like PEEK or ULTEM, trademarked names for mechanically sound, temperature-resistant materials used on aircraft.

Companies are eagerly exploring commercial

uses for 3D printing

In each class, there are only a few available

materials for 3D printing

3D printing offers a more efficient, less

costly means of producing custom items




Figure 103 Figure 104

Objet Eden 250 3D printer PRUSA 3D Printer

Source: Vigh Előd, John Abella (Wikipedia)

There is a wide variety of materials for use in 3D printing, including thermoplastics, thermoplastic elastomers, ferrous alloys, non-ferrous alloys, sand, ceramic, paper, glass, electrical inks and even biological materials. Available materials are limited for each class and there is no standardized way to categorize or evaluate them, no standardized codex exists for defining their characteristics. By contrast, traditional injection molding enjoys greater variety of materials. With injection molding, not only is there more selection of thermoplastic families, but within each one of those families, there may be 50 or 100 different choices from 10 different vendors. Not so with 3D printing. Industries that lead manufacturing adoption are aerospace, dental, hearing aids and motor sports. Aerospace seeks to drive out as much weight as possible and that typically means either advanced materials or sophisticated designs that maintain strength and have much less weight. There are challenges for additive-manufactured parts on an aircraft (eg, regulations for safety). Companies like Boeing are manufacturing only non-critical components. Motor sports (eg, Formula 1, motorcycles, race cars, boats) love additive manufacturing because of their low volume and high complexity.

The dental industry is a high-profile poster child for success. There are smaller components, which are good for additive manufacturing because they are typically time dependent and cost is size dependent. It’s not feature dependent like traditional manufacturing. Align Technologies’ InvisAlign braces for teeth straightening are printed and work for a week or two at a time. This is facilitated by additive manufacturing. The hearing-aid industry has been using additive manufacturing for 12 to 15 years; over 90% of all ear-hearing-aid shells are produced through additive manufacturing.

Bioprinting offers a lot of promise There is no potential application more exciting than in the emerging subfield of 3D bioprinting. Medical researchers and startups are already developing methods of printing human organs using genetically identical tissue from patients. Organovo (ONVO) is developing techniques to print human tissue for skin grafts and surgery.

Industries that lead adoption are aerospace,

dental, hearing aids and motor sports

Dental is a high-profile poster child for success

Researchers have developed ways to

fabricate genes using new types of chips




3D Bioprinting Solutions is a Skolkovo, Russia-based company headed by Vladimir Mironov that announced in March 2015 that it printed a thyroid gland for an animal (mouse) using the company’s patented bioprinting process. Researchers at Princeton University have developed a proof of concept for a 3D-printed bionic ear. The scientists seeded a culturing cartilage with human cells in the shape of a human ear, and then embedded an inductive coil made of polymer with silver nanoparticles to allow signal processing from cochlea-shaped electrodes. Joseph Jacobson, head of the MIT Media Lab's Molecular Machines research group, is bio-fabricating genes using CMOS chips with integrated sending electrodes on silicon. Although implementation of printed organs remains some years off, replacement parts for humans - as there already are for machines - may someday be readily available at the touch of a button.

Key issues around 3D printing include intellectual property (IP) rights. The question is who owns the designs to objects if they can be scanned and converted into a 3D file? Current law does not protect the digital rendering of a physical object, only the associated brands and trademarks. Without established IP laws, we expect continuing skirmishes to arise over ownership and rights for 3D schematics. 3D printing marketplaces like Shapeways or Makerbot’s Thingverse include digital renderings of copyrighted characters, but under the Digital Millennium Copyright Act, marketplaces have a safe harbor provision as long as they have a posted policy that says they will remove infringing material if the copyright holder requests it.

Print me up a car One of the most impressive demonstrations of the technology is 3D printing a car. Local Motors designs “kit cars” that are crowd-sourced designs, and the components can be machined for easy assembly by customers who are also hobbyists. Local Motors’ Strati is an electric vehicle with just 25 parts. In the future, customers may be able to walk into a mall, choose the configuration and style and have it complexly printed within a few hours. While this may seem very far off, the technology is on a trajectory that could realize this vision within a few years.

Figure 105

Assembling the Strati

Source: CLSA

The Strati electric car has 25 discrete parts versus

1,700 for the average automobile

Local Motors’ Strati is an electric vehicle

with just 25 parts

Key issues around 3D printing include

intellectual property rights




Falling prices stimulate growing consumer market The consumer market for 3D printing continues to attract dozens of startups as prices fall as key patents held by Stratasys and others expire. Designers have been using use printers like the MakerBot to start consumer-oriented businesses selling shoes and jewelry (often on sites like Etsy) and we are seeing 3D printing devices and services developing a consumer market. The MakerBot Replicator Mini is priced at US$1,375, less than half the price of the full-size Replicator at US$2,899. There is also a plethora of startups offering sub-US$500 devices including Solidoodle 2 Pro, XYZ Printing Da Vinci 1.0 and PrintrBot Simple Maker Edition Kit.

Future breakthroughs There are a number of exciting developments that could prove disruptive to a number of different industries. A new generation of 3D printers that print edible products promises to enable personalized nutrition, for instance creating a personalized nutrition bar with the precise amount of proteins, carbs, vitamins and supplements needed in that moment. Over time this could give rise to personalized pharmaceuticals compounded and created specifically for the end user.

Fashion is another area where 3D Printing is gaining ground, particularly in shoes and garments. Nike and other shoe makers will be able to create shoes customized to size, posture, stance, and your arch. Accessories will be customizable and immediately printable. There’s a lot of interest in printing accessories and clothing, for example Manufacture NY is a new initiative in Brooklyn’s DUMBO neighborhood collocating designers and 3D fabricators for all manner of apparel and accessories. Technology is advancing to enable printing of mixed-material devices that may include rubber, structure and wiring so that we may be able to print electronics, cars and even houses.

The combination of 3D Printing (additive manufacturing) with traditional subtractive manufacturing promises to increase production speeds by orders of magnitude. We will increasingly see 3D printing combined with CNC machining, additive layering, and injection molding into “factory in a box” systems.

As we highlighted in Section 1 of this report, 3D printing stocks have experienced a cycle of hype followed by disappointment. Although there are limited investment opportunities at present, the technology itself promised far greater impact among users of the technology.

Figure 106

3D-printing related stocks


EPS (US$) PE (x) Market cap (US$m) FY15CL FY16CL FY15CL FY16CL

3D Systems DDD-US N-R US$ 12.56 0.27 0.33 46.52 38.24 1,407

Stratasys SSYS-US N-R US$ 21.45 0.19 0.33 112.89 65.74 1,118

ExOne XONE-US N-R US$ 10.10 (1.79) (0.90) na na 162

Voxeljet VJET-US N-R US$ 4.73 (2.81) (1.17) na na 88

Materialise MTLS-US N-R US$ 7.19 (0.07) (0.05) na na 340

Autodesk ADSK-US N-R US$ 55.71 0.84 (0.82) 66.32 na 12,516 Source: CLSA, FactSet

3D printing devices and services are rapidly

developing a consumer market

Fashion is another area where 3D Printing is

gaining ground




The Internet of Things hits the mainstream Hype around the Internet of Things reached fever pitch in the mainstream media in 2015, as technology and industrial raced to articulate their out IoT strategies. Our 2014 report Deep Field: Discovering the Internet of Things focused on growing relevance and opportunities across consumers and businesses in a full spectrum of industries. Interest in wearable computing and smart home also reached fever pitch in 2015 with a flood of new market entrants resulting in a highly fragmented market. In 2016 we are seeing industrial proof of concept projects chart paths to sustainable ROI, but some businesses are taking a wait-and-see approach. While we do not expect a major inflection point in IoT spending, the success of small projects paves the way for greater investments down the road.

Figure 107

Innovation What it means Who could benefit Potentially at risk Related companies Connected Everything (IoT, eHealth, Sharing economy)

Myriad implications for both industrial and consumer

Consumers, businesses, manufacturers, logistics, military, public safety, wireless sensor network providers, analytic software vendors

Companies with high reliance on manual processes

IBM (IBM), Cisco (CSCO), GE (GE), PTC (PTC), National Instruments (NI),Google (GOOG), Intel (INTC), AMD (AMD), Siemens (SI), Oracle (ORCL), Salesforce (CRM), Amazon (AMZN), Teradata (TDC), SAP (SAP), Splunk (SPLK), Broadcom (BCOM), Qualcomm (QCOM); Apple (AAPL), Samsung (5930), Sony (6758), Nike (NKE), Intel (INTC), Microsoft (MSFT), GoPro (GPRO), FitBit (FIT) , wireless network, sensor and analytic software vendors

Source: CLSA

The term, Internet of Things, broadly encompasses a class of devices that ‘can monitor their environment, report their status, receive instructions and even take action based on the information they receive’, according to a definition by The McKinsey Global Institute.

2015 was the year that the Internet of Things as a term went mainstream. Our September 2014 report Deep Field: Discovering the Internet of Things focused on the growing relevance and opportunities across consumers and businesses in a full spectrum of industries. At the 2016 Consumer Electronics Show over 900 companies out of 3,800 at the show said they had Internet of Things products.

The promise of IoT results from the confluence of powerful technology innovations (eg, sensor networks, ubiquitous connectivity, cloud computing, Big Data, Internet Protocol version 6 [IPv6] and other communications protocols, declining chip and compute costs). Miniaturization and declining cost of sensors make it increasingly easy for devices and physical assets to be instrumented; declining cost of connectivity makes it easier to transmit data so it can be collected, analyzed and acted upon based on context and business need - in many cases automatically.

The scope of what’s referred to as the Internet of Things is extraordinarily broad, touching nearly every segment of the economy. Much of the early focus has centered on industrial uses (eg, GE’s Industrial Internet), public infrastructure (IBM’s Smarter Cities) and energy (smart-grid initiatives). New services are democratizing access for individuals as well. The automotive and transportation industries are actively adopting connected applications. There is a growing range of products in self-tracking and healthcare (fitness bands and other wearables) and home automation.

Market interest around the Internet of Things is building, but the landscape is still taking shape and remains difficult to delineate precisely. Initial multitrillion-dollar forecasts have focused on economic value-add. Various

The promise of IoT results from the confluence of

powerful technology innovations

New services are democratizing access

for individuals

We know it’s big, but it’s still early

A world connected - the Internet of Everything





sources forecast dedicated technology spend on IoT to range between US$300-500m by 2020. Investment is focused around consumer technologies (home automation and health & fitness) and convergence industrial/operational and information technologies.

We had anticipated that hype would outpace reality in 2015, and what’s happened so far in 2016 suggests that IoT may have entered what Gartner refers to as the “trough of disillusionment” - the disenchantment with the hard realities of implementing new technologies following a “peak of inflated expectations”.

On the consumer side, we’ve see some air come out of the proverbial bubble. 2015 was the year that hype around wearable computing reached peaked, and a flood of new market entrants has resulted in a highly fragmented market. FitBit’s (FIT) IPO in July 2015 was initially greeted with a lot of investor enthusiasm but growing concerns over competition and commoditization caused the shares to trade down. As of April 27, 2016, the shares are trading at US$18.10, a 65% decline from annual high of US$47.60. We expect Quantified Self, medical monitoring and vertical applications to give rise to a plethora of specialized wearable devices over time, but business value will be created from applications and data rather than devices themselves.

The home automation sector did see Alarm.com IPO in June 2015 shares gaining 36% as of 27 April 2016. However, Google’s Nest business unit has seen some unfavorable public commentary. After Google announced it would discontinue any support for its Revolv automation hub in March 2016, the former CEO of Dropcam spoke out publicly criticizing management of the Nest product line. Apparently Dropcam accounted for over half of the US$340m in annual revenues where the unit was losing over US$100m per year. Samsung’s SmartThings unit also faced negative blowback in the summer of 2015 when a faulty software update created problems for users.

Figure 108

Consumer IoT application categories and sample devices

Source: IDC, March 2015

On the commercial/industrial side, there’s been a lot of steady progress with proof of concept projects. IBM pledged to invest US$3bn into IoT technologies and solutions. Cisco and GE have oriented their market messaging and product strategy around IoT and connected products. Salesforce, Amazon and Microsoft have launched IoT technology platforms. Intel reports IoT-related revenues, and carriers like ATT, Verizon, Telefonica and others are investing in IoT businesss units.

2015 was the year that hype around wearable

computing reached peak hype

Google’s Nest business unit has seen some unfavorable public

commentary

IoT may have entered what Gartner refers to as

the “trough of disillusionment”

The consumer IoT is the most visible in the market

today




Forecasts anticipate robust growth IDC estimates that worldwide Internet of Things (IoT) market spend will grow from US$591.7bn in 2014 to US$1.3tn in 2019, a Cagr of 17%. The installed base of IoT endpoints is expected to grow from 9.7 billion in 2014 to over 25.6 billion in 2019, reaching 30 billion in 2020. Devices/hardware, connectivity, and ongoing services are estimated to account for approximately 60% of worldwide IoT market spend in 2019. The Asia/Pacific region leads the market with roughly 38% of spending in 2019, followed by North America with 28% and Europe with 22%.

ABI Research forecasts that IoT-related value-added services will grow 1t a 16% Cagr from US$50bn in 2012 to US$120bn in 2018. BI Intelligence forecasts IoT software and services will grow at a 44% Cagr to reach US$600bn by 2019, with number of IoT connected growing at a 35% Cagr through 2019. A survey by Vodaphone found that 27% of respondents have a Machine to Machine (M2M) project in place, up from 22% in 2014 with a further 37% reporting projects ready to go live within two years.

Figure 109

Internet of Things market forecast to grow at a 17% Cagr

Source: IDC

Gartner forecasts Internet of Things endpoint spending will reach US$1.5tn in consumer, US$566bn in cross-industry business, and US$911bn in vertical business. Gartner forecasts 6.4 billion connected things in use in 2016, growing a 30% YoY to reach 20.8 billion by 2020. 5.5 million new things will get connected every day in 2016. Gartner also estimates that through 2018, 75% of IoT projects will take up to twice as long as planned. GrandView Research estimates the IoT market at US$606bn in 2015 growing at a 15%

IDC estimates IoT market spend will grow from

US$591.7bn in 2014 to US$1.3tn in 2019

IoT market expected to grow to US$1.3tn




Cagr to US$1.9tn by 2022. There are also forecasts that look at the number of connections. Ericsson, the telecommunications company, has forecast 50bn connected devices by 2020 including M2M and other devices. Cisco estimates that every second 80 “things” are newly connected to the internet, and this is expected to reach 250 per second by 2020, reaching 50 billion connected “things” and five billion connected people. Gartner estimates there will be 21 billion IoT devices connected by 2020.

Manufacturing is where the biggest IoT opportunity resides According to IDC, process manufacturing (US$162bn) and discrete manufacturing (US$163bn) are the largest market opportunities. There are 110m cars on the road with massive opportunity for sensors. By 2025 there could be 80bn sensors connected globally. Agriculture is a multi billionsensor market. There are massive opportunities for factories - some of which have 50,000 sensors. The home is a big opportunity as are human beings. Over 4,800 devices are connected to networks every minute. By 2025 IDC estimates there will be 152,200 devices connected every minute.

There is an enormous amount of data being created as well. IDC’s Digital Universe project has broken out types of data: classically created (transactions), IoT relevant data (sensor data) and IoT actionable data. 44 Zettabytes in 2020 becomes 180 ZettaBytes. IoT actionable data will grow in 2025 at a rate eight times that and be as big as the digital universe in 2020.

Estimates of aggregate impact are even larger Different approaches have estimated IoT’s value potential in the trillions of dollars. GE forecasts that the Industrial Internet will add US$15tn to the world economy over the next 20 years. McKinsey Global Institute estimates that the Internet of Things has the ability to create potential annual economic impact of between US$2.7-6.2tn annually by 2025. Most current solutions target cost savings and efficiency gains. New revenue-generating and Everything-as-a-Service apps will create the greatest lasting value in our view. Platform economics will power the long-term winners.

Cisco prominently publicized its view that there is US$14.4tn of “value at stake” globally over the next decade, driven by connecting people to people, people to machines and machines to people. Value at stake represents potential profits that can be created from improved asset utilization, employee productivity, supply chain and logistics efficiencies, improved customer experiences and innovation.

The pace of Internet-of-Things M&A is accelerating M&A of Internet of Things continues as a torrid pace, with Cisco’s acquisition of Jasper Technologies for US$1.4bn in February 2016 the most significant deal of late. PTC has spent over US$500m on several IoT assets including Coldlight, Axeda and ThingWorx. According to M&A advisory firm Hampleton Partners, there was over US$39.9bn in total disclosed IoT transaction value over the past three years, Transaction volumes increased by 110% YoY in the period between November 2014-2015, with a median disclosed revenue multiple of 3.5x; multiples ranged from 0.5x revenue through to 51.6x. According to Hampleton Partners’ analysts, the initial wave of M&A focused on lower levels of the tech stack as players sought to establish ownership of the basic building blocks for IoT solutions. Business, consumer and government demand is driving adoption and currently M&A activity is most active around monitoring.

Ericsson and Cisco forecast 50bn connected

devices by 2020

Over US$7.5bn in total disclosed IoT transaction

value from Jan 12 to Sep 14

Massive opportunity defined by new

business models

Cisco believes there is US$14.4tn of “value

at stake”

Process and discrete manufacturing are the

largest IoT market opportunities




Capital is readily accessible for startups By one estimate, 10% of VC investments are IoT related. There are a growing number of IoT-dedicated efforts from Intel Capital, Frost Capital, McRock, Motus and many other firms. According to CB Insights, IoT companies raised over US$2.0bn across 221 deals in 2014, with US$1.5bn 141 for the first nine months of 2015. Crowdfunding is also playing a key role, with IndieGoGo and Kickstarter helping startups test the market for their ideas.

Asia leads adoption of IoT solution According to IDC’s the US and China will account for around 65% of the market opportunity together. Every region is moving ahead for different reasons. In the US adoption is focusing on Digital Transformation. In China, there’s a surge of investment focusing on bringing a middle class online. Japan is interested in creating a connected society around manufacturing. In Europe adoption is driven by the auto industry and Industry 4.0 initiatives.

Internet of Things is accelerating innovation In the view of IDC, the IoT is the most advanced and most important of innovation accelerators what the analyst firm refers to as the “3rd Platform” - a set of technologies encompassing cloud, big data, social and mobile. The cloud is necessary for scale - variable workloads from different devices that need to scale. IoT needs to have analytics engines that enable real-time decision making.

Spending will target applications, services and analytics

We see parallels to the enterprise-resource-planning (ERP) market in the late 1980s and early 1990s as benefits of declining compute and storage costs are amplified by near-ubiquitous connectivity. For the near term, IoT relies on customized software and services. Over time, we see growth of SaaS businesses and connected products.

Software lies at the heart of IoT solutions, with value accruing to applications, the analytics “stack” and platforms. Concerns over heterogeneous standards are overdone in our view. As IoT-specific standards emerge, interoperability is key to success, facilitated by evolving software platforms and cheaper compute cost.

Unique requirements of IoT applications are driving innovations in hardware, components, wireless networks and networking architecture. IoT-native devices will benefit from cheap open-source components, improving power management and lean communication protocols. Next-generation architectures will power IoT-designed wireless networks and “fog computing”.

Intelligence at the core of IoT value The idea that machines can be instrumented with sensors and connected to the internet to transmit data and updates to business has been around for many years. The intersection of information technology and operational technology is the key and there is no dominant vendor at this point. An intelligent system increases the value of business decisions. The value of IoT solution is proportionate to the speed and number of times IoT data is analyzed. Value and speed are the two vectors to measure potential value creation. The City of Melbourne has a highly sensor-enabled tram system that gives citizens apps to see what’s going on. The trams are free and the business district never has problems with parking.

Software and services are the big opportunities

IoT companies raised over US$1.5bn across 153

deals in 2014

Native innovations plumbing the future

While fragmented, software platforms will

knit the IoT together

The US and China will account for around 65%

of the IoT market

Intersection of information technology

and operational technology is the key




Cost reduction, incremental revenues and customer satisfaction are the main areas for value creation from data and predictive or preventive maintenance. Replenishment automation is both a revenue enhancer and customer satisfaction driver.

Creating a high value IoT system of intelligence The IoT ecosystem is enormously complex. For devices the main focus is energy consumption and density. There’s a lot of discussion of who will build the network. There’s an aggregation point called the IoT gateway (this is key for ownership and mindshare). There needs to be an open-data platform as the input funnel for solutions. IoT platforms need to manage at an enterprise level. Most IoT contracts are in the US$5-10m range. There are not a lot of big paydays right away for most vendors.

Figure 110

Benefits accrue broadly from IoT adoption Category Applications Benefits

Smart cities Street lights that dim when roads are empty, traffic sensors that optimize lights to traffic flow, monitoring bridges and other public infrastructure to monitor maintenance

Energy savings, lower congestion, cost savings from proactive maintenance

Healthcare Remote monitoring of patients, personal health monitoring, diagnostic and surgical procedures assisted by ingestible computing devices

Better quality care, more frequent access to professionals, better health

Automotive Accident alerts (GM OnStar), automated traffic-flow optimization, proactive maintenance alerts

Regulatory requirement

Supply chain and logistics Fleet optimization, supply-chain tracking and product tracking Cost savings, improved logistics

Intelligent buildings Automated monitoring and adjustments of HVAC based on human activity Energy cost savings

Retail/commerce Wireless payments, indoor "geo-fencing" applications, enhanced omni-channel experience

Higher sales, improved customer satisfaction and loyalty

Construction, mining and transportation

Sensors to monitor location, usage of vehicles, anticipate preventive maintenance

Better efficiency, lower repair costs, less downtime

Emergency services Faster, better coordinated disaster response Improved safety, response time

Source: CLSA

A market at the early stages of exponential growth Much of the activity around Internet-of-Things startups, device adoption and value-add is expected to accelerate in the 2018-20 period. Currently, there are early adopters in retail, energy (notably lighting and smart grid), healthcare and industrial equipment but we expect a broad range of products to be sold into different markets. In healthcare, there will be an increasing range of wearable devices, monitors, surgical tools and even ingestible devices. In physical infrastructure, we will see devices for monitoring all types of transportation (air, rail, shipping and trucks) as well as sensors on roads, bridges and in public spaces. In agriculture, there will be soil, temperature and weather sensors. Factories will increasingly deploy sensors on industrial machinery and robotics. Some of the key industries likely to see massive changes and benefits from adoption of connected devices include healthcare, manufacturing, insurance, banking, retail, computing services, government, transportation, utilities and others.

Retailers are the most advanced industry in terms of adoption. The retail industry is smaller (around US$70bn according to IDC) and that’s leading the way for IoT customers. Burberry currently tracks 80% of all transactions and everything in the supply chain. One of the biggest challenges is getting people to opt in for retail. In manufacturing there’s a far more spread out view of IoT deployments. Harley Davidson makes a new motorcycle every 86 seconds. Healthcare is a laggard, but adoption is still at the learning phase, because its risk averse, highly regulated and security conscious.

Activity around the Internet of Things is

expected to accelerate in the 2018-20 period

Retailers are the most advanced industry in

terms of IoT adoption

IoT ecosystem is enormously complex




Business model shift - From buying and selling to renting Perhaps the most subtle yet significant benefit in a world of connected “things”. While the trend of attaching maintenance and monitoring services is not new, what’s increasingly possible is the business model transition from buying and selling products to renting assets. Internet-of-Things technologies enable a new class of applications that deliver goods and services on demand to customers (mostly consumers) via the smartphone.

Figure 111

Technologies enable asset-sharing, rental and context-based services

Source: Steven Schlafman, RRE Ventures

According to a Deloitte survey, most current IoT applications target cost savings, visibility, efficiency gains and risk reduction with only 15% focused on innovation and new business models. We believe an inflection point will follow from new services that harness incremental value from data and analytics. Importantly, the Internet of Things allows buying-and-selling businesses to transform to rental businesses, while enabling cities, homes and cars to become platforms at the center of further innovation.

Uber is a prime example. Its smartphone app allows customers to order a car service while being able to see where cars are in real time on a map. The ride is paid for on the rider’s account through Uber, so no money changes hands, and both the driver and rider are rated (leveraging social technologies). In many respects, this is a simple location-aware ecommerce app; what could make this truly an IoT experience is if the car were self-driving.

Daimler makes Cars To Go, an IoT app that allows a customer to get a car based on a local app. Launched in several German cities and San Francisco initially, the app allows customers to rent from a fleet of cars by the minute or the hour. The cars can be parked in reserved street spaces and users can

The rise of on-demand mobile services is an

outgrowth of connectivity and location services

Internet-of-Things technologies enable a

new class of applications

Cars To Go is an IoT app that allows a

customer to get a car based on a local app

New business models and innovation will

provide the catalyst




monitor the electric charge or gas tank levels, mileage and other diagnostics. This is a way for an automaker to become an operator of a rental fleet. This cuts out fleet providers, distributors, insurance companies and other intermediaries.

Continental automotive subsystems are developed so that devices in autos are an IoT package that enables connected vehicles. Cummins makes industrial diesel engines for tricks and combines, IBM enables them to collect information and potentially rent it to customers. Pratt & Whitney’s jet engines have 16% more fuel economy from thousands of sensors, with 1TB of data per flight analyzed to help improve reliability and reduce fuel cost. The company runs this as a service, with airlines renting engines. Whirlpool has connected devices for home appliances. The value is not from connections and instrumentation, it is the value of the data and the experience it creates for customers. The entire delivery model changes in a connected products environment.

Figure 112

IoT technology elements

Source: IDC, 2015

Crowd Companies founder Jeremiah Owyang estimated that sharing economy companies have received over US$21bn in funding over the last 10 years. A study by PWC estimates that five main sharing-economy sectors (peer-to-peer lending and crowdfunding, online staffing, car sharing, peer-to-peer accommodations and music and video streaming) accounted for US$15bn in revenue in 2013, growing to US$335bn by 2025. The traditional rental sector, including equipment, bed & breakfast (B&B) and hostels, books, cars and DVDs, accounted for US$240bn in 2013 and are also expected to reach US$335bn in 2025. There has been quite a bit of pushback though, car-for-hire service Lyft announced it would pay US$12m to settle a class action suit along with extending some benefits for drivers. Uber and Airbnb have also faced lawsuits over employee benefits and customer safety.

IoT technologies span the full stack plus

cloud services and vertical expertise

The value is not from connections and

instrumentation, it’s the value of the data

PwC estimates that five sharing-economy

sectors will reach US$335bn by 2025




For investors, the Internet of Things poses a bewildering array of potential winners and losers with a sprawling universe of companies exposed to different layers in the technology stack and different segments in the value chain. There is a broad range of technologies with applicability and varying stages of maturity. There’s connectivity, which includes cellular, fixed broadband, WiFi, protocols like ZigBee, ZWave and other networking technologies. There are hardware requirements such as at end points there’s need for modules and sensors, gateways or propagator nodes where data is aggregated and real-time analytics performed. There’s need for security hardware, server, storage and other hardware including switches, routers and industry-specialized hardware.

Software includes analytics, for real-time analysis and event-triggers as well as deeper trend and predictive analytics. Application software can extract data from physical systems and translate to workflow, or provide industry specific functions. Platforms can handle the flow of data from device through the network, enabling collection, integration and analysis of data. Services may include consulting, IT implementation and installation, managed services and outsourcing and delivery of security, hardware and software as a service. For large-cap investors, companies like GE and Cisco have made IoT central to their messaging and business strategy, so to some extent these companies provide a proxy for adoption of IoT solutions more broadly. In software, we favor PTC as a platform play - its ThingWorx offering is arguably the broadest and deepest in the market, and while it represents just 8% of software revenues for the firm, we’d expect robust growth to continue for the foreseeable future. Splunk also represents a promising and unique play on machine data analytics and the company has a dedicated IoT practice in early stages of evolution headed by a former GE Capital CIO.

Figure 113

Internet of Things related stocks Company Ticker Rating Currency Last


(US$m) FY15CL FY16CL FY15CL FY16CL IBM IBM-US O-PF US$ 151.02 14.92 13.44 10.12 11.24 145,021 PTC PTC-US O-PF US$ 37.75 2.23 1.59 16.93 23.74 4,327 Google GOOGL-US BUY US$ 732.77 29.58 34.47 24.77 21.26 462,020 Oracle ORCL-US U-PF US$ 38.84 2.62 2.80 14.82 13.87 161,181 Salesforce.com CRM-US BUY US$ 81.05 0.75 0.95 108.07 85.32 54,938 Amazon AMZN-US O-PF US$ 719.15 1.25 6.39 575.32 112.48 339,386 Teradata TDC-US BUY US$ 27.30 2.06 2.43 13.25 11.23 3,549 Splunk SPLK-US BUY US$ 56.64 0.18 0.31 314.67 182.71 7,526 Samsung 005930-KR BUY won 1,380,000 121,491 152,165 11.36 9.07 188,491 Sony 6758-JP BUY ¥ 2941.00 119.40 90.60 24.63 32.46 35,055 Microsoft MSFT-US O-PF US$ 49.81 2.63 2.70 18.94 18.45 391,687 GoPro GPRO-US N-R US$ 9.71 0.76 (0.91) 12.78 na 989 Fitbit FIT-US N-R US$ 13.00 1.07 1.17 12.15 11.10 2,018 Qualcomm QCOM-US N-R US$ 52.70 4.66 4.10 11.31 12.87 77,412 Intel INTC-US N-R US$ 32.14 2.33 2.39 13.79 13.44 151,765 GE GE-US N-R US$ 30.45 1.31 1.50 23.24 20.28 279,916 Cisco CSCO-US N-R US$ 28.96 2.21 2.33 13.10 12.42 145,660 FleetMatics FLTX-US N-R US$ 42.76 1.58 1.76 27.06 24.33 1,667 CalAmp CAMP-US N-R US$ 14.42 1.15 1.24 12.54 11.66 529 Numerex NMRX-US N-R US$ 7.45 (0.12) (0.17) na na 145 Digi International DGII-US N-R US$ 11.12 0.26 0.39 42.77 28.33 288 AT&T T-US N-R US$ 40.40 2.71 2.86 14.91 14.14 248,721 Verizon VZ-US N-R US$ 52.99 3.99 3.93 13.28 13.50 216,003 Source: CLSA, FactSet

The Internet of Things poses a bewildering array

of potential winners and losers


Important disclosures 2020 innovation


Companies mentioned 3D Bioprinting Solutions (N-R) 3D Systems (N-R) Accenture (ACN US - US$116.47 - OUTPERFORM)¹ Acxion (N-R) Adobe Systems (N-R) Aerovironment (N-R) Airbnb (N-R) Alcatel-Lucent (N-R) Align Technologies (N-R) Alphabet (GOOGL US - US$721.71 - BUY)¹ Amazon (AMZN US - US$702.80 - OUTPERFORM)¹ Ambarella (N-R) AMD (N-R) Amex (N-R) Apple (AAPL US - US$95.22 - BUY)¹ Applied Micro (N-R) Arcam (N-R) ArcSight (N-R) Aruba Networks (N-R) AT&T (N-R) Audi (N-R) Autodesk (N-R) AVG (N-R) Baidu (BIDU US - US$170.05 - BUY)² Barclays (N-R) Barracuda Networks (N-R) BBVA (N-R) BlackDuck Software (N-R) BlackRock (BLK US - US$354.01 - OUTPERFORM)¹ Bloomberg (N-R) BMW (N-R) Boeing (N-R) Bosch (N-R) Box (N-R) Broadcom (AVGO US - US$146.73 - OUTPERFORM)¹ Broadcom (N-R) BYD Electronic (N-R) CA Technologies (N-R) Canon (7751 JP - ¥3,115 - BUY)² CBA (CBA AU - A$77.96 - SELL)² Checkpoint Systems (N-R) Cisco (N-R) Cleversafe (N-R) Cloudera (N-R) Cognex (CGNX US - US$40.11 - UNDERPERFORM)¹ Cognitive Scale (N-R) Cognizant Tech (CTSH US - US$62.94 - BUY)¹ Coin Desk (N-R) Coinbase (N-R) Coldlight (N-R) Composite Software (N-R) Credit Suisse (N-R) CSC (N-R) Cvent (N-R)




CyberArk (N-R) Cycorp (N-R) DataStax (N-R) Dell (N-R) Didi-Kuaidi (N-R) Digital Reasoning (N-R) DISH Network (N-R) Dominos Pizza (N-R) Dropcam (N-R) Dupont (DD US - US$65.75 - UNDERPERFORM)¹ EMC (EMC US - US$27.81 - OUTPERFORM)¹ Emerson (EMR US - US$50.65 - UNDERPERFORM)¹ EnerNoc (N-R) EOS (N-R) Ethereum (N-R) Eucalyptus (N-R) ExOne (N-R) Expedia (N-R) Experian (N-R) Fab.com (N-R) Facebook (FB US - US$117.35 - BUY)¹ FICO (N-R) FireEye (N-R) Fitbit (N-R) Flipkart (N-R) Ford Motor (F US - US$13.19 - OUTPERFORM)¹ Fortinet (N-R) Fujitsu (6702 JP - ¥417 - BUY)² Gemalto (N-R) General Dynamics (N-R) General Electric (N-R) General Motors (GM US - US$30.57 - OUTPERFORM)¹ GitHub (N-R) Goldman Sachs (GS US - US$154.51 - OUTPERFORM)¹ GoPro (N-R) Greenplum (N-R) Gridpoint (N-R) Hitachi (6501 JP - ¥495 - BUY)² Honeywell (N-R) Hortonworks (N-R) HP (HPQ US - US$11.66 - UNDERPERFORM)¹ HTC (2498 TT - NT$67.3 - BUY)³ IBM (IBM US - US$147.25 - OUTPERFORM)¹ Imperva (N-R) Imprivata (N-R) Infosys (INFO IB - RS1,190.8 - OUTPERFORM)² Ingram Micro (IM US - US$33.79 - UNDERPERFORM)¹ Intel (INTC US - US$30.15 - UNDERPERFORM)¹ Interbrand (N-R) Intuit (N-R) Intuitive Surgical (N-R) Invensys (N-R) IP Softcom (N-R) iTron (N-R) IXYS Corp. (N-R)




Jasper (N-R) Jaunt (N-R) Joulex (N-R) JPMorgan Chase (JPM US - US$63.51 - BUY)¹ Kayak (N-R) Kensho (N-R) Lenovo (992 HK - HK$4.88 - BUY)² Lexmark (N-R) LinkedIn (LNKD US - US$125.56 - BUY)¹ Lockheed Martin (N-R) Looking Glass (N-R) Loop AI (N-R) Lyft (N-R) Mashery (N-R) Materialise (N-R) McAfee (N-R) Meituan-Dianping (N-R) Mercedes-Benz (N-R) Microsoft (MSFT US - US$50.62 - OUTPERFORM)¹ Mind Meld (N-R) Mobiletron (N-R) Mobileye (MBLY US - US$36.84 - BUY)¹ MongoDB (N-R) Moog (N-R) Motorola Solutions (MSI US - US$68.75 - UNDERPERFORM)¹ National Instruments (N-R) Nest (N-R) Netflix (N-R) New Egg (N-R) Nice Systems (N-R) Nike (N-R) Nissan Motor (7201 JP - ¥1,050 - BUY)² Northrop Grumman (N-R) NQ Mobile (N-R) Numenta (N-R) Nvidia (N-R) OpenTable (N-R) Oracle (ORCL US - US$39.41 - UNDERPERFORM)¹ Palantir (N-R) Palo Alto (N-R) Panasonic (6752 JP - ¥963 - BUY)² Panaya (N-R) Pinterest (N-R) Pivotal Labs (N-R) Polycon (N-R) Priceline (N-R) Proofpoint (N-R) Proto Labs (N-R) PwC (N-R) Qihoo 360 (N-R) QlikTech (QLIK US - US$30.95 - BUY)¹ Qualcomm (QCOM US - US$54.52 - BUY)¹ Qualys (N-R) Quest Software (N-R) Quirky (N-R)




Rackspace (RAX US - US$23.53 - SELL)¹ Rapid7 (N-R) Raytheon (N-R) Red Box (N-R) Renishaw (N-R) Rio Tinto (RIO AU - A$44.00 - UNDERPERFORM)² Ripple (N-R) Risk Metrics (N-R) Rockwell (ROK US - US$111.17 - UNDERPERFORM)¹ Royal Bank of Scotland (N-R) RSA (N-R) Saffron Technology (N-R) SAIC (N-R) Salesforce.com (CRM US - US$81.02 - BUY)¹ Samsung Electronics (005930 KS - 1,286,000 WON - BUY)² Secret (N-R) SecureWorks (N-R) Sentient Technologies (N-R) ServiceMax (N-R) Shapeways (N-R) Siemens AG (N-R) Silverspring Networks (N-R) Skyworks (N-R) Snapchat (N-R) Softlayer (N-R) Sony (6758 JP - ¥2,904 - BUY)² Sourcefire (N-R) SpaceX (N-R) ST Microele (N-R) Staples (N-R) State Street (STT US - US$61.43 - UNDERPERFORM)¹ Stratasys (N-R) Symantec (N-R) Tableau (N-R) Tata Comm (TCOM IN - RS449.5 - BUY)² Teledyne (N-R) Teradata (TDC US - US$26.44 - BUY)¹ Tesla (TSLA US - US$220.28 - UNDERPERFORM)¹ Texas Instruments (TXN US - US$58.53 - UNDERPERFORM)¹ Textron (N-R) Textura (N-R) Theranos (N-R) Tiger Direct (N-R) Time Publishing (N-R) Toyota Motor (7203 JP - ¥5,506 - BUY)² Twitter (TWTR US - US$14.43 - UNDERPERFORM)¹ Uber (N-R) UBS (N-R) UPS (N-R) Varonis (N-R) Verisk Analytics (N-R) Vertica (N-R) Virgin Galactic (N-R) VMware (VMW US - US$59.76 - OUTPERFORM)¹ Vocal IQ (N-R)




Volkswagen (N-R) Voltage (N-R) Volvo (N-R) Voxel Jet (N-R) Vuforia (N-R) Walmart (N-R) WebEx (N-R) Whetlab (N-R) XenSource (N-R) Xiami (N-R) Yahoo! (YHOO US - US$36.50 - OUTPERFORM)¹ Zenefits (N-R) Zipcar (N-R) Zynga (ZNGA US - US$2.56 - UNDERPERFORM)¹ ¹ Covered by CLSA Americas; ² Covered by CLSA; ³ Covered by CAST Analyst certification The analyst(s) of this report hereby certify that the views expressed in this research report accurately reflect my/our own personal views about the securities and/or the issuers and that no part of my/our compensation was, is, or will be directly or indirectly related to the specific recommendation or views contained in this research report.

Important disclosures

The policy of CLSA (which for the purpose of this

disclosure includes subsidiaries of CLSA B.V. and CLSA Americas, LLC ("CLSA Americas")), and Credit Agricole Securities (Taiwan) Company Limited (“CA Taiwan”) is to only publish research that is impartial, independent, clear, fair, and not misleading. Analysts may not receive compensation from the companies they cover. Regulations or market practice of some jurisdictions/markets prescribe certain disclosures to be made for certain actual, potential or perceived conflicts of interests relating to a research report as below. This research disclosure should be read in conjunction with the research disclaimer as set out at www.clsa.com/disclaimer.html and the applicable regulation of the concerned market where the analyst is stationed and hence subject to. This research disclosure is for your information only and does not constitute any recommendation, representation or warranty. Absence of a discloseable position should not be taken as endorsement on the validity or quality of the research report or recommendation.

To maintain the independence and integrity of CLSA’s

research, our Corporate Finance, Sales Trading and Research business lines are distinct from one another. This means that CLSA’s Research department is not part of and does not report to CLSA Corporate Finance (or “investment banking”) department or CLSA’s Sales and

Trading business. Accordingly, neither the Corporate Finance nor the Sales and Trading department supervises or controls the activities of CLSA’s research analysts. CLSA’s research analysts report to the management of the Research department, who in turn report to CLSA’s senior management.

CLSA has put in place a number of internal controls

designed to manage conflicts of interest that may arise as a result of CLSA engaging in Corporate Finance, Sales and Trading and Research activities. Some examples of these controls include: the use of information barriers and other information controls designed to ensure that confidential information is only shared on a “need to know” basis and in compliance with CLSA’s Chinese Wall policies and procedures; measures designed to ensure that interactions that may occur among CLSA’s Research personnel, Corporate Finance and Sales and Trading personnel, CLSA’s financial product issuers and CLSA’s research analysts do not compromise the integrity and independence of CLSA’s research.

Neither analysts nor their household

members/associates/may have a financial interest in, or be an officer, director or advisory board member of companies covered by the analyst unless disclosed herein. In circumstances where an analyst has a pre-existing holding in any securities under coverage, those




holdings are grandfathered and the analyst is prohibited from trading such securities.

Unless specified otherwise, CLSA/CLSA Americas/CA

Taiwan did not receive investment banking/non-investment banking income from, and did not manage/co-manage a public offering for, the listed company during the past 12 months, and it does not expect to receive investment banking compensation from the listed company within the coming three months. Unless mentioned otherwise, CLSA/CLSA Americas/CA Taiwan does not own a material discloseable position, and does not make a market, in the securities.

As analyst(s) of this report, I/we hereby certify that the views expressed in this research report accurately reflect my/our own personal views about the securities and/or the issuers and that no part of my/our compensation was, is, or will be directly or indirectly related to the specific recommendation or views contained in this report or to any investment banking relationship with the subject company covered in this report (for the past one year) or otherwise any other relationship with such company which leads to receipt of fees from the company except in ordinary course of business of the company. The analyst/s also state/s and confirm/s that he/she/they has/have not been placed under any undue influence, intervention or pressure by any person/s in compiling this research report. In addition, the analysts included herein attest that they were not in possession of any material, nonpublic information regarding the subject company at the time of publication of the report. Save from the disclosure below (if any), the analyst(s) is/are not aware of any material conflict of interest.

Key to CLSA/CLSA Americas/CA Taiwan investment rankings: BUY: Total stock return (including dividends) expected to exceed 20%; O-PF: Total expected return below 20% but exceeding market return; U-PF: Total expected return positive but below market return; SELL: Total return expected to be negative. For relative performance, we benchmark the 12-month total forecast return (including dividends) for the stock against the 12-month forecast return (including dividends) for the market on which the stock trades.

In the case of US stocks, the recommendation is

relative to the expected return for the S&P500 of 10%. Exceptions may be made depending upon prevailing market conditions. We define as “Double Baggers” stocks we expect to yield 100% or more (including dividends) within three years at the time the stocks are introduced to our “Double Bagger” list. "High Conviction" Ideas are not necessarily stocks with the most upside/downside, but those where the Research Head/Strategist believes there is the highest likelihood of positive/negative

returns. The list for each market is monitored weekly. Overall rating distribution for CLSA/CLSA Americas

only /CA Taiwan only Universe: Overall rating distribution : Buy / Outperform - CLSA:

63.23%; CLSA Americas only: 61.07%; CA Taiwan only: 73.97%, Underperform / Sell - CLSA: 36.77%; CLSA Americas only: 38.93%; CA Taiwan only: 26.03%, Restricted - CLSA: 0.00%; CLSA Americas only: 0.00%; CA Taiwan only: 0.00%. Data as of 31 March 2016.

Investment banking clients as a % of rating category:

Buy / Outperform - CLSA: 2.48%; CLSA Americas only: 0.00%; CA Taiwan only: 0.00%, Underperform / Sell - CLSA: 2.03%; CLSA Americas only: 0.00%; CA Taiwan only: 0.00%, Restricted - CLSA: 0.00%; CLSA Americas only: 0.00%; CA Taiwan only: 0.00% . Data for 12-month period ending 31 March 2016.

There are no numbers for Hold/Neutral as CLSA/CLSA

Americas/CA Taiwan do not have such investment rankings.

For a history of the recommendations and price

targets for companies mentioned in this report, as well as company specific disclosures, please write to: (a) CLSA Americas, Compliance Department, 1301 Avenue of the Americas, 15th Floor, New York, New York 10019-6022; (b) CLSA, Group Compliance, 18/F, One Pacific Place, 88 Queensway, Hong Kong and/or; (c) CA Taiwan Compliance (27/F, 95, Section 2 Dun Hua South Road, Taipei 10682, Taiwan, telephone (886) 2 2326 8188). © 2016 CLSA Limited, CLSA Americas, and/or CA Taiwan.

© 2016 CLSA Limited, CLSA Americas, LLC (“CLSA Americas”) and/or Credit Agricole Securities Taiwan Co., Ltd. (“CA Taiwan”)

This publication/communication is subject to and

incorporates the terms and conditions of use set out on the www.clsa.com/disclaimer.html. Neither the publication/communication nor any portion hereof may be reprinted, sold, resold, copied, reproduced, distributed, redistributed, published, republished, displayed, posted or transmitted in any form or media or by any means without the written consent of CLSA, CLSA Americas and/or CA Taiwan.

CLSA, CLSA Americas and CA Taiwan have produced

this publication/communication for private circulation to professional, institutional and/or wholesale clients only. This publication/communication may not be distributed or redistributed to retail investors. The information, opinions and estimates herein are not directed at, or intended for distribution to or use by, any person or entity in any




jurisdiction where doing so would be contrary to law or regulation or which would subject CLSA, CLSA Americas and/or CA Taiwan to any additional registration or licensing requirement within such jurisdiction.

The information and statistical data herein have been

obtained from sources we believe to be reliable. Such information has not been independently verified and we make no representation or warranty as to its accuracy, completeness or correctness. Any opinions or estimates herein reflect the judgment of CLSA, CLSA Americas and/or CA Taiwan at the date of this publication/communication and are subject to change at any time without notice. Where any part of the information, opinions or estimates contained herein reflects the views and opinions of a sales person or a non-analyst, such views and opinions may not correspond to the published view of CLSA, CLSA Americas and/or CA Taiwan. This is not a solicitation or any offer to buy or sell. This publication/communication is for information purposes only and does not constitute any recommendation, representation, warranty or guarantee of performance. Any price target given in the report may be projected from one or more valuation models and hence any price target may be subject to the inherent risk of the selected model as well as other external risk factors. This is not intended to provide professional, investment or any other type of advice or recommendation and does not take into account the particular investment objectives, financial situation or needs of individual recipients. Before acting on any information in this publication/communication, you should consider whether it is suitable for your particular circumstances and, if appropriate, seek professional advice, including tax advice. CLSA, CLSA Americas and/or CA Taiwan do/does not accept any responsibility and cannot be held liable for any person’s use of or reliance on the information and opinions contained herein.

To the extent permitted by applicable securities laws

and regulations, CLSA, CLSA Americas and/or CA Taiwan accept(s) no liability whatsoever for any direct or consequential loss arising from the use of this publication/communication or its contents. Where the publication does not contain ratings, the material should not be construed as research but is offered as factual commentary. It is not intended to, nor should it be used to, form an investment opinion about the non-rated companies.

Subject to any applicable laws and regulations at any

given time, CLSA, CLSA Americas, CA Taiwan, their respective affiliates or companies or individuals connected with CLSA/CLSA Americas/CA Taiwan may

have used the information contained herein before publication and may have positions in, may from time to time purchase or sell or have a material interest in any of the securities mentioned or related securities, or may currently or in future have or have had a business or financial relationship with, or may provide or have provided investment banking, capital markets and/or other services to, the entities referred to herein, their advisors and/or any other connected parties. As a result, investors should be aware that CLSA, CLSA Americas, CA Taiwan and/or their respective affiliates or companies or such individuals may have one or more conflicts of interest. Regulations or market practice of some jurisdictions/markets prescribe certain disclosures to be made for certain actual, potential or perceived conflicts of interests relating to research reports. Details of the disclosable interest can be found in certain reports as required by the relevant rules and regulation and the full details are available at http://www.clsa.com/member/research_disclosures/. Disclosures therein include the position of CLSA, CLSA Americas and CA Taiwan only. Unless specified otherwise, CLSA did not receive any compensation or other benefits from the subject company covered in this research report.

If investors have any difficulty accessing this website,

please contact [email protected] on +852 2600 8111. If you require disclosure information on previous dates, please contact [email protected].

This publication/communication is distributed for and

on behalf of CLSA Limited (for research compiled by non-US and non-Taiwan analyst(s)), CLSA Americas (for research compiled by US analyst(s)) and/or CA Taiwan (for research compiled by Taiwan analyst(s)) in Australia by CLSA Australia Pty Ltd; in Hong Kong by CLSA Limited; in India by CLSA India Private Limited (formerly CLSA India Limited) (Address: 8/F, Dalamal House, Nariman Point, Mumbai 400021. Tel No: +91-22- 66505050. Fax No: +91-22-22840271; CIN: U67120MH1994PLC083118; SEBI Registration No: INZ000001735); in Indonesia by PT CLSA Indonesia; in Japan by CLSA Securities Japan Co., Ltd; in Korea by CLSA Securities Korea Ltd; in Malaysia by CLSA Securities Malaysia Sdn Bhd; in the Philippines by CLSA Philippines Inc (a member of Philippine Stock Exchange and Securities Investors Protection Fund); in Thailand by CLSA Securities (Thailand) Limited; in Taiwan by CA Taiwan; in Singapore by CLSA Singapore Pte Ltd and in United Kingdom by CLSA (UK).

India: CLSA India Private Limited, incorporated in

November 1994 provides equity brokerage services (SEBI Registration No: INZ000001735), research




services (SEBI Registration No: INH000001113) and merchant banking services (SEBI Registration No.INM000010619) to global institutional investors, pension funds and corporates. CLSA and its associates may have debt holdings in the subject company. Further, CLSA and its associates, in the past 12 months, may have received compensation for non-investment banking securities and/or non-securities related services from the subject company. For further details of “associates” of CLSA India please contact [email protected].

United States of America: Where any section of the

research is compiled by US analyst(s), it is distributed by CLSA Americas. Where any section is compiled by non-US analyst(s), it is distributed into the United States by CLSA solely to persons who qualify as "Major US Institutional Investors" as defined in Rule 15a-6 under the Securities and Exchange Act of 1934 and who deal with CLSA Americas. However, the delivery of this research report to any person in the United States shall not be deemed a recommendation to effect any transactions in the securities discussed herein or an endorsement of any opinion expressed herein. Any recipient of this research in the United States wishing to effect a transaction in any security mentioned herein should do so by contacting CLSA Americas.

Canada: The delivery of this research report to any

person in Canada shall not be deemed a recommendation to effect any transactions in the securities discussed herein or an endorsement of any opinion expressed herein. Any recipient of this research in Canada wishing to effect a transaction in any security mentioned herein should do so by contacting CLSA Americas.

United Kingdom: In the United Kingdom, this research

is a marketing communication. It has not been prepared in accordance with the legal requirements designed to promote the independence of investment research, and is not subject to any prohibition on dealing ahead of the dissemination of investment research. The research is disseminated in the EU by CLSA (UK), which is authorized and regulated by the Financial Conduct Authority. This document is directed at persons having professional experience in matters relating to investments as defined in Article 19 of the FSMA 2000 (Financial Promotion) Order 2005. Any investment activity to which it relates is only available to such persons. If you do not have professional experience in matters relating to investments you should not rely on this document. Where the research material is compiled by the UK analyst(s), it is produced and disseminated by CLSA (UK). For the purposes of the Financial Conduct Rules this research is prepared and intended as substantive

research material. Singapore: In Singapore, research is issued and/or

distributed by CLSA Singapore Pte Ltd (Company Registration No.: 198703750W), a Capital Markets Services license holder to deal in securities and an exempt financial adviser, solely to persons who qualify as institutional investor, accredited investor or expert investor, as defined in Section 4A(1) of the Securities and Futures Act (Cap 289). Pursuant to Regulations 33, 34, 35 and 36 of the Financial Advisers (Amendment) Regulations 2005 of the Financial Advisers Act (Cap 110) with regards to an accredited investor, institutional investor, expert investor or overseas investor, Sections 25, 27 and 36 of the Financial Adviser Act (Cap 110) shall not apply to CLSA Singapore Pte Ltd. Please contact CLSA Singapore Pte Ltd (telephone No.: +65 6416 7888) in connection with queries on the report. [MCI (P) 013/11/2015]

The analysts/contributors to this

publication/communication may be employed by any relevant CLSA entity, CA Taiwan or a subsidiary of CITIC Securities Company Limited which is different from the entity that distributes the publication/communication in the respective jurisdictions.

MSCI-sourced information is the exclusive property of

Morgan Stanley Capital International Inc (MSCI). Without prior written permission of MSCI, this information and any other MSCI intellectual property may not be reproduced, redisseminated or used to create any financial products, including any indices. This information is provided on an "as is" basis. The user assumes the entire risk of any use made of this information. MSCI, its affiliates and any third party involved in, or related to, computing or compiling the information hereby expressly disclaim all warranties of originality, accuracy, completeness, merchantability or fitness for a particular purpose with respect to any of this information. Without limiting any of the foregoing, in no event shall MSCI, any of its affiliates or any third party involved in, or related to computing or compiling the information have any liability for any damages of any kind. MSCI, Morgan Stanley Capital International and the MSCI indexes are service marks of MSCI and its affiliates. The Global Industry Classification Standard (GICS) was developed by and is the exclusive property of MSCI and Standard & Poor's. GICS is a service mark of MSCI and S&P and has been licensed for use by CLSA.

EVA® is a registered trademark of Stern, Stewart &

Co. Unless otherwise noted in the source, "CL" in charts and tables stands for CLSA/CLSA Americas estimates and “CT” stands for CA Taiwan estimates.
