Valuing Dot Coms

7/31/2019 Valuing Dot Coms

1/11


2/11

high-growth, high-uncertainty, high-loss firmshas been a challenge, ro say the least; some prac-

titioners have even described it as a hopeless one.In this article, we respond to that challenge by using a classic

discounted-cash-flow (DCF) approach to valuation, buttressed bymicroeconomic analysis and probability-weighted scenarios. A lthough


3/11

THE McKINSEY QUARTERLY 2000 NUMBER 1

years, partly because of the high marketing costs (aimed at attracting cus-tomers) that they must write off against current earnings. Second, thesecompanies are growing at very high rates: successful ones will increase theirrevenues by 100 times or more in the early going. Finally, the fate of thesecompanies is quite uncertain.Shorthand valuation approaches, including price-to-earnings and revenuemultiples, are meaningless when there are no earnings and revenues are

growing astronomically. Some ana-lysts have suggested benchmarkssuch as multiples of customers oru rd if t h e r e a r e n o e a r n i n g s multiples of revenues three yearsout. These approaches are funda-mentally flawed: speculating abouta future that is only three or even

five years away just isn't very useful w he n h igh gro w th will co nti nu e for anadditional ten years. More important, these shorthand methods can'taccount for the uniqueness of each company.The best way of valuing Internet companies is to return to economic funda-mentals with the DCF approach, which makes the distinction betweenexpensed and capitalized investment, for example, unimportant becauseaccounting treatments don't affect cash flows. The absence of meaningfulhistorical data and positive earnings to serve as the basis for price-to-earningsmultiples also doesn't matter, because the DCF approach, by relying solelyon forecasts of performance, can easily capture the worth of value-creatingbusinesses that lose money for their first few years. The DCF approach can'teliminate the need to make difficult forecasts, but it does address the prob-lems of ultrahigh growth rates and uncertaint}^ in a coherent way.In this discussion, we assume that the reader has a basic knowledge of the


4/11

VALUING DOT-CO MS 151

$25 billion as of mid-November 1999. Yet Amazon has never turned a profitand is expected to lose at least $300 million for the year, so it has becomethe focus of a debate about whether Internet stocks are greatly overvalued.

Start from the futu reIn forecasting the performance of high-growth companies like Amazon,don't be constrained by current performance. Instead of starting from thepresentthe usual practice in DCF valuationsstart by thinking aboutwhat the industry and the company could look like when they evolve fromtoday's very high-growth, unstable condition to a sustainable, moderate-growth state in the future; and then extrapolate back to current perfor-mance. The future growth state should be defined by metrics such as theultimate penetration rate, average revenue per customer, and sustainablegross margins. Just as important as the characteristics of the industry andcompany in this future state is the point when it actually begins. SinceInternet-related companies are new, more stable economics probably lieat least 10 to 15 years in the future.But consider what Amazon has already achieved. Its ability to enter anddominate categories is unprecedented, both in the off- and the on-lineworlds. In 1998, for example, it took the company only a bit more thanthree mon ths to banish C DN OW to second place amon g on-line purveyorsof music. In early 1999, Amazon assumed the leadership among on-line


5/11

THE McKINSEY QUARTERLY 2000 NUMBER 1

What operating profit margin could Amazon.com earn on that $60 billion?The superior market share of the company is likely to give it significant pur-chasing power. Remember too that Amazon will earn revenues and incur fewassociated costs from other retailers using its site. In this optimistic scenario,Amazon, with an average operating margin in the area of 11 percent, wouldmost likely do a bit better than most other retailers.

And what about capital? In theis not to det ine precisely optimistic scenario, Amazon maywill h a p p e n to Internet well need less working capital andbut to offer a r igorous fewer fixed assets than traditionalo t wha t could h a p p e n retailers do. in almost any scenario,

it should need less inventory becauseit can consolidate its stock-in-trade

in a few warehouses , and it won't need retail stores at all. We assume thatAmazon's 2010 capital turnover (revenues divided by the sum of workingcapital and fixed assets) will be 3.4, compared with 2.5 for typical retailers.Combining these assumptions gives us the following financial forecast for2010: revenues, $60 billion; operating profit, $7 billion; total capital, $18 bil-lion. We also assume that Amazon will continue to grow by about 12 percenta year for the next 15 years after 2010 and that its growth will decline to 5.5percent a year in perpetuity after 2025, slightly exceeding the nominal growthrate of the gross domestic product.^ To estimate Amazon's current value, wediscount the projected free cash flows back to the present. Their present value,including the estimated value of cash flows beyond 2025, is $37 billion.H ow can we credibly forecast ten or more years of cash flows for a companylike Am azon? We can't. But our goal is not to define precisely what willhappen but instead to offer a rigorous description of what could.


6/11

153VALUING DOT-COMS

oroff-line) based in the United States.it is

an on-line operation. It captures much higher operating marginsit is the on-line retailer of choice; even if its prices are comparable to

of other on-line retailers, it has more purchasing clout and lower oper-in the

of 1999.io B has Am azon captu ring revenues almost as large as it does in

for capital fall in the range betweenof the first scenario and the margins and capital requirements of a

a valueof the fourth quarter of 1999.

a large retailer inScenario C, though not as largeit does inScenario B, and the company's economics are closer to those

a value for Amazon of

inScenario D, Amazon becomes a fair-sized retailer with traditionalof the busi-in each field. Competition transfers most of the

ofgoing on-line to consumers. This scenario implies that Amazon was

inwhich the company's value ranges fromto $79 billion. Although the spread is quite large, each scenario

1


7/11

4 THE McKINSEY QUARTERLY 2000 NUMBER 1

I B I T 2com: Expected value

Discounted-cash-flowvalue, $ billion7937153

3

Probability,percent5

353525

.com: Volatility of expected

Expectedvalue, $ billion3.9

13.05.30.8

$23.0 billion

values

Low probability Base probability High probabilityof outcome of outcome of outcomeABCD

0

125I35I40

35

16 23 32Discounted-cash-flow value, $ billion

is plausible.^ Now comes the criticalphase of assigning probabilities andgenerating the resulting values forAmazon (Exhibit 2). We assigna low probability, 5 percent, toScenario A, for though the companymight achieve outrageously highreturns, competition is likely toprevent this. Amazon's currentlead over its competitors suggeststhat Scenario D too is improbable.Scenarios B and C both assum ingattractive growth rates and reason-able returnsare therefore the mostlikely ones.When we weight the value of eachscenario, depending on its proba-bility, and add all four of thesevalues, we end up with $23 billion,which happened to be the com-pany's market value on October 31,1999. It therefore appears thatAmazon's market valuation can besupported by plausible forecasts andprobabilit ies.

Now, however, look at the sensitivity of this valuation to changing probabili-ties. As Exhibit 3 shows, relatively small variations lead to big swings invalue. Indeed, the volatility of the share prices of companies like Amazon


8/11

VALUING DOT-COMS 155

T he average revenue per cus tom er per year from purcha ses by its cus-tomers, as well as revenues from advertisements on its site and fromretailers that rent space on it to sell their own products

Th e total num ber of custom ers Th e con tribution m argin per custom er (before the cost of acquiringcustomers) The average cost of acquiring a custome r Th e custom er churn rate (that is, the pro po rtion of custom ers lost

each year)Let us see how Amazon could achieve the financial performance predictedby Scenario B and com pare this with the company's curren t performance.As Exhibit 4 shows, the biggest changes over the next ten years involve thenumber of Amazon's customers andthe average revenue for each. In EXH i B iT 4Scenario B, Am azon's cus tom er baseincreases from 9 million a year in1999 to about 120 million worldwideby 201084 million in the UnitedStates and 36 million outside it. Weassume that Am azon will remain thenumber-one US on-line retailer andachieve an attractive position abroad.Scenario B also calls for Amazon's average revenue per customer to rise to$500 by 2010, from $140 in 1999. That $500 could be accounted for bytwo CDs at $15 each, three books at $20 each, two bottles of perfume at

Amazon.com: Customer economics, Scenario B

Average revenue per customer, $Customers, millionContribution margin, percentAcquisition cost per customer, $Customer churn rate, percent

1999140

9142925

2010500120

145025


9/11

56 THE McKINSEY QUARTERLY 2000 NUMBER 1

can achieve brand dominance and advertising economies of scale. The costof acquiring new customers is closely linked to the customer churn rate,which at 25 percent suggests that once Amazon acquires customers it willkeep them four years. This implies a truly world-class (or addictive) cus-tomer offer and a deeply loyal (or lazy) customer base.

I B I T 5

percent

percente s d i s c o u n t r a te o f 1 2 % i n Y e a r 2 .

exampleLoyalty.com

250157520

Turnover.com'342159346

Looking at customer economicsin this way m akes it possible togenerate the kind of information thatis needed to assess the probabilitiesassigned to various scenarios.Consider how two hypotheticalyoung companies, Loyalty.comand Turnover.com, with differentcustomer economics might evolveover time (Exhibit 5). Each had

$100 million in revenues in 1999 and an operating loss of $3 million.On traditional financial statements, the two companies look very muchthe same. Deeper analysis, however, using the customer economics model,reveals striking differences.The lifetime value of a typical Loyalty.com customer is $50 over an averageof five years; the typical Turnover.com custo m er is wo rth - $ 1 over tw o years.The difference in the value of a customer reflects the churn rate (20 percentattrition each year for Loyalty.com versus 46 percent for Turnover.com) andTurnover.com's higher acquisition costs.E X H I B I T 6Long-term performance: An example$ m i l l i o n


10/11

VALUING DOT-COMS 157

rev-per customer than Loyalty.com does and hassus-

Loyalty.com will find it much easier to grow because it doesn't

tax (EBIT) will turn

certainty is here to stayy using the ad apted DCE approach outlined here, we can generate reasonable

C plays out, the investor will earn about 7 percent


11/11

Documents

Valuing Dot Coms