9

Current Asset Allocation

Ass~

~ I CaP

I large I Sma!

~ ~

FlXeCI ... _._

US Credit Fixed Higl hYleld~ Ban Ik~

Lc:g . Debt

TIPS

Real

RNI

~ ~

Total

FRESNO COUNTY EMPLOYEES' RETIREMENT ASSOCIATION FCERA i1rVe3tment Policy Statement (IPS)

APPENDIX A

STRATEGIC ASSET ALLOCATION

1~ " 7% 27% I 3000 Index 14~ ~ 3'l1 1" .,. 29% IISCI ACWI ex US Index 9% 3% 7%

31% 21% 41% IndeX-'

' ....... _w

7% 5%

4%

11% 3% 0% 6% E 'Index ~ 0% 10% 'OOCElndu

.3% JI% ~. 'DDCl:lndu

~ " 3% HFRlFoF Iinclex 0% I 3000 + 2110 bps 3% 1+250 bps

100%

Notes: 1) The cash requirements of the Fund will be classified under Fixed Income. 2) Asset allocation targets. and any rebalancing transactions. should give first priority to maintenance of broad asset group targets. and then sub-class targets. Target ranges are set broadly for flexibility when needed. Rebalancing actions will nonmally occur before portfolio weights are displaced to those levels.

-8 -

Asset Allocation .'; .....

What is Asset Allocation? .. ;: :.:. : ',.:: .. ;, .:

'. "Asset aliocationN is the analysis that investors apply in deciding how to distribute mvestm.e.nJ:s~mongNarjous ~asses,of jnvestmentNehicles ,(e,g" , : stocks"bQ{lds{tomm()ditie~;,reatestate;·,alternatiVesre~) •. · . .": i·.,,, . ",;, .:' ;:

:":.;.. : ;- :., . <- ):.:; ' .. :; '.( '.\'-I~ ; . -. ,:. 'k"; :'! ; ;"::;t:;; (: ':~ ··· ···: :> :: ~; ·i :·'. :i ·:·:; ~ : -·~;,, · !

Thechoic~,oN~sset;class;Weight,ngS(ha$-;a ,'majorjm,pact,on~retitrris··in,times when the different.assetscla$ses.haile significantly different results. , .....

Assetallocatlon;isbased:o!l ·~he idea thatdjffe~ent.assetdaS5es :wlll .· · ... , .. !

perform· differentiyJn. different.periods and thafitis,impossible to1<!entify .. thebest ·aJ1d~worSU!s$etclasseSireliabIY'<itrea·d::of ,til'Yle:·O)Vi!rsifyingassets c" a mOhg :severa las~eN:la~ses :accordi rrg~o iai:onsistent a lIoecition. pia n:helps '. protect. investors:to ,al1oid,disaStrousperformanci! through over~'; ' .... ' .. ;,.; ,.':: concentra.tion"Diversifieatioti,has ,b~en'desttibedca$~." the ;only,free htnth ~.; you will find.iO<the'jn~tmerit!game,~ ,", ;;:,:; '.j ' : . .• . , . ,': ., . ...

long term (strategic) asset allocation involves setting asset class targets, and re-balancing to those targets periodically.

Short term "tactical" allocation - sometimes referred to as market timing­attempts to move allocation to asset classes up or down depending on predictions of short term asset class returns. While short term allocation has its proponents, attempts to time the markets short term can easily lead to worse results over a given period than the results achieved by Investors who consistently adhere to their predetermined asset allocation plan over the same period.

Asset Allocation

Long-Term Asset Allocation

The following discussion involves long term asset allocation.

long-term aSsetia11ocatioJiistypitally practiced 1:Ising,thequantitatiiletbols of "mean variance optimization'!,whichis an capplicatinn ,of ~Modern Portfolio Theory" or MPT. Those who consistently practice strategic asset allocationbelievethatby'carefuIlY'l:hoosinga;portfolio ·of.di·fferent assets, an investormay,beiabletomaximiz.e retutnwhile l1iinimizingii~k -, .

Because different asset classes'teturn'S over any given investment period' are notperfectlycorrelatedwith'eachother.-diver'Slfyingassetsamong differenusset classes' shouldi'helpreduce-theoverall riskin,the'-portfolio:' _, (expressed as the variability,orvolatilitY,'Of returns) for- a,givenJevel of. expected overall returni< Having a-mixture of different asset-classes- in-a - -, portfolio may help investor5'r'neetreturns goals-while also keeping portfolio risk within the parameters of their investment policy guidelines,

.,":" .... '.

. .,.

-'

"' . ..

\ , ,.

\

Asset Allocation

Long-Term Asset Allocation .... . ', { ..

The "Markowitz mean-varia~ce optimization model" Is the technical name for this-as~etaUbcatiOi'i apprO'act\;'1t ii1v6h;~smak'jng a~sllmpti6ns alidtif ; expected total return for asset classes, risk (standard deviation ar6uindth~ expected return) of each asset class, and correlations between the various asset' classes '(I\bw rt.'litt(the-Y'rriiSlktagetli'~r); Tne' ()litto.neof1l1emod~1 is a serlef()fpris~tfleassefilrijXalojdfs;eaffi '6fwhichha~'1h'e hlttlist ': expect€d+etuj:;ffdl:iits 'liibjeci~'dHsIC(Th'ls 'se-Hes:iof#osslble ;i'riixeSis : ' ' ,, ', knowrtib 'th'e""efflcleni'fi'o'iltief':)i "" ,, ",";.,,:' i',: ,,' ; i ;< :;:, :" ,',

' •..• , )., .• , !' ':- >", r ,"'0 ~f .::·:.:.:-~: ~:· lJ':.<'~t: ... :::";-:": .. ;: :·':;·. ~ i: : .~·.': .:.: .. ~.~ .<:.r~;: .... --.-

There' 'dlii'6e'iid guaf~rifee1h~f pasffellitibnsl1lps' w1ii 'CbntinJeln'thi! i;, ""

future; t'lierefOi"e; ith~' ab()v~'~spect '6f'the hl6defis'geiierally t:bnsld~recf one of the "wearcljn~n intfaCiitiorlat ;i$sefali(j'catibi1'$1r\ltegi~ 'Ui'iithave " been derived from MPT.

: ·.f ; : ~: . ' . . : ,:'.

AnothEiFIssOewith'tM 'mdd~1 'is·thiit5ee'mitigr'y hiln()i ~rr()rsihforecaSting ' may lead tdfec6mrtu:!riif~d alioclitiOn~ 'thllfare rrnpractlcal'ari(/Ynii'yvioiate "common' seris~ii : Tili~ 'may r~$.:Iit irta d!rtain'iili'ioui'it oftWliakihg'of ' , ', .. ' assumptionstb dbtalii reasona'ble results: ' ' " ".'; <'

". ;"i .. . ", ,' : . F, " , :'.; '. ",'

. ••.• • ". <: ~

, . , . . ,

. , .

. : .,.'.:',

.- .. , '. . .. ', .:~ -" , .

Asset Allocation

How Important is Asset Allocation?

Academic ~~~earch has pair,s~kingIY examined t~~'ir'JPQrt~~~~':(ia~set allocation.. .. .., .. .. .. .' . ..... , .. , :. ;: ::"

In 1986;8;i~son,H'~~d:~nd~~eb~w~r{~~ B}~Ublis~~d)~st4dy~tioJt ~h~ asset~llocation of9iJargepE!nsionf~~dsmeasi.Jr.ed .fr~ma.974 t~ 1983 . .. The authorsreplilted .the~~~si.on:fu~ds'~t~ck,.~pnd,and cas~~election~ '. with cOrresp6ndingmarket indexes .. Theind~xedqu~rteily~etur,ris were ' .. found to be higher than the pension plans' aggreg~tea~al 'quarterly" . returns. The two q.uarterly.return series' linear. correlation wa.s,Ineas!lred at 96.7%,.withshar~ va~j<lQce .of93;6%.·A199itPilow~upstuciyby .. ... . Brinson; Singer,'~ndBe;;~~~erme~sl!n!d 'li. vati~~~e of~l.~';£: .·, . . .. .., ..

: .,' .. '~ " ; ' 'l:' : ' ;:." .. " '.;-

The key lesson of these studies taken together was that, for the period studied, the choice of w\:lich asset cI~~ses to invest in had a bigger impact on thevo!atilitYofthe!IJllds iilan ~id~he chc;>ice ofanypariicl!lar . ' . '.' investment mariager.Somealso r,ead the studiesi~'~ea~ that .more time should be~pent on a~et~I~c:atio'n comp~r~d to ~~~kingacilve ' . " .. management (and that replacing actiiie in~estnient~ariagement choices with index funds might make sense.)

Later papers pointed out that It wasn't necessarily the effect on volatility that investors cared about, but rather the effect on returns. While asset allocation may drive the market sensitivity of returns, excess returns (plus or minus) delivered by investment managers still have a major Impact on total returns achieved. In 2000, Ibbotson and Kaplan used 5 asset classes in their study "Does Asset Allocation Policy Explain 40, 90, or 100 Percent of Performance?" Their conclusion was that Asset Allocation determines 90% of risk but 40% of the difference In returns.

.,

' .. ,

/

. /

Assei: Allocation

PredlctablRty of Manager Out-and,Under-Perf~nce based.on:Histl:nlc . Results

. '. '.' ".;,;:: 0:';. '. "; " . , 'i.:' " ; '. ,"' :""-., .~

McG~!iiaWd~~i~rb~,d :~'~ , ~xajhiri~ti~h cij'6i~d~ 't~~tw~~~:Wi' the' .t~ph!!a~Jle " of p~.1orir;aiiC:e' ci~ri~'g 1983' tci'1993: ' j)liii~gth~':se~hl;in~as~~ein~~t ' ' .

. ' . ..',: , .. --:,.,,, : ::~ ,/ ., .. : _ :.:.: ":;0.-_ :;:".! ; . ,'1.

period of 1993 to 2003, only 28.57% of the funds remained in the top . quartile. 33.33% of the funds dropped to the second quartile. The rest of the funds dropped to the third or fourth quartile. In fact, data suggested that low cost was a more reliable indicator of performance .

Asset Allocation

Using Fees to Predlct:Manager:Performance ,- ~: .- . ~ ;: : .".

Jack Bogle of Vanguard noted that an examination of 5 year performance data of large-cap blend funds revealed that funds in the lowest quartile of cost had the best ji'erf6rmaiice; and' thadJnds in the 'highest quartile of . cost had the:woi'St ~ifOtmiinc~> :; . . . " . ;,

\

I

.. Asset Allocation

References

J\ Gary P. Brinson, L. Randolph Hood, and Gilbert L. Beebower, Determinants of Portfolio Performance, The Financial Analysts Journal, July/August 1986. J\ Gary P. Brinson, Brian D. Singer, and Gilbert L. Beebower, Determinants of Portfolio Performance II: An Update, The Financial Analysts Journal, 47, 3 (1991) J\ William Jahnke, The Asset Allocation Hoax, Journal of Financial Planning, February 1997 J\ Roger G. Ibbotson and Paul D. Kaplan, Does Asset Allocation Policy Explain 40%,90%, or 100% of Performance?, The Financial Analysts Journal, January/February 2000 J\ Thomas P. McGuigan, The Difficulty of Selecting Superior Mutual Fund Performance, Journal of Financial Planning, February 2006 J\ James Dean Brown, The coefficient of determination, Shiken: JALT Testing

& Evaluation SIG Newsletter, Volume 7, No.1, March 2003 J\ Meir Statman, The 93.6% Question of Financial Advisors, Journal of Investing, Spring 2000 1\ L. Randolph Hood, Response to letter to the Editor, The Financial Analysts Journal 62/1, January/February 2006 1\ L Randolph Hood, Determinants of Portfolio Performance - 20 Years Later, The Financial Analysts Journal 61/5 September/October 2005 J\ Bekkers Niels, Doeswijk Ronald Q. and Lam Trevin W.,

[http://papers.ssrn.com/soI3/papers.cfm ?abstracUd=1368689 Strategic Asset Allocation: Determining the Optimal Portfolio with Ten Asset Classesl, Working Paper Series, March 2009>

Vl

w I-

< U

0 Vl

Vl

< ,... ,...

0 ,...

0" 0'" ,...~

~o Vl N - E N M

I- '" - 0 c8~ ~

a:: ~ Lrl t w l!) ::I

~ ...J Z ~ ~ « .

~ « V) ~ wO V)...J>

At the October 16, 2013 Retreat, the Board gave further consideration to the asset liability study, deciding to pursue a more risk-diversified portfolio with a dedicated tail risk-strategy.

The objective today is to review specific portfolio mixes in order to arrive at an allocation that meets the following objectives:

Reduces the portion of total risk coming from equities (i.e. more risk-diversified) Achieves a similar expected return as the mixes that were reviewed in the asset-liability study It is critical that the Trustees have a firm understanding and high level of comfort with whatever mix is ultimately decided.

The mixes under consideration move progressively further away from the current mix and closer to the risk-diversified portfolio that was presented previously.

It will be up to the Trustees to decide how much "peer risk" they are willing to take on, the primary trade-off being a better reduction in equity risk.

What follows is several different analyses aimed at highlighting the relative differences amongst the mixes through various lenses.

I I WURTS (lfJ ASSO C I A TE S

24 17 14 10 4.5 63 II/Mid Cap US Equity 5 4 3 3 6.9

Total Domestic Equity 29 21 17 13 4.5

International Large 15 11 9 7 4.5 8.0 International Small 4 3 3 3 8 ,3

Emerging Markets 5 7 7 8 15 96 Totollnt'l Equity 24 21 19 18 19.5

US Core Fixed Income 19 2.0 US Credit Fixed Income 8 5 30 High Yield Fixed Income 4 5 5 5 4.9 Bank Loans 4 5 6 5 41

obal Sovereign I 5 7 8 10 2.2 I

Emerging Market Debt Local 3 I 4 5 8 10 57 4 ! I 3 4 4 5 2.2

Total Fixed Income 26 n 28 31 31 35

Commodities 4 Ii 3 3 3 5 4.3

Real Estate 6 I ' 5 5 5 5 5.6

Infrastructure ! 3 3 3 56

Total Real Assets 10 .-.. I l 11 11 11 10

Liquid Alts/HFoF 4 j - r 6 8 11 15 5.4 Private Equity!VC 7 6 6 6 6 9.9 Private Credit - Meuanine

I 4 5 5 5 5.9

Private Credit - Distressed 3 3 5 5 6.0 11 19 22 27

Total Allocation 100 I 100 100 100 100

2 1 wU RTs (I1J ASSO CI A TE S

Total Equity Allocation 60 ,-.,<'",,----------------.

50

40

30

20

10

o Current Portfolio

MlxA MixB Mix C Risk Average Diversified SACRS'

Total Domestic Equity Allocation

35 ,--------------------------------------29 28 30 _ _

25

20

15

10 •

5

o Current Mix A Portfolio

MixB Mix C Risk Average Diversified 5ACRS'

* SACRS data obtained from RV Kuhns Public Fund Universe Analysis. as of 6/3011 2.

The equity allocation is reduced under all scenarios.

Most of that reduction is coming from domestic equities.

Total International Equity Allocation

30 ~

25 I 24

20 - 21 1Q

15

10

5

o Current Portfolio

Mix A MixB Mix C Risk Average Diversified SACRS'

31 WU RT S O ASSOC I A T E S

Total Fixed Income Allocation

40 35

35

30

25

20

15

10

5

0 Current Mix A MlxB Mix C Risk Average Portfolio Diversified SACRS'

Composition of Fixed Income Allocation

40 ,------------------------

35 1 J I 30 n 2S L 20

15 +--10 +--

5 +--­o +---

Current Portfolio Mix A MixB MixC

While the total allocation to fixed income is similar to the median SACRS allocations, the composition is very different.

Median SACRS fixed income allocations are 1.5% global, 1.8% International, and 23.98% U.S. fixed, of which a large portion is likely core bond strategies.

TIPS

Emerging Market Debt Local I

• Global Sovereign

Bank Loans

• High Yield Fixed Income

US Credit Fixed Income

Risk Diversified III US Core Fixed income

41 WU RTS ASSO C IA TE S

Real Assets 12

10

8

6

4

2

0 Current Mix A MixB Mix C Risk Portfolio Diversif ied

Alternatives Allocations 35 _

30 -l-----------r,t 25 - I 20 D I 15 I 10

5

o Current Mix A Portfolio

MixB Mix C Risk Diversified

11:1 Real Estate

• Commodities

Infrastructure

Liquid Alts/HFoF

• Private EquityNC

• Private Credit -Dist ressed

Private Credit -Mezzanine

Within the real assets allocation, Mix A-C introduce a 3% allocation to infrastructure.

The allocation to alternatives increases significantly.

As part of the PE allocation, the current portfolio has -2.5% in private credit strategies.

5 I W U RTS O ASSO C I ATES

100

90 80

70

60

50

40

30

20

10

Liquidity Analysis Annual Operating Cash Flows as a % of Assets

Current Portfolio

Mix A

\I Very Liquid

Very liquid

Semi liquid

Not liquid

MixB MixC Risk Diversified

Semi-Liquid • Not Liquid

10% 8% 6%

4% -~--------------------------------------

2% ·r~-------------------------------------0% .............. <' ----.

-2% 201-2 .'2.....: 2032 2042 20S2~---, -4% ""-------.,;;

::~ +1----------------------10% ~1==========================================9

Net cash flows (employer/employee contributions less benefit payments) were forecasted to remain positive until 2020, after which the annual outflows never exceed 5%.

Assets that trade on highly liquid public exchanges. Domestic, international, EM stocks, domestic and global bonds, commodit ies

Assets that could likely be fully liquidated between several weeks and one year from request.

Assets that generally have mUltiple-year lock-up periods

Bank loans, emerging markets debt, liquid a Iternatives/HFoF

Private equity, mezzanine & distressed debt, infrastructure, real estate

"The above classifications are broad generalizations. Liquidity Is driven by market conditions (which can fluctuate significantly) and investment vehicles used. Certain asset classes classified above as not liquid can be accessed through open-ended fund structures whereby assets can generally be liquidated within one year.

61 WURTS ASSO C I A TE S

Volatility (as measured by standard deviation or the range of

Mea!'! Variance Optimizer Analysis scenario forecasts) decreases as

Forecast 10 Year Return 6.5 6.6 6.5 6.5 6.5 you progress from the current

Standard Deviation 11.0 I I 10.3 9.5 9.3 8.9 portfolio to risk-diversified portfolio.

95% cVAR 27.6 25,6 23.8 23.2 124

Sharpe Ratio 0.44 0.48 0.50 0.52 0.53 1"'- There are multiple ways to

Economic Scenario Anal}<sis -. measure tail risk. One of those is "conditional value at risk" - this

10 Year Return Forecast measure takes the weighted

Stagflation 4.0 4.1 4.1 4.2 4.3 average loss that can be expected Weak Economy 3.3 3.5 3.5 3.5 3.6 under a tail event (defined here as Average Economy 5.2 5.2 5.0 5.0 5.0 the worst 5% of all possible Strong 7.3 7.1 6.8 6.8 6.5 outcomes).

of Scenano Forecast 4.0 36 3.4 3.2 29

Shock (/ year) -34.5 -349 -334 -330 -303 This will be the starting point for what the tail-risk hedging manager attempts to mitigate.

71 WU RTSO ASSO C I AT E S

15

10 +1---

5

o

~ -5 ~

c .. ~ ~ -10

-15

-20

-25

-30 I -27.6

11.0

Current Portfolio

10.3

-25.6

Mix A

. 95%cVAR

9.5 9.3 8.9--

-22.4

MixB MixC Risk Diversified

Standard Deviation

Volatility

Downside Tail Risk

81 WURTS O A S SO C IA TES

120%

100%

80% - - - , • Currency ,

60% • Inflation

40% t---. • • • Rates

Credit

20%

• Equity - - - - -0%

-20% Current Mix A MixB MixC Risk Diversified

91 WU RT S O ASSOCIAT E S

-40% -30% -20% -10% 0% 10"..6 20%

i

1 • Current

Mix A

MixB

. MixC

• Risk Diversified

30%

1972 - 1974 Oil Crisis (Dec. to Sep.)

1987 Market Crash (Oct. 14 to Oct. 19)

1989 - 1990 Nikkei Stock Price Correction

1992 - 1993 European Currency Crisis

1994 US Rate Hike

1997 - 1999 Oil Price Decline

1998 Russian Financial Crisis

2001 Dot-com Slowdown

2007 - 2009 Subprime Mortgage Meltdown

2009 July - January

40%

10 I W U R T 5 I) ASS 0 C IA T E S

-14%

• Current

Mix A

MixB

. MixC

• Risk Diversified

-12% -10% -8% -6%

Global Equities -20%

GloballR +200bps

Global Credit Spreads +100bps

USD+20%

Commodity -20%

-4% -2% 0%

I I I WURTS O ASSO CI ATES

Act5V2 Risk 1!S Fres,'l l'.) Po lic.y 3.0

2.S 2.5

2,Q

1.5

0.9 1.0 1.0

O.S

0.0 Current MrxA M;x B Mi~C 'Risk Diversified

Active Risk vs Average SACRS 3.0

2.S +1----

2.0 +------..-.-- ---'-'';-;;-1.6 " "

1.S

1.0

O.S

0.0 Current Mix A

1.5

I

MixB MlxC

2.5

Risk Diversified

Active Risk: (annualized tracking error) a measlire of holY cfose/y each portfolio tracks the berchmcrk (defined above as either t.he current Fresno Policy or tl,e Median SACRS allocator;.

,

I 10.

107 0.6

0.5

10.1\

'

10.3

O. I

10.1 I iO.O

I.

Beta

V·PD -----0. o. ~---'-61 0.60

Current MlxA MhcB MixC Risk Diversified

Equity Beta: a measure describing the sensitivity of portfolio returns with returns of the equity market.

~~~----~-------------------

Effective Duration (Portfolio) 2.0 .,---- ... -- 1.8

il.S ~ .,

i I I il.0

I 10.5

10.0 Current

1.6 1.7

Mix A MixB MI,C Risk Diversified

Effective Duration: a measure describing the percent change in price for a given change in yields.

12 1 WURTS O ASSO C I A TES