Anda di halaman 1dari 18

125HighStreet OliverStreetTower Suite802 Boston,Massachusetts02110 617.742.

4226

HedgeFundRegional PerformanceStudy
AStatisticalReviewofHistoricalPerformance ByU.S.City

ExecutiveSummary ThispaperrepresentsBalterCapitalManagementsinauguralregionalstudyofhedgefundmanagerreturns. Wearenotawareofanysimilarstudiesthathavebeendonetodateandweexpectthattherewillbeanumber ofitemsthatwillneedtobeupdatedandimproveduponaswemoveforwardinpublishingthisreviewonan annualbasis.Wehavelongbelievedtherearebroadbasedregionaldifferentialsinhedgefundmanager returns.Thebasisofthisthinkingcamefromthefollowingfactors: 1. Managersinthesamelocaleoftenseemtotrafficinasimilarsetofnames/securities/themes. 2. Thepresenceofsocialevents,ideadinnersandfamiliarityseemtoleadtofurtherconcentrationofideas. 3. Managerfeestructurestendtodifferacrosslocales. 4. Therelationshipandtoleranceforriskmaydifferacrossgeographicalregions. 5. Hedgefundstrategysubgroups(e.g.,long/shorttechnology)mayhaveheavierconcentrationinsome areasversusothers. Methodology Theobjectiveofthisresearchpaperistoidentifyandinvestigatewhetherthereweredifferencesinhedgefund returnsbasedontheirgeographicallocationwithintheUnitedStates.Forourfirststudy,weidentifiedseven cities;thecriterionfortheinclusionofacitywasourabilitytoidentifyatleast30hedgefundfirmswithvalid longtermdatasetsforthe12yearperiodfromJanuary2000throughDecember2011.Ourfindingsindicate thatmanymajorhedgefundindexprovidersofferindiceswithfewerthan30managers,includingDowJones CreditSuisseandHFR.ThecitiesweidentifiedforthisfirststudywereBoston,Chicago,Dallas,Greenwich,Los Angeles,NewYorkandSanFrancisco.GoingforwardwewouldlooktoincludeotherareassuchasDenver, Philadelphia,MiamiandMinneapolis,alongwiththepotentialtoaddinternationallocationssuchasLondon, HongKong,andToronto.Themostimportantfactorindeliveringcredibleconclusionsistheneedtoconfirm setsofdatathathavesufficientdepthandlongevity. Weusedourproprietaryinternaldatabaseforthedatasetwiththefollowingevaluationmethodology: 1. Onlyonesetofreturnswasincludedfromeachhedgefundfirm,asopposedtomultipleproducts. 2. Weselectedthefundatthefirmthathadthelongestreturnhistory. 3. Whenthereturnhistorywasthe Managersbyregion sameforboththeonshoreand 140 131 offshorevehicles,wedefaulted 120 totheonshorefund. 96 100 4. Weexcludedfundsoffunds 86 83 80 70 fromourstudy. 66 60 5. Theresultingreturnscompiled 33 areequallyweighted. 40 6. Themonthlyreturnforeach 20 indexisasimpleaverageofall 0 Boston Chicago Dallas Greenwich LosAngeles NewYork SanFrancisco returnsavailableforthatmonth.

Aswithanyhedgefundindex,structuralweaknessesexist.Therewillbethetypicalbiasesthatmostindustry participantsareawareexistinalmostallhedgefundstudies.Furthermore,becauseweonlyincludedlive managersinourstudy,thesebiasesarelikelytobemorepronouncedinourcitycompositeindices.These weaknessesaremitigatedtosomedegree,aseachindexcontainsthesamebiases;thereforethecomparisons toeachothermaycarrymoreweightthancomparisonsagainstothermarketorhedgefundindices.Biases includebutarenotlimitedto: Selectionbias:Notallhedgefundmanagersreportresultsconsistentlytoadatabase. Survivorshipbias:Hedgefundsthatgooutofbusinessaretypicallydroppedfromhedgefund databases.Asaresultsurvivorshipbiastendstooverstatehedgefundreturns. Backfillbias:Alsoknownasinstanthistorybias,thisoccurswhenahedgefundisaddedintotheindex anditshistoricalreturnsarebackfilledintothereturnhistory.Ourcityindicesareentirelybackfilled giventhattheywerecreatedforthepurposeofthisstudy. PerformanceReview January2000December2011;12YearPeriod OverthefulltwelveyearperiodwestudiedfromJanuary2000toDecember2011annualizedreturnsforthe variouscitiesrangedfrom+9.33%to+13.64%,withChicagoatthetopendandSanFranciscoatthebottom. WhileBostonwasthesecondbestperformingcityovertheaggregateperiod,itwasoutperformingChicagobya narrowmarginthroughearly2007beforeexperiencinglossesthatledtomaterialunderperformanceonwards. LosAngeles,SanFranciscoandGreenwichwereconsistentlytheworstperformingcitieswithNewYorkCity remainingatthemiddleofthepackforthedurationofthestudy.

Growthof$1,000
Jan2000 Dec2011
5,000

4,000

3,000

2,000

1,000 2000 2001 2002 Chicago 2003 2004 2005 Greenwich 2006 2007 LosAngeles 2008 2009 NewYork 2010 2011 Boston Dallas SanFrancisco

December2012 3of18

January2002December2011;10YearPeriod Duringthelasttenyears,ChicagoandDallaswerethebestperformingcitiesat+12.80%and+11.53%, respectively,onanannualizedbasis.Whiletheremainingcitiesweretightlyclustered,theworstperformerswere SanFranciscoandGreenwichat8.42%and8.83%,respectively.Themostmaterialchangefromthetwelveyear performancerecordtothetenyearperformancerecordwasBoston,whosereturndecreased2.15%onan annualizedbasis.Thedropinperformancewasduetotheyears2000and2001rollingoffthereturnhistory whichwerebothparticularlystrongyearsforBoston.Amongstthe70fundsthatweutilizedinconstructingour Bostonindex,thereareanotablenumberof ANNUALIZEDPERFORMANCE(Jan2000Dec2011) City 1Year 2Years 3Years 5Years 10Years 12Years long/shorthedgefundsthatutilizeavalue Boston 1.63 4.90% 13.08% 5.49% 9.75% 11.90% investmentapproach.Giventherelative Chicago 1.08 7.33% 12.62% 8.82% 12.80% 13.64% outperformanceofvaluestocksversus Dallas 0.25 6.97% 10.62% 9.34% 11.53% 11.53% Technology,Media,andTelecom(TMT) Greenwich 1.28 4.96% 11.07% 6.72% 8.83% 9.89% stocksoverthisperiod,itcomesasno 2.61 5.20% 13.81% 5.01% 8.90% 9.35% surprisethatBostonsawsuchaprecipitous LosAngeles 2.48 5.15% 12.59% 6.46% 9.57% 10.49% declineintheannualizedreturnoftheindex NewYork SanFrancisco 0.07 7.71% 12.69% 5.37% 8.42% 9.33% oncetheseyearswerenolongerincluded. January2007December2011;5YearPeriod Thelastfiveyearsisaninterestingperiodtoanalyzebecauseitincludesboththepreandpost2008financial crisisperiods.OverthisperiodDallasandChicagohavebeenthebestperformingcitiesat+9.34%and+8.82%, respectively,outperformingtheremaininggeographiesbymorethan200basispointsonanannualizedbasis. Bostonsperformancedegradationcontinuedduringthisperiodwiththeirannualizedreturnsfallingan additional4.26%(versustrailing10yearperiod),morethananyothercity.TheworstperformingcitywasSan Francisco,withaspreadofnearly400basispointslessthanDallas;aspreadthatisfairlyconsistentbetween thesetwoareasfora5,10,and12yearperiod.Notsurprising,poorrelativeperformancein2007and2008was theprimarycontributingfactorthatdroveannualizedreturnsovera5yearstretch.Aswebegintoassess1,2, and3yearperformancerecords,returnsbegintomateriallyconvergewhichisfairlyobviousinthestatistical chartsthatyouwillseeoverthecourseofthisreport. PerformanceDuringMarketStress Greenwichwasconsistentlythebestperformingcityduringthethreeperiodsofsubstantialmarketstressthat waswitnessedoverthedurationofthestudywhileNewYorkwastheworstperformingcityintwoofthethree periods.Inaggregate,bothSanFranciscoandLosAngelescapturedahigherpercentageofmarketdrawdown thantheothercities.WhiletophonorsgotoChicagoasthebestoverallperformeroverthe12yearperiod,it onlydistinguisheditselffromacapitalpreservationstandpointin20072009.GreenwichandBostonwerethe bestperformingcitiesduringthemarketcorrectionin2011,experiencingadrawdownofslightlymorethan40% ofthebroadmarket.NewYorkhasdonethepoorestjobofpreservingcapitalovertime,capturing17.9%ofthe drawdownin2000/2002and62.6%ofthedrawdownin2011.

December2012

4of18

Weobservedthat,asagroup,hedgefundsseemtobehavingmoredifficultyprotectingcapitalduringmarket correctionsthantheydidintheearlieryearsofthestudy.Duringthe20002002internetbubblemarket correction,hedgefunds(asdefinedbytheHFRICompositeIndex)wereabletoshieldinvestorsfrommostofthe marketcorrection,capturingonly14.3%ofthedrawdown. DrawdownCapturePercentage Thismetricincreasedto42.0%in20072009and55.2%in 8/009/02 10/072/09 5/119/11 2011.Weattributetheworseningdownmarketprotection Boston 7.1% 39.5% 41.8% torisingcorrelationsamongstassetclassesglobally.The Chicago 10.4% 26.9% 43.7% 20002002drawdownwaslargelydrivenbyaselloffin Dallas 9.1% 29.7% 44.2% technologystocksbuttherewereareasofthemarket, Greenwich 3.4% 24.4% 40.9% particularlyvaluestocks,thatwereupduringtheperiod. LosAngeles 8.4% 43.9% 51.6% Thedrawdownin20072009and2011werecharacterized byfearsofasystemiccollapsewithallriskassetssellingoff NewYork 17.9% 39.5% 62.6% indiscriminately.Nocity,includingtheHFRIComposite SanFrancisco 13.0% 47.0% 49.1% Index,improvedtheirdrawdowncapturefrom20072009 HFRIComposite 14.3% 42.0% 55.2% to2011. CorrelationtoMajorAssetClasses ChicagoandDallashadthelowestcorrelationtoglobalequitymarketswhileNewYorkandSanFranciscohad thehighest.Chicagohadthelowestcorrelationswitheveryequityorientedindexandthehighestcorrelation withlowriskassetssuchastheBarclaysAggregateIndex(fixedincome)andtheBarclaysU.S.TreasuryIndex. WeattributethistothelargenumberofCommodityTradingAdvisorfunds(CTA)inthecity.Dallashadlow correlationstoallmarketindices,whichwebelieveisduethediversityofhedgefundstrategiesemployedby managersinthearea.CorrelationsforalllocationswerehigherfortheRussell2500(midcapequities)thanthey werefortheS&P500(largecapequities).Webelievethatthisisbecauselong/shortequityhedgefundstendto targetmidcapequitiesforinclusionintheirportfolios.NewYorkhada0.88correlationtotheRussell2500 Index,whichisextremelyhighandimpliestousthatmanagersinthecityareactivelyinvolvedintradingmid capsecurities,perhapstothepointwheretheymaybeinfluencingthebehavioroftheunderlyingindex.Los Angeleshadthehighestcorrelationtohighyieldbondsat0.80,whichislikelyduetotheconcentrationoflarge creditmanagersthatarebasedinthearea.Webelievetheirpresenceinthecitycontributestotradinginnon
December2012 5of18

investmentgradecreditthroughbothideasharingandnaturalattritionofskilledanalyststostarttheirown firmsorjoinsmallerfirmswheretheymayhavehighercompensationpotential.Thedifferenceincorrelation betweengrowth(+0.85)andvalue(+0.74)stockswasgreaterinSanFranciscothaninanyothercity.The presenceofventurecapitalfirmsandtechnologycompaniesintheBayareaandtheSanJosepeninsulalikely fosteramentalityofgrowthinvestinginthehedgefundcommunity.


CorrelationtoMajorAssetClasses(Jan2000Dec2011)
Index Boston Chicago Dallas Greenwich LosAngeles NewYork SanFrancisco

S&P500 Russell2000 Russell2500 Russell2500Growth Russell2500Value MSCIEAFE MSCIWorldexU.S. S&PGSCI BarclaysHighYield BarclaysAgg. BarclaysU.S.Treasury

0.71 0.42 0.55 0.65 0.69 0.78 0.72 0.73 0.45 0.57 0.73 0.75 0.86 0.84 0.77 0.48 0.60 0.75 0.78 0.88 0.85 0.72 0.38 0.56 0.69 0.74 0.87 0.85 0.75 0.52 0.58 0.72 0.72 0.77 0.74 0.76 0.54 0.65 0.75 0.78 0.81 0.78 0.77 0.55 0.67 0.77 0.79 0.82 0.79 0.45 0.42 0.50 0.48 0.49 0.42 0.49 0.72 0.40 0.52 0.63 0.80 0.72 0.68 0.06 0.16 (0.04) 0.03 0.08 (0.04) (0.01)

MSCIEmergingMarket 0.79 0.51 0.70 0.77 0.81 0.84 0.81

(0.25) (0.01) (0.24) (0.24) (0.26) (0.33) (0.29) CorrelationtoHedgeFundStrategiesasaWhole WiththeexceptionofChicagoandDallas,allcitieshadveryhighcorrelationswiththeHFRICompositeIndexand boththelong/shortequityandeventdrivenstrategies.NewYorkandSanFranciscohadthehighestcorrelation tohedgefunds,implyingthatthemanagersinthosecitiesareundifferentiatedandperforminginlinewiththe overallhedgefundindustry.NewYorkandSanFranciscoscorrelationtolong/shortequityhedgefundswas closeto1inbothcases,whichisfairlylogicalgiventheconcentrationofpurelong/shortequityhedgefundsin eachcity.ChicagohadthehighestcorrelationwithGlobalMacro(+0.63)andCTA(+0.62)strategieswhichisnot unexpectedgiventhattherearealargenumberofCTAmanagersinthecity.Greenwichalsohadarelatively highcorrelationwithGlobalMacro(+0.62)comparedtoothercities.Intuitivelythismakessensetousassome ofthelargestGlobalMacromanagersinthecountryarebasedintheGreenwicharea.
CorrelationtoHedgeFunds(Jan2000Dec2011)
Index HFRIComposite HFRIEquityHedge HFRIEventDriven HFRIMacro BarclaysCTA
Boston Chicago Dallas Greenwich LosAngeles NewYork SanFrancisco

0.89 0.55 0.77 0.89 0.89 0.96 0.95 0.88 0.55 0.75 0.88 0.89 0.96 0.96 0.88 0.53 0.76 0.87 0.89 0.94 0.92 0.45 0.63 0.49 0.62 0.44 0.47 0.48 0.12 0.62 0.18 0.35 0.11 0.04 0.08

MarketCaptureRatios Upmarketcaptureratiosforallcitiestrendedtogetherduringtheperiod,startingrelativelylow,peakinginthe middleof2008,andthenretreatingtonewlowsfollowingthe2008financialcrisis.NewYorkmaintainedthe highestupmarketcaptureratiowhileGreenwichmaintainedthelowestonaverageduringtheperiod.As


December2012

6of18

noted,upmarketcaptureratiosareatalltimelowsforeachcityandwebelievethereareafewexplanations. Mostnotably,equitieshaveseenastrongresurgencefollowingthefinancialcrisisandhedgefundshaverarely participatedfullyinsuchmarketenvironments.Hedgefundshavealsobeencarryinglowernetexposurethan theyhaveinpastperiodswithMorganStanleyPrimeBrokeragereportingnetexposureattheendof2011was ~40%netlongcomparedto~60%netlongin2005andapeaknorthof75%inlate2007.Thisiscertainlya contributingfactorforthegeneralunderperformanceofhedgefundsandlikelyreflectsbroadbasedconcerns thatrangefromeconomicuncertaintytoalackofclarityonhowtheregulatoryenvironmentmaychangeforthe assetclass.

WiththeexceptionofDallas,downmarketcaptureratiosforallcitiestrendedinthesamedirection.Themost notabledeviationtothegeneraltrendwasin2007whenDallasexperiencedasubstantialnegative(positivefor investors)downmarketcaptureratio.Webelievethatthisiscanbeattributedtotheshortsubprimetradethat variousDallasbasedmanagershadonatthetime.Thisreinforcesourbeliefthatideasharingiscommonplace inallgeographies.Intheperiodsincethe2008financialcrisis,citiesappeartohaveclusteredattwodifferent levels.NewYork,SanFrancisco,BostonandLosAngelescurrentlyhavedownmarketcaptureratiosinthe30% to40%rangewhileDallas,GreenwichandChicagoareinthelow20%range.

December2012 7of18

RollingCorrelation Insomecities,correlationtotheS&P500Indexshifteddramaticallyoverthecourseoftheperiodwestudied. ThemostnotableshiftwasinChicagowhichstartedwitharolling48monthcorrelationof0.07andendedthe periodat0.80.Dallashadalargecorrelationshiftofitsown,startingtheperiodat0.60,fallingtoalowof0.18 inthemiddleperiodandrisingto0.67attheendoftheperiod.Asitstands,Dallashasthelowestrolling48 monthcorrelationasofDecember31,2011whileNewYorkhadthehighestcorrelation,onaverage,totheS&P 500duringthisperiod.Correlationsforvirtuallyallriskassetshavebeenrisingsincethe2008financialcrisis, whichhasalsobeenthecaseforthevariouscityindiceswehavecompiled.

CorrelationtotheHFRICompositeIndexhasbeensteadilyrisingforallcitiesovertheperiod.Themost dramaticrisecanbeseeninChicago,wherecorrelationsrosefrom0.14atthestartoftheperiodto0.90bythe endoftheperiod.Fortheaggregateperiod,DallasmaintainedrelativelylowcorrelationwhileNewYorkand SanFranciscobothhadcorrelationsinthehigh0.90s.Correlationsforallcitiesbegantoconvergeinlate2008, whichisconsistentwithvirtuallyallriskassetsglobally.CurrentlyDallasandChicagoexhibitthemost differentiatedbehaviorfromtheoverallhedgefundindustry.

December2012 8of18

StandardDeviation Thelevelofstandarddeviation(acommonmeasureofvolatilityandrisk)betweenthecitieshasbeenquite differentovertheperiodwhiletheoveralltrendhasbeenquitesimilar.Chicagostandsoutasitinitiallyhadthe highestvolatilityofanycitybutsince2008hasfallentothemiddleofthepack.Greenwichconsistentlyhadthe loweststandarddeviationonaverage,andiscurrentlywellbelowanycityaswellastheHFRICompositeIndex. Weattributethistoglobalmacroexposure,astrategythattendstoexhibitlowerlevelsofvolatility,whichis prevalentinthecity.LosAngeleshadthebiggestswinginvolatilityreachingalowof3.5%inearly2007and toppingallcitiesat10%attheendof2011.Wesuspectthatthisisduetocreditmanagersshiftingfromthe utilizationofinternalpricingmodelstomarketbasedpricingmodelsin20072008.

BetatotheS&P500Index Therangeofbetas(ameasureofmarketsensitivity)forthecitieswasquitewide.Chicagohashadthegreatest variationovertimerangingfromnearzeroatthestartoftheperiodandrisingtoanorthof0.60duringthefirst partof2006.ThisdoesnotcomeasasurprisetousaswevediscussedtheconcentrationofCTAsinChicago thatusetrendfollowingmodels.Thesemodelstendtolooktoincreaseexposuretothemarketasitisgoingup andreduceexposureasitisgoingdown.Dallas,onaverage,hadthelowestbetaduringtheperiod.NewYork hadthehighestbeta,onaverage,duringtheperiodwhichisconsistentwiththecityhavingthehighestupand downmarketcaptureratios.Since2008,betaforallcitieshasleveledoff.Webelievethisisduetotwofactors; risingcorrelationsamongstallassetclassesanddeliberateintentionsbyhedgefundstomaintainmore conservativenetexposures.ThebetasforNewYork,LosAngelesandSanFranciscohavebeenvirtuallyidentical since2008.TheHFRICompositesbetafallsinthemiddleofallthecities,whichistobeexpectedgiventhe broaderscopeoftheindex.

December2012

9of18

Conclusions Ourfindingssuggestthathedgefundperformancebehaviorvarieswidelydependingonthegeographical locationofthefirm.Amajorcontributingfactortodifferencesinperformanceislikelydrivenbythestyleand strategyofhedgefundsthataremostprevalentinaparticularcity.Webelievethatcitiesdevelopareasoffocus relatedtotheexistenceoflargeassetmanagementfirmsthatinevitablyseetalentleavetostarttheirownfirms. Giventhemannerinwhichtheyaretrained,eachnewlyestablishedfirminagivenareaislikelytohavesimilar portfolioexposures,intentionallyornot.Wehaveseenmultipleincidencesofsharedtradesinagivencity, mostnotablytheshortsubprimetradewitnessedinDallasin2007.Whilevariousmeasuresofriskandreturn produceddistinctresultsacrossoursampleset,mostcitiesexperiencedsimilartrends. Boston:Bostonwasthesecondbestperformingcityduringtheperiod,returning+11.90%annualized.Despite itsoverallstanding,performancetrailedoffmateriallyafter2001,whenvaluestocksoutperformedgrowth orientedequities.BostonhadthehighestcorrelationtotheRussell2500ValueIndexwhichimpliesthat managersinthecityareactivelyinvestinginvaluestocksandthatisconsistentwiththefactthatthereare numeroushighprofilevalueorientedhedgefundsinBoston.Withtheexceptionof20082009,Bostondida betterjobthanmostcitiespreservingcapitalduringmarketcorrections.Correlationtothebroadmarketand hedgefundindiceswasgenerallyinthemiddleofthegroup.Similarly,marketcaptureratios,betaandstandard deviationtrendedwiththeoverallgroupandwereundistinguished. Chicago:Chicagowasthebestperformingcityinthestudy,returning+13.64%onanannualizedbasis,and outpaceditsclosestcompetitorsbymorethan1.5%peryear.Chicagoperformedwellduringperiodsofmarket stressandwasthesecondbestperformingcityduringthe2008financialcrisis.TheprevalenceofCTAsmay explaintheperformancedifference,whichisconsistentgiventhemeaningfulcorrelationtotheCTAIndexat 0.62.CorrelationtotheHFRIGlobalMacroIndexwasalsohighat0.63.Chicagoalsohadthelowestcorrelation tothebroadmarketandtotheHFRICompositeIndexduringtheperiod.Ontheotherhand,wealsonotethat Chicagohadthehigheststandarddeviation,whichreducestheirriskadjustedreturnsasillustratedbytheir SharpeRatioof1.26,themedianofthegroup. Dallas:Dallaswasthe3rdbestperformingcityduringtheentireperiod,producingannualizedreturnsof +11.53%,whilehavingthebestperformanceoverthelast5years.Thesuperiorperformancecouldbepartially impactedbythefactthatfundsinDallashavelowercompensationstructuresthanfundsinotherpartsofthe
December2012 10of18

country.Dallaswasamongstthebestatpreservingcapitalduringperiodsofmarketstressandexhibitedlower correlationwiththebroadmarketandhedgefundindicesthananyothercity,withtheexceptionofChicago. Weattributethispartlytoadiversegroupofunderlyingstrategieswithinthecity.Inthe2007timeperiod, Dallasexhibitedmuchlowerdownmarketcaptureandbetathanalloftheothercities,whichwebelieveisdue totheshortsubprimetradethatsomehighprofilemanagershadonatthetime.Dallasalsohadthesmallest samplesizeofanyofthecitiesinourstudywhichcouldalsobeacontributingfactortotheirperformance characteristics. Greenwich:Amiddleofthepackperformerduringtheperiod,returning+9.89%onanannualizedbasis,trailing themedianperformer(NewYork)by60basispointsannually.Thatsaid,Greenwichwasthecitythatperformed thebestduringeveryperiodofmarketstresswestudied.Greenwichalsohadamongstthelowestbetaand standarddeviationofanyofthecities,resultinginstrongriskadjustedreturnswhichcanbeseenintheirSharpe Ratioof1.32.ThisisonlymarginallylowerthanDallasandBostonat1.34and1.33,respectively.Greenwich hadlowercorrelationwithbroadmarketindicesthanmostothercitiestrailingonlyChicagoandDallas,while correlationwiththehedgefundindiceswasinthemiddleofthepack.Greenwichhadhighcorrelationwiththe HFRIGlobalMacroIndexwhichmakessensegivensomeofthehighprofileglobalmacrohedgefundsbasedin thecity. LosAngeles:LosAngeleswasthesecondworstperformingcityduringtheperiodat+9.35%onanannualized basis,outpacingSanFranciscobyjust2basispoints.Theywerealsoamongsttheworstcitiesinprotecting capitalduringperiodsofmarketstress.LosAngeleshadthelowestupmarketcaptureandthehighestdown marketcaptureratioofallthecitiesinourstudy.Correlationstomajormarketandhedgefundindiceswere relativelyhighforLosAngeles.LAhadthehighestcorrelationtohighyieldbondswhichweattributetothe existenceofseverallargecreditfocusedfirmswithincitylimits.LosAngeleshadthelargestshiftinstandard deviationofanycityinourstudy.Inouropinion,thischangewasduetocreditmanagersshiftingtheirportfolio pricingpracticesfrominternalpricingtoexternalpricing,likelydrivenbythedemandsofinvestorsfollowingthe 2008creditcrisis. NewYork:NewYorkwasthemedianperformingcityduringtheperiodat+10.49%annualized.NewYorkdid notprotectcapitalwellonarelativebasisasitwastheworstperformingcityduringtwooutofthethree periodsofmarketstresswestudied(20002002and2011)andhadthehighestcorrelationwithmarketand hedgefundindices.CorrelationtotheRussell2500Indexwas0.88whichimpliesthatfundsinthecityare activelytradingmidcapequities.Correlationwasalsoveryhigh(+0.87)totheRussell2500GrowthIndexwhich mayimplythatNewYorkbasedmanagersactivelytradetechnologystocks.Thepoorrelativedrawdownin2000 2002alsopointstothisastechnologystocksperformedpoorlyduringthatperiod.Furthermore,our experiencehasshownthatNewYorkfundstradeintechnologynamesbecausetheytendtoofferoutsized returnpotentialovershorterperiodsoftime.NewYorkwasalsoconsistentlyamongstthehighestwhenitcame toupanddownmarketcaptureratios,betaandstandarddeviationwhichallimplyhedgefundsinthecityare managingtheirportfoliosinanaggressivemanner. SanFrancisco:SanFranciscowastheworstperformingcityduringtheperiodat+9.33%annualizedandwasa particularlypoorperformerduringperiodsofmarketstress.SanFranciscowastheworstperformingcityin 20072009,thesecondworstperformerin20002002,andthethirdworstperformerin2011.SanFrancisco hadthesecondhighestcorrelationtomarketindicesofallthecitieswestudied.SimilartoNewYork,theyhad veryhighcorrelationwiththeRussell2500GrowthIndex(+0.85)whichimpliestheyaretradingtechnology
December2012 11of18

stocks.Thiscouldbeduetothepresenceofventurecapitalfirmsandtechnologycompaniesintheregionwhich maybeinfluencingmanagerbehavior.TheywerealsothecitywiththesecondhighestcorrelationtotheHFRI CompositeIndex.SanFranciscowasalsoamongstthehighestrankingcitiesinupanddownmarketcapture ratios,betaandstandarddeviation.Thisisimplyingthathedgefundsinthecityaremanagingtheirportfolios aggressively,withoutsizedexposuretotechnologyasawhole. NextSteps Weintendtoincreasethesizeandthescopeofthisstudynextyear,updatingandreevaluatingthe performancehistoryandcharacteristicsofeachregionin2012.Inadditiontopotentiallyaddingnewcities,we willlooktoreportonthenumberofunderlyingstrategiesineachcityandcompileassetsundermanagement andaveragefeestructure. AboutBalterCapitalManagement Formedin2006,BalterCapitalManagement,LLCisaregisteredinvestmentadvisorfocusingpurelyonhedge fundresearchandinvesting.Webelieveourdedicatedfocusonhedgefundinvestingprovidesuswiththe highestlevelofinsightanddifferentiationbyapplyingourrigorousapproachtosourcing,analyzing,monitoring andinteractingwithhedgefunds.Weutilizebothestablishedfundsaswellascontinuouslysourcingnonbrand namemanagersthathavenotyetbeenidentifiedbythebroaderinvestmentcommunity. PleasecontactBenDeschaineviaemailatbdeschaine@baltercap.comwithquestionsorcomments. BalterCapitalManagement,LLC

December2012

12of18

Glossary Beta:Calculatedusingregressionanalysis,andmeasuresthetendencyofasecurity'sreturnstorespondto swingsinthemarket.Abetaof1indicatesthatthesecurity'spricewillmovewiththemarket.Abetaofless than1meansthatthesecuritywillbelessvolatilethanthemarket.Abetaofgreaterthan1indicatesthatthe security'spricewillbemorevolatilethanthemarket.Forexample,ifastock'sbetais1.2,it'stheoretically20% morevolatilethanthemarket. Correlation:Astatisticalmeasureofhowtwosecuritiesmoveinrelationtoeachother.Resultsrangebetween 1and+1.Acorrelationof+1meansthatthetwosecuritiesareperfectlypositivelycorrelatedandmovein lockstep.Acorrelationof1meansthatthetwosecuritiesareperfectlynegativelycorrelatedandmovein oppositedirectionofeachother.Acorrelationof0meansthattwosecuritiesmoveindependentofeachother. RollingStatistics:Wechosetoevaluatecertainperformanceandriskstatisticson48monthrollingperiods.This approachlooksatstatisticsona48monthperiodandtherollsforwarddroppingthelastmonthandaddingthe nextmonth.Thisallowsustoevaluateriskandreturnoveralongerperiodoftimewhilesmoothingoutofsome noiseinthedata. StandardDeviation:Astatisticalmeasureofthevolatility(ameasureofrisk)ofthehistoricreturnsofasecurity. Avolatileorriskysecuritywillhaveahighstandarddeviationcomparedtoaconservativesecuritythatwillhave alowerstandarddeviation. IndexDefinitions BarclaysCapitalU.S.AggregateBondIndex:Abroadbasedbenchmarkthatmeasurestheinvestmentgrade, U.S.dollardenominated,fixedratetaxablebondmarket,includingTreasuries,governmentrelatedand corporatesecurities,MBS(agencyfixedrateandhybridARMpassthroughs),ABS,andCMBS.TheU.S. AggregaterollsupintootherBarclaysflagshipindices,suchasthemulticurrencyGlobalaggregateIndexand theU.S.UniversalIndex,whichincludeshighyieldandemergingmarketsdebt.TheU.S.AggregateIndexwas createdin1986withindexhistorybackfilledtoJanuary1,1976. BarclaysCapitalU.S.HighYieldIndex:MeasuresthemarketofUSDdenominated,noninvestmentgrade, fixedrate,taxablecorporatebonds.SecuritiesareclassifiedashighyieldifthemiddleratingofMoodys,Fitch, andS&PisBa1/BB+/BB+orbelow,excludingemergingmarketdebt.TheU.S.CorporateHighYieldIndexwas createdin1986withhistorybackfilledtoJuly1,1983,androllsupintotheBarclaysU.S.UniversalandGlobal HighYieldIndices. BarclaysU.S.TreasuryIndex:TheU.S.TreasuryIndexincludespublicobligationsoftheU.S.Treasury.Treasury billsareexcludedbythematurityconstraintbutarepartofaseparateShortTreasuryIndex.Inaddition,certain specialissues,suchasstateandlocalgovernmentseriesbonds(SLGs),aswellasU.S.TreasuryTIPS,are excluded.STRIPSareexcludedfromtheindexbecausetheirinclusionwouldresultindoublecounting. SecuritiesintheindexrolluptotheU.S.Aggregate,U.S.Universal,andGlobalAggregateIndices.TheU.S. TreasuryIndexwaslaunchedonJanuary1,1973.

December2012

13of18

MSCIEAFEIndex:Afreefloatadjustedmarketcapitalizationindexthatisdesignedtomeasuretheequity marketperformanceofdevelopedmarkets,excludingtheUS&Canada.TheMSCIEAFEIndexconsistsofthe following22developedmarketcountryindices:Australia,Austria,Belgium,Denmark,Finland,France,Germany, Greece,HongKong,Ireland,Israel,Italy,Japan,theNetherlands,NewZealand,Norway,Portugal,Singapore, Spain,Sweden,Switzerland,andtheUnitedKingdom. MSCIEMIndex:Afreefloatadjustedmarketcapitalizationindexthatisdesignedtomeasureequitymarket performanceofemergingmarkets.TheMSCIEmergingMarketsIndexconsistsofthefollowing21emerging marketcountryindices:Brazil,Chile,China,Colombia,CzechRepublic,Egypt,Hungary,India,Indonesia,Korea, Malaysia,Mexico,Morocco,Peru,Philippines,Poland,Russia,SouthAfrica,Taiwan,Thailand,andTurkey. MSCIWorldIndexexUSA:Afreefloatadjustedmarketcapitalizationweightedindexthatisdesignedto measuretheequitymarketperformanceofdevelopedmarkets.TheMSCIWorldIndexconsistsofthefollowing 24developedmarketcountryindices:Australia,Austria,Belgium,Canada,Denmark,Finland,France,Germany, Greece,HongKong,Ireland,Israel,Italy,Japan,Netherlands,NewZealand,Norway,Portugal,Singapore,Spain, Sweden,Switzerland,andtheUnitedKingdom. Russell2000Index:MeasurestheperformanceofthesmallcapsegmentoftheU.S.equityuniverse.The Russell2000isasubsetoftheRussell3000Indexrepresentingapproximately10%ofthetotalmarket capitalizationofthatindex.Itincludesapproximately2000ofthesmallestsecuritiesbasedonacombinationof theirmarketcapandcurrentindexmembership. Russell2500Index:MeasurestheperformanceofthesmalltomidcapsegmentoftheU.S.equityuniverse, commonlyreferredtoas"smid"cap.TheRussell2500isasubsetoftheRussell3000Index.Itincludes approximately2500ofthesmallestsecuritiesbasedonacombinationoftheirmarketcapandcurrentindex membership. Russell2500GrowthIndex:MeasurestheperformanceofthesmalltomidcapgrowthsegmentoftheU.S. equityuniverse.ItincludesthoseRussell2500Indexcompanieswithhigherpricetobookratiosandhigher forecastedgrowthvalues. Russell2500ValueIndex:MeasurestheperformanceofthesmalltomidcapvaluesegmentoftheU.S.equity universe.ItincludesthoseRussell2500Indexcompanieswithlowerpricetobookratiosandlowerforecasted growthvalues. S&P500Index:RegardedasthesinglebestgaugeoftheU.S.equitiesmarket,thisindexincludes500leading companiesinleadingindustriesoftheU.S.economy.AlthoughtheS&P500focusesonthelargecapsegment ofthemarket,withapproximately75%coverageofU.S.equities,itisalsoanidealproxyforthetotalmarket. S&P500ispartofaseriesofS&PU.S.indicesthatcanbeusedasbuildingblocksforportfolioconstruction.S&P 500ismaintainedbytheIndexCommittee,ateamofS&PDowJonesIndiceseconomistsandindexanalysts, whomeetonaregularbasis.ThegoaloftheIndexCommitteeistoensurethattheS&P500remainsaleading indicatorofU.S.equities,reflectingtheriskandreturncharacteristicsofthebroaderlargecapuniverseonan ongoingbasis.

December2012

14of18

S&PGSCI:Providesinvestorswithareliableandpubliclyavailablebenchmarkforinvestmentperformanceinthe commoditymarkets.Theindexisdesignedtobetradable,readilyaccessibletomarketparticipants,andcost efficienttoimplement.TheS&PGSCIisthefirstmajorinvestablecommodityindex.Itisoneofthemostwidely recognizedbenchmarksthatisbroadbasedandproductionweightedtorepresenttheglobalcommoditymarket beta.TheS&PGSCIiscalculatedprimarilyonaworldproductionweightedbasisandiscomprisedofthe principalphysicalcommoditiesthatarethesubjectofactive,liquidfuturesmarkets. HedgeFundIndexDefinitions HFRIMonthlyIndices(HFRI):Equallyweightedperformancebenchmarks,HFRIindicesarebrokendowninto4 mainstrategies,eachwithmultiplesubstrategies.AllsinglemanagerHFRIIndexconstituentsareincludedin theHFRIFundWeightedComposite,whichaccountsforover2200fundslistedontheinternalHFRDatabase. DuetomutualagreementswiththehedgefundmanagerslistedintheHFRDatabase,wearenotatlibertyto disclosetheparticularfundsbehindanyindextonondatabasesubscribers.FundsincludedintheHFRIMonthly Indicesmust,reportmonthlyreturns,reportnetofallfeesreturns,reportassetsin$USD,andhaveatleast$50 Millionundermanagementorhavebeenactivelytradingforatleasttwelve(12)months EquityHedge:Fundswhomaintainpositionsbothlongandshortinprimarilyequityandequityderivative securities.Awidevarietyofinvestmentprocessescanbeemployedtoarriveataninvestmentdecision, includingbothquantitativeandfundamentaltechniques;strategiescanbebroadlydiversifiedornarrowly focusedonspecificsectorsandcanrangebroadlyintermsoflevelsofnetexposure,leverageemployed,holding period,concentrationsofmarketcapitalizationsandvaluationrangesoftypicalportfolios.EquityHedge managerswouldtypicallymaintainatleast50%exposureto,andmayinsomecasesbeentirelyinvestedin, equitiesbothlongandshort. EventDriven:Fundswhomaintainpositionsincompaniescurrentlyorprospectivelyinvolvedincorporate transactionsofawidevarietyincludingbutnotlimitedtomergers,restructurings,financialdistress,tender offers,shareholderbuybacks,debtexchanges,orsecurityissuance.Securitytypescanrangefrommostsenior inthecapitalstructuretomostjuniororsubordinated,andfrequentlyinvolveadditionalderivativesecurities. EventDrivenexposureincludesacombinationofsensitivitiestoequitymarkets,creditmarketsand idiosyncratic,companyspecificdevelopments.Investmentthesesaretypicallypredicatedonfundamental characteristics,withtherealizationofthethesispredicatedonaspecificdevelopmentexogenoustotheexisting capitalstructure. Macro:Fundswhotradeabroadrangeofstrategiesinwhichtheinvestmentprocessispredicatedon movementsinunderlyingeconomicvariablesandtheimpactthesehaveonequity,fixedincome,hardcurrency andcommoditymarkets.Managersemployavarietyoftechniques,bothdiscretionaryandsystematicanalysis, combinationsoftopdownandbottomuptheses,quantitativeandfundamentalapproachesandlongandshort termholdingperiods.AlthoughsomestrategiesemployRVtechniques,MacrostrategiesaredistinctfromRV strategiesinthattheprimaryinvestmentthesisispredicatedonpredictedorfuturemovementsinthe underlyinginstruments,ratherthanrealizationofavaluationdiscrepancybetweensecurities. BarclayCTAIndex:Aleadingindustrybenchmarkofrepresentativeperformanceofcommoditytradingadvisors. Therearecurrently602programsincludedinthecalculationoftheBarclayCTAIndexfortheyear2012,whichis unweightedandrebalancedatthebeginningofeachyear.ToqualifyforinclusionintheCTAIndex,anadvisor
December2012 15of18

musthavefouryearsofpriorperformancehistory.Additionalprogramsintroducedbyqualifiedadvisorsarenot addedtotheIndexuntilaftertheirsecondyear.Theserestrictions,whichoffsetthehighturnoverratesof tradingadvisorsaswellastheirartificiallyhighshorttermperformancerecords,ensuretheaccuracyand reliabilityoftheBarclayCTAIndex.


December2012 16of18

AdditionalChartsandTables
Statistics(Jan2000Dec2011)
Boston Chicago Dallas Greenwich LosAngeles NewYork SanFrancisco S&P500 MSCIWorld exU.S. HFRIComp

Ann.CompoundReturn CumulativeReturn StandardDeviation BestMonthlyReturn WorstMonthlyReturn MaxDrawdown SharpeRatioGeom.(tbill) SortinoRatio(5.00%) GainDeviation LossDeviation DownsideDeviation(5.00%) Skewness Kurtosis NumberofPeriods NegativePeriods(%) PositivePeriods(%)

11.90% 285.24% 6.77% 5.97% 7.49% 20.10% 1.33 1.49 4.31% 5.57% 4.29% 0.79 2.98 144 26.39% 73.61%

13.64% 363.88% 8.38% 11.56% 4.94% 13.73% 1.26 1.81 6.21% 4.40% 4.40% 0.48 1.99 144 35.42% 64.58%

11.53% 270.45% 6.48% 5.66% 7.45% 15.11% 1.34 1.45 3.95% 5.21% 4.19% 0.87 2.42 144 25.00% 75.00%

9.89% 210.08% 5.48% 4.98% 4.92% 12.41% 1.32 1.36 3.61% 3.66% 3.35% 0.38 1.13 144 29.17% 70.83%

9.35% 192.19% 7.02% 6.10% 9.96% 22.36% 0.96 0.81 3.91% 6.66% 5.06% 1.49 6.40 144 29.17% 70.83%

10.49% 230.95% 8.09% 7.78% 7.37% 20.11% 0.96 0.97 5.01% 5.76% 5.29% 0.49 1.32 144 31.25% 68.75%

9.33% 191.59% 7.80% 8.97% 8.67% 23.94% 0.86 0.77 4.79% 5.86% 5.26% 0.58 3.03 144 32.64% 67.36%

0.55% 6.80% 16.32% 10.93% 16.80% 50.95% 0.10 0.34 9.38% 11.61% 12.60% 0.43 0.65 144 41.67% 58.33%

0.98% 12.38% 18.39% 12.90% 20.80% 56.63% 0.07 0.28 9.77% 13.29% 14.18% 0.60 1.29 144 44.44% 55.56%

5.69% 94.27% 6.88% 6.16% 6.84% 21.42% 0.48 0.13 3.93% 4.83% 5.08% 0.57 1.31 144 34.03% 65.97%

CalendarYearPerformance
Year Boston Chicago Dallas Greenwich LosAngeles NewYork SanFrancisco S&P500 MSCIWorld exU.S.

HFRI Composite

2011 2010 2009 2008 2007 2006 2005 2004 2003 2002 2001 2000

1.63% 11.87% 31.40% 17.15% 9.03% 15.51% 12.59% 11.79% 22.33% 9.12% 21.66% 24.92%

1.08% 16.46% 23.99% 4.88% 12.33% 13.21% 12.62% 11.06% 29.29% 19.32% 10.20% 26.27%

0.25% 14.72% 18.27% 10.26% 28.65% 19.71% 11.93% 11.22% 22.87% 4.08% 11.49% 11.59%

1.28% 11.61% 24.37% 10.27% 12.57% 12.92% 10.41% 9.69% 18.76% 3.64% 11.43% 19.44%

2.61% 13.65% 33.21% 20.61% 9.12% 13.32% 7.28% 11.10% 23.88% 9.74% 10.37% 12.89%

2.48% 13.39% 29.07% 17.57% 16.23% 15.90% 9.97% 13.38% 28.17% 1.53% 14.85% 15.54%

0.07% 16.09% 23.35% 22.15% 16.63% 10.91% 10.48% 11.05% 22.60% 3.57% 10.94% 17.09%

2.11% 15.06% 26.45% 37.00% 5.45% 15.81% 4.90% 10.88% 28.68% 22.10% 11.88% 9.11%

12.18% 8.94% 33.66% 43.56% 12.44% 25.71% 14.47% 20.38% 39.42% 15.80% 21.40% 13.37%

5.25% 10.25% 19.98% 19.03% 9.96% 12.89% 9.30% 9.03% 19.55% 1.45% 4.62% 4.98%

AnnualizedPerformnace(Jan2000Dec2011)
Year Boston Chicago Dallas Greenwich LosAngeles NewYork SanFrancisco S&P500 MSCIWorld exU.S.

HFRI Composite

1Year 2Year 3Year 4Year 5Year 6Year 7Year 8Year 9Year 10Year 11Year 12Year

1.63% 4.90% 13.08% 4.62% 5.49% 7.10% 7.86% 8.35% 9.82% 9.75% 10.78% 11.90%

1.08% 7.33% 12.62% 7.96% 8.82% 9.54% 9.98% 10.11% 12.09% 12.80% 12.56% 13.64%

0.25% 6.97% 10.62% 4.98% 9.34% 11.00% 11.13% 11.14% 12.39% 11.53% 11.53% 11.53%

1.28% 4.96% 11.07% 5.30% 6.72% 7.73% 8.11% 8.30% 9.42% 8.83% 9.06% 9.89%

2.61% 5.20% 13.81% 4.01% 5.01% 6.36% 6.49% 7.05% 8.80% 8.90% 9.03% 9.35%

2.48% 5.15% 12.59% 4.15% 6.46% 7.98% 8.26% 8.89% 10.88% 9.57% 10.04% 10.49%

0.07% 7.71% 12.69% 2.73% 5.37% 6.28% 6.87% 7.38% 8.98% 8.42% 8.65% 9.33%

2.11% 8.39% 14.10% 1.64% 0.26% 2.25% 2.63% 3.62% 6.15% 2.91% 1.47% 0.54%

12.18% 2.19% 8.54% 7.83% 4.09% 0.33% 2.24% 4.35% 7.76% 5.14% 2.39% 0.98%

5.25% 2.20% 7.82% 0.37% 2.22% 3.92% 4.68% 5.21% 6.71% 5.87% 5.75% 5.69%

December2012

17of18


December2012

18of18