January 27, 20
Groupement d’intérêt Scientiﬁque ENSAE - HEC - X
The 3rdAnnual Conference on Hedge Funds “Market intermediaries, ﬁnancial markets, and the real economy”
Paris - January 27th- 28th, 20
Organizers: erge arollesyor A and CT Christian ouriérouCT and niersity of Toronto aid ThesarHC Paris
Session 1: Equities and Leverage Chair: R. Sadka (Boston College) T. Adrian (Federal Reserve Bank of New York), E. Etula (Federal Reserve Bank of New York), . uir (Northwestern niversit) Broker-Dealer Leverage and the Cross-Section of Stock Returns isussant: . ilson (niversit of ford)
. ng (Coluia niversit),S. Gorovyy (Columbia niversiy), S. an nwegen (v sset anageent) Hedge Fund Leverage isussant: R. Savona (niversit of Bresia)
. Benavid (hio State niversit), F. Franoni (niversit of ugano),A. andier (Toulouse Sool of onomis), R. oussawi (niversit of ennslvania) Stock Price Manipulation b Hedge Funds isussant: .ero (Evr niversit and CRES) eaCoffee
Keynote Speaker Chair: . Cordell (NYSE iffe) . Aarya (New York niversiy) Sovereign Credit Risk
Session 2: arkets and oatiity Chair: C.ourierou (niversit of oronto and CRES) T. Ramadorai (niversiy of ford) nvestor nterest and Hedge Fund Returns isussant: . oert (EC aris) B. ovanovi (New York niversit),A. enkveld ( niversiy Amserdam) Middleen in Liit-rder Markets isussant: . e Fol (auhine niversit and CRES) N. ang (Boston College), . ondor (Central Euroe niversit),R. Sadka (Boson Collee) diosncratic Return olatilit in the Cross-Section of Stocks isussant: R. aria (EEC) eaCoffee
Session 3: noration rodution and edge unds Chair: R. aria (EEC) R. Sadka (Boston College),G. ik (C) Media and nvestent Manageent isussant: C. érignon (EC aris) S. Chung (Singaore anageent niversit),. Teo (Sinaore anaemen niversiy) Hedge Funds and nalst ptiis isussant: . eilethe (oard dier) Keynote Speaker Chair: : Seer (or )Supported by . . Bouaud (Caial Fund anaemen)
January 28, 20
Groupement d’intérêt Scientiﬁque ENSAE - HEC - X
9.00-11.00 Session 4: edge und isk and Surviva Chair: . outain (oods) N. Bollen (anderbil niversiy) Hedge Fund R hat’s under the Hood isussant: . ui (Erasus niversit Rotterda) . ragon (riona State niversit),. Nanda (Georia Te niversiy) n ournaent Behavior in Hedge Funds High ater Marks Fund Liuidation and the Backﬁlling Bias isussant: . Nag (niversit of Neuhâtel) S. arolles (or and CRES), . agliardini (Swiss Finane nstitute and niversit of ugano),C. Gourierou (niversiy of Torono and CRST) Survival of Hedge Funds Frailt vs Contagion isussant: . Sion (Caital Fund anageent)
Session : unds o edge unds Chair: S. arolles (or and CRES) . Billio (niversiy of enie), . ao (niversit of oronto), . elion (niversit of enie) Hedge Funds ail Risk and Marginal Risk Contribution in Fund of Hedge Funds isussant: .. Feranian (ENSE and CRES) G. Crion (ombard dier), . Saillet (Swiss Finane nstitute at ECniversit of eneva) nsupervised Risk Factor Clustering Construction Fraeork for Funds of Hedge Funds isussant: R. ouad (Riskdata)
Session : Liquidity and te inania risis Chair: . hesar (EC aris) C. Cao (enn Sae niversiy)iang (niversit of assahusetts at herst), . o ( Sloan Shool of anageent), Y. Chen (irginia eh niversit), B. Can Hedge Funds ie Market Liuidit isussant: . oenväärä (niversit of ulu) C. Chua (Singaore anageent niversit),. au (NSA and CR), S. ai (Singaore anageent niversit) Role of uit Funds in the Financial Crisis Propagation isussant: . Siegann ( niversit sterda) C. Cella (Stokhol Shool of Eonois), . Ellul (ndiana niversit),. Gianei (Sokolm Sool of onomis CR and CG) nvestors’ Horions and the pliﬁcation of Market Shocks isussant: . seselidakis (oethe niversit Frankfurt)
Scientiﬁc Committee: . Brav (uke niversit), S. arolles (or and CRES), . oldstein (harton Shool), C. ourierou (niversit of oronto and CRES), . atton (uke niversit), R. Sadka (Boston College), . hesar (EC aris).
he udated rogra is availale on the CRES wesite:thr.rest.ft:www
ATACT Adrian, Etula and uir ‘Leverage and the Cross-Section of Stock ReturnsBroker-Dealer ’ e ﬁnd that a stoks’ eeted returns is largel elained its ovariane with rokerdealer leverage. ur single leverage fator odel oetes with standard ultifator odels on the sie and ooktoarket ortfolios and outerfors suh odels when onsidering the industr and oentu ortfolios. e show how to tie leverage to the ﬁrst order onditions of ative investors’ ortfolio hoie role and derive leverage as a state variale in an C fraework. ng,Gorovyyand van nwegen ‘Hedge Fund Leverage’ e investigate the leverage of hedge funds using oth tieseries and rosssetional analsis. edge fund leverage is ounterlial to the leverage of listed ﬁnanial interediaries and dereases rior to the start of the ﬁnanial risis in id. edge fund leverage is lowest in earl when the leverage of investent anks is highest. Changes in hedge fund leverage tend to e ore reditale eonowide fators than fundseiﬁ harateristis. n artiular, dereases in funding osts and inreases in arket values foreast inreases in hedge fund leverage. ereases in fund return volatilities also inrease leverage. Benavid, Franoni,Landierand oussawi ‘Stock Price Manipulation b Hedge Funds’ e ﬁnd evidene for signiﬁant rie aniulation at the stok arket hedge funds. e douent that stoks that are held hedge funds eeriene anoral returns at the end of the onth (tie at whih hedge funds reort to investors), and a reversal in the following da. sing intrada data, we trae anoral uing ativit in these stoks ust efore the ell. e douent siilar atterns, although soewhat weaker, in short selling ativit. Finall, we reort that investors do not distinguish etween real returns and aniulated ones. Ramadorai ‘nvestor nterest and Hedge Fund Returns’ Eloing a new dataset of over , eressed deands for over hedge funds fro a seondar arket for hedge funds, this aer ﬁnds evidene that hedge fund investors rationall antiiate future hedge fund erforane. Both deand and sul indiations of interest arrive following eriods of fund outerforane. eand (sul) indiations foreast inreases (dereases) in strategadusted hedge fund returns over the suseuent ear, and large dollaraount indiations are etter foreasters than salldollar aount indiations. ndiationased aital alloation strategies are otentiall useful to realworld investors: the ield high alhas when ileented using alendartie ortfolios. ovanovi andenved Middleen in Liit-rder Markets’ ‘ liitorder arket enales an earl investor to trade with a late investor. But, uli news in the interarrival eriod reates adverseseletion ost and haers trade. high freuen trader (F) ight restore trade through its uniue ailit to uikl udate its uote on news arrival. But, F entr ight in itself worsen adverse seletion if seed is used to adversel selet investors’ uotes. his aer studies F entr oth theoretiall and eiriall. he entr of an Ffriendl new arket is used as an instruent. iddleen arrival oinides with a redution in the idask sread and a dro in volue. ang, ondor andSada ‘diosncratic Return olatilit in the Cross-Section of Stocks’ his aer unovers tie trends of idiosnrati volatilit in the rosssetion of stoks over the eriod . ighidiosnrati volatilit stoks have eoe ore volatile over tie while lowidiosnrati volatilit stoks have eoe less volatile, relative to total arket idiosnrati volatilit. he trends are roust to industr, liuidit, and sie, as well as sign of rie hange. e test whether atterns in idiosnrati ashﬂow volatilit, hedgefund assets under anageent, and overall institutional ownershi an elain the effet. he uward trend in highidiosnrati volatilit an e elained the inrease in idiosnrati ashﬂow volatilit, while the downward trend in lowidiosnrati volatilit an e attriuted to ongShortEuit hedgefund ativit. here is soe evidene that ongShortEuit funds at as liuidit deanders while nonongShortEuit funds and other institutions at as liuidit roviders. iand Sadka ‘Media and nvestent Manageent’ e lassif news ites aout euit hedge funds over into three soure grous: eneral newsaers, Seialied againes, and Cororate Couniation. ling a tetual analsis to news ites, we unover three tes of edia iases. First, a reorting stle ias, that is, when a fund is overed ultile soures at the sae tie, the sentient is ost ositive in Cororate overage and least in eneral overage. he differenes in soure sentient are ore signiﬁant in ases of elusive overage, indiating a seond ias, editorial seletion. Finall, eaining ostoverage erforane, we douent that Cororateovered funds outerfor and eneralovered funds undererfor, with a erfor ane differene of aout annuall. his result suggests a ontent ias, onsistent with fund anagers resenting onservative foreasts, while reorters resent relativel ore lenient views. owever, investor fund ﬂow does not reat to this inforation, whih suggests that investors do not see to eloit valuale inforation eedded in edia overage.
Chung andeo ‘Hedge Funds and nalst ptiis’ his stud elores the relationshi etween hedge fund euit holdings and sellside analst reoendations. e ﬁnd that analsts are ore likel to issue favorale reoen dations for stoks redoinantl owned hedge funds. oreover, these otiisti reoendations translate into oorer stok erforane over the net three to si onths. edge funds take advantage of these ﬂattering reorts onurrentl ofﬂoading their stok holdings. he ﬁndings are stronger for seasoned analsts and for high dollar turnover hedge funds. ur results suggest that analsts are relutant to downgrade stoks held their ost iortant lients. oen ‘Hedge Fund R hat’s under the Hood’ Returns of individual hedge funds generall feature low orrelation with oth risk fators and stle fators, onsistent with the goal of offering an alternative investent. e identif “low R” hedge funds for whih the otial fator odel R is insigniﬁantl different fro ero. hen live funds are sorted ased on alha, over of the to deile onsists of low R funds. roortional haard rate odel indiates that low R funds feature an inreased roailit of suseuent fund failure. hese results indiate that low R funds a ose an iortant risk to investors who selet funds ased on rior erforane. ragon andNanda ‘n ournaent Behavior in Hedge Funds High ater Marks Fund Liuidation and the Backﬁlling Bias’ e anale risk shifting oorl erforing hedge funds and test reditions on the etent to whih risk hoies are related to the fund’s inentive ontrat, risk of fund losure and disseination of erforane inforation. Consistent with theoretial arguents, we ﬁnd that the roensit for losing funds to inrease risk is signiﬁantl weaker aong those that tie the anager’s inentive a to the fund’s highwater ark () suggesting a ossile eneﬁt fro suh inentive strutures and aong funds that fae little iediate risk of liuidation. Risk shifting ehavior is affeted oth asolute and relative fund erforane and is found to e ore revalent in the akﬁlled eriod, when soe funds a e at an inuation stage. verall, the oination of fators suh as highwater ark rovisions, low risk of fund losure and the reorting of erforane to a dataase aear to ake oorl erforing funds ore onservative with regard to riskshifting. arolles, agliardini andGourierou ‘Survival of Hedge Funds Frailt vs Contagion’ he rather short lifeties of a aorit of hedge funds and the reasons of their liuidation elain the interest of investors and aadeis in hedge fund survival analsis. n this aer we are interested in the deendene etween liuidation risks of individual hedge funds. his deendene an result fro oon eogenous shoks (frailt) or e due to ontagion henoena when an endogenous ehaviour of a fund anager iats the Net sset alues of other funds. e introdue dnai odels ale to distinguish etween frailt and ontagion henoena, and to test for the resene and agnitude of suh deendene effets, aording to the age, anageent stle, registration ountr of the fund. iio, ao and elion ‘Hedge Funds ail Risk and Marginal Risk Contribution in Fund of Hedge Funds’ his aer studies hedge fund tail risk fro the investor oint of view, ore seiﬁall for funds of hedge funds. s funds of hedge funds riaril invest in hedge funds, we develo a risk anageent tehniue that atures the risk eosure of the individual hedge fund strategies and the tails risk ontriution of these hedge fund strategies toward the overall ortfolio risk, seiall in light of the and rises. e show that, ostl during risis, there is no sae to hide, all the different hedge fund strategies are ontriuting to the tail risk of the ortfolio of hedge funds, ostl eause of the eosure hedge fund strategies fae to liuidit and redit risk. Even the dediated short ias strateg is ontriuting to tail risk rather than reduing it during risis.
Critonand Saillet ‘Risk Factor Clustering Construction Fraeork for Funds of Hedge Fundsnsupervised ’ e suggest to lassif edge Funds ased on risks instead of returns. e use a ultifator odel and otain a set of estiated eta. hen, we al a lustering aroah to irove on edge Fund lassiﬁation. e show that our ethodolog outerfor standard lassiﬁation ased on returns. Sine edge Funds are often arket neutral, we assess the luster risk aling the ethodolog of atton and sine the erforane of a edge Fund is related to “alha», we assess the erforane of our luster aling the False isover Rate aroah of Barras, Saillet and erers. ith these ethodologies, we uild of Funds of edge Funds whih are ver erforative with a liited risk eosure. e show that the risk oniresent in edge Funds is harateried the interest rate risk and the redit sread risk. oreover we show that the ost aggressive lusters otained a igger roortion of true alha. ur reliinar results show that hedge funds aear to onsue liuidit rather than to rovide liuidit. t ties of low liuidit in the arket, hedge funds redue their euit holdings, and in artiular in illiuid stoks. his result is roust to liuidit deﬁnitions, oth at the arket level and at the stok level.
Cao, Chen, iang and o ‘Can Hedge Funds ie Market Liuidit’ his aer eaines how hedge funds anage their arket risk resonding to hanges in aggregate liuidit onditions. sing a large sale of euitoriented hedge funds during the eriod of –, we ﬁnd strong evidene that hedgefund anagers ossess the ailit to tie arket liuidit at oth the stle ategor level and the individual fund level. he inrease (derease) their ortfolios’ arket eosure when euitarket liuidit is high (low). his liuidit tiing ailit is asetri, and deends on arket liuidit onditions: hedge funds redue their ortfolios’ arket eosure signiﬁantl when arket liuidit is etreel low, ut the do not inrease their arket eosure when arket liuidit is unusuall high. Finall, we ﬁnd that investing in to liuidit tiing funds an generate eonoiall signiﬁant roﬁts. ur results ersist after ontrolling for alternative enhark odels, various data iases, and returntiing and volatilittiing ailities.
Chua,Hauand ai ‘uit Funds in the Financial Crisis PropagationRole of ’ he earl stage of the reent ﬁnanial risis was arked large value losses for ank stoks. his aer identiﬁes euit funds ost affeted this valuation shok and eaines its onseuenes for nonﬁnanial stoks owned the sae funds. e ﬁnd that stoks with a large ownershi share eosed funds eeriened onsideral lower risis returns relative oarale stoks in the sae industr and ountr. Based on a orehensive gloal fund level holding data, we also show that the risis was roagated to nonﬁnanial stoks through euit ﬁre sales the ost ank eosed euit funds. owever, noneosed euit funds ontriuted to ﬁnanial stailit as stoks with a large share of fund owners generall erfored uh etter throughout the risis.
Cella, Ellul andGiannetti ‘nvestors’ Horions and the pliﬁcation of Market Shocks’ fter negative shoks, investors with short trading horions are inlined or fored to sell their holdings to a larger etent than investors with longer trading horions. his a alif the effets of arketwide shoks on stok ries. e test the relevane of this ehanis eloiting the negative shok aused ehan Brothers’ ankrut in Seteer . Consistent with our oneture, we ﬁnd that shortter investors sell signiﬁantl ore than longter investors around and after the ehan Brothers’ ankrut. ost iortantl, we show that stoks held short ter institutional investors eeriene ore severe rie dros and larger rie reversals than those held longter investors. Sine the are otained after ontrolling for the stoks’ eosure to volatilit, various ﬁrs’ and investors’ harateristis, inluding the oentu effet or the roensit of institutional investors to follow an inde, our results annot e elained harateristis of the institutions’ investent stles other than their investent horions. e also show that the effet of shareholder trading horion eerges during other large arket delines. verall, the eirial evidene strongl indiates that investors’ short horions alif the effets of arketwide negative shoks.