P.

Joion-Risl( Nlanagement Lessons from I'TCM

14

In retmspect, the firm nade major slcnunagement bhurders LTCNI had rclied or recent history to estimate risk, srrrely assigning a zero probability io events slnrh as the Rrrssian clefarrlt and major market disruptions such a,s the 1987 crash, which also led to a
flight to liqrddity. Sincc credit_scnsitivc instrumcnts arc by Datruc affccted by rarc cvcnts, ihis approach misscd the tnlc sks of the portfolio Going back to Figrrrc 2, for cxamplc, it sccms cleal ihat the credit sprcads of the 1995_97 Period lvere low by histoical standards' Any slowclown in the economy \\'ould widen these spreads. Box 1 gives another exampie of a famorls hed8e frmd sPeculator lvho was catght short when the rece t data failed him' This type of strafegy exposes the portfolio to catastrophic dsk. Note how this interpre tation frmdanentally differs from LTC[4's vierv of the cvents of 1998. LTCM claims the strategy is frmdame tally sotrnd, aDd that the events of 199E were an abcrration lDstcad, the cxplanatioD here is that these "axbitrage' trades are doomcd to take largc losses orr e ilr
a while

ard that the truc dsks canrot be pi&ed trp

ea-sily by

VAR systems

BOX 1 Victor Niederhoffer: The Education of a Speculator Victor Niederlroffer outlined his investment Philosphy iu hls book Educatton o;f a Specrldior, r.hich qldckly bccame a best-seller' An ecccntri. ard blilliant investor,
hc rvas a legend of the hedge firnd brrsiness. Ltdeed he ha,J compilcd an outstanding track rccord a 32% compound annual leturn since 1982. Nicderhoffer's mission was to "apply scienrie" to the market. Althorrgh he had a PhD ir brNiness from the UniveNity of Chicago, he did not believe in efiicient markets and traded oD statistical alromalies. He believed, for iristance, that the market wor d nel,cr drop by more than 5 percent in a siDgle day. PuttiDg this theory into practice,

Nieclerhofier sold naked out of the-money plts on stock index firtlues When the stock nrarket plummeted by 7 pcrcerlt on October 27, 1997, Ile was ruable fo meet margin calls for some $50 nitlior. His brokers liqrridated the positions, wipirg oui his funds. Cornparirg his dovnfall to LTCNI'S bailout, he later wished he llad goticn "similar help." Apparently, his views wcrc nar:rorvly based on re(ent history' It is true that the volst loss had been 3.1% iD thc pre-vio]ls 5-year period. Larger loss do occur once in a v'hile, howe.r'er. Most notablS the market lost 20.4% on October 19, 1987

Admittedly, the losses werc exacerbated due to the sizc of the positions and the dlying ilp of liqrdditv in turbl ent firrancial markets. LTCXI atso claims that the selling of securities was made worse by brokeN "ftont-rurning" the LlCIvt portfolio'2 It is hard to verify these claims. Nlore findameDtally, the firm had badly underestimated its risk and did not have enough long term capital to ride out the iurbulence of 199E(1999). "Sc{: Le".is

Risk Management Lessons from Long-Term Capital Management

Phi,li,ppe Jorion

Draft: June

1999

Thanls are due to Neil Peaxon and

Nea.l Stoughton

for usetul comments

Corespondence can be addre"ssed to: Philippe Jo on, Graduate School of M&nagemnt, Unive$ity of California at lrvine,

Irvine, CA 92697-3125,
(949) 824-5245, (01999 P. Jorion

FAx: (949) 8248469,

F-mail: pjorion@uci.edu