RISK MANAGEMENTFinance

Inside this issueStringentcontrolsThe collapseof LehmanBrothersprompted hedgefunds tointroduceexactingpractices Page 2

Relying on one toolWe look at the use of Var, acontroversial measurementmethod Page 2

Correlation remains crucialA personal view by Paul J DaviesPage 2

Follow the debt The scale ofborrowing is auseful tool toexplain whyJapan went popin the 1980s,companies atthe end of the1990s, andbanks in the

past decade, while Greece tookon huge loans Page 3

Central counterparties Clearingreforms aim at OTC expansionPage 3

A real problem for regulatorsShadow banks provide specialchallenges Page 4

The poverty of controlWhy risk management providesmisleading comfort about riskyactivities and the true level ofexposure Page 4

Tech firmskey strokeHigh valuationsfor plannedflotations ofcompaniessuch asFacebookshould be aworry Page 4

Unknown unknownsSatyajit Das explainsthe perils of trying tocontrol the systemPage 4FINANCIAL TIMES SPECIAL REPORT | Tuesday March 22 2011

Brave new world of uncertainty

Financial risk managementbefore the crisis was the goosethat laid the golden eggs. But,to mix fables, in the cauldron

of 2008, that goose was well and trulycooked.

The highly technical, quantitativeside of financial risk management –and the financial theories that sup-ported it – came to dominate the disci-pline and was backed in earnest byregulators and policymakers.

And no wonder. It seemed to bringthe most efficient allocation of capitaland turbocharged the returns onequity for banks. It also apparentlydispersed credit and investment risksaround the world, so that a little localdifficulty ought never again lead toregional banks failing.

But when defaults on US mortgagesstarted to increase in 2006, the ele-gance and efficiency of the risk frame-work in place was found wanting.

“Lots of financial business had whatthey thought was an effective riskmanagement framework in place, butthe crisis showed that it did not dowhat it said on the tin,” says CliveMartin, an Ernst & Young partnerfocused on risk in financial services.

There were certainly problems withthe underlying quantitative modelsthemselves.

This was illustrated more thanamply in August 2007 in the famousquote from David Viniar, GoldmanSachs’s chief financial officer, aboutlosses taken by one of the banks’ equi-ty-focused “quant” funds. It had suf-fered market moves with probabilitiesso infinitesimally small that millionsof universes should not statisticallycontain them.

But there are three broad elementsto financial risk management.

There is the technical, which gov-erns all links in the chain from Valueat Risk models for individual tradersup to capital requirements for anindustry. Then there is the judgmen-tal, or qualitative, which should havemade its presence felt in the decisionsof individual traders, their immediatebosses and everyone else up to aninstitution’s board of directors and itsregulatory supervisors.

The third element is the systemic,which was entirely lacking in any for-mal sense before the crisis.

The technical side certainly becamedistorted and relied upon too heavily,but does that mean it was fundamen-tally wrong? The normal statisticaldistributions used in financial riskmodels have attracted heavy criticismfor placing too much faith in a rela-tively stable world and playing downthe chances of very bad events.

The stock market crash of 1987 andthe influence of futures markets inthose events led the Chicago-basedOptions Clearing Corporation to intro-duce a different statistical base – aLévy distribution – to manage the col-lateral risks for companies tradingderivatives, according to Donald Mac-Kenzie, an academic. The move betterenabled the clearing house to accountfor large, sudden market moves.While it made trading more costly, itmeant extreme “tail events” would beless disruptive.

Nothing so radical appears to behappening in the worlds of banking,insurance or asset management.Rather, the focus is on trying to main-

tain a shift in the balance of poweraway from highly technical practicestoward greater oversight and qualita-tive judgment. Or even more simplyto just a much bigger cushion ofsafety.

Hugo Bänziger, chief risk officer atDeutsche Bank, says that banks canonly be sure of being resilient with asignificant shock absorber in the formof tier one capital of 10 per cent.

“Technical risk management works,but only to a certain degree,” he says.“Bad risk management means you getwiped out at the beginning, but goodrisk management is not enough.”

However, Mr Martin of Ernst &Young reckons that arguments for alarge safety cushion are more abouthow to deal with a loss – not aboutthe risk of a loss in the first place.

Models are being improved throughtweaks and changes to make themmore reactive to operational controlsand more integrated with decision-making. “Scenario analysis is beingused a lot more alongside models toadd a sense check to their outputs,”he says. “There are also improve-ments in the quality of data and theassimilation of recent experience.”

Finance may have something tolearn from the pharmaceutical indus-try, according to Steven Culp, wholeads the global risk managementpractice for Accenture. Drug compa-nies learnt long ago about how toassess when to stop investing in aproduct at the research and develop-ment stage, he says. Financial serv-

ices are now beginning to learn thatthey should think more about thefuture profitability of certain businesslines than about modelling every sin-gle exposure and potential outcome.

“Rather than letting models leadthe way, institutions are now lookingto turn that quantitative talent to thetask of validating certain questions,”he says, adding that this is part of abroader conversation about risk.

“There is less of a view that riskassessment comes at the end of thechain to say yes – or no – to a trade orproduct line, but that it comes at thebeginning to set the direction.”

This is in line with the kind of re-commendations made by policyreviews such as that conducted in theUK by Sir David Walker on govern-ance and the role of risk managers.He recommended that the chief riskofficer be a board-level position and anumber of companies have alreadymoved in that direction.

The new European capital rules forinsurers due to come into force in 2013also contain provisions about howcompany managements and boardsmust demonstrate that they properlyunderstand their risk managementframeworks and properly apply themin making business decisions.

This will be much more importantin the future than it is now. At somepoint, markets and the economy willhave recovered and be moving intoanother boom, which will very likelycontain the seeds of its own downfall.Then, the pressure will be on to relaxrestrictions, trim capital bases andtighten up the parameters of models.

Paul Evans, chief executive of AxaUK, says the changes being boughtabout in insurance by the comingrisk-focused capital rules are alreadyleading companies to conduct properrisk analysis on their investments.

“One-in-200 year analysis is great,but if you think something is unthink-able, you won’t include it in youranalysis,” he says.

And this is where the importance ofmaintaining the balance with the

qualitative, judgmental side of riskoversight will be most crucial.

“It is the job of the board to ensurethere is that robustness behind quali-tative judgements,” Mr Evans adds.

However, judgement cannot func-tion welll if the financial infrastruc-ture of clearing and settlement, orlegal elements such as bankruptcyproceedures are not up to the task. Mr

Bänziger, says these are areas thatneed investment and suggests thatexisting levies could be useful if theywere ring-fenced for such investment.

“The principal reason we have traincrashes is a lack of investment in railinfrastructure – and the reason wehave systemic crises is a lack ofinvestment in financial infrastruc-ture.”

Paul J Davies explainswhy there is now a greaterunderstanding that there islittle guidance to be foundfrom the past whenpreparing for the future

The beginning of the end: when defaults on US mortgages started to increase in 2006, the risk framework in place was found wanting Getty

‘Financial businesseshad what they thoughtwas an effectiverisk managementframework in place butthe crisis showed it did notdo what it said on the tin’

Near­death experiencehas left deep scars

Of all the changes that haveswept through the financialindustry since the crisis, itis perhaps the area of riskmanagement where theshake-up has been mostacute – and nowhere moreso than at the world’s big-gest banks.

The alarming failure bybanks to spot unsustainablecredit risks building in thesystem in the years beforethe near-collapse of the sec-tor triggered a thoroughoverhaul of the way institu-tions predict and analysepotential problems.

Consultants say that inthe worst cases before thecrisis, banks’ boards treatedrisk management as some-thing of an afterthought toa potential deal or productlaunch – a hindrance totheir aggressive growthplans, rather than a pri-mary consideration.

At HBOS, one of the big-ger casualties of the finan-cial crisis, for example, theformer head of risk claimedhe was sacked for express-ing concerns over the prob-lems brewing.

Now, risk managementhas been repositionedfirmly at the forefront ofbanks’ operations. Chiefrisk officers have a moreprominent role, typicallywith a seat on the mainboard, and a stronger voicewithin their organisation.

Previously, these execu-tives often reported to thefinance director, which

meant their observationsmay not have filteredthrough to the most seniorexecutives. But they nowtend to have a direct line tothe chief executive and areinvolved in discussionsabout strategic decisionssuch as acquisitions orproduct developments froman early stage.

“There has definitelybeen a beefing up of therisk officer role,” saysVishal Vedi, a partner ofrisk and regulation atDeloitte. “There has been achange in reporting lines,the kind of informationthey are expected to pro-vide and who they provideit to.”

The changes follow far-reaching recommendationsfrom Sir David Walker, theCity grandee, about the roleof chief risk officers, whichincluded establishing “totalindependence” from the

business lines they monitor.New financial regulation

has to a certain extentforced banks to act: theirriskier activities now carrygreater constraints, such astougher capital and liquid-ity requirements.

Consultants say regula-tors have also been pushingparticularly hard on liquid-ity and stress testing.

“Boards are more focusedon stress testing,” saysPatricia Jackson at Ernst &Young. “There has been sig-nificant improvement inthis area but banks havequite a long way to go tomake risk management

BanksSharlene Gofflooks at howthe industryhas changed

HSBC has a new risk officer

Continued on Page 2

www.ft.com/risk­management­finance­march2011 | twitter.com/ftreports

2 ★ FINANCIAL TIMES TUESDAY MARCH 22 2011

Risk Management | Finance

Correlation remains crucial

A large solar flare erupts.Less than an hour later, ina busy, urban area, railnetwork controllers losetrack of trains, whileclocks and communicationsfail in automated signalboxes. Sat-navs in cars andlorries start givingerroneous positions,distracting drivers.

Most dangerous of all,though, is the interruptionto dispatchers of theemergency services.

The flare has causedsignificant disruption to asingle system: the satellite-based Global PositioningSystem. The RoyalAcademy of Engineeringwarned this month aboutthe ubiquity of this systemamong a dizzying list ofindustries and applications.It is cheap and efficientand has solved a lot of co-ordination problems. Butits very success means it isa single connection amongotherwise unrelatedactivities. Its failure couldbe catastrophic.

This is a problem ofcorrelation – and it is oneof the key problems of anefficient, modern society.

Misjudgments about, orfailures to spot, correlationin various forms wereamong the main causes ofthe recent financial crisis.The biggest miss wasmemorably summed up by

Andy Haldane, the Bank ofEngland’s head of financialstability, who comparedboom-era banks toTolstoy’s happy families:they were all alike.

Banks’ individual questfor diversity of earningsmade them ever moresimilar – and correlated.But expectations aboutcorrelation, however badlyunderstood, are still crucialacross financial riskmanagement.

There are a number oflessons relevant to thefinancial system from theRoyal Academy’s GPSwarning.

First, the problem ofsuccessful technologies.From railways to theinternet, new technologieshave inspired speculativebubbles and consequentcrashes. The over-exploitation of thetechnology of securitisationwas a primary cause of themost recent bubble. Bothits pervasiveness andoverblown ideas of itscapabilities caused a hugebuild-up of poorlyunderstood risk.

Second, the problem oflots of people doing thesame sort of thing in thesame sort of way. Oneresponse to the crisis hasbeen a significant overhaulof regulations. Thisoverhaul is not only aboutthe amount of capital thatinstitutions must hold.More importantly, thechanges look to define

more closely, and on amore global basis, the risksagainst which capital isheld – and how those risksare assessed and managed.

In a competitiveenvironment with a singletargeted outcome – thebest available return oncapital – and whereadvantage is uncoveredquickly, this seems likelyto result in more similarbehaviour among the biginstitutions that survivedthe crisis.

However, there are alsolessons in the differencesbetween GPS and finance.With GPS malfunctions, it

can be easy to spot bigerrors, for example if aship is shown to be manymiles inland and travellingnear the speed of sound.

But are such errors soeasy to detect in financialmarkets? The meaningcontained in the words“ship”, “inland” and “speedof sound”, makes theirconjunction absurd.However, the same cannotbe said of terms such as“mortgage”, “default” and“very high loss rates”.

The reason it cannot besaid is not because amortgage never defaults

(indeed, a ship can beinland – though it wouldnot be performing a ship-like function), but becausethe financial terms aresignificantly vaguer andtheir meaning can shift inthe short term.

Further, people react andadjust their behaviour toperceived changes in whatthese terms mean.Simplistically, a mortgagewas not a high-riskinvestment before late 2006– two years later, anymortgage seemed toxic.

The economist HymanMinksy said ideas of“liquidity” change withnew financial technologies.Innovation increasesefficiency and apparentlyboosts liquidity – at timesexponentially – withoutany gain in productivity orprofit. But the liquidityillusion can vanish easily.

The differences betweenfinance as a social activityand the physical sciencesare intuitively obvious.The odds on a 100-yearsolar flare do not changejust because people thinksuch a flare is more likely.

Correlation amongvarious systems connectedby use of GPS is relativelystraightforward to assessand counter once it hasbeen identified – itsriskiness is not affected byour awareness or lack of it.In finance, correlation isamong the most slippery ofrisks and remains alwaysjust beyond our grasp.

Personal ViewPAUL J DAVIES

Contributors

Paul J DaviesInsurance Correspondent

Jane CroftLaw Courts Correspondent

Telis DemosUS Markets Reporter

Sharlene GoffRetail BankingCorrespondent

Jeremy GrantEditor, FT Trading Room

Sam JonesHedge FundCorrespondent

Brooke MastersChief RegulationCorrespondent

Richard MilneCapital Markets Editor

Martin BriceCommissioning EditorSteven BirdDesignerAndy MearsPicture EditorFor advertising contact:Ceri Williams on:+44 (0)20 7873 6321;ceri.williams@ft.com

Traumaleavesdeepscars

quicker and more inte-grated.”

Although the overhaul ofmonitoring by banks hasbeen driven by the regula-tory agenda, the experienceof losing billions of poundsby failing to spot the dan-gers of excessive risk hasprompted them – particu-larly those bailed out by thestate – to recognise theimportance of restructuringthis part of their business.

A recent survey by Hed-ley May, the executivesearch firm, showed nineFTSE-listed financial serv-ices companies, includingHSBC and Prudential,appointed new chief riskofficers in the past year.

The heightened impor-tance of the role meansbanks are also looking toattract a higher calibre ofindividual.

Chief risk officers areexpected to be more proac-tive – spotting problems inadvance rather than dealingwith them as they arise – aswell as more involved instrategic decisions from anearly stage and willing tochallenge those responsiblefor making them.

Consultants say one bigfailing in risk managementbefore the financial crisiswas the quality of informa-tion collected and passed onto senior management.

Banks made sweepingassumptions based on datathey collected about, forexample, the ability of cus-tomers to repay debt, andwere slow to react whenproblems, such as loanimpairments, surfaced.

“Without exception, allbanks have done some anal-ysis of the quality of dataand the integrity and accu-racy of processing thatdata. All found to somedegree a need for improve-ment,” says Andrew Gray,UK banking leader at PwC.

While the problems didnot tend to reflect a lack ofinformation – in fact con-sultants say risk officerswere frequently producingtoo much data – the qualityand transparency was poor.

“It was difficult to navi-gate to where the problemswere,” adds Mr Gray.“Work has been done toimprove the quality of data,to make it more accurate,granular and predictive.”

Banks are now looking tocompile more succinct butmore useful data that betteranalyses exposure to risks.

In particular, biggergroups such as globalinvestment banks havebeen examining how toprice illiquid assets andmanage the risks of individ-ual businesses and marketsrather than across the insti-tution as a whole.

However, while progresshas been made, consultantssay banks have more workto do. Ms Jackson at E&Ysays a significant culturalchange is needed for banksto move away from thesales-driven environmentthat permeated organisa-tions in the boom years.

“Now risk officers have aseat at the table when itcomes to product develop-ment or strategy. But thatdoesn’t stop it swingingback again in the nextboom to a more sales-drivenculture,” she says.

Continued from Page 1

People doingthe same sortof thing inthe same sortof way is aproblem

The danger of relying too much on only one tool

Since the near collapse ofthe banking system morethan two years ago, the roleplayed by value at risk(Var) – used by banks as ameasurement too – has trig-gered fierce debate.

Supporters maintain Varwas only ever intended tobe used with measures suchas scenario testing andstress testing.

Var was developed toshow risks that are run ona daily basis by banks. Themetric was developed, butnot originated, at JPMorganin the early 1990s inresponse to a demand fromthen chairman Dennis

Weatherstone to be told at4.15pm every day the bank’srisk.

He was concerned tradingdesks and business unitsmight be taking correlatedbets without knowing it.

After the market crash ofOctober 1987, value at riskbecame a management toolin financial firms and sincethen 200 books – almost onea month – have been pub-lished on the subject.

At the height of theboom, so much reliance wasplaced on it that bankstrumpeted Var numbers.But the crisis badly shookconfidence in it.

Indeed, Pablo Triana, for-merly a derivatives traderand academic at the Univer-sity of Madrid, argues in hisbook Lecturing Birds onFlying that financial modelssuch as Var have donemore harm than good.

“I blame mathematicalmodel value at risk for thecredit crisis,” he wrote in

2009, arguing that as “a con-struct that borrows frompast data and from im-proper probabilistic as-sumptions, Var can danger-ously deliver numbers thatare too low for comfort”.

But three years after thecrisis, Var is still beingused – albeit with greaterawareness of its limitations.

Simon Bray-Stacey, headof investment risk for AvivaInvestors London, says:“When Var was promotedby JPMorgan all thoseyears ago, I think the ideaof having a single risk sta-tistic was alluring for eve-ryone, particularly seniormanagement.

“Although the limitationsand assumptions werealways kept in mind by riskprofessionals, it became lessobvious that Var was not acure-all for risk problems.

“Var is still being used asa tool by institutions and itis still very useful for meas-uring exposures.

“But the assumptionsaround the way Var is pro-duced and the concept thatit is only a single point ofthe distribution of returnshas to be borne in mind.”

Certainly Var can provideuseful data. Its figures canbe produced very rapidly

and monitored virtuallyreal-time – crucial whenmanaging complex posi-tions in volatile markets.

Var is also seen as bestsuited to instruments, suchas equities, that displaydaily changes in risk –although, at the height ofthe markets boom, Var was

applied to other instru-ments such as credit.

Miles Kennedy, partner atPwC, says: “Var was a use-ful measure then and it’s auseful measure now, but onits own it can be mislead-ing. The industry got a hostof things wrong and onewas the over-reliance on asingle measure or set ofmeasures, including Var.

“The wrong response is todispense with Var. Theright response is to under-stand it for what it is andsupplement it with otherinformation, such as usingstress testing or scenarioanalysis, to give a morecomplete picture of risk.”

Part of the problem isthat the Var model dependson the data fed into it: itassumes tomorrow will bebroadly similar to today.

Benign economic datafrom before the crisis mayhave presented a rosier pic-ture than was the case.

Mathematical sophistica-

tion of models may alsohave given executives afalse sense of security aboutits accuracy.

Mr Kennedy of PwCpoints out: “A measure ofrisk driven by historicaldata assumes the futurewill follow the pattern ofthe past. You need tounderstand the limitationsof that assumption. Moreimportantly, you need tomodel scenarios in whichthat pattern breaks down.

“In financial services,there is a tendency to placegreater confidence in riskinformation that is data-driven, in the belief thatthis confers objectivity andtruth.

“Objectivity is fine, but itdoesn’t equate to truth. Ithas to be remembered thatrisk is about the future, andthere are no facts about thefuture. Var gives a usefulindication of what thefuture may hold, but nomore.”

Banks are now likely tolook at stressed Var, whichlooks at how the positionchanges in extreme stressand improves transparency.

Mr Bray-Stacey at Avivasays that professionals nowlook at other measures toprovide a much clearer pic-ture of risk.

“To say you can sum uprisk in a portfolio or aninvestment bank in onemeasure is quite daft,” hesaid.

“Now people look at othermeasures besides Var, suchas stress testing and thiscan bolster the understand-ing of the tails of the returndistribution.

“Something else we use islooking at different his-torical periods as modelinputs.

“From a risk manage-ment standpoint it is goodthat people are more waryof Var. It is not the be-alland end-all.” Mr Bray-Stacey adds.

Value at riskJane Croft analysesthe use of acontroversialmeasurementmethod

‘To say you cansum up the risk ina portfolio or aninvestment bankin one measureis quite daft’

Stringentcontrols onlosses andinvestment

Try though they might –perhaps rarely very hard– hedge funds have nevershaken the image of

being high-rolling, high-octane,high-stakes market gamblersthat they won after 1998, whenthe collapse of the fund LTCMcame close to triggering a WallStreet-wide panic.

Since that debacle, there hasbeen a hedge fund blow-up moreor less every couple of years tokeep that perception alive.

It is, however, a damaging onefor an industry that prides itself,above all else, on risk manage-ment. Hedge funds – in theory atleast – are low risk. They usetrading strategies and instru-ments such as derivatives tohedge away the multitude ofrisks that playing the marketssupposedly entails so they canmake money with a minimalchance of losing it.

Estimations of hedge fund lev-erage – the money borrowed tojuice up investment returns on

client capital – vary, but mostput it at between one and twotimes for the average hedgefund.

“If anything, leverage hasfallen quite significantly,” saysAnthony Kirby, director of regu-latory and risk management atErnst & Young. Even at its peakin 2007, the average hedge fundprobably used leverage of aboutthree times, at most.

Hedge funds have perhapsunfairly been “clobbered with ahigh degree of regulatory over-sight,” says Mr Kirby. Althoughmany hedge funds – thanks pri-marily to their typically smallsize – do not deploy the samegold-plated risk practices asbanks and big asset managers –they are far from being free-wheeling.

Indeed, big hedge funds havesome of the most sophisticatedand exacting risk managementpractices anywhere in asset man-agement.

Brevan Howard, for example,Europe’s largest hedge fund, is abyword for caution and savvy ininvesting, even if risk is appar-ent concentrated at the firm inthe hands of a very smallnumber of people. While Brevanemploys hundreds, an estimated60 per cent of the $36bn it man-ages is traded by Alan Howard,its founder, alone. Another 30

per cent of the assets under man-agement are controlled by thesmall team of about seven toptraders around him.

The fund has precise limits onacceptable loss levels, however.Anything more than 4 per centresults in a visit to the firm’schief risk officer, Aron Landy.An 8 per cent fall leads to a

trader’s money being cut. A 12per cent fall is likely to lead to atimeout and them being shownthe door. There are no excep-tions.

The firm is not alone. Mostlarge hedge fund managersemploy similarly stringent risklimits and investment controls.

The reality for most is that theysimply do not have the luxury oftaking big risks – investors can,and do, pull money fast in theevent of losses.

To boot, “founder syndrome” –where risk management is oftena function of the top-trader orfounder’s own views – is rarely,if ever, an issue for big fundsthese days, says Mr Kirby. “Theidea of the founder chasing alphaat the expense of risk complianceis very much fading,” he says.

If anything, the events of 2008have only led managers to havemore conviction in the kind ofrisk management proceduresthat began to be put in placepre-crisis.

“We saw headcount in risk andcompliance rise quite signifi-cantly in 2009,” says Mr Kirby.

In many ways it is little won-der why. Fund managers havebecome increasingly institution-alised as they chase more institu-tional money.

Investments from pensionfunds or insurance companies

are prized because they tend tobe “stickier” and less prone towithdrawal in times of trouble,but they also come with greaterdemand for transparency andsolid, risk management proce-dures.

“Institutional investors aren’tnecessarily interested in the spe-cifics of the quantitative riskmanagement models we use,”says the chief risk officer of onelarge multibillion-dollar fund,“But they want to know that wehave rules we follow and inter-nal procedures, and people topolice them.”

Since 2008, investors have beendigging much deeper into theoperational side of the hedgefund business. Due diligenceprocesses now involve questionsabout the prime brokerages thathedge funds use, the arrange-ments for custody of their assets,and the structures of the fundsthemselves.

The collapse of Lehman Broth-ers, the US investment bank, inparticular alerted many inves-

tors – and hedge fund managersthemselves – to counterpartyrisks at the portfolio and busi-ness level.

Even top-flight managers suchas GLG Partners were forced towait for years before they wereable to claim back assets fromthe collapsed bank, thanks to alack of understanding about theway Lehman sequestered assetsfrom its prime brokerage unit inLondon back to New York on anightly basis.

Testament to the changes inthe industry is the growing seri-ousness with which the HedgeFund Standards Board – a self-regulatory body set up in 2008 –is being taken. The standards,which cover everything frommarketing to risk management,are increasingly coming to beseen as the industry’s bench-mark.

As money flows back andmemories of 2008 fade and inves-tors’ appetite for high returnsgrows., it remains to be seenwhether the rules stick.

Hedge fundsSam Jones says thatthe sector has some ofthe most sophisticatedand exacting practices Box clever: the collapse of Lehman Brothers alerted investors to counterparty risks at the portfolio and business level AP

‘Institutional investorswant to know we haverules and internalprocedures, andpeople to police them’

FINANCIAL TIMES TUESDAY MARCH 22 2011 ★ 3

Risk Management | Finance

Central counterparties eyea wave of opportunities

Hong Kong is not an obvi-ous place to start if you arelooking for evidence thatthe G20 reforms aimed atcleaning up the financialsystem, post-crisis, haveprompted profound changesin the way the marketsfunction.

But last month HongKong Exchanges and Clear-ing (HKEx) laid out ambi-tions for expanding into theprovision of clearing serv-ices for over-the-counter(OTC) derivatives.

It plans to establish anOTC clearing house by theend of 2012 – possibly inconjunction with partners –“to support global regula-tory initiatives and takeadvantage of businessopportunities in OTC deriv-atives clearing”.

The move showed thatreforms enshrined in theDodd-Frank act, passed lastJuly by the US administra-tion, are pushing clearingto the forefront of theagenda of many in thefinancial markets wellbeyond the US and Europe.

A clearing house standsbetween parties to a trade,taking on the financial riskif one party defaults. It usesfunds posted by members ofthe clearing house – knownas margin, or collateral – toensure deals are completedin the event of default.

Clearing houses, alsoknown as central counter-parties (CCPs), are requiredunder Dodd-Frank to acceptfor clearing swathes ofover-the-counter (OTC)derivatives, such as theinterest rate swaps thatHKEx plans to clear.

These contracts used notto be cleared. The banksthat bought and sold themfrom each other trustedeach other’s creditworthi-ness when it came toassessing whether the otherside might default. The lackof a safeguard such as aCCP was starkly exposed

when Lehman Brothersdefaulted in September2008, since many counter-parties were left holdingopen positions with thefailed bank.

The Dodd-Frank act andsimilar reforms in Europeknown by the acronym“Emir” are set to change allthat by trying to ensurethat as many OTC deriva-tives go through clearinghouses as possible.

That means not only thatCCPs see a business oppor-tunity in handling this newwave of derivatives. It hasalso raised concerns amongregulators and centralbanks that pushing moreactivity into CCPs could beconcentrating risks intothese institutions, makingthem a single point of fail-ure.

In its Emir proposal, theEuropean Commission says:“In view of their systemi-cally important role and inview of the proposed legisla-tive requirement to clear all‘standardised’ OTC deriva-tives through CCPs, theneed to subject them tostrict prudential regulationat EU-level cannot be over-emphasised.”

For central banks, thequestion is whether, and towhat extent they shouldand can provide a backstopto CCPs, should any ofthem get into difficulty. Noone yet has a clear answer.

Regulators and centralbankers have been workingto ensure that, at least, thestandards under which

CCPs are governed and runare as far as possible har-monised globally. The sameapplies to ensuring robustrisk management, which isa clearing house’s mostimportant function.

The concern is to avoid asituation where clearers’commercial priorities incompeting with each otherover the spoils of OTCderivatives might temptthem into relaxing thefinancial requirements theyhave of their members inorder to attract more busi-ness.

Regulators have tried

twice before – in 2001 and2004 – to put in place a setof harmonised standards forCCPs. But the G20 reforms,and the eagerness of mar-kets in Asia – principallyHong Kong, Japan and Sin-gapore – to build their ownOTC clearing infrastruc-tures, have made it moreimportant that regulatorssucceed this time.

The fear is that otherwisea patchwork of standardswill emerge, making it hardfor regulators to get a holis-tic view of risk at CCPs.

This month the Bank for

International Settlements(BIS) proposed in a report“new and more demandingstandards” for payment,clearing and settlement sys-tems.

“A [clearing house]should maintain additionalfinancial resources, such asadditional collateral or apre-funded default arrange-ment to cover credit expo-sures from participantdefaults in extreme butplausible market condi-tions,” it said.

The BIS report, compiledby the its Committee onPayment and SettlementSystems and the Interna-tional Organisation of Secu-rities Commissions, sets outa set of proposed principlesfor “financial market infra-structures (FMIs)”.

It is the first attemptsince the financial crisis toestablish global standardsfor governance, funding andmanagement of post-tradestructures such as CCPs.

However, users of CCPs,such as banks, are anxiousthat there is not just afocus on making clearinghouses financially morerobust – not least becausethe cost of doing so willultimately fall on users.

Stephen Burton, directorat the Association forFinancial Markets inEurope, says: “Our mem-bers want CCPs to continueto apply robust and sophis-ticated risk managementpolicies rather than simplyapplying additional collat-eral or increasing defaultarrangements.”

Simon Gleeson, UK part-ner at Clifford Chance, callsthe BIS paper “a missedopportunity”.

He notes it does not men-tion a separate Decemberpaper by the BIS aboutbanks’ exposures to CCPs.In that paper, banks aresupposed to calculate theircapital requirements for theexposures by first figuringout how much capital theclearing house would needif it were a bank.

But the Iosco paper usesan entirely different calcu-lation. “There doesn’t seemto have been a lot ofjoined-up thinking,” hesays.

ClearingReforms aim atOTC expansion,says Jeremy Grant

‘Our memberswant CCPs toapply robust andsophisticated riskmanagementpolicies’

Togetherness: HKEx plans to move into clearing Reuters

Follow the line of debt to spot a coming crisis

Mark Thomas, a busi-ness strategy spe-cialist at PA Con-sulting, has a single

sheet of paper he shows to cli-ents that could strike fear intoany chief executive, investor orregulator.

It shows a radar screen withthe dozen or so events he thinkscould over the next decadepotentially be the next landminefor the global economy.

Broken into four areas repre-senting risks in the consumer,corporate, banking or govern-ment sector, the chart showsworries stretching from a com-modity price shock to the healthof financial institutions in theeurozone and China.

But many of the biggest“bubbles” on the radar screenare located in the governmentsector.

Few countries are spared frombeing a potential problem:alongside the usual suspects ofthe peripheral eurozone coun-tries of Greece, Ireland, Portu-gal, Spain and Italy, Mr Thomashas placed Japan, the US, theUK and even China.

“There has been a huge trans-fer on to government balancesheets because of the crisis,” hesays.

As regulators and companiespeer on to their own personalradar screens to see what thenext bubble might be, theoptions can seem bewildering.

Even when it can be identi-fied, working out when it willpop is almost impossible. Theprevious bubble was no differ-ent. While several economistswarned of the dangers of sub-prime lending in the US or prop-erty prices in Spain, many ofthem did so for years before theevent.

Only four years after thefinancial crisis first broke, thedanger of complacency isalready back.

Analysts at Citi noted in earlyMarch how deep out-of-the-

money equity options – whichwould make large amounts ofmoney only if stock markets fellprecipitously – were at theircheapest since before the col-lapse of Lehman Brothers in2008. Steven Englander, a cur-rency strategist, noted signs of“Black Swan fatigue” – weari-ness with the idea that anextremely unlikely event could

take place and cause havoc inthe markets.

Difficult as it is to spot thenext bubble, some strategistsand investors say the answershould be relatively simple: fol-low the debt.

“It is very easy to tell wherethe next bubble is forming: it iswhichever sector is taking onthe most debt. You can get

bubbles that are not associatedwith debt, but nobody reallyworries when they burst,” saysMatt King, head of creditresearch at Citi.

Under his analysis, debt helpsexplain why Japan went pop inthe 1980s, companies at the endof the 1990s, and banks in thepast decade.

Now the debt – as demon-

strated by the eurozone crisis –is all at the government level.

For some analysts, this repre-sents potentially the final act ina decade-long drama that hasseen companies, consumers andbanks all have debt problems.

“If you think all the bubblesof the past 10-15 years were con-nected, then government bondsare the last shoe to drop,” says

Jim Reid, credit strategist atDeutsche Bank. “Governmentsare the last chain in the rollingsupercycle of bubbles.”

The peripheral eurozone coun-tries look most immediatelychallenged, and their problemsalso fuel concerns about banksin France and Germany.

But as Mr Thomas’ radarscreen demonstrates, few coun-tries can be excluded.

Some investors argue thatJapan, the UK and the USshould be excluded as they canstart the printing presses, butthe potential for intense con-cerns about the level of theirdebt in the coming yearsremains. “It is all a big confi-dence trick. It has worked in theUS so far. But the debt is stillthere and that leverage createsmuch more downside risk andfragility than people think,”says Mr King.

Mr Thomas even believesChina could be at risk, joiningsome hedge fund managers infretting about the fast-growingeconomy. His argument is thatinvestment in infrastructure issuch a big part of growth ratesthat, coupled with the diffi-culty in trusting economic dataand a belief that it is a new erafor China, makes a bubble a pos-sibility.

So what should people doabout potential bubbles? Ignorethem or worry so much they donothing?

Mr Thomas argues the firstthing is simply to acknowledgethey exist.

“Don’t run your business onthe basis that everything will beperfect. Some landmines willexplode,” he says.

But just because a bubblemight be brewing in, say, China,it does not mean investorsshould shun the country.

He adds: “If China is a bubbleyou don’t want to invest in Chi-nese real estate. But you mighttake a long-term view thatunless something goes funda-mentally wrong it should still bea good growth story, albeit notquite as good as in the recentpast, for the next 20 years.”

The next bubbleRichard Milneexplains why largeincreases ingovernment balancesheets are worrying

The Jianwai SOHO complex inBeijing: some fear China maybe entering a bubble becausespending on such infrastructureis a large part of its economicgrowth Reuters

There are signs of‘Black Swan fatigue’ –weariness with theidea that an extremelyunlikely event couldcause havoc

4 ★ FINANCIAL TIMES TUESDAY MARCH 22 2011

Risk Management | Finance

The poverty of theentire idea of control

The late Peter Bernstein inAgainst The Gods: TheRemarkable History ofRisk, argued thatcapitalism would beimpossible withoutquantification and hedgingof risk.

He wrote: “Therevolutionary idea thatdefines the boundarybetween modern times andthe past is the mastery ofrisk: the notion that thefuture is more than awhim of the gods and thatmen and women are notpassive before nature.”

Unfortunately, successivecrises, beginning with thefailure of portfolioinsurance in the 1987 stockmarket crash andculminating in the globalfinancial crisis of 2008,have illustrated thepoverty of riskmanagement technology.

The flawed empiricismunderlying riskmeasurement washighlighted in August 2007by David Viniar, GoldmanSachs’s CFO: “We wereseeing things that were 25-standard-deviation moves,several days in a row.”

When in October 2008,the Dow Jones IndustrialAverage moved more than10 per cent on two days,economists Paul DeGrauwe, Leonardo Ianiaand Pablo RoviraKaltwasser used a normaldistribution to estimatethat such moves shouldoccur only every 73 to 603trillion billion years –

making this “a trulymiraculous event”.

Following the crisis, riskmanagers and regulatorshave recalibrated theirabacuses. The changes arerefinements of existingapproaches rather thanany fundamental shift. Thereason offered is that thereare no “real” alternatives.

Problems remain. Allrisk management startswith accurate valuations ofpositions. But there aredisagreements aboutstandard models forvaluing even conventionalderivatives, such asinterest rate and currencyswaps. For exotic products,these problems are greater.

Key changes intechnology have entailedusing higher volatility,more rigorous correlationassumptions and alsostress tests to supplementrisk measurement. But thevolatility of volatility hasalso increased. Large shiftsin correlations betweendifferent assets – or“regime changes” – arenow an increasing factor inmarkets. All this makesrisk measurement andmanagement more difficult.

The problem is thatmethodologies still do notrecognise that the real riskin markets is driven byboth external events(government policy or oilprice shocks for example)and, increasingly, thestructure of markets andtrading themselves.

The design of markets,flawed regulatory regimesand large, cross-bordercapital flows contributegreatly to risk. Tightly

coupled markets withcomplex linkages betweenparticipants createcomplexity andinterdependence.

Credit enhancementtechniques, primarily theuse of collateral, facilitategreater participation intrading and higherleverage. But their effectson the demands for cashand the need to liquidatepositions can exacerbateprice moves and volatility.

The arcane effects ofhighly technicaldocumentation andoperational risks are notadequately captured byrisk systems.

Consolidation within thefinancial services industryhas created higherconcentration of trading

and risk among a smallgroup of large dealers.

In theory, participants,such as investors andhedge funds, provideliquidity and help disperserisk. In reality, tradingstrategies of key playersare often poorly diversified.They involve large bets onthe same event usingdifferent instruments,causing higher volatilityunder certain conditions.

The roles of largemarquee hedge funds andprime brokers (who finance

these hedge funds) alsocreate significant riskconcentrations.

The models used toprice, risk-manage andvalue instrumentsfrequently do not capturethe underlying marketdynamics. Assumptionsabout trading, liquidity andfunding are unsustainable.The use of broadly similarrisk models createsdangerous feedback loops.

Trading behaviour andtrader interactions are alsopoorly understood. Moralhazards are prominent incompensation systems fortraders. As some tradersrecognise, “there is a newrisk factor – and it is us.”

Many of the real riskelements do not lendthemselves to fancifulmathematical modellingand the exactness belovedof risk mangers. In lovewith the beauty of theirmodels, risk professionalsare reluctant to admit theflaws in quantitative riskmanagement. Recognitionof the difference betweenrisk and pure uncertaintyand the need for morequalitative measures islargely rejected as Luddite.

The lack of progress ispredictable. Financialmarkets, investors andtraders have becomewedded to increased risk-

taking as an importantsource of profit.

Risk managementcontinues to be the figleafbehind which seniormanagers, directors andregulators shelter.

Testifying before the USFinancial Crisis InquiryCommission, Sandy Weill,the former Citigroupchairman, provided insightinto the practices of majorbanks. “If you lookat . . . what happened onWall Street, it became,‘well, this one’s doing it, sohow can I not do it – if Idon’t do it, then people aregoing to leave my placeand go some place else’.”

Risk managementcontinues to provideimprecise and misleadingcomfort to financialinstitutions and regulatorsabout risky activities andthe true level of exposure.

Risk managementincreasingly exposes thefirms and, ultimately thewhole economy, to the sumof all our fears.

Satyajit Das is author ofExtreme Money: TheMasters of the Universe andthe Cult of Risk (to appearin September) and Traders,Guns & Money: Knownsand Unknowns in theDazzling World ofDerivatives

Guest ColumnSATYAJIT DAS

’Risk managers and regulators recalibrated their abacuses’

A real problemfor regulators

Now that regulators have moved toimpose tougher capital and liquidityrequirements on banks, attention isturning to other sources of systemicrisk, especially fast-growing entitiesmuscling in on bank business thathave escaped the same level of scru-tiny.

The $16,000bn “shadow banking“system includes everything fromhedge funds and private equity tomoney market funds, clearing housesand special-purpose vehicles that holdcomplex securities. Though it hasshrunk somewhat since the 2008financial crisis, it is still larger thanthe $13bn banking system proper,according to research by the FederalReserve Bank of New York.

These shadow banks have taken onpart or all of the maturity transforma-tion role of banks – matching shortterm depositor funds with long-termlending – and much of the attendantrisk. But they do not have the samerequirements to hold capital and liq-uid assets against losses or a rash ofcustomer withdrawals or failures.

This spring, the Financial StabilityBoard, a group of global regulatorsand central bankers, is due to issue areport and come up with recommen-dations on how the world’s largesteconomies can get a better handle onthe risks shadow banking systemposes. The US has a new FinancialStability Oversight Council chargedwith identifying and regulatingshadow banks that have become dan-gerously important and the new UKFinancial Policy Committee will havea similar mandate.

In one measure of the regulators’determination to get to grips with theshadow sector, the FSA started run-ning a twice a year survey of hedgefunds to assess the risks they pose tothe broader system. The most recentfound that they play relatively largeroles in the convertible bond, interestrate and commodity derivatives mar-kets but are otherwise relatively smallforces that pose little risk.

The watchdog noted the vast major-ity of hedge fund borrowing is withfive banks, suggesting concentrationissues, but noted with approval thosebanks have tightened lending termsand are holding more capital.

Other parts of shadow banking arelikely to receive similar demands forinformation; US and European moneymarket funds are a likely target. Cen-tral counterparties – particularlyclearing houses – are also potentialtargets for regulation because theyare critical to day-to-day trading.

Michael Raffan, partner at Fresh-fields, says: “Central counterpartiesare the mother of too-big-to-fail prob-lems, because, if one goes, it is likelyto take the system down.”

Global regulatory groups recentlyissued a report urging governments toset “new and more demanding stand-

ards” for payment, clearing and settle-ment systems – the post-trade infra-structure that underpins markets.

“A [clearing house] should maintainadditional financial resources . . . tocover credit exposures from partici-pant defaults in extreme but plausiblemarket conditions,” according to thereport from the International Organi-sation of Securities Commissions andthe Bank for International Settle-ments. That suggests banks wouldhave to hold more collateral.

Analysts point out that movingrisky activities such as proprietarytrading and high-risk lending out ofthe banking sector proper is not sucha bad thing, as long as regulators actto limit the impact of any failures onthe larger system.

“In many ways, transferring higherrisk activities into vehicles wheredepositors’ money is not at risk is atthe core of the current wave of bank-ing regulation. However the real chal-lenge will be . . . systemic risks thatare detected too late. For examplehow many LTCM ‘time bombs’ willexist in the shadow banking system,”says Bill Michael, UK head of finan-cial services at KPMG

But some analysts doubt that globalauthorities will be able to regulatealternatives to banks. They say impos-ing tighter controls on banks andtheir near competitors will drive bor-rowers to less regulated sources.

Simon Gleeson, UK partner at Clif-ford Chance, the law firm, says:“Credit is like internet content. It canbe created by anyone anywhere.Imposing restrictions on bank creditcreation would be like trying to con-trol the size of the internet by impos-ing word limits on newspaper web-sites.

“Not only would it be unsuccessful,it would result in less good contentand more bad content.”

He adds: “Since banks are requiredto cover the costs of financing pluscapital requirements, while shadowproviders only have to cover cost offinancing, it is likely that the shadowbanking system will provide credit atlower prices than banks – thus a pol-icy aimed at reducing credit growthmay end up stimulating it.”

With financiers around the globeitching to exploit loopholes in regula-tion, analysts say, the best that regu-lators can hope for is to contain,rather than eliminate, the dangerposed by the unregulated part of thefinancial sector

“You can move the risk around, butyou can’t get rid of it,” says Mr Raf-fan. “Better to set the [regulatory]perimeter at a reasonable place andmonitor it to protect what is inside.

“If you get it right, the failure ofsomething outside it shouldn’t havesystemic implications.”

Shadow banksBrooke Masters explainswhy the sector providesunique challenges

‘Centralcounterparties arethe mother oftoo­big­to­failproblems’Michael Raffan,partner atFreshfields

Tech firms to unveil key stroke

Around of hot technology com-panies — names such asFacebook, Groupon andLinkedIn — set to go public

makes some investors very happy.But it gives others a few bad flash-backs.

A study of the US initial publicoffering pipeline by Ernst & Youngfound 150 companies waiting to gopublic as of February, seeking to raise$42bn. Of those, 22 per cent were tech-nology companies backed by venturecapital firms.

Typically, those issues launch afterthe end of the first quarter, whenauditors can thoroughly scrutinise thebooks of younger companies.

“Assuming markets are robust,we’re about to enter a period ofincredible activity on smaller deals,”said Mark Hantho, global co-head ofequity capital markets at DeutscheBank.

In some respects, this is a veryhealthy sign: investors are confidentin the long term, and are willing totake a chance on riskier companiesthat could deliver huge returns. TheFTSE Renaissance IPO Compositeindex rose 20.3 per cent last year, ver-sus 12.8 per cent for the S&P 500.

The FTSE Renaissance index holds

companies for two years after theyhave gone public.

That is a significant turnround from2008, when investors were dumpingsuch groups. The index fell 50 per centthat year, versus a 23 per cent drop inthe S&P.

“The bar for going public continuesto come down. When this market firstreopened, people asked, what’s theprice-to-earnings multiple – is it prof-itable?

“But right now, investors are will-ing to look further out, and are moreconcerned about growth than profita-bility,” says Will Bowmer, head oftechnology equity capital markets atBarclays Capital.

“Fund manager who want to takemore risk are going straight to theIPO market.”

But it also raises the spectre of atechnology bubble. Memories of theIPO-crazed late 1990s, when cab ridesand backyard barbecue parties werefilled with talk of “hot stocks”, arestill fresh for many investors.

“It’s a Web 2.0 craze and bubblemania, no question,” says HaroldBradley, chief investment officer atthe Kauffman Foundation, whichworks to foster entrepreneurship.

“We’ve gone from the ‘eyeballs’craze to the ‘viral’ craze. These com-panies are marginally profitable butare supported by huge valuations.”

Certainly, one can be scepticalabout the estimated valuations for thelikes of Facebook, put at more than$70bn, if deals to sell its shares onprivate markets are reliable, or Twit-ter, valued at $4.5bn by a privateJPMorgan Chase fund, or Groupon,said to be valued at $5bn by a take-over overture from Google.

Many people who frequently deal insuch companies are a doubtful aboutthose eye-popping numbers. “It’s hardto know who’s buying these shares,and for what reasons,” says a seniorWall Street banker. Barry Diller, aseasoned technology investor,recently called those valuations“insane”.

However, that is not diminishingbankers’ enthusiasm for this group ofcompanies. Many of them see a silverlining in the timing of the financialcrisis for this group technology com-panies: they had more time to growand mature.

“The average length of a companyin our pipeline waiting to go public isseven to eight years. That’s longer

than the four or five years we wereseeing earlier in this decade, and theincredibly short turnround during thedotcom boom,” says Jackie Kelley,Americas IPO leader at Ernst &Young.

In the past year, she says, a strikingnumber of companies have come toher practice for pre-public training inhow to talk to investors, deal withauditors and publish regular reports.“These companies, as a group, are

going to be unusually ready for theirIPOs,” she says. “These social net-works may have become householdnames only a year or two ago, butthey started a few years before that.”

There are several themes that aresaid to be compelling to investors atthe moment. Companies that sell“software as a service”, or “cloud”software, are highly prized.

Cornerstone OnDemand, a software-as-a-service company, is expected toprice its IPO this week, with morethan the offered shares said to be indemand. Smart Technologies, based inCalgary, which creates digital white-boards that can be shared remotely,last year was the largest venture-backed technology IPO to price in theUS.

Social networking and other “web2.0” services are also set to be a signif-icant theme. LinkedIn, the business-centric network, and Skype, the web-based telephone service, have filed togo public this year.

Facebook, the world’s largest socialnetwork, and Groupon, which offersdiscounts to groups of people, areexpected to file in the next year ortwo, though they have been able todelay that filing by raising moneythrough private offerings.

Mr Bowmer believes that those pri-vate valuations, even if not entirelyreliable, are making it possible forinvestors to evaluate individual offer-ings rather than rely on a wave ofcompeting IPOs — much like whathappened in the 1990s — to supportvaluations.

“You don’t have to wait to see howa similar company will trade, you cannow speculate based on private valua-tions,” he says.

Dotcom sectorHigh valuations areringing warning bells forsome investors, even thosewith short memories,writes Telis Demos

The world in your hands: as investors focus on growth prospects the suggested price for companies such as Facebook has reached very high levels Bloomberg

‘We’ve gone from the‘eyeballs’ craze tothe ‘viral’ craze.These companies aremarginally profitable butare supported byhuge valuations’

The problem isthat methodologiesstill do notrecognise the realrisk in markets