2012

Investment banks are highlylevered and undercapitalized,thanks in part to regulatory rulesthat lowered capital requirements.So argues David Einhorn, who alsocontends that value-at-risk (VaR)is a flawed model that encouragesexcessive risk-taking and thatrating agencies have done a poorjob of analyzing bank balancesheets.Aaron Brown (see pg. 19)cautions that too much capital canbe dangerous, explains why VaR isuseful and defends the trackrecord of regulators.

GLOBAL ASSOCIATION OF RISK PROFESSIONALS10 JUNE / JULY 08 I SSUE 4422

PrivateProfits and

Socialized Risk

C OV E R S TO RY: P O I N T / C O U N T E R P O I N T

GLOBAL ASSOCIATION OF RISK PROFESSIONALS 11

Afew weeks ago, the financial world was present-ed with the imminent failure of a publicly trad-ed entity called Carlyle Capital Corporation.You see, it had leveraged itself more than 30 toone. The press scoffed about what kind ofinsanity this was. Who in their right mind

would take on such leverage?The fact was that the Carlyle portfolio consisted of gov-

ernment agency securities. Historically, after treasuries,these have been among the safest securities around.Carlyle’s strategy was to take relatively safe securities thatgenerate small returns and, through the magic of leverage,create medium returns. Given the historical safety of theinstruments, Carlyle and its lendersjudged 30 times leverage to beappropriate. One could check thebackward-looking volatility andcome to the same conclusion.

Of course, the world changed,and the models didn’t work.Carlyle’s investors lost most of theirinvestment and the world, withnormal 20-20 hindsight, haslearned that investment companieswith 30 times leverage are not safe.

It didn’t take long for investors torealize that the big investment banks sport similar leverage. Infact, the banks count things such as preferred stock and sub-ordinated debt as equity for calculating leverage ratios. Ifthose are excluded, the leverage to common equity is evenhigher than 30 times.

And I’ll tell you a little secret: these levered balancesheets hold some things that are dicier than governmentagency securities. They hold inventories of common stocksand bonds. They also have various loans that they hope tosecuritize. They have pieces of structured finance transac-tions. They have derivative exposures of staggering notion-al amounts and related counterparty risk. They have realestate and private equity.

The investment banks claim that they are in the “mov-ing” business rather than the “storage” business, but thevery nature of some of their holdings suggests that this isnot true. And they hold this stuff on tremendously leveredbalance sheets.

Origins of Levered Balance SheetsThe first question to ask is, how did this happen? Theanswer is that the investment banks outmaneuvered thewatchdogs, as I will explain in detail in a moment. As aresult, with no one watching, the management teams at theinvestment banks did exactly what they were incentivized to

do: maximize employee compensation. Investment bankspay out 50% of revenues as compensation. So, more lever-age means more revenues, which means more compensation.

In good times, once they pay out the compensation,overhead and taxes, only a fraction of the incremental rev-enues fall to the bottom line for shareholders. Shareholdersget just enough so that the returns on equity are decent.Considering the franchise value, the nonrisk fee-generatingcapabilities of the banks and the levered investment result,in the good times the returns on equity should not bedecent — they should be extraordinary. But they are not,because so much of the revenue goes to compensation.

The banks have also done a wonderful job at public rela-tions. Everyone knows about the 20% incentive fees in thehedge fund and private equity industry. Nobody talksabout the investment banks’ 50% compensation structures,which have no high-water mark and actually are exceededin difficult times in order to retain talent.

The Trouble with VaRThe second question is, how do the investment banks justi-fy such thin capitalization ratios? And the answer is, inpart, by relying on flawed risk models, most notably value-at-risk (VaR). VaR is an interesting concept. The idea is to

tell how much a portfolio stands to make or lose 95% ofthe days or 99% of the days or what have you. Of course,if you are a risk manager, you should not be particularlyconcerned how much is at risk 95% or 99% of the time.You don’t need to have a lot of advanced math to knowthat the answer will always be a manageable amount thatwill not jeopardize the bank.

A risk manager’s job is to worry about whether the bankis putting itself at risk in the unusual times — or, in statisti-cal terms, in the tails of distribution. Yet, VaR ignores whathappens in the tails. It specifically cuts them off. A 99%VaR calculation does not evaluate what happens in the last1%. This, in my view, makes VaR relatively useless as a riskmanagement tool and potentially catastrophic when its usecreates a false sense of security among senior managers and

Investment banks pay out 50% ofrevenues as compensation. So, moreleverage means more revenues,which means more compensation.

DavidEinhorn


watchdogs. This is like an airbag that works all the time,except when you have a car accident.

By ignoring the tails, VaR creates an incentive to takeexcessive but remote risks. Consider an investment in acoin-flip. If you bet $100 on tails at even money, your VaRto a 99% threshold is $100, as you will lose that amount50% of the time, which obviously is within the threshold.In this case, the VaR will equal the maximum loss.

Compare that to a bet where you offer 127 to 1 oddson $100 that heads won’t come up seven times in a row.You will win more than 99.2% of the time, whichexceeds the 99% threshold. As a result, your 99% VaR iszero, even though you are exposed to a possible $12,700

loss. In other words, an investment bank wouldn’t haveto put up any capital to make this bet. The math whizzeswill say it is more complicated than that, but this is thebasic idea.

Now we understand why investment banks held enor-mous portfolios of “super-senior triple A-rated” whatever.These securities had very small returns. However, the riskmodels said they had trivial VaR, because the possibility ofcredit loss was calculated to be beyond the VaR threshold.This meant that holding them required only a trivialamount of capital, and a small return over a trivial amountof capital can generate an almost infinite revenue-to-equityratio. VaR-driven risk management encouraged accepting alot of bets that amounted to accepting the risk that headswouldn’t come up seven times in a row.

In the current crisis, it has turned out that the unluckyoutcome was far more likely than the backtested modelspredicted. What is worse, the various supposedly remoterisks that required trivial capital are highly correlated; youdon’t just lose on one bad bet in this environment, you lose

on many of them for the same reason. This is why in recentperiods the investment banks had quarterly write-downsthat were many times the firmwide modelled VaR.

Ratings: Flawed Opinions fromUnderstaffed AgenciesWhich brings us to the third question, what were the watch-dogs doing? Let’s start with the credit rating agencies. Theyhave a special spot in our markets. They can review non-public information and opine on the creditworthiness of theinvestment banks. The market and the regulators assumethat the rating agencies take their responsibility to stay ontop things seriously. When the credit crisis broke last sum-mer, one of the major agencies held a public conference callto discuss the health of the investment banks.

The gist of the rating agency perspective was, “Don’tworry.” The investment banks have excellent risk controls,and they hedge their exposures. The initial reaction to thecredit crisis basically amounted to “everyone is hedged.” Afew weeks later, when Merrill Lynch announced a big loss,that story changed. But initially, the word was that every-one was hedged. Securitization had spread the risk aroundthe world, and most of the risk was probably in Asia,Europe, Dubai or at the bottom of the East River. Thebanks were in the “moving” business, not the “storage”business, so this was no big issue. I wondered whether any-one saying this had actually looked at the balance sheets.

Of course, this raised the following question: how dideveryone hedge and who were the counterparties holdingthe bag? I pressed star-1 during the conference call andasked the rating agency analyst how everyone hedged themassive apparent credit risks on the balance sheets. Theanalyst responded that the rating agency had observedenormous trading volumes on the Chicago MercantileExchange (“the Merc”) in recent days.

The Merc offers products that enable one to hedge inter-est rate risk, not credit risk. I called the rating analyst backto discuss this in greater depth. At first he told me that youcould hedge anything on the Merc. When I asked how tohedge credit risk there, he was less familiar. I came to sus-pect that the rating agency analyst viewed his role as one torestore confidence in the system, which the rating agencycall did do for a while, rather than to analyze risk.

I later had an opportunity to meet a recently retired seniorexecutive at one of the large rating agencies. I asked him howhis agency went about evaluating the creditworthiness of theinvestment banks. By then, Merrill had acknowledged largelosses, so I asked him what the rating team found when itwent to examine Merrill’s portfolio in detail.

He answered by asking me to refocus on what I meant by“team.” He told me that the group covering the investmentbanks was only three or four people and they have to cover


Everyone knows about the 20%incentive fees in the hedge fund and

private equity industry. Nobodytalks about the investment banks’

50% compensation structures,which have no high-water mark and

actually are exceeded in difficulttimes in order to retain talent.


all of the banks. So they have no team to send to Merrill for athorough portfolio review. He explained that the agencydoesn’t even try to look at the actual portfolio, because itchanges so frequently that there would be no way to keep up.

I asked how the rating agencies monitored the balancesheets so that when an investment bank adds an asset, theagency assesses a capital charge to ensure that the bankdoesn’t exceed the risk for the rating. He answered thatthey don’t and added that the rating agencies don’t evenhave these types of models for the investment banks.

I then asked what they do look at. He told me they lookmostly at the public information — basic balance sheetratios, pretax margin and the volatility of pretax margin.They also speak with management and review managementrisk reports. And, of course, they monitor VaR.

I was shocked by this, and I think that most market partic-ipants would be surprised as well. While the rating agenciesdon’t actually say what work they do, I believe the marketassumes that they take advantage of their exemption fromRegulation FD to examine a wide range of non-public mater-ial. A few months ago, I made a speech where I said that rat-ing agencies should lose the exemption to Regulation FD, sothat people would not over-rely on their opinions.

The market perceives the rating agencies to be doingmuch more than they actually do. The agencies themselvesdon’t directly misinform the market, but they don’t dis-abuse the market of misperceptions — often spread by therated entities — that the agencies do more than they actual-ly do. This creates a false sense of security, and in times ofstress, this actually makes the problems worse. Had thecredit rating agencies been doing a reasonable job of disci-plining the investment banks — which unfortunately hap-pen to bring the rating agencies lots of other business —then the banks may have been prevented from takingexcess risk and the current crisis might have been averted.

The rating agencies remind me of the department ofmotor vehicles: they are understaffed and don’t pay enoughto attract the best and the brightest. The DMV is scary, butit is just for mundane things like driver’s licenses. Scarydoes not begin to describe the feeling of learning that thereare only three or four hard-working people at a major rat-ing agency judging the creditworthiness of all the invest-ment banks; the agency, moreover, does not even have itsown model for evaluating creditworthiness.

Insufficient Capital RequirementsThe second watchdog to talk about is the SEC. In 2004, theSEC instituted a rule titled, “Alternative Net CapitalRequirements for Broker-Dealers That Are Part ofConsolidated Supervised Entities.” In hindsight, as you willsee, an alternative name for the rule might have been the“Bear Stearns Future Insolvency Act of 2004.”

The purpose of the new rule was to reduce regulatorycosts for broker-dealers by allowing large broker-dealers touse their own risk management practices for regulatorypurposes. According to the SEC’s website, very large bro-ker-dealers had the opportunity to volunteer for additionaloversight and confidential disclosure to the SEC and, inexchange, would be permitted to qualify for “the alterna-tive capital computation method.”

While the SEC did not say that the alternative capitalcomputation method would increase or decrease the capitalrequirements, the rule says that “deductions for marketand credit risk will probably be lower under the alternativemethod.” Obviously, since this appears to be the carrotoffered to accept additional supervision, and I believe thatall of the largest broker-dealers have elected to participate,I think it is reasonable to speculate that the rule enabledbrokers to lower their capital requirements.

Under this new method, the broker-dealer can use“mathematical modelling methods already used to manage

their own business risk, including VaR models and scenarioanalysis for regulatory purposes.” It seems that the SECallowed the industry to adopt VaR as a principal method ofcalculating regulatory capital.

Unfortunately, it gets worse. In the new rule, the SEC alsosaid, “We are amending the definition of tentative net capi-tal to include securities for which there is no ready market.…This modification is necessary because … we eliminatedthe requirement that a security have a ready market to qual-ify for capital treatment using VaR models.” Without themodification, the “no ready market” securities would havebeen subject to a 100% deduction for capital purposes.

Is it any wonder, then, that over the last few years theindustry has increased its holdings of no ready marketsecurities? In the rule itself, the SEC conceded, “… inclu-sion in net capital of unsecured receivables and securities

According to the SEC’s website,very large broker-dealers had theopportunity to volunteer for addi-tional oversight and confidentialdisclosure to the SEC and, inexchange, would be permitted toqualify for “the alternative capitalcomputation method.”


that do not have a ready market under the current net capi-tal rule will reduce the liquidity standards.”

These adjustments reduced the amount of required capi-tal to engage in increasingly risky activities. The SEC esti-mated at the time the rule was proposed that the broker-dealers taking advantage of the alternative capital compu-tation would realize an average reduction in capital deduc-tions of approximately 40%. From my reading, the finalrule appears to have come out even weaker, suggesting thatthe capital deductions may have been reduced even further.

Obviously, since the rule was implemented, the broker-dealers have modified their balance sheets to take advan-tage of it. They have added lots of exposure to low-returnbonds with credit risk perceived to be beyond the VaRthreshold, and they have added more no ready marketsecurities — including whole loans, junior pieces of struc-tured credit instruments, private equity and real estate.

If this wasn’t enough, the 2004 rule also changed whatcounts as capital: “In response to comments received, theCommission has expanded the definition of allowable capi-tal … to include hybrid capital instruments and certaindeferred tax assets,” the SEC explained.

The rule also permits the inclusion of subordinated debtin allowable capital. The SEC permitted this because subor-dinated debt “has many of the characteristics of capital.” Ifind this one particularly amazing; apparently, it doesn’tactually have to be capital. For everyone else except thebroker-dealers, subordinated debt is leverage. The commis-sion considered but stopped short of allowing the broker-dealers to count all long-term debt as capital.

In reading through the rules and the SEC’s response tocomment letters, it seems that the SEC made concessionafter concession to the large broker-dealers. I won’t boreyou by describing how the rule eased the calculations ofcounterparty risk, maximum potential exposures and mar-gin lending or how the rule permitted broker-dealers toassign their own credit ratings to unrated counterparties.

My impression of this is that the large broker-dealers con-vinced the regulators that the dealers could better measuretheir own risks, and with fancy math, they attempted to showthat they could support more risk with less capital. I suspectthat the SEC took the point of view that these were all large,well-capitalized institutions, with smart, sophisticated riskmanagers who had no incentive to try to fail. Consequently,they gave the industry the benefit of the doubt.

In the cost-benefit analysis of the rule, on the benefitside, the SEC estimated the “value” to the industry by tak-ing advantage of lower capital charges to earn additionalreturns. In the “cost” part of the analysis, the SEC carefullyanalyzed the number of hours and related expense of themonitoring and documentation requirements, as well as ITcosts. It did not discuss the cost to society of increasing the

probability that a large broker-dealer could go bust.

The Fall of Bear StearnsI don’t know what effect the new rules had on Bear Stearns.The information the broker-dealers provide the SEC to showtheir compliance with these regulatory capital requirementsis confidential. It would be interesting to know how ade-quately capitalized Bear and other large broker-dealerswould have been under the rules as they existed before 2004.

In response to this possible regulatory failure,Christopher Cox, the SEC chairman, said recently that thiscurrent voluntary program of SEC supervision should bemade permanent and mandatory. Reuters reported thatCox said that the current value of the SEC supervisory pro-gram “can never be doubted again.”

Rather than looking at its own rules — which permittedincreased leverage, lower liquidity, greater concentrationsof credit risk and holdings of no ready market securities —the SEC is conducting an investigation to see if any short-sellers caused the demise of Bear by spreading rumors.

Of course, Bear didn’t fall because of market rumors. Itfell because it was too levered and had too many illiquidassets of questionable value and at the same time dependedon short-term funding. With the benefits of the reducedcapital requirements and reliance on flawed VaR analysis,Bear — like the other investment banks — increased its riskprofile over the last few years.

While VaR might make sense to the quants, it has led torisk-taking beyond common sense. If Bear’s only businesswas to have $29 billion of illiquid, hard-to-mark assets,supported by its entire $10.5 billion of tangible commonequity, in my view, that by itself would be an aggressiveinvestment strategy. However, as of November 2007, thatsliver of equity was also needed to support an additional$366 billion of other assets on Bear’s balance sheet.

When Bear’s customers looked at the balance sheet andalso noticed the increased cost of buying credit protection onBear, they had to ask themselves whether they were beingcompensated for the credit risk and counterparty risk indoing business with Bear. Many decided that they weren’t anddid the prudent thing to protect their own capital by curtail-ing their exposure. Bear suffered a classic “run on the bank.”

Perils of a Reverse-Robin Hood SystemWhen I came up with the title for this discussion, it wasbefore Bear Stearns failed. I was going to point out that wewere developing a system of very large, highly levered,undercapitalized financial institutions — including theinvestment banks, some of the large money center banks,the insurance companies with large derivatives books andthe government-sponsored entities (GSEs). I planned tospeculate that regulators believe all of these are too big to



fail and would bail them out if necessary. The owners, employees and creditors of these institu-

tions are rewarded when they succeed, but it is all of us —the taxpayers — who are left on the hook if they fail. Thisis called private profits and socialized risk. Heads, I win.Tails, you lose. It is a reverse-Robin Hood system.

In any case, with the actual failure and subsequentbailout of Bear Stearns — and regardless of what our lead-ers told Congress last week, it is a bailout under any defini-tion — I am shifting the subject of this talk from a potentialbailout to the real, live thing.

Some would say that it wasn’t a bailout, because theshareholders, including the risk-taking employees, lost mostof their money, so they were properly punished and the sys-tem is intact. However, the problem is that we don’t have anequity bubble. (In fact, the equity markets seem to be func-tioning fine, with a good number of excellent companies atreasonable valuations.) What we do have is a credit bubble,and the Bear Stearns bailout has reinforced the excessiverisk-taking and leverage in that arena. Specifically, thebailout preserved the counterparty system. The governmentappears to have determined that the collapse of a single sig-nificant player in the derivatives market would cause somuch risk to the entire system that it could not be permittedto happen. In effect, the government appears to have guar-anteed virtually the entire counterparty system.

The message is that if you are dealing with a major player— anyone in the “too big to fail” group — you don’t have toworry about that player’s creditworthiness. In effect, yourrisk is with the US Treasury. The government does not wantcustomers of the next Bear Stearns to have to evaluate itscreditworthiness, find it lacking and determine that exposureneeds to be curtailed, creating a run on another bank.

The Bailout ImpactThe next question is whether the bailout was a good idea.It really comes down to Coke vs. water. If you are thirsty,you have choices. Coke tastes better and provides an imme-diate sugar rush and caffeinated stimulus, while quenchingthirst. Water also quenches thirst, but it isn’t as stimulating.It purifies your body. It doesn’t make you fat and is muchbetter for your long-term health.

One of the things I have observed is that American finan-cial markets have a very low pain threshold. Last fall, withthe S&P 500 only a few percent off its all-time high pricesafter a multi-year bull market, certain TV commentatorsand market players were having daily tantrums demandingthat the Federal Reserve System (“the Fed”) give them thefinancial equivalent of Coke. Other parts of the worldendure much greater swings in equity values withoutdemanding relief from central planners.

The Fed responded by providing liquidity and lower rates.

Even so, the crisis deepened. So now they have introducedthe “Big Gulp,” also known as the Bear Stearns bailout, andan alphabet soup of extraordinary measures to support thecurrent system. If that doesn’t turn the markets, they arethreatening the financial equivalent of having the water utili-ties substitute Coke for water throughout the system.

In early April, Ben Bernanke told Congress that he hopesthat Bear Stearns is a one-time thing. In the short-term, itmight be. If market participants accept as an article of faiththat the Fed will bail them out, it reinforces risk-takingwithout the need for credit analysis. As night follows day, itis certain that in the absence of tremendous governmentregulation, this bailout will lead to a new and potentiallybigger round of excessive risk-taking. If Mr. Bernanke isunlucky, the payback may come later in this cycle; if he islucky, it will come in the next cycle.

Since the government is now on the line for the losses, thereis a strong public interest in increased supervision, whichshould result in dramatically higher capital requirements forthe major players. Additionally, regulators should considerdismantling the counterparty system so that the market cansurvive the failure of a big player. One step could be to requirethe posting of all derivative trades, clearing them through acentral system and regulating margin requirements.

Lehman Brothers vs. Bear Stearns: A Quick ComparisonFinally, I’ll offer a few words about Lehman Brothers.(Greenlight Capital is short the stock.) Lehman’s manage-ment is charismatic and has almost cult-like status. It getstremendously favorable press for everything from handlingthe 1998 crisis to supposedly hedging in this crisis to notplaying bridge while the franchise implodes.

From a balance sheet and business mix perspective,Lehman is not that materially different from Bear Stearns.Lehman entered the crisis with a huge reliance on US fixedincome, particularly mortgage origination and securitiza-tion. It is different from Bear in that it has greater exposureto commercial real estate and its asset management franchisedid not blow up. Incidentally, neither Bear nor Lehman hadenormous on-balance-sheet exposure to CDOs.

At the end of November 2007, Lehman had Level 3assets and total assets of about 2.4 times and 40 times itstangible common equity, respectively. Even so, at the end ofJanuary 2008, Lehman increased its dividend and autho-rized the repurchase of 19% of its shares. In the quarterended in February, Lehman spent over $750 million onshare repurchases, while growing assets by another $90 bil-lion. I estimate Lehman’s ratio of assets to tangible com-mon equity to have reached 44 times.

There is good reason to question Lehman’s fair value cal-culations. It has been particularly aggressive in transferring


mortgage assets into Level 3. Last year, Lehman reported itsLevel 3 assets actually had $400 million of realized andunrealized gains. Lehman has more than 20% of its tangi-ble common equity tied up in the debt and equity of a singleprivate equity transaction — Archstone-Smith, a real estateinvestment trust (REIT) purchased at a high price at the endof the cycle. Lehman does not provide disclosure about itsvaluation, though most of the comparable company tradingprices have fallen 20-30% since the deal was announced.The high leverage in the privatized Archstone-Smith wouldsuggest the need for a multibillion-dollar write-down.

Lehman has additional large exposures to Alt-A mort-gages, CMBS and below-investment-grade corporate debt.Our analysis of market transactions and how debt indicesperformed in the February quarter would suggest Lehmancould have taken many billions more in write-downs than itdid. Lehman has large exposure to commercial real estate.Lehman has potential legal liability for selling auction-ratesecurities to risk-averse investors as near cash equivalents.

What’s more, Lehman does not provide enough trans-parency for us even to hazard a guess as to how they haveaccounted for these items. It responds to requests forimproved transparency grudgingly, and I suspect thatgreater transparency on these valuations would not inspiremarket confidence.

Instead of addressing questions about its accounting andvaluations, Lehman wants to shift the debate to where it ison stronger ground. It wants the market to focus on its liq-uidity. However, in my opinion, the proper debate shouldbe about Lehman’s asset values, future earning capabilitiesand capital sufficiency.

In early April, Lehman raised $4 billion of new capitalfrom investors, thereby spreading the eventual problemsover a larger capital pool. Given the crisis, the regulatorsseem willing to turn a blind eye toward efforts to raise capi-tal before recognizing large losses; this holds for a numberof other troubled financial institutions.

The problem with 44 times leverage is that if your assetsfall by only a percent, you lose almost half the equity.Suddenly, 44 times leverage becomes 80 times leverage andconfidence is lost. It is more practical to raise the newequity before showing the loss. Hopefully, the newinvestors understand what they are buying into, eventhough there probably isn’t much discussion of thisdynamic in the offering memos. Some of the sovereignwealth funds that made these types of investment last year

have come to regret them.Lehman wants to concentrate on long investors; in fact,

it went to great lengths to tell the market that it sold all ofits recent convert issue to long-only investors. Putting asidethe fact that some of the clearing firms have told us thatthis wasn’t entirely true, companies that fight short sellersin this manner have poor records. The same goes for com-panies that publicly ask the SEC to investigate short selling,as Lehman has done. There is good academic research tosupport my view on this point.

As I have studied Lehman for each of the last three quar-ters, I have seen the company take smaller write-downsthan one might expect. Each time, Lehman reported amodest profit and slightly exceeded analyst estimates thateach time had been reduced just before the publicannouncement of the results. That Lehman has not report-ed a loss smells of performance smoothing.

Given that Lehman hasn’t reported a loss to date, there islittle reason to expect that it will any time soon. Even so, Ibelieve that the outlook for Lehman’s stock is dim. Anydeferred losses will likely create an earnings headwindgoing forward. As a result, in any forthcoming recovery,Lehman might underearn compared to peers that have beenmore aggressive in recognizing losses.

Further, I do expect the authorities to require the broker-dealers to de-lever. In my judgment, a back-of-the-envelopecalculation of prudent reform would require 50-100% cap-ital for no ready market investments; 8-12% capital forwhat the investment banks call “net assets”; 2% capital forthe other assets on the balance sheet; and an additionalcharge that I don’t know how to quantify for derivativeexposures and contingent commitments.

Only tangible equity, not subordinated debt, shouldcount as capital. On that basis, assuming that Level 3 assetsare a good proxy for no ready market investments —assigning no charge for the derivative exposure or contin-gent commitments and assuming its asset valuations arefairly stated — Lehman, based on its November balancesheet, would need $55-$89 billion of tangible equity, whichwould be a three- to-five-fold increase.

So what do I expect to happen? I just finished a book onAllied Capital and the lack of proper and effective regula-tory oversight. Based on my book and the current regulato-ry environment, the pessimistic side of me says that regula-tors will probably decide to send me a subpoena and sendLehman a Coke. ■

✎ DAVID EINHORN is the president and founder of Greenlight Capital, a long-short value-oriented hedge fund. Since its inception in 1996,Greenlight has generated a greater than 25% annualized net return for its partners. Einhorn is the chairman of Greenlight Capital Re, Ltd.(Nasdaq: GLRE), and serves on the boards of the Michael J. Fox Foundation for Parkinson’s Research and Hillel:The Foundation for JewishCampus Life. Einhorn is the author of Fooling Some of the People All of the Time:A Long Short Story (Wiley, May 2008).

This article is a reprint of a speech given April 8, 2008, at Grant’s Spring Investment Conference


Counterpoint:Capital Inadequacy


Iagree with David Einhorn that the financial dealer sys-tem is undercapitalized, and I think he has identified theimportant links in the chain: management incentives,financial regulation and risk management. However, Ihave different views about each of those, and put themtogether in a different order. Therefore, it makes sense

to me to start at the end, with his recommendation for“prudent reform”:

“50-100% capital for no ready market investments; 8-12% capital for ... net assets; 2% capital for all other assetson the balance sheet; and an additional charge that I don’tknow how to quantify for derivative exposures and contin-gent commitments.”

This, of course, is Basel I, the standard created in 1988and put into law in 1992. Basel I had five buckets from 0%(for home country sovereign debt) to 8% (for corporatedebt), with an add-on for the notional amount of deriva-tives. The prudent reform proposal has about 50% highercapital levels (based on a quick estimate applying bothrules to Lehman’s 2007 balance sheet) and different cate-gories, but the principles are the same.

There were three problems with Basel I, all of whichapply to prudent reform as well. First, everything in abucket had the same capital charge, which encouragedinstitutions to fund the riskiest things in each bucket andignore the safest things. The most notorious example wasthe 0% charge for sovereign risk, which encouraged over-exposure to emerging markets. This distorted markets andled to excessive risk-taking.

Second, Basel I did not encourage risk management. Inmost cases, hedging a risk — except for certain nearly per-fect hedges — increased rather than decreased the capitalcharge. Two assets had the same capital charge whetherthey were highly correlated or negatively correlated. Somemajor risks, like counterparty credit in an era when over-the-counter transactions were rarely collateralized, did notappear at all. At a higher level, there was no requirement tomonitor risk, nor to report it to regulators, nor to discloseit to investors.

Finally, under Basel I, derivatives and commitments areafterthoughts. This was inadequate in 1988, and 20 years

later is less adequate. Basel I treated every asset like a loan,which made it impossible to treat derivatives. Basel II treatsevery asset like a derivative, and loans are easy to fit intothat. Unfortunately, Basel II also requires more assump-tions, flaws that can lead to disaster. Still, I prefer assump-tions, which might be right and can be revised as circum-stances change, to embedded contradictions, which cannotbe either.

One point I concede immediately: David Einhorn is a farbetter equity analyst than I am. If he tells you what he thinksof a Lehman common stock investment, no one should con-sider asking my opinion. My portfolio consists of low-costindex funds and stock I received as compensation (all fromoverleveraged financial firms, but not including Lehman),which I hold out of loyalty rather than calculation. So I’mnot going to touch the question of whether Lehman is over-leveraged from the standpoint of shareholders.

Building a Better BaselOn the other hand, I suspect Iknow more about the nuts andbolts of calculating risk-sensitivecapital levels for big financial insti-tutions. I’ve done it for four ofthem, and I know a lot of the peo-ple who did it for the rest. Twentyyears ago, I looked at Basel I andquit my job as head of mortgagesecurities to become an advocatefor what came to be called VaR.

How could I have been so stu-pid as to help develop a “useless” and “potentially cata-strophic” measure that, to quote Mr. Einhorn, is “like anair bag that works all the time, except when you have a caraccident”? It wasn’t for lack of alternatives. There were allsorts of candidates for a risk-sensitive measure in thosedays, including ones that looked deep in the tails. VaR hasonly one virtue, but it was enough to win out over all theothers. VaR is observable.

VaR is defined as the loss amount such that there is aspecified probability (e.g., 1%) that the loss from a portfo-lio at a specified future time (say, close of business tomor-row) will be greater than the VaR amount, assuming fixed

By Aaron Brown

AaronBrown


positions and normal markets. You evaluate VaR by back-test; you compute it every night on the portfolio, and checkthe next day how much that fixed portfolio would havegained or lost (this is not the same as the actual gain or lossif there is trading during the day). You check that the lossexceeds VaR on 1% of the days (plus or minus expectedstatistical variation) and test that the break days are bothindependent in time and independent of market factors(most importantly, the level of the VaR).

VaR is only as good as its backtest. When someoneshows me a VaR number, I don’t ask how it is computed, I

ask to see the backtest. If I think I could make money bet-ting either side at 100 to 1 on whether or not a break willoccur tomorrow, I disregard the VaR. If anyone argues withme, I challenge them to take my bet over the next year.Also, it’s essential that VaR is actually calculated before thestart of trading. That means there will be errors: mis-booked trades, missing feeds, stale prices and other prob-lems. The noise from these errors has to be included inVaR, because they are real risks.

If you report a $50 million VaR, then revise it to $200million after a $150 million loss, you did nothing usefuland I don’t trust either the original or the revised number. Iwant to see a VaR backtest based on the actual values pub-lished before the start of the period they reference.Hypothetical backtests based on scrubbed data are suspect,and tests with after-the-fact corrections to the VaR for dataor processing errors are worthless.

Why not go deeper into the tails? Because you won’thave enough data for convincing validation. A 99% one-

day VaR has to operate for about three years before youtrust it. A 99.97% one-year VaR, which some people usefor economic capital, requires 26,000 years for the samelevel of confidence. That makes deep tail VaR a matter offaith and assumptions, not something you can observe withreasonable statistical certainty over a moderate time inter-val. Moreover, assumptions like fixed positions and normalmarkets make VaR far less relevant over longer periods atsmaller probabilities.

Most important, some days have undefined losses dueto events like market closures, extreme liquidity events,legal uncertainties and so forth. And however careful youare, there is some chance your VaR estimate is significantlymisstated due to data or calculation errors (including thoseinduced by rogue traders and embezzlers and third-partycrooks). These might happen two or three times a decade,about 0.1% of the time. VaR excludes these days, which isreasonable in a 99% measure. But in a 99.9% measure,you would be excluding as many days as you had VaRbreaks, making the number pretty useless; 99.97% isentirely pointless.

New and Improved!A once-popular alternative to VaR, which has enjoyed aresurgence among academics but not practitioners, isexpected shortfall (ES) — defined as how far you expect togo beyond VaR, given that you exceed it. ES has somedesirable mathematical properties and tells you more aboutthe tail than VaR. However, expected loss, conditional orunconditional, is not observable because there is always thepossibility of a small-probability, large-impact event that ismissing from historical data. Standard deviation, whichdepends on the square of P&L movements, is even moresensitive to this problem.

Another alternative is to consider maximum loss withinthe VaR interval. But this is hard to define with securitiestrading in different markets and time zones. You can esti-mate it with a parametric model, but you have to trust yourmodel; you can’t prove the results with transactions.

Useful numbers are compromises between what wewant to know and what we can measure well. VaR can bemeasured better than alternatives, but what does it tell us?It’s not the worst-case loss: in fact, we expect to lose morethan VaR two or three times a year. Sometimes peopleassume that a multiple of VaR, say three times VaR, is agood almost-worst-case number. This is a terrible assump-tion on both theoretical and empirical grounds.

VaR is a tool of risk management, not risk measure-ment. For one thing, putting in a VaR system always leadsto tremendous improvements in information systems andcommunication within a firm and always turns up someimportant surprises. You could throw away the final num-


VaR is only as good as its backtest.When someone shows me a VaRnumber, I don’t ask how it is com-puted, I ask to see the backtest. If Ithink I could make money betting

either side at 100 to 1 on whetheror not a break will occur tomor-row, I disregard the VaR. If anyone

argues with me, I challenge them totake my bet over the next year.


ber and still reap great benefit from the improvementsforced by the discipline of publishing a VaR number ontime every day and passing backtests. Just as anyone can bea great investor at a cocktail party, but not when you lookat three-year Sharpe ratios, anyone can be a risk expert,until you demand rigorous statistical evidence. You onlylearn from models that sometimes surprise you.

It also makes sense to use VaR on the trading desk andin the executive suite. In 1988, a trader was less likely tohave a PhD in physics than to have been kicked out of ele-mentary school for running a craps game in the principal’soffice after school. But anyone who could add up the spotson two dice could understand that his losses exceededyour published VaR estimate three times in the last 300trading days.

Today, it’s a safe assumption that at least one member ofthe executive committee of a big financial institution hassome derivative pricing experience — but even the formertreasury salespeople and investment bankers could under-stand a graph with a line showing VaR over the last year,points showing the daily P&L, with 1% of the pointsbelow the line. VaR was the first true communicationbetween bank executives and traders.

VaR defines a perimeter. Inside the VaR limit, we haveplenty of data to optimize decisions. Outside the VaR limit,we cannot rely on data. Instead, we use other things tojudge risk. We look at longer-term history and other mar-kets to guess what might happen, knowing the compar-isons are not exact. We think about plausible extreme sce-narios, and we concentrate on the things VaR ignores.

What if, for example, markets are not normal? What iftrading exacerbates losses? What if the losses occur over ashorter or longer period than the VaR horizon? What ifP&L does not measure the real risk? The VaR point rele-

gates 99% of the data to be handled by rote, so that riskmanagers can focus on the 1% that matters. My rule ofthumb for financial businesses is that no plausible scenario,either hypothetical or based on history, should result in aloss more than 10 times 99% one-day VaR. If it does,either the business plan should be adjusted or that scenarioshould be hedged or the VaR should be increased.

If foreseeable events can cause losses more than 10 timesthe size of what you can assign accurate probability esti-mates to, you have no idea if your business has a positiveexpected value. You can’t manage what you can’t measure.People will be afraid to take risks up to the VaR point,because exceeding VaR can lead to uncontrollable losses.Timidity will hurt profits, which means you have lessreserve to survive the surprises. Half of risk management ispersuading people to take more risk the 99% of the timeit’s safe to do so, half is surviving the other 1%.

Of course, when the surprises occur, they will not be theplausible scenarios you envisioned. They might be muchworse. But the discipline of preparing for what you canforesee greatly improves your chances of dealing success-fully with what actually happens.

VaR of the JungleVaR is like a fence built around a village in the jungle.There are all kinds of dangers in the jungle: tigers,snakes, poisonous plants, swamps, unfriendly humansand other things you don’t know about. The fence isbuilt in a relatively safe place, and remaining dangersinside the fence are cleared out. We collect lots of statisti-cal data about dangers in the village; we notice problemsquickly and fix them.

But people spend 1% of their time outside the fence,where dangers are both greater and less well-known. Ifpeople disappear in the jungle, we don’t know if they wereeaten by a tiger or stuck in a swamp or if they just decidedto find a new home. So we don’t have the kind of data weneed either to evaluate or to fix problems.

The first decision of risk management is where to buildthe fence. Inside, we have a well-defined population andlots of observations, so we can monitor risk with statisticsand adjust it through rules. Outside, we have to rely onanecdotes and imagination, knowing both have manyerrors. We equip people leaving the village with defensesagainst known dangers and general-purpose survival gear.And we wish them good luck.

It would be foolish to use any extrapolation of statisticscollected inside the fence to predict things outside it.Statistics collected outside are noisy and sparse and arefatally flawed by survivorship bias (the dangers we hearabout are the ones that people survived). In my friendNassim Taleb’s (another VaR-hater) terms, inside the fence

Just as anyone can be a greatinvestor at a cocktail party, but

not when you look at three-yearSharpe ratios, anyone can be a

risk expert, until you demand rig-orous statistical evidence.You only

learn from models that some-times surprise you.


is Mediocristan, outside is Extremistan. Tools that work inone place are useless in the other.

I’m not talking about the VaR you find in the “risk man-agement” section of “management’s discussion and analy-sis” of the financial statements. That considers only themarket risk of the firm’s trading and portfolio positionsand usually has an unimpressive backtest. Instead, I meanthe VaR concept applied to all firm risks: define the riskprecisely enough to specify an observable measure, calcu-late and backtest the measure, then use scenario and stresstesting to ensure foreseeable risks are within an order ofmagnitude of the observable risk.

Regulators were wise, not foolish, to allow firms to useinternal models to estimate risk. The quality of the risk pre-dictions can be judged without looking at the internalworkings (the regulators check under the hood as well). Noset of rules could possibly cover the complexities of a mod-ern diversified financial institution, and if they could, theywould be out of date before they were implemented.

The regulators offer simpler fixed alternatives for risk-challenged institutions, but the top financial institutions allinvested in Basel II “advanced” approaches. Basel II doesnot only require risk estimation for capital adequacy: mod-els and results are reviewed by supervisors and exposed tomarket participants.

How Much is Enough?I’ve argued that capital should be risk-sensitive and thatVaR is the appropriate first step to measure risk. Given theVaR, how much capital should a bank hold? No multipleof VaR is a safe worst-case loss estimate. But it’s silly to set

capital equal to worst-case loss. Capital does not protectagainst all risks.

The classic example is more cash in the vault doesn’t pro-tect against the danger of the vault being robbed. Physicalrobbery is not a significant risk for modern financial insti-tutions, but there are other scenarios in which the morecapital you have, the more capital can be taken away. Whatif the Justice Department goes after top managers, as hap-pened to Drexel? What if the US government defaults ontreasuries, as almost happened in 1995? What if BearStearns had collapsed and frozen all financial institutiontransfers for a week?

Other risks can be measured in dollars, but plausibly canscale larger than any reasonable level of capital. If a roguetrader can cost Societe Generale $7 billion, why can’t thenext one cost five or 10 times as much? The 9/11 terroristattack caused the stock market to close for a week duringwhich stocks fell 5%, and what financial institution canhold enough capital to guard against markets being closedfor a month while prices move 25% or 50%?

The simplest view of capital is to consider what happensif a firm gets into trouble and decides to liquidate. If it cansell its assets to repay its liabilities, it can have an orderlyliquidation, with all claimants paid in full. Therefore, capi-tal is the excess in value of assets over liabilities, and thecapital requirement should be set such that there is verylow probability of a firm going from well capitalized toinsolvent (negative capital) before a regulator can force liq-uidation.

Of course, this does not mean that the preferred reactionto losses is liquidation. If you can liquidate and repayeveryone — that is, if you have positive equity value —then you have other options, such as raising new capital orselling yourself. If your assets are worth less than your lia-bilities, you are no longer in control of your bank, andsome third parties are going to take losses.

This seems to be what David Einhorn is thinking of withhis prudent reform proposal. Consistent with the liquida-tion view, he wants to count only tangible equity as capital.That assumes intangible assets are worthless in a liquida-tion: 100% capital for illiquid assets assumes they cannotbe sold in a liquidation; 50% assumes a steep discount.

An 8% to 12% write-down — about $30 billion to $45billion for Lehman — seems to be in the ballpark for the equi-ty a prime broker would demand to finance the portfolio ofLehman’s marketable assets. The collateralized lending assetsare considerably safer: 2% (or about $6 billion) for Lehmanseems like a conservative but not absurd potential loss figure.If Lehman meets this standard one month, a regulator canhave an appropriate level of confidence that the next month’sreport will show enough equity cushion for an orderly liqui-dation with full payout to creditors.


I’ve argued that capital should berisk-sensitive and that VaR is the

appropriate first step to measurerisk. Given the VaR, how much cap-ital should a bank hold? No multi-ple of VaR is a safe worst-case lossestimate. But it’s silly to set capitalequal to worst-case loss. Capitaldoes not protect against all risks.


Dynamic HedgingA more common view of capital, and a more realistic onein my opinion, assumes a going concern and dynamic capi-tal management. The question isn’t what can happenbefore the regulators force a liquidation, but whether man-agement can raise enough new capital to stay in business. Iunderstand why a stockholder, like David Einhorn, wouldbe suspicious of such a view. Raising capital might takeadvantage of the new investors if, as he writes, regulators“turn a blind eye toward efforts to raise new capital beforerecognizing large losses.” If this fools investors into over-paying for the new stock, it’s good for the solvency of thefirm and the interests of old shareholders, but bad for thenew shareholders and the market in general.

Now the relevant measure of capital is market capital-ization, not book-tangible equity. In a highly theoreticalanalysis, a company should be able to raise new equity cap-ital up to the existing amount of equity divided by theprobability of bankruptcy. Let E be the market capitaliza-tion before selling new equity and p be the probability ofbankruptcy. To a first approximation, raising X additionaldollars of equity adds pX to the wealth of creditors (it’sactually less than this, because the equity infusion will

reduce the probability of bankruptcy). Therefore, it adds(1-p)X to the equity value of the firm, which becomes E +(1-p)X. As long as this is greater than X, which is equiva-lent to X < E/p, the company can dilute existing sharehold-ers enough to raise X. Better yet, the company can do itagain and again, to raise infinite equity capital startingfrom any positive market capitalization.

Of course, real equity investors do not behave as this theo-retical analysis suggests. If there is any suspicion of trouble ata firm, asking for new equity will depress the stock price —and the repeated need for equity infusions, each one injuring

existing shareholders, will rapidly destroy any marketappetite for the stock (nevertheless, people really did try toexploit the unlimited equity effect, through securities popu-larly known as “toxic” or “exploding” convertibles).

Still, a company that recognizes problems early, main-tains a solid balance sheet and has credibility with investorsshould be able to raise the equity it needs to continue inbusiness, even if tangible assets are worth less than liabili-ties. The equity still has going concern and option value inthis situation, and that should be enough to save it.

Loaded for Bear

Consider Bear Stearns, for example. By mid-March 2007,HSBC had announced a $10.5 billion write-off from sub-prime, New Century Financial stock had been suspendedfor bankruptcy concerns and Donald Tomnitz, CEO of thelargest US homebuilder, got headlines warning of “hugesubprime losses.” Bear Stearns stock traded at over $150per share ($36 billion market capitalization) for more thanthree months after that. The stock took a hit in June 2007,when Bear announced bailouts of two of its hedge funds,but the price was still over $125 per share ($30 billion mar-ket cap). Three weeks before Bear failed, the price wasalmost $90 per share ($20 billion market cap).

Clearly, the problem wasn’t that Bear was undercapital-ized going into the crisis; it’s that Bear management chosenot to raise significant equity capital long after the problemwas obvious to the public. This is not 20/20 hindsight: Iwrote in October 2007 (not referencing Bear in particular)that banks hit by the crisis were talking about “raisingenough capital to keep their credit ratings, not the amountsneeded to reassure their customers.” Of course, many otherpeople made similar points.

This is not intended to be a criticism of the managementof Bear Stearns or any other firm. Perhaps their actionswere the best for their shareholders — or at least the bestgiven the information available at the time. Nor is it a criti-cism of regulators who had to work within their mandates.


Clearly, the problem wasn’t thatBear was undercapitalized goinginto the crisis; it’s that Bear manage-ment chose not to raise significantequity capital long after the prob-lem was obvious to the public.

A company that recognizes prob-lems early, maintains a solid bal-

ance sheet and has credibility withinvestors should be able to raise

the equity it needs to continue inbusiness, even if tangible assets are

worth less than liabilities.


But the outcome was clearly suboptimal, except possiblyfor JPMorgan shareholders.

The lesson learned should not be to make financial insti-tutions hold more capital in general. A big stack of cashheld for disasters is going to be used up; a bigger stackmight last longer, but the outcome is the same, only moreexpensive. Too much capital insulates the firm from marketdiscipline and leads to waste and sloppiness. It kills return-on-equity, making investors unwilling to provide additionalcapital when the need arises.

In poker terms, Wall Street is a tournament with infiniterebuys. A good player will survive, even with a short stack,because he or she can always find backers and rebuy. Aplayer with too large a stack will never learn to play, keepstoo much at risk on the table and won’t have backers. Heor she can only win with an unbroken run of luck.

The proper risk management lesson is to address prob-lems early and aggressively, so that they never grow tofirm-threatening size and block your ability to raise addi-tional capital. It’s painful to sell new shares on pricedeclines, and you’ll take heat from analysts and existingshareholders (or, if you manage to hide the bad news untilafter the offering, heat from new shareholders). Most ofthe time, the precaution will be unnecessary, and you’ll endup buying back the shares at a higher price. This may ormay not maximize shareholder wealth; I don’t know.

A shareholder with a diversified portfolio might preferbanks to risk failure. But if regulators feel compelled to bailout losers, they should make firms pay equity-market insur-ance premiums through creating synthetic puts on their shares.

Regulatory DefenseIt is not fair to label the SEC’s Basel Capital rule, “BearStearns Future Insolvency Act of 2004.” It’s true that Bear’sleverage ratio increased during the period of ConsolidatedSupervisory Entity regulation from 27 to 33, but it’s unlike-ly this was due to reduced capital standards.

Bear always held more than the regulatory minimumnet capital, with ratios comfortably over the “well-capital-ized” threshold of 10%. You can argue the regulatory min-imums are too low or the definition of capital too lenient,but not that Bear took advantage of the lowered levels andfailed as a result.

In any event, the amounts were too small to matter forthe events that overtook Bear. With $20 billion of liquiditydisappearing in three days, a billion or two extra capitalwould not have helped directly (it might have helped indi-rectly by preventing the run on the bank).

Bear was brought down by a crisis in market confidence,not lax regulation. Excessive leverage certainly contributedto the loss of confidence, but Bear’s leverage was much lessthan the maximum allowed under the old rules. The mar-

ket didn’t suddenly change its mind about the value of Bearin an orderly liquidation; it lost confidence that Bear hadthe immediate cash on hand to pay day-to-day bills, whichis a self-fulfilling fear. That’s not a capital problem, nor is ita traditional concern of broker-dealer regulation.

If a broker-dealer fails but all customers get their cashand securities back in a reasonable amount of time, it usedto be considered a success. You can argue that regulatorsshould have realized months earlier that Bear could not beallowed to fail and forced the firm to raise additional equi-ty capital at that time (assuming that fell within their regu-latory authority) — and that such a move, in turn, mighthave sustained market confidence in Bear’s liquidity. Butit’s excessive to push the blame back to 2004.

I offer my personal opinion that the SEC did not make“concession after concession to the large broker-dealers.” Iwas there. There was only one SEC and five broker-dealers.We cooperated in the effort, so I had a good picture of theentire process. There was extensive consultation andreview on both sides. In almost all cases, issues wereresolved the right way, sometimes after years of discussion.

In a few cases, the SEC insisted on unreasonably strictrule; in a few cases, the industry might have gotten away

with cutting a corner. But in none of those cases was it aresult of begging or pressure by the broker-dealers. Weakrules were oversights, not concessions. Anyone who con-ceives of the size and complexity of the project will not caststones over a few oversights, all of which can be correctedwith ongoing revision.

It’s absolutely true that “dealers could better measure andmonitor their own risks with fancy math” than with old-style, fixed-percentage-of-notional capital rules. Financial

Bear always held more than theregulatory minimum net capital,with ratios comfortably over the“well-capitalized” threshold of 10%.You can argue the regulatory mini-mums are too low or the definitionof capital too lenient, but not thatBear took advantage of the low-ered levels and failed as a result.


risk requires fancy math. Math, plain or fancy, is no guaran-tee against error, but at least you have a chance of getting theright answer. And it’s mostly true that “these were all large,well-capitalized institutions, with smart, sophisticated risk-managers who had no incentive to try to fail.” The onlyarguable point is “well-capitalized,” and the SEC did notstart by assuming that these institutions had enough capital.

Back to the StartFinally, I end up at David Einhorn’s starting point: leverageratios. Comparing Lehman’s leverage ratio to Carlyle Capitalis a great rhetorical point, but it needs some adjustment.Carlyle Capital borrowed on margin, meaning any decline inits asset value triggered immediate cash demands fromlenders. I don’t know the margin terms it negotiated, but theymay have allowed loans to be called or terms changed withvery short notice. Lehman’s borrowings include long-termbonds and loans with more stability than demand financing.It’s prudent to run a higher leverage ratio if you have time toaddress problems and wait out unfavorable markets.

Lehman also has access to more capital sources thanCarlyle Capital. It has more ability to sell stock in publicmarkets and to arrange private financing. It has strongfranchise and going concern values that can be monetizedthrough sale or other transactions. It has thousands of peo-ple working to sell its assets, so things illiquid to Carlylemight be liquid for Lehman.

Depending on the time of year, Lehman will have up toabout $5 billion on its balance sheet accrued for employeebonuses. Its more diversified portfolio is unlikely to sufferlosses in all segments at once and is likely to have at leastsome liquid securities.

Despite those caveats, it is a sobering thought that a pub-lic fund (hence one that did not have to worry about investorredemptions) holding only AAA vanilla US governmentagency mortgage-backed securities couldn’t survive at aleverage ratio significantly lower than the investment banks,which hold much more volatile, complex and illqiuid assets.I have to agree that leverage is too high, especially during acredit crunch with unheard-of levels of credit spread.

As a risk manager, I would like to see firms managingrisk by raising equity capital when risks increase. I wouldlike to see investors enforcing discipline on the companiesthey own instead of demanding bail-outs. If firms won’traise capital and investors won’t accept pain, then regula-tors have no choice but to force leverage down. If everythirsty person demands Coke from the government, theonly sensible solution is to make sure there’s enough water


A Quick Rebuttal from David Einhorn

Aaron Brown and I actually agree more than one mightthink. He writes, "The VaR point relegates 99% of thedata to be handled by rote, so risk managers can focuson the 1% that matters." I couldn't agree more. To me,the normal 99% is portfolio management. Risk manage-ment is all about the tails, and probably begins aboutwhere VaR ends. Risk management is the airbag thatmust always work, but only in the multi-sigma eventwhere you have an accident — an event that happensmuch less often than once in every hundred driving days.

Aaron and I have sharply different opinions about thetolerance for blowups. He equates blowups to needingto "rebuy" in a poker tournament. This suggests a will-ingness to put capital to such risk — i.e., that it’s anacceptable fact of life that, from time to time, investorsget wiped out or require a bailout.

This thinking favors the interests of those who benefitasymmetrically from a high risk tolerance: investmentbank employees being paid 50% of revenue or hedgefund managers earning 20% incentive fees. The victimis the shareholder who suffers most of the loss whenthose inevitable tail events happen. Over the long haul,winning investors will be those that preserve most oftheir capital on the downside and never have to sufferthe permanent dilution and loss from a "re-buy" event.

Of course, having "too much" capital, as Aaronwrites, might diminish returns on equity to shareholdersof the big financial institutions. There are three possibleways to fix that: first, if everyone de-levers to safer lev-els, spreads will widen and the cost will be shifted toborrowers. Second, institutions can move their businessmix toward non-capital, non-risk taking, fee-based rev-enue streams. Thirdly, and perhaps unthinkably, theshare of revenues allocated for employee compensationcould be reduced.

In a paradigm of private profits and private risks, it isacceptable — though maybe not wise — for investors toenter into asymmetric arrangements. However, nowthat we have entered the era of private profits andsocialized risks, the public has an interest in making surethat the "too big to fail" financial institutions don't dojust that.

The taxpayer gets no benefit from the upside of thespeculation. As a result, it is in the public interest to insiston capital standards that fully consider the entire rangeof outcomes and that don't cut-off the tails.

✎ AARON BROWN is a risk manager at AQR Capital Management and the author of The Poker Face of Wall Street (John Wiley & Sons,2006). He is also serves on the GRR editorial board and is the former executive director and head of credit risk architecture at Morgan Stanley.He can be reached at [email protected] article expresses the personal opinions of the author, which are not necessarily shared byhis employer or any other entity.

2012

Documents

Transcript of 2012