Recently a distinguished jurist asked me, "How is it possible that the
financial industry - the smartest guys out there - did so many stupid things?"
In fact, the financial industry is full of people who know perfectly well that
they are mediocre, but who nonetheless want to make a great deal of money. So
A few months after the energy trading giant Enron filed for bankruptcy in
December 2001, I went up to the partners' dining room at the New York
headquarters of Credit Suisse. Enron, which Fortune had named "America's most
innovative company" for six years in a row, turned out to be one of the biggest
in American history, and several of its senior managers went to prison.
The bank used to offer an excellent free lunch to its managing directors to
promote collegiality across divisions. Next to me were a group of investment
bankers, and I eavesdropped on their gossip while I ate alone.
"Who was the team captain for the Enron relationship last year?," one of them
asked. Wasn't it [censored], who arranged the share trust financing?"
"I thought it was [censored], who set up the off-balance sheet financing
through the holding company subsidiaries," said another.
"No, it had to be [censored], who did all the derivatives trades with offshore
entities," offered a third.
"What you're telling me," said the fourth (and most senior) of the group, "is
that there wasn't any team captain for Enron. Everybody was doing their deals
without telling anyone else, because they were afraid someone else would take
away their business."
In a nutshell, all the department heads doing business with Enron deliberately
occulted their actions from their colleagues and from the management of Credit
Suisse, in order to maximize their personal gain at the risk of the bank's
reputation and its customers' money.
That sort of cupidity and backstabbing helps explain why Enron was able to
defraud the public on such a grand scale. It wasn't just Credit Suisse, which
was a minor player in the grand deception. Enron's creditors sued 11 other
banks for abetting Enron's fraud about its true financial condition; the banks
ultimately paid $20 billion in damages.
Only one of the Enron transactions slithered past my desk on the fixed income
trading floor, where I ran the bank's Credit Strategy group: a so-called "share
trust" financing. Enron would borrow money by issuing bonds, and if it didn't
have the money to pay the bondholders, it would issue common stock and give it
It might seem obvious that the stock of a company that can't pay its debt can't
be worth very much; this kind of arrangement used to be known as a death-spiral
convertible bond. But the capital market types pitched it with a straight face.
We ignored it.
In some ways, the 2008 financial collapse was Enron writ large. The ratings
agencies - Moody's, Standard and Poor's, and Fitch - agreed that it was
inconceivable that more than a third of a pool of subprime mortgages could
default. If the banks got one group of investors to accept the first 35% of
losses on the pool, the ratings agencies would label the rest of the pool
default-proof, and give it a triple-A rating.
The banks then went to the Federal Reserve and asked permission to increase
leverage on what the ratings agencies called ultra-safe securities. Normally
banks can hold about $12 of loans or securities for every $1 of their own
money, but the Fed allowed them to own $70 of the phony subprime triple-A's for
every $1 of shareholders' capital. Some of those bonds are now trading at 33
cents on the dollar.
That's one reason the banks had massive losses, but not the only one. Banks
(and insurance companies) were writing huge amounts of guarantees on phony
triple-A-rated debt, generating up-front fee income in return for turning the
banks' balance sheet into a toxic waste dump.
By the time that Lehman Brothers went under in September 2008, there is no way
that its chairman, Dick Fuld, could have calculated the volume of losses for
which his bank was on the hook. Every derivatives and structuring desk was
taking in all the fees and back-loading all the risk it could, telling the risk
managers as little as possible.
What do we want from Wall Street now?
Under the new regime banks have radically reduced proprietary trading, that is,
speculating for their own account. The entire business of packaging mortgages
or loans into trusts and structuring derivative securities - so-called
collateralized debt obligations - has vanished. The banks have radically
reduced risk on their books.
Global issuance of collateralized debt obligations by quarter
The volume of ordinary credit-default swaps (insurance contracts on individual
corporations or corporate indices) has fallen by half since the crisis (the net
total is much smaller because many of these contracts cancel out).
Notional amount of credit-default swaps outstanding
The exotic securities are extinct, except for a shrinking volume of older
issues locked away in bank or hedge-fund portfolios, and even the plain-vanilla
securities are at half their previous volume.
Banks, meanwhile, are not making any money. They can't make money. They can't
find earning assets to put on their books. Their share prices languish not too
far above the flat-line levels of late 2008. They can't raise capital. And they
have no one to lend to. Federal Reserve policy is ineffective because no matter
what the central bank does, the banks won't take on risk.
We don't want the banks to bulk up on risk again and threaten the financial
system. But we want them to take on risk to promote economic growth. Bank
management has no idea what to do, except to hold on to their jobs while the
legislature and the regulators decide what they want. The price of bank shares,
meanwhile, gyrates wildly, because investors have no way to gauge the risks
that might lurk on bank balance sheets.
Central banks have proposed any number of solutions, few of which seem
convincing. On October 24, Andrew Haldane of the Bank of England suggested that
the problem lay in the incentives offered to bank managers, who would always
choose to maximize leverage.
He proposed financing banks with instruments that penalize excessive risk, such
as contingent convertible securities (instruments that pay a coupon like debt
until bank capital falls below the safety point, at which point they convert
into common stock). That sounds like the old death-spiral convertible bonds.
The root of the problem, I believe, lies in the measurement of risk. The
incentive to cheat always will be there as long as bankers can represent a
sow's ear as a silk purse. Both managers as well as the public need to measure
risk, such that they understand the way that investments or innovations add to
or reduce risk.
Some popular finance writers insist that risks are inherently impossible to
measure, because conventional risk models based on the normal distribution of
returns don't assign enough weight to the likelihood of extreme outcomes. In
fact, the point of risk models is to estimate the likelihood of an extreme
outcome, and the banks have reasonably good models of borrower defaults which
do just that.
The ratings agencies became the arbiters of risk not because they had good
models (they did not) or because they employed particularly skilled analysts,
but because they were eminently corruptible. In October 2008, congressional
investigators found e-mails from Moody's credit analysts warning management
that they had "sold our soul to the devil for revenue".
For every Collateralized Debt Obligations, Moody's and Standard and Poor's
received a fee in the low six figures, and these fees made up the bulk of their
revenues. They acted as a adjunct to the investment banks' structuring teams,
advising them on the best way to game their own models.
Why hasn't the government prosecuted the rating agencies for fraud? Incredibly,
the ratings agencies take the position that their ratings are "opinions" with
the same legal status as a newspaper editorial. Newspaper editorialists,
though, don't take money from big advertisers for endorsing their products. But
whether or not the ratings agencies are held accountable for past malfeasance,
they have become a liability. They need to be replaced. But by whom, and what?
There is a way to do it, in a way that investors will come to trust within a
reasonable time frame. Make risk modeling a transparent business with universal
access to data and models. The regulators should require large financial
institutions to provide their internal data on defaults and delinquencies to
Many of the large banks have enormous amounts of data describing the
characteristics of defaulting borrowers and the circumstances under which they
defaulted. The Federal Reserve should collate this data into a single set, and
make it available for download on its website.
The large banks also should be required to publish their internal risk models.
To require a private company to divulge proprietary information is a burden,
but in light of the 2008 catastrophe, the public has a right to know how the
banks are measuring their own risk.
Once the data and models are available, independent analysts (including the
rating agencies) will have an independent capacity to measure the riskiness of
bank portfolios. Public scrutiny and third-party analysis will compel bank
senior management to improve risk measurement in order not to appear less
transparent than the competition.
None of this is particularly difficult. It is a matter of sorting data and
crunching numbers and comparing results. It will take the investing public a
bit of time to learn what actually occurs in bank portfolios. But this approach
could produce robust measures of risk, and allow banks to take on the kind of
risk that promotes economic growth, while giving the market the means to punish
banks that take on excessive risk.