The Rational Argumentator

JPMorgan Chase & Co., one of the nation’s leading banks, revealed in May that a London trader racked up losses reportedly amounting to $2.3 billion over a 15-day period. The losses averaged over $150 million per day, sometimes hitting $200 million daily. The bank originally stated the trades were done to hedge possible losses on assets that might suffer due to Europe’s economic woes. There is now doubt whether it was a hedge or just a risky financial bet.

A hedge is a financial transaction designed to offset possible losses in an asset or good already owned. The classic hedge occurs when a farmer sells his crop in a futures market for delivery at a specified date after harvesting. He sells today what he will only produce tomorrow, and locks in the price. If the price at harvest time is lower than today’s price, he makes money on the forward contract, while losing a corresponding amount of money on the crops in the ground. In a perfect hedge the gains and losses should exactly offset each other.

How did JPMorgan suffer such large losses on its hedges, and what are the lessons?

It appears the London trader entered into financial transactions on the basis of observed relationships among various bond indices. The market relationships broke down. The indices moved differently from what historical patterns or financial models predicted. Such a breakdown has been at the heart of a number of spectacular financial collapses, notably that of Long-Term Capital Management (LTCM) in 1998 and a number of others during the financial meltdown of 2007–08.

LTCM invested the money of rich clients in financial bets based on the expected relationships among the prices of various assets. According to Nicole Gelinas in After the Fall: Saving Capitalism from Wall Street—and Washington, at the time of its collapse LTCM had $2.3 billion of client money. By borrowing, it leveraged that investment 53 to 1. Further, it employed derivatives to further magnify its bets so that its total obligations were a fantastic $1.25 trillion.

A derivative is any security whose price movements depend on (are derived from) movements in an underlying asset. “Puts” and “calls” on equity shares are relatively simple derivatives familiar to many. Asset prices, like various bonds, move in predictable ways with respect to each other, and values of derivatives linked to the assets similarly move in a predictable fashion with respect to the prices of the underlying assets—in normal times.

But the summer of 1998 was not a normal time. There was turmoil in Asian financial markets, then Russia threatened to default on its domestic debt. Global credit and liquidity dried up, and LTCM could not fund itself. It collapsed spectacularly.

A decade later there was turmoil in housing finance. The housing bubble was bursting. Mortgage lenders were under pressure, and some were failing. Many mortgages had been packed together in mortgage-backed securities, which were sold to or guaranteed by Fannie Mae and Freddie Mac. Fannie and Freddie, allegedly private entities but in reality guaranteed by the government, were failing. Lehman Brothers, an investment bank, was heavily involved in housing finance; it borrowed short-term, even overnight, to finance long-term holdings; it employed heavy leverage; and it made liberal use of derivatives contracts. It declared bankruptcy on September 15, 2008.

The specifics varied between 1998 and 2008, and between LTCM and Lehman. But the reliance on certain asset prices moving in predictable fashion was one shared element. So, too, was the heavy use of borrowed money (leverage) and the reliance on derivatives contracts. The volatility of complex derivatives contracts led legendary investor Warren Buffett to characterize them as “financial weapons of mass destruction.”

The Usual Suspects

In short there is nothing new in what happened to JPMorgan. It claimed it was not trying to make risky financial bets, but hedge risks already booked on its balance sheet. While details of the trades that led to losses are sketchy at this writing, they apparently employed both leverage and derivatives. As documented here, these are elements present in major financial blowups and collapses going back decades (and further). LTCM, Lehman, and Fannie and Freddie all thought they had at least some of their risks hedged. But hedges have a tendency to unravel just when needed most: in times of financial turmoil. Even so, financial institutions permit their traders to make the same kinds of dangerous bets over and over again. We used to have financial crises every decade or so. Now the cycle seems to be halved.

In the past I have dubbed today’s banking practice of placing dangerous financial bets “casino banking.” It differs little from the activities conducted at gaming tables in Las Vegas and has little or no reference to the fundamentally healthy activity of matching viable businesses with capital and credit.

In a Cato Policy Analysis, “Capital Inadequacies: The Dismal Failure of the Basel Regime of Bank Capital Regulation,” Kevin Dowd and three coauthors examined some of the technical problems with standard risk models used by large banks. It is an exhaustive analysis, and I commend it to those interested. The authors delve into many issues, but concentrate on the many flaws of the complex mathematical models used by banks to control risks.

In August 2007 Goldman Sachs Chief Financial Officer David Viniar puzzled over a series of “25-standard deviation moves” in financial markets affecting Goldman. (Returns deviated from their expected values by 25 standard deviations, a measure of volatility.) Such moves should occur once every 10-to-the-137th-power years if the assumptions of the risk model were correct (a Gaussian, or “normal,” distribution of returns). As Dowd and his coauthors put it, “Such an event is about as likely as Hell freezing over. The occurrence of even a single such event is therefore conclusive proof that financial returns are not Gaussian—or even remotely so.” And yet there were several in a matter of days. In Dowd & Co.’s telling, the models lie, the banks swear to it, and the regulators pretend to believe them. All of this goes to answer how the losses at Morgan might have happened. Traders rely on flawed models to execute their trades.

Now to the Lessons

Major financial institutions continue to take on large risks. Why? Assume the trades made by Morgan really were to hedge the bank’s exposure to events in Europe. That implies, of course, that risky investments had already been put in place (since they then needed to be hedged). Additionally, the risks were so complex that even a highly skilled staff (which Morgan certainly employs) could not successfully execute hedges on them.

Reports indicate that senior management and the board of directors were aware of the trades and exercising oversight. The fact that the losses were incurred anyway confirms what many of us have been arguing. Major financial institutions are at once very large and very complex. They are too large and too complex to manage. That is in part what beset Citigroup in the 2000s and now Morgan, which has until now been recognized as a well-managed institution.

If ordinary market forces were at work, these institutions would shrink to manageable sizes and levels of complexity. Ordinary market forces are not at work, however. Public policy rewards size (and the complexity that accompanies it). Major financial institutions know from experience that they will be bailed out when they incur losses that threaten their survival. Morgan’s losses do not appear to fall into that category, but they illustrate how bad incentives lead to bad outcomes.

Minding Our Business

Some commentators have argued that politicians and the public have no business in Morgan’s losses. Only Morgan’s stockholders, who saw its share price drop over 9 percent in one day, and senior management and traders who lost their jobs should have an interest. But in fact losses incurred at major financial institutions are the business of taxpayers because government policy has made them their business.

Large financial institutions will continue taking on excessive risks so long as they know they can offload the losses onto taxpayers if needed. That is the policy summarized as “too big to fail.” Let us not forget the Troubled Asset Relief Program (TARP), signed into law by President George W. Bush in October 2008. It was a $700 billion boondoggle to transfer taxpayer money to stockholders and creditors of major banks—and to their senior management; don’t forget the bonuses paid out of the funds.

Banks may be too big and complex to close immediately, but no institution is too big to fail. Failure means the stockholders and possibly the bondholders are wiped out. Until that discipline is reintroduced (having once existed), there will be more big financial bets going bad at these banks.

Changing the bailout policy will not be easy because of what is known as the time-inconsistency problem. Having bailed out so many companies so many times, the federal government cannot credibly commit in advance not to do so in the future. It can say no to future bailouts today, but people know that when financial collapse hits tomorrow, government will say yes once again. The promises made today will not match the government’s future actions. There is inconsistency between words and deeds across time.

What to do in the meantime? The Volcker Rule was a modest attempt to rein in risk-taking. Former Fed Chairman Paul Volcker wanted to stop banks from making risky trades on their own books (as opposed to executing trades for customers). Industry lobbying has hopelessly complicated the rule and delayed its issuance.

Morgan’s chief executive officer, James Dimon, asserted the London trades would not have violated the rule. If true, it suggests that an even stronger rule needs to be in place. Various suggestions have been made to address excessive risk-taking by financial firms backed by the taxpayers. It is time to take them more seriously.

Gerald O’Driscoll is a senior fellow at the Cato Institute. With Mario J. Rizzo, he coauthored The Economics of Time and Ignorance.

This article was published by The Foundation for Economic Education and may be freely distributed, subject to a Creative Commons Attribution United States License, which requires that credit be given to the author.