Pragmatique, rationnel, indépendant

The banking crisis

In recent years, banks have taken more risk, have relied on contingent and unstable funding, have stretched their balance-sheet to the limit, and have generally based their operations on the assumption of the continuation of business as usual. The subprime crisis has triggered a painful process of unwinding.

In a previous article, we analyzed the causes and manifestations of the current financial crisis. In the present article, we try to understand why banks, and especially investment banks, find themselves in a mess unprecedented since the Great Depression.

First, banks have taken more risk. Whereas stress tests to calculate their value-at-risk used to involve market moves deviating from the mean by as much as 5 standard deviations, it is said that stress tests performed in 2007 could only go as far as 2 standard deviations. Anything more would have virtually bankrupted them. Basically, after a period of extremely stable macroeconomic and financial conditions (known as "the great moderation") accompanied by steadily rising asset prices, banks have started to ignore risk. The asymmetric compensation structures of their front office employees (those who make trading and investment decisions) typically protects them against downside risk and gives them a large portion of the profits they generate. This has clearly led to risk-seeking behaviors.

Second, the banks' balance sheets had generally been stretched to benefit from all the opportunities offered by the economic boom. A few months before resigning under pressure, Citigroup's CEO had thus declared: "As long as the music is playing, you've got to get up and dance". After the crisis, not only did the value-at-risk of the banks' portfolios rose, but they also had less capital to cushion losses and depreciations. In effect, a combination of factors dangerously diminished their capital ratios. After July 2008, investors stopping buying the loans that the banks had already originated. At the same time, the banks' funding dried up, as confidence and liquidity evaporated. To make things worse, they could not refinance their off-balance-sheet vehicles anymore. They therefore had to absorb these assets onto their balance sheets, further degrading their capital ratios. They were therefore left with tons of loans which they could not refinance, a situation which cannot last forever. Finally, their losses and write-offs completed the bleak picture, and explained the bailing-out of Bear-Sterns and the bankruptcy of Lehman Brothers in 2008. Banks such as Bank of America and Barclays which take deposits tended to have more capital to support losses, and a more prudent investment philosophy. Most have survived the current crisis. On the contrary, stand-alone investment banks such as Merrill Lynch and Morgan Stanley have generally faced huge difficulties, and are now seeking to merge with big commercial banks.

However, as The Economist observed in a recent editorial, no banking model was either totally vindicated or discredited. The quality of the management of different banks made a difference. While Credit Suisse was run by a CEO with an investment banking background, Citigroup was run by a lawyer, Chuck Prince, would presumably did not fully understand the intricacies of today’s financial markets and instruments.

Fourth, the promises of the originate-and-distribute lending model based on the securitization of loans were not fulfilled. In 2007, banks held on and off their balance sheets as much as 50% of AAA asset-backed securities. They were therefore still exposed to any systematic shock affecting the value of these securities.

Fifth, banks used extremely elaborated models to evaluate their exposure to risk. The problem is that these models were created and calibrated on the basis of recent historical experience, i.e., a period of unprecedented stability. Mathematical and econometric models are very useful and accurate in the short-run, as their parameters are relatively stable. But since they rely on the available data, they are typically not adapted to predict the future direction of the economy and the financial markets, and will be worthless in the context of a regime shift.

Sixth, several perverse feedback loops are in place. Banks which restrict lending impose that consumers and companies reduce their expenses, and restrict their borrowing abilities. This in turn adversely affects the economic outlook and potentially leads to a rise in defaults. Marking to market is also to blame: the values of assets is temporarily depressed by fire-sales from struggling institutions, which makes other institutions vulnerable. The vicious circles goes on. Lastly, capital adequacy ratios are pro-cylical under Basel 2: as the riskiness of assets (as measured by market spreads or ratings) rises in a downturn, capital ratios deteriorate at a time when raising capital is most difficult.

The exuberant gold rush of the past few years will in all likelihood give way to more prudent investment strategies and more robust banking models, especially with respect to funding. Compensation structures will be revised to account for the incentives they generate. Last but not least, a clearer division of labor will emerge in the financial system, in which hedge funds engage in speculation and risky arbitrage, while banks provide liquidity and hold solid long-term investments. In conclusion, past excesses do not invalidate the indispensability of banking and finance. They simply call for reform.

Pierre Chaigneau

Editeurs: Pierre Pâris, Bruno Lannes, Pierre Chaigneau.

Bank strategies, no size fits all. The Economist, August 16 2008.

[2] According to Citigroup, in “A re-emerging market”, Financial Times, July 1 2008.  
[3] Marking to market consists in equating the value of assets to their market price.