Paul Strebel and Hongze Lu asked why, at the end of 2008, Citigroup, Merrill Lynch and UBS had well over $40 billion in sub-prime write-downs and credit losses, while some of their competitors were much less exposed. Their research into the situation revealed correlations of great import to today’s firms.
Where were the boards of directors of the world’s largest financial firms when their managements were loading up on risky collateralized debt obligations and structured investment vehicles? What happened to the riskmanagement procedures laid out in their corporate governance guidelines? We have found a strong correlation between the size of the sub-prime writedowns and losses, the lack of financial market expertise on bank boards, and the negative role played by dominating CEOs and board chairmen. Most importantly, we have found that risk management at the top of corporations, especially financial firms, is not about computer modelling; it’s about executive judgement.
Risk modelling can be very misleading It’s widely recognized that much of the current economic mayhem can be traced to banks and other financial institutions that allowed themselves to become subjected to ever-riskier loan portfolios. Bundles of loans tied to unwise mortgage lending (also known as sub-prime loans) became collateralized debt obligations (CDOs) and structured investment vehicles (SIVs) – names for huge collections of debts that were without reliable debtors and/or without properties whose intrinsic value equalled the amounts of the loans. As an enormous imbalance grew between assets and debts, many financial firms either needed (and received) a government bailout – or they went out of business. Lehman Brothers, a company with a history going back to 1850, is a prime example of a major financial firm that ultimately went bankrupt when it failed to garner government support. There are many other examples. How could it happen – especially when firms had developed systems and computer models to avoid such a calamity?
Even firms with sophisticated computer models can be led astray if they fail to adapt their underlying assumptions to changing conditions. For example, UBS had the reputation of being one of the world’s best managed banks with the most prudent risk management. Its risk-management philosophy was backed up by the latest quantitative modelling techniques, which determined the risk limits associated with individual assets and investments, enforced by the head office in Zurich. The separation of front-line operations and control was in line with best practices in risk management. But the mathematical sophistication of the models did not lend itself to judgement based on common sense. The models encouraged a mechanistic application of the risk assessments they produced. Moreover, the models were unreliable because they depended on data from rating agencies developed during the period of booming housing prices from 2000 to 2005.
Continue Reading in PDF Format . . .