The crisis engulfing the global banking sector continues to worsen. The looming downgrades and potential bankruptcies of the Jerome Kerviel bond insurers could be the straw that breaks the financial system’s back. Many banks have used monolines to wrap their products and also bought credit-default swaps through them.
The power of rewards that lead to repeated actions and the flawed compensation structure that led to misaligned incentives could be one mental model to understand what has happened.
As Raghuram Rajan pointed out in the Financial Times (9 January), the compensation practices in the financial sector are deeply flawed. Compensation is based on the so-called “alpha” that a manager of financial assets generates. There are three sources of “alpha”:
1) Truly special abilities in identifying undervalued assets (e.g., Warren Buffett).
2) Activism: using financial resources to create, or obtain control over, real assets and to use the control to change the payout obtained on the financial investment.
3) Financial engineering: financial innovation or creating securities that appeal to particular investors.
Many managers create “fake alpha.” They appear to create excess returns but are taking on “tail” risks. These produce a steady return most of the time, as compensation for the very rare, very negative returns (“black swans”). The AAA-rated collateralized debt obligations (CDOs) generated higher returns than similar AAA-rated bonds. The “tail risk”, so evident in hindsight, of the CDO defaulting was not as small as perceived. The excess return was compensation for that.
The credit rating agencies that rated these securities AAA because of their “insured” status were themselves wrongly incentivized. They were compensated by the issuers of the securities rather than investors. Furthermore, the ratings war as to who assigned the highest ratings for junk became a classic prisoners’ dilemma game.
The managers themselves seem to have been suffering from self-serving bias (i.e., an overly positive view of their own abilities and an overly over-optimistic view of the future). They have been hard-wired with other defects.
First, there was self-deception and denial, for they seem to have indulged in collective wishful thinking. Bias from consistency tendency (looking for evidence to confirm their optimistic beliefs even after problems surfaced) led to status quo bias or the do-nothing syndrome. Impatience in valuing the present more highly than the future again caused by incentives that made them so myopic.
Bias from envy from managers who were making large returns with apparently no extra risk leads to distortion by contrast comparison, because the steady escalation of commitments must have seemed incrementally small, caused by anchoring to what seemed like small relative numbers. Social proof, which led to imitating the behaviour of their peers.
Bias from overinfluence by authority, in that the CEOs of the banks which have suffered the most seem to have been run by people who did not have a “trading” or “market” background and they were swayed by the “experts” they were overseeing. This bias led them to sensemaking, in that they were too quick to draw conclusions and may have become reason-respecting, in that they complied with requests from their subordinates merely because they had been given some reason leading to a do-something syndrome, all caused by mental confusion from stress.
Competition for business must have led to the winner’s curse, i.e., overestimating the value of the securities and overestimating predictive ability. Collectively they did not foresee their actions would have adverse systemic consequences and the implications to their balance sheets if things went wrong. They failed to consider the increasing instability due to their actions caused a phase change as a tipping point was reached, and that a system is only as strong as its weakest link.
In factoring the odds, the managers seemed to have underestimated risk exposure where the frequency and magnitude of consequences was unknown because of the novelty of the securities. They seem to have underestimated the number of possible outcomes for the unwanted events currently being witnessed. They certainly do not seem to have correctly calculated expected values or else they would not be in the hole they are currently in. They, in fact, did not consider the consequences of being wrong. They probably worked in an “illusion of control” over what were probabilistic events and thus did not factor in a “margin of safety”. In the limited history of the securities, the managers overestimated the evidence from the small sample of data.
As Gillian Tett noted in her column in the Financial Times, one should look at the political structures of the survivors. CEOs of the relatively unscathed banks “tended to be meddlers—very hands on.” They had a direct career experience in trading and managing market risk. Thus, the mindset was different from being a lawyer or a salesman. Furthermore, the losers had more hierarchical structures in which the different “business lines have existed like warring tribes, answerable only to the chief. Moreover, the most profitable tribe has inevitably wielded the most power—and thus was untouchable and inscrutable to everyone else”.
This is proving to be a game of chicken between the regulators and the players (banks and monoline insurers). In a game of chicken, two cars drive towards each other. The first driver who turns loses. Of course, if neither car swerves then there is a crash. The best outcome for each player results when he goes straight, while his opponent turns. Insane players have a massive edge in a game of chicken. At this point of time, the jury is out given the level of insanity in the system.
(Chetan Parikh is director of Jeetay Investments Pvt. Ltd. Comment at firstname.lastname@example.org)