Saturday, September 14, 2013

Wisdom Of The Week

This week, I read Nassim Taleb's new paper (coauthored with Constantine Sandis) called  The Skin In The Game Heuristic for Protection Against Tail Events. The paper was brilliant and it's sort of a follow up (an implementation guide per se) of his recent book Antifragile: Things That Gain from Disorder.

Standard economic theory makes an allowance for the agency problem, but not the compounding of moral hazard in the presence of informational opacity, particularly in what concerns high-impact events in fat tailed domains. But the ancients did; so did many aspects of moral philosophy. We propose a global and morally mandatory heuristic that anyone involved in an action which can possibly generate harm for others, even probabilistically, should be required to be exposed to some damage, regardless of context. While perhaps not sufficient, the heuristic is certainly necessary hence mandatory. It is supposed to counter risk hiding and transfer in the tails. We link the rule to various philosophical approaches to ethics and moral luck.

The problems and remedies are as follows:
  • First, consider policy makers and politicians. In a de- centralized system, say municipalities, these people are typically kept in check by feelings of shame upon harming others with their mistakes. In a large centralized system, the sources of error are not so visible. Spreadsheets do not make people feel shame. The penalty of shame is a factor that counts in favour of governments (and businesses) that are small, local, personal, and decentralized versus ones that are large, national or multi-national, anonymous, and centralised. When the latter fail, everybody except the culprit ends up paying the cost, leading to national and international measures of endebtment against future generations or "austerity ". These points against "big government " models should not be confused with the standard libertarian argument against states securing the welfare of their citizens, but only against doing so in a centralized fashion that enables people to hide behind bureaucratic anonymity. Much better to have a communitarian municipal approach:in situations in which we cannot enforce skin- in-the game we should change the system to lower the con- sequences of errors.
  • Second, we misunderstand the incentive structure of corporate managers. Counter to public perception, corpo- rate managers are not entrepreneurs. They are not what one could call agents of capitalism. Between 2000 and 2010, in the United States, the stock market lost (depend- ing how one measures it) up two trillion dollars for in- vestors, compared to leaving their funds in cash or treasury bills. It is tempting to think that since managers are paid on incentive, they would be incurring losses. Not at all: there is an irrational and unethical asymmetry. Because of the embedded option in their profession, managers received more than four hundred billion dollars in compensation. The manager who loses money does not return his bonus or incur a negative one4.The built-in optionality in the com- pensation of corporate managers can only be removed by forcing them to eat some of the losses5.
  • Third, there is a problem with applied and academic economists, quantitative modellers, and policy wonks. The reason economic models do not fit reality (fat-tailed reality) is that economists have no disincentive and are never penalized by their errors. So long as they please the journal editors, or produce cosmetically sound "scientific" papers, their work is fine. So we end up using models such as port- folio theory and similar methods without any remote em- pirical or mathematical reason. The solution is to prevent economists from teaching practitioners. Again this brings us to decentralization by a system where policy is decided at a local level by smaller units and hence in no need for economists.
  • Fourth, the predictors. Predictions in socioeconomic domains don’t work. Predictors are rarely harmed by their predictions. Yet we know that people take more risks after they see a numerical prediction. The solution is to ask — and only take into account— what the predictor has done (what he has in his portfolio), or is committed to doing in the future. It is unethical to drag people into exposures without incurring losses. Further, predictors work with bi- nary variables (Taleb and Tetlock, 2013), that is, "true" or "false" and play with the general public misunderstanding of tail events. They have the incentives to be right more of- ten than wrong, whereas people who have skin in the game do not mind being wrong more often than they are right, provided the wins are large enough. In other words, pre- dictors have an incentive to play the skewness game. The simple solution is as follows: predictors should be exposed to the variables they are predicting and should be subjected to the dictum "do not tell people what you think, tell them what you have in your portfolio" (Taleb, 2012, p.386) . Clearly predictions are harmful to people as, by the psychological mechanism of anchoring, they increases risk taking.
  • Fifth, to deal with warmongers, Ralph Nader has rightly proposed that those who vote in favor of war should subject themselves (or their own kin) to the draft.
We believe Skin in the game is a heuristic for a safe and just society. It is even more necessary under fat tailed environments. Opposed to this is the unethical practice of taking all the praise and benefits of good fortune whilst disassociating oneself from the results of bad luck or mis- calculation. We situate our view within the framework of ethical debates relating to the moral significance of actions whose effects result from ignorance and luck. We shall demonstrate how the idea of skin in the game can effectively resolve debates about (a) moral luck and (b) egoism vs. altruism, while successfully bypassing (c) debates between subjectivist and objectivist norms of action under uncertainty, by showing how their concerns are of no pragmatic concern.

No comments: