Saturday, 14 March 2009

Behavioural financial regulators

Robert Peston has an approving discussion of Hector Sants' comments about irrationality in financial markets.

Sants says that "Markets have shown not to be rational" but (at least in Peston's quotes) does not distinguish between the agency problem, and the phenomena of behavioural irrationality at the individual level. It's an important distinction.

The agency problem is amenable to more conventional solutions: aligning the interests of shareholders and employees through share schemes, some types of options, long-term performance based bonuses. These problems and the simple solutions are well known and have been discussed in the management literature for decades. This is obviously not enough. Perhaps the focus of investors slipped a bit in recent years (Peston discusses the inclination of shareholders to sell rather than reform the companies they invest in). Alternatively the effectiveness of some solutions might have diminished as employees learned to game the system and ignored the cultural influences that had previous controlled them, becoming more rather than less rational. But broadly, I believe companies already have the tools to deal with this.

Individual behavioural irrationality is different. It is not at all well understood in companies; there is an assumption that people are broadly rational except when, occasionally, they're not. Even behavioural economic theory tends to operate in this way - on the idea that there are a number of arbitrary individual phenomena which don't fit into the rational agent model.

But a more systematic approach to irrationality would make a big difference. And if Hector Sants is genuinely proposing that behavioural issues, herding, miscalibration of risk and hyperbolic discounting can be taken into account by macroeconomic institutions, I applaud him.

Dani Rodrik has a brief discussion this week of the incorporation of behaviour into macroeconomics, and my article in VoxEU earlier this year discusses similar themes. I am developing the view that behavioural theory needs to be taken into account by existing macroeconomic institutions, or that a new body should be set up to focus, especially, on the management of systemic risk perception and asset valuation. We know something about ways to influence perceptions of risk - through framing or by recasting questions in a way that makes it easier for us to make rational decisions - and we have some ways to measure perception of risk too. So there is the option for an institution to make macro-level choices to control risk pricing at systemically stable levels.

Peston's final conclusion does not get into this thorny area, but simply says that we should act as responsible owners - as if that would solve the agency problem once and for all. Maybe it would make an impact, but I suspect this will address a fraction of the problem.

p.s. The FT has a feature "The Future of Capitalism" along the lines of my posting last week. Seems to be a good review of some aspects of the present and future, though it doesn't cover behavioural issues either. I suppose the future of economics is different to the future of capitalism. But capitalists would be well advised to really understand how their customers, employees and investors actually think and behave, if they want the future of capitalism to include them.

No comments: