Friday, 14 November 2008

Bounded rationality and agency

I have been working recently on an exploration of bounded rationality. This post from Robert Peston gives an interesting example of the insidious overlaps between this and other problems in economic theory - in this case the agency problem.

Barclays management preferred to accept a private £2.8bn investment from the Qatar and Abu Dhabi states, rather than take UK taxpayers' money, to the apparent detriment of existing shareholders. Peston suggests (in an earlier post which he links to in this one) that the reason may be that top Barclays management want to preserve their freedom to pay big salaries and bonuses to themselves (and each other).

Without addressing the accuracy of this suggestion, as I have no data either way, economic theory does shed some light on how this could happen. Assume for now that the facts are as Robert Peston says.

Classical theory says that firms act in their own interest (equivalently, the interests of their shareholders). In this case, that would mean taking the Treasury's money.

However, the agency problem, based on the fact that firms are not controlled by their shareholders but by professional managers, says that these managers will act in their own interests instead of that of the owners. Apart from the question of fairness, this also presents a problem for the proper workings of the economy because it means capital will not be allocated efficiently.

This is well-known, and so there are some defenses against it. With most major decisions such as a capital dilution, the shareholders have to vote on the proposal. Thus there is a vote on 24th November to determine whether it should go ahead.

However, this is where bounded rationality comes in. Because of the psychological phenomenon of anchoring, and the resulting question of credibility, it becomes critical for an organisation to follow through on what it says. In a purely rational world, any firm could re-evaluate at any time the expected utility of any decision, and change its mind.

But in real life, people become attached to ideas which have been decided or announced; and so to back down on anything may cause a crisis of confidence. Particularly in the financial markets, which rely so heavily on trust - it's no accident that credit and credibility have the same root - and where trust is in such short supply right now.

This has two consequences. One is that shareholders are being urged to recognise the danger of voting against the deal. Even though they are being mistreated, if they reject the proposal the value of the bank will be substantially damaged and they will lose out even more.

The other is that the Treasury has its own face to save, and is no longer likely to offer Barclays money on the same terms that were available a few weeks ago - because the original proposal itself cast a shadow on the Treasury's cash and thus on the other banks which have accepted it.

So because of anchoring, Barclays shareholders and British taxpayers are both potentially stuck with a deal which is in the interest of neither. Barclays management and the Qatari and Abu Dhabi royal families will do OK, though.

There are game-theoretic explanations for this behaviour which are distinct from the idea of anchoring. Both concepts - game theory and anchoring - are likely to be a factor, and it would certainly be interesting to explore whether the psychological phenomenon is what gives rise to the game theory solution; or indeed whether game theory explains why our psychology evolved this way.

No comments: