Despite the best efforts of the profession to move away from the imprecision of psychological concepts, many of the paradigms explaining economic behaviour failed to stand up to rigorous testing. Whilst these were initially explained away as anomalies which did not negate the underlying assumptions, developments in cognitive psychology from the 1960s began to be seen in some quarters as better explanations of certain forms of economic behaviour. Over the last 20-30 years, a number of these insights, derived from experimental psychology, have been applied to economic and financial decision making as better explanations of behaviour than the standard model. The new field of behavioural economics, for which Richard Thaler this week won the 2017 Nobel Prize for economics, examines what happens when we relax the assumptions of rationality and perfect information which underpin much of modern macroeconomics.
Amongst the range of judgement and decision biases which clearly violate the principle of rationality, behavioural economists have focused on factors such as overconfidence, wishful thinking, conservatism, belief perseverance, availability biases and anchoring (estimates based on an initial, often random, value). Using a combination of empirical evidence and thought experiments, academic researchers have demonstrated that some of these characteristics are at work in driving the expectations formation process. For example, evidence for the overconfidence hypothesis suggests that the confidence intervals assigned to outcomes tend to be too narrow. In a famous 1974 paper, Kahneman and Tversky[1] find evidence that whilst individuals often start off with an initial value in making estimates of future values, they are often reluctant to make big adjustments to this estimate when revising their assessment (the anchoring problem). This might go some way towards explaining why economists are reluctant to radically change their forecasts on a regular basis.
We could go on, but the point is made that there is enough empirical evidence to challenge the rational expectations assumption and thereby the idea that markets are efficient. This is a problem for many economists to deal with, for they have often spent years learning to deal with the sophisticated mathematics underpinning their stochastic models, which use rational expectations as a convenient simplifying assumption. It is an even bigger problem for the finance industry which spent many decades convincing itself that prices adequately reflect all available information.
One of the great ironies of a trading environment is that if rationality is common knowledge, there ought to be relatively little trading since a rational investor should be reluctant to buy if another investor is willing to sell. But the converse is true since the trading volume on world exchanges continues to rise. Indeed, much of the empirical evidence suggests that traders would make higher returns if they trade less frequently. Moreover, the same body of research indicates that investors are unwilling to sell assets which trade at a loss relative to the price at which they were purchased – behaviour which may well reflect an irrational belief in mean-reversion.
There are also clear patterns in purchasing decisions where there is evidence to suggest that investors buy stocks which have previously been big winners (in the hope that this performance will be repeated) or big losers (in the expectation of mean reverting performance). Neither of these is consistent with rational behaviour, but one reason why investors may follow such strategies is that they do not have time to systematically analyse the whole range of stocks. The choice of which to sell is limited to the range of stocks currently owned, but the range of stocks from which investors can choose to buy is enormous, and they are attracted to the outliers in what is known as the attention effect.
Clearly, markets display characteristics at odds with efficiency and expectations are not always formed rationally. The world thus owes a debt to Thaler and his colleagues for pointing out some of the absurdities in conventional economic thinking. Behavioural economic does not have all the answers. In the minds of many people it is just a collection of theories which can only ever be tested on small samples and thus its wider applicability is limited. But to the extent that it makes us think about some of the reasons why economics has not always come up with the right answers, Thaler’s award is well deserved.
[1] Kahneman,
D. and Tversky, A. (1974) 'Judgement Under Uncertainty: Heuristics and Biases,'
Science, 185, 1124-1131
No comments:
Post a Comment