The idle ramblings of a Jack of some trades, Master of none

Well, what do you know? According to a study done by a couple of European academics, finance professionals are not as good as undergraduate students at forecasting exchange rates over the ensuing period. While other professionals are generally better than Joe Public at predicting features of their domains of expertise (e.g. meteorologists generally get the weather right more often than the hoi polloi), it turns out that financiers have biases that mess up their abilities to forecast foreign exchange levels accurately.

Here's the abstract of the paper:
Professional forecasters in foreign exchange markets are not able to beat naive forecasts. In order to find reasons for this phenomenon we compare the empirical forecasts of experts with the experimentally generated forecasts of novices of the EUR/USD exchange rate in three different forecast horizons. Although the subjects are only provided with the realizations of the exchange rate and are not supported by any statistical procedures they outperform experts in accuracy. Professionals consistently expect a reversal of foregoing exchange rate changes whereas novices extrapolate trends. The judgemental forecasts appear to be unbiased and professionals appear to be biased. We demonstrate that professionals are influenced by the fundamental value - an irrelevant anchor in speculative exchange markets. The poor performance of the experts is not a common failure of human decision-making in market environments but caused by misleading information.
This should be dispiriting to those of us who work in the business. Why is it that a bunch of students who have no clue about the way foreign exchange markets behave outperform a bunch of highly-paid City types?

The reason appears to be two-fold. City economists have a cognitive bias - they 'know' that currencies tend to revert to a level given by what is called purchasing power parity. They, therefore, tend to predict a fall in a given exchange rate if they think it is above that parity level, or a rise if it is below. The students, on the other hand, merely project the visual trend they observe given the recent performance of the currency. Because there are brief periods of trending behaviour among currencies (although overall they tend to follow a random walk), the students get their predictions right more often than the professionals.

The other reason is that City professionals set much store by all the masses of data they encounter on their Bloomberg or Reuters terminals, and subconsciously assume that this gives them some sort of inside view into the markets. Unfortunately, over the periods they are asked to forecast - one to six months, say - the most recent exciting news (which usually has the greatest effect on their thinking) usually has little bearing.

The study can be criticised in a number of ways, though. Whereas the researchers recruited their pool of students, they didn't do so with the professionals. Instead, they used a consensus forecast as provided by some third party. The consensus forecast, in our experience, has little practical relevance. What is generally more informative is the spread in the forecast - how far apart are the individual forecasts? - or how many economists predict a level higher than the median as opposed to how many predict a level lower than the median. Also, it is true that these economists are rarely brought to task with respect to the accuracy of their forecasts. After all, no money is invested based on them. The ones who are tasked with making money are portfolio managers and the like. They are not really interested in forecasts of exchange rate levels - what is more important to them is the accuracy of the direction in which the exchange rate is likely to move.

A colleague recently conducted a brief study into the long-term dynamics of currency rates. He found that at any given time, a third of the exchange rates under consideration tend to be utterly random. He joked that we might as well stay at home a third of the time. But of course the problem is that we don't know which of these currencies will stop behaving randomly in the next period (a week or a month or a quarter). If we did, we could invest in those rates and listen to the 'kaching, kaching' noises as the profits poured in.

It is this constantly changing behaviour of currencies that makes them so difficult to model. All we can strive to do is to capture the brief periods of inefficiency and hope they last long enough to be monetised. It's all good fun.

- J. Leitner, R. Schmidt, "A systematic comparison of professional exchange rate forecasts with the judgmental forecasts of novices."


Post a Comment