The perils of forecasting

Economic forecasts play an essential part in policy making. Why are they so often wrong?

It's probably not much fun being a forecaster, whether it's the weather or the economy which you're trying to predict. When you get it right, no-one seems to notice; but when you're wrong, you may face ridicule or vilification.

Typically, central banks need forecasts of the path of the economy in order to steer their monetary policy decisions. Hong Kong is something of an exception here, in that the Currency Board mechanism functions in accordance with pre-set rules, with little scope for discretionary action. So, in the monetary field we are tied very much to US policy - and thus influenced more by the consequences of what the Fed's forecasters think than by our own ability to forecast.

The situation is rather different for other areas of government. The budgetary process, for example, requires, as its very foundation, an assessment of the economic outlook some distance ahead. There are two basic components to such an assessment. One is an understanding of developments specific to Hong Kong and of the ways in which economic forces within Hong Kong interact. We can reasonably expect that there should be a reliable pool of local expertise in those spheres. The other is a forecast of what is likely to happen in the rest of the world, which is immensely important for an economy so externally oriented as Hong Kong's. In practical terms we cannot, however, be expected to have the resources to produce entirely independent forecasts of the prospects for the USA, Japan, Europe and so on. Instead we must rely largely on other sources - official forecasts in the respective countries, or the regular assessments made by organisations such as the International Monetary Fund, or the views of private sector analysts which are often conveniently summarised in publications providing "consensus" overviews.

Downward revisions to growth forecasts for Hong Kong in the course of this year have mainly been the result of progressive downward revisions to the global outlook. It may seem something of a cop-out to make such excuses, but it is a fact that forecasts for Hong Kong are very sensitive to the outlook in the wider world, over which we certainly have no control and for which we almost certainly don't possess any special predictive insight.

Although we are unavoidably driven to a certain dependence on forecasts, it is nevertheless surprising how readily we seem to accept a revised forecast, even though the very act of revision has revealed the frailty of the process. Our submissiveness may be explained to some extent by the fact that there is usually a seemingly persuasive consensus among forecasters; they exhibit strong herding tendencies; their predictions tend to cluster together. This in turn reflects partly the fact that they are mostly applying the same sort of model as to how the economy works, and partly the instinct for "safety in numbers": if you are all wrong together you may survive, whereas, if you are an outlier, the risks (e.g. to your career) of being wrong are probably greater than the rewards of being right. It's also worth noting that forecasters often experience particular difficulty in pinpointing turning points in the economic cycle.

For these various reasons, despite our dependence on them, we would be wise to exercise a degree of caution in adopting forecasts. And we must be ready for surprises, although we can reasonably hope that these turn out to be pleasant as often as unpleasant.

Tony Latter
Deputy Chief Executive

6 September 2001

Click here for previous articles in this column.

Document in Word format