The trouble with predictions: Forecasting the future is the easiest job out there
Will the economy go into recession this year? Will the British electorate vote for Brexit? These are just two of the questions everyone would love to know the answer to and there is an army of people willing to forecast them. But is their track record at forecasting any good? For some of the big picture questions, would it help with investment decisions if the future was known with any certainty? The answer, alas, is that pundits frequently get it wrong.
To become a forecaster, you have to make two choices. First, choose the area you want to specialise in to become an “expert”. Second, decide if you are a conventional or an extreme forecaster. Conventional forecasters make forecasts that are not far off consensus, whereas extreme forecasters make wild predictions.
The key to success for conventional forecasters is to make sure their prediction is not far from the herd. Keynes, the British economist, captured this when he said “worldly wisdom teaches that it is better for reputation to fail conventionally than to succeed unconventionally”. The industry is full of consensus forecasts.
Extreme forecasters make wild and controversial predictions. But the key to success, as one economist once put it, is to “never put a timeframe on it. One day you will be right.” Another trick often employed by extreme forecasters is the use of the words “might” and “could”. They say “the oil price might go to $10”. That is a wild forecast. But it has no timeframe and it might happen, just as one day you might win the lottery.
So, becoming a forecaster is not that difficult: familiarise yourself with a topic, decide whether you want to hug the consensus or go out on a limb.
Pundits make predictions on a range of topics, but how good is their track record?
Take economic growth. Growth in the UK is expected to be, on average, 2.1 per cent in 2016 according to 50 economists. The highest forecast is 2.7 per cent and the lowest is 1.5 per cent. Not one is predicting a recession. Conventional forecasters are in good company in the field of forecasting growth. A study by the IMF shows that, in 2008 and 2009, not a single prediction was made for a recession, yet there were 13 and 49 recessions globally in 2008 and 2009 respectively. Ironically, the study was conducted by the IMF, a serial forecaster of economic growth.
Even if growth could be forecast with accuracy, how important is it in making investment decisions? The short answer is that it is not. Research shows economic growth has a poor relationship with equity market performance. The fact that an economy is growing is no guarantee that the equity market will perform well. So economic growth is not only impossible to predict, but knowing what growth will be is not helpful.
What about the oil price? Oil analysts are predicting that oil will finish the year at $47 in 2016. The highest prediction is $70 and the lowest is $38. Not a single one of them is predicting a decline. Further, it is illuminating that the difference between the highest and lowest prediction is as small as $32 for an asset class that has suffered huge swings. Those same oil analysts predicted the oil price would end 2015 at $70. In fact it came in at $37, but they live to make another prediction.
There are many examples in the world of forecasts that have gone wrong or events that nobody forecast. Forecasting is inherently difficult, yet there is a market for it. As long as there are unanswered questions, there will be lots of interest in what might happen.
I am not intending to bash forecasters. They are among the first to acknowledge how difficult it is to predict outcomes, and those who do not should be dismissed outright. The good ones frequently present arguments for and against a particular view and talk about the probability of outcomes. That is the way the world works. The only thing we know with any degree of certainty is what has happened in the past, and even that, at times, is impossible to know for sure.