Now that the jingoistic gush fest that is the Olympics is over we can all recover from the notion that every Australian athlete who actually performed up to expectations was a genius and that those that failed to do so still had a big future in front of them (most likely in hospitality) we can sit down at look at the notion of predictions. One of the interesting things about social science style events such as the Olympics is that they bring forth a number of predictions and as the games were approaching I collected a number of these predictions for future reference. Knowing full well that most would miss the mark – some spectacularly as you can see by the chart below.
Event such as politics, economics and sport are extraordinarily hard to make predictions about because of their extraordinarily chaotic nature and in the case of sport the reliance upon humans not being human. Consider the recent Brexit vote; forecasters had the Stay vote as being on track for winning. Right up until the time it didn’t. Likewise predictions surrounding the FTSE were equally off the mark. When it became apparent that the Leave vote was going to win the common refrain from experts was that the FTSE would drop through the floor. It has now moved higher than before the referendum.
The utter failure of forecasters to do better than guessing has been known for a long time and was first brought to widespread attention by Phillip Tetlock in the 1980’s. He found that forecasters performed little better than groups of knowledgeable amateurs and it is the loudness of their argument that carries the day. Not only are expert forecasters wrong most of the time but they are also unable to self correct or learn from their failures. One of the great things about science, particularly quantitative fields such as mathematics and physics is that it is self correcting and this self correction brings about a unique form of internal consistency. Social sciences because they are in part driven by personality lack this functionality – there is no capacity for self correction. To get a sense of this consider the graph below which tracks the stockmarket predictions of Marc Faber.
As you can see Faber has been predicting a stockmarket crash of biblical proportion every year since Moses played full back for the Mount Sinai Under 11B’s. What is interesting about this sort of prediction and its absolute failure is that if you Google Marc Faber you get an endless stream of news sites reporting breathlessly on his latest prognostication. Yet none report on his consistent failure, failures that have undoubtedly cost investors a small fortune. What is more intriguing is that Faber has no mechanism for correcting his internal failure, there seems to be no acknowledgement of previous errors. This is not a failing unique to him, virtually everyone in the public arena who makes predictions that are consistently wrong fail to acknowledge their errors. This failure to acknowledge error is a natural human phenomenon – we all act to defend our egos and in the case of public guesswork we tend to fall in love with our guesses. This is why you will never see a politician admit that they are wrong, leaving aside that most are self serving, self obsessed marginal sociopaths.
As the market reminds me everyday i actually know very little about trading and I rely upon the market to tell me what it knows. Trouble for traders occurs when they start to think they actually know something.