Predictions of recent elections have been wide of the mark, and delving into the reasons why can teach us lessons about the nature of investment.
It was not too long ago that online betting and prediction markets like Betfair and PredictIt were being held up as beacons of unerring accuracy in the otherwise murky world of political forecasting. But having seemingly been wrong-footed by the results of both the UK’s referendum on the European Union on 23 June, and the US presidential election on 8 November, faith in their omnipotence is being challenged. On the eve of the election, leading US prediction markets implied that Donald Trump had only a 22% chance of becoming president, and yet he was elected. How could the wisdom of crowds have failed so spectacularly when the stakes were so high?
Firstly, a disclaimer is required. Claiming prediction markets “got it wrong” is in itself a potential error. Saying there’s a 20% chance of an event occurring clearly is not the same thing as saying it definitely will not happen. For prediction markets to be well-calibrated, events that are deemed relatively unlikely should occur every so often: outsiders with a 20% chance should win in one in every five elections over the long-run. It remains possible that both Trump’s election and the vote for Brexit are simply unfortunately high-profile and freakishly clustered examples of non-consensual, low frequency outcomes.
But there is an intriguing alternative explanation: widespread faith in the power of online betting markets might, paradoxically, be undermining their accuracy. As prediction market guru David Rothschild has argued, one sign that predictions markets have too much faith placed in them is when they are suspiciously stable. This happened in the run up to both the Brexit vote and the US election.
Prediction markets have derived their reputation for accuracy by cleverly harnessing the ‘wisdom of crowds’ principle –that the collective opinion of a large group is more reliable than a single expert; but itis important to remember that this principle only applies to aggregating individual estimates that are truly independent. In the real world, markets rarely conform to this ideal of independence. As betting markets have grown in prominence and popularity, they may have come to be held in such high regard that they have started to influence the direction of future bets. If individual bets cease to be independent as a result of this positive feedback loop, then the market’s ability to incorporate new information is likely to be seriously impaired. Increasingly widespread faith in the infallibility of prediction markets may have ultimately proven self-defeating.
THE PROBLEM WITH PASSIVE
For active investors worrying that the rising popularity of passive investing represents a serious threat to their approach, this should be comforting news. Passive investing is, by definition, a clear violation of the assumption of independence that underpins the wisdom of crowds. Ironically, as indexing grows in prevalence, it becomes ever less likely that the conclusions of the efficient market hypothesis will be borne out in reality. If more and more investors would rather blindly channel their money into passive index trackers than make independent decisions based on fundamental analysis, the equity market will inevitably become less efficient at pricing individual companies. Unchecked faith in efficiency will breed inefficiency.
For truly active managers with the temperament and time-horizon to stand apart from the crowd, the rising popularity of indexing strategies should therefore present more of an opportunity than a threat. Just as prediction markets break down if too many participants lazily assume the odds are already correct, equity markets that are dominated by passive flows are highly unlikely to conform to the Platonic ideal of the efficient market. As markets become increasingly dominated by passive funds, the scope for active managers to exploit unsustainable discrepancies between market prices and real world values should increase. Amid all the noise around a passive revolution, the seeds of an active renaissance are already being silently sown. As Mr Trump might say, “it’s gonna be a beautiful thing”.