Are Betting Markets Dead?
In class we talked about the wealth of knowledge prediction markets can provide and their impressive historical accuracy. Prediction markets are so powerful because they combine the knowledge, and feelings of many different individuals with accuracy that polls can’t obtain. This is because of the nature of prediction markets: Since money must be placed on an outcome, each better is trying to pick an outcome they truly believe is the most likely, in order to maximize their payoff (like how the betters in auctions try to maximize their own payoffs in our study of auction theory). However, most major prediction markets failed to predict the outcome of “Brexit” and the 2016 United States Presidential election which has caused politicians and economists to start to doubt modern prediction markets.
There are a number of theories as to why the prediction markets failed during Brexit and the US elections. Some believe that a small number of wealthy betters placed large bets on the wrong outcomes and swayed the markets. Once one outcome has gained an advantage, the markets often tend to “cascade” towards extreme probabilities. This theory seems plausible because there are large wealthy groups that supported Hillary and Britain staying in the UK (Wall Street and the LadBrokes gambling firm respectively). This theory also works well with other theories claiming that confirmation bias played a large roll in leading the prediction markets astray. Many betters we’re very emotionally invested in the outcomes of Brexit and the US elections. This may have led them to look for data that confirmed their beliefs, even if it wasn’t exactly appropriate like the 2014 referendum for Scotland’s independence and early polling information in the election. Confirmation bias may have also caused betters to stick to their original beliefs even after new information that conflicted with their beliefs came out, like election polls after the FBI re-opened their investigation into Hillary’s emails.
Some economists believe that it is possible that the prediction markets we’re correct or not very far off. They believe that the results of the betting markets are being improperly evaluated. When the betting markets give Hillary an 85% chance of winning, they do not say that Hillary will win, but rather that Hillary is more likely to win. It is possible that Hillary did have an 85% chance of winning the day of the election, and that Donald Trump happened to get lucky and win with a 15% chance. Whichever side you believe, if you’re a gambler it may be worth your while to look at recent prediction markets and betting against them.
Source: http://www.economist.com/blogs/graphicdetail/2016/06/polls-versus-prediction-markets