Link to Google Doc
Today I am going to introduce the blog to the concept of prediction markets, and the superior knowledge of markets over expert guesses. This post will be part of a miniseries of posts in which I am going to introduce another way in which one could use policy dollars and prediction markets to improve politics.
Prediction markets is a way to gather information regarding uncertain events in an accessible and precise way. They are set up around an event which can be answered in two ways, either the event happens, or it doesn’t. For example, a prediction market could be set up around whether or not Donald Trump is elected president in the next U.S election.[1] Market participants can then buy contracts for either yes or no, depending on what they think will happen in the election. The contract gives the owner the right to a payout of one dollar, given that the outcome is what the buyer predicted. If I buy a contract which says that I believe that Donald Trump will win the election, then I get one dollar when he wins. If he doesn’t win, the contract is worthless.
The price of the contracts depends on the probability that the market deems to each outcome, every contract will be worth between 0 and 1 dollar. Which means that the price of a contract can be interpreted as a market probability for the outcome of the event. For example, the current price for a yes-contract for Donald Trump winning the 2020 election is 46 cents. This would mean that the aggregated beliefs in this market give Trump the odds of 23:27 of winning the election, i.e. he wins 23 out of 50 elections. Depending on your own belief about the event, you can either buy or sell if the price is not reflecting your belief. Say the price of a contract is 35 cents, but you believe that the actual probability of the event happening is 50. Then you should buy the contract at any price under or at 50 cents, because you are making an expected profit at all those prices according to your beliefs regarding the event. At the same time, if the price of the contract is higher than your belief, then you should sell contracts or buy the contract that gives a payoff in the opposite outcome. This mechanism makes each buyer put the money where her mouth is, the buyer believes that she has enough information to make an accurate prediction.
The market mechanism of “putting your money where your mouth is” ought to lead to increased information search to not lose money, leading to relatively better-informed decisions and predictions compared to polling. This becomes relevant and important in the coming posts, where I will discuss the uses for prediction markets in society, especially in politics.
I find the subject of market mechanisms and information gathering along with the psychological aspects to be quite interesting, which is why I would want to elaborate on the topic and to go on a journey to find ways to use them productively. Prediction markets can also teach us a very important lesson about uncertainty and decisions; never judge a decision based on its outcome. Especially when uncertainty is involved.
[1] Many prediction markets are used for this exact purpose, to predict presidential elections. It has been shown that prediction markets are more accurate than polls in this area.
1 thought on “Prediction Markets”
Pingback: Predicting the Unverifiable – Samuel Svensson