What Are Predictions Markets?

In its simplest form, a market is where services or goods, tangible or intangible, that have value are bought and sold. Prediction markets work similarly, but the product is a position on the outcome of an unknown future event. These are open markets where specific outcomes on real world events can be predicted via financial incentives, whereby market prices (or odds) indicate what the crowd thinks the probability of an event is. Prediction markets utilize the ‘wisdom of the crowd’ with results that are often more accurate than the most accurate expert forecasts and estimates.

Prediction markets can be thought of as belonging to the more general concept of crowdsourcing which is specially designed to aggregate information on particular topics of interest. They function as a decentralized oracle, where participants take positions on the likelihood that a certain outcome(s) will occur. From political outcomes to sports matches, weather events, or crypto prices, these positions are then used to create an aggregate prediction about what will happen – and this market value can then be used by others to make decisions about their own behavior.

Wisdom of the crowd is a theory (Surowiecki, 2004) that assumes large crowds are collectively smarter than individual experts. It believes that the collective knowledge and opinions of a group are better at decision-making, problem-solving, and innovating than an individual. By incentivizing correct predictions with profit, prediction markets provide an incentive for people to gather information, evaluate it and make accurate forecasts. Studies show that market based prediction methods are proving to be more accurate at predicting inflation than traditional methods of forecasting. For example, while 2/10 economists accurately predict inflation, 6/7 prediction markets accurately predict inflation (Dye, 2008)

While prediction markets date back to the 16th century (Snowberg et al., 2013), more recently they have developed an interesting problem: a feedback mechanism has arisen whereby odds reify themselves. Traders are treating market odds as correct probabilities and not updating enough based on outside information: belief in the correctness of prediction markets causes them to be too stable (Rothschild, 2016), limiting the opportunity to capitalize on market inefficiencies. Faced with this dynamic, many prediction markets attempted to address and alleviate the concerns by centralizing odds making and clearing. A classic example would be a Las Vegas Casino or online sports book: odds and payouts are calculated based on proprietary logic, while positions are aggregated.

While this may seem to be a good solution for the centralized entities themselves, further inefficiencies are introduced for the market participant. The opaque, black box pricing mechanisms of centralized prediction markets introduce significant fees so that participants do not receive optimal market-derived pricing for their positions. Furthermore, counterparty risk is magnified within these centralized entities because their operations are often undercapitalized relative to a worst-case scenario (where they have to make considerable payouts all at the same time). We see this happen consistently –- not only with prediction markets, casinos and sports books, but also with bank runs, insurance claims following natural disasters, and more.