WHAT IS IT?

Prediction markets in their most basic form are exchange-traded markets that allow participants to trade (i.e. take a position) on the probability of a future event occurring. Every one of us has encountered these markets in some form or another with the most obvious example being financial markets. Market prices for stocks, bonds, commodities, etc. trade on the basis of different participants taking divergent views on the value of these financial instruments.

The execution levels of very liquid markets occur at prices that (in theory) represent the aggregated probability estimate of every single market participant.

WHY DOES IT MATTER?

Financial markets are obviously a massive industry employing millions of people, affecting (almost) every single one of our lives to varying degrees. But the incentive system behind them is what makes them so powerful. Consider that:

• Prediction markets incentivize those with good information to act on it& (over time) weed out those who either (i) act on poor information or (ii) struggle to interpret the information they have to make accurate predictions. For the latter group, after continued prediction failures/losses (& in turn continued financial losses), there is no financial incentive to keep participating. Those actors fall away from the stage. That leaves only those who are able to continue making good predictions, and promotes a steadily improving market from an information perspective. This system also encourages those with good information to come forward & participate, while simultaneously requiring those without it to go out and find it. Analysts pour over 10K’s, attend conferences, speak to customers/suppliers because they’re searching for important insights that might tell them something they otherwise wouldn’t know. The same structure applies to all prediction markets: those lacking relevant information are encouraged to seek it out.
• The degree of information strength matters. That is to say that those with information they deem highly reliable & relevant are incentivized to act more forcefully (i.e. put more capital at risk) than those who feel their information is not as strong. The cliché, but in this case apt description, “put your money where your mouth is” applies.
• Real-time data is more important today than ever before. Prediction markets are real-time. This is the fastest form of information aggregation readily available; prices change continuously based on an ever-changing environment, which brings us to the last & most important characteristic…
• ***Disparate information aggregation.***Every participant is acting with their own, unique information set. Even those with identical access will process its meaning differently, which is where prediction markets really shine. They aggregate all available data, in real-time & assign probability to the likelihood of an event outcome based on the actions of all underlying participants. A great example of this is the market action following the Challenger Crash in 1986. The full paper can be found here but the short-story is it took the market just minutes to accurately assign blame to Morton-Thiokol.

SO WHAT?

Why aren’t there more prediction markets? Or a centralized platform that allows for different prediction markets to exist? Sports betting is the only prediction market that’s gained real traction recently in the United States. More than $20 billion has been bet with US sportsbooks since May 14, 2018 (the day the Supreme Court struck down a law that essentially banned commercial sports betting in most states). Keep in mind that’s with only 18 states & ~30% of the population having access to regulated sports betting markets in some form (think DraftKings, FanDuel, etc.). Also keep in mind, not everyone is interested in sports! But everyone is interested in *something! *Prediction markets are inherently voluntary, speculative markets. As such, a prediction market could ask *literally anything *if enough participants are interested in engaging with that market. A few examples:

• Over/under 245.5 electoral college votes for Donald Trump in the 2020 US Presidential Election
• What date will the latest iPhone be released
• Will a Category 4 hurricane hit New Orleans in August 2021
• Which of these players will score more fantasy football points this week
• Will TikTok still be accessible to US users in January 2021
• Which of the following locations will be the site of the 2028 US Open

All of these (and almost any other probability-based scenario) should be possible & publicly available for consumers. A centralized platform that makes it easy for users to both participate in & propose additional event futures seems to be an untapped opportunity. Think about what happened during the pandemic — people flocked to similar areas of interest: Netflix binging, baking sourdough, & learning to cook. Importantly, they also flocked to sports betting apps & the stock market. Robinhood user growth was explosive. DraftKings & FanDuel user growth was explosive. These trends only reinforce my opinion that there’s untapped potential for prediction markets that cover “regular” events.

COULD I BE WRONG?

The best argument against the existence of this type of service in my opinion, is that it answers a question nobody is asking. Price discovery is an important component of prediction markets and for everyday events, is there really demand for this? Are enough users going to participate such that the markets become efficient & the platform becomes scalable? These are fair questions and certainly would be important to work through but I would offer anecdotally (dangerous, I know), that the internet has birthed niche communities of people who otherwise would have never found each other.

Even so, if there are communities of people who have diverse interests, that doesn’t necessarily mean that (a) these people possess information they feel leads them to be better informed than others and (b) they are willing to back that thinking with financial resources (i.e. bet).

The shortest rebuttal? There’s a reason these keep going up.