Image generated by AI.

A wager, a trade, a contract – can you tell the difference? Each one involves a calculated risk predicated on a hoped-for outcome. Prediction markets blur the boundary further.

Even before Polymarket correctly called the winner of last year's US presidential election the line was pretty fuzzy. Then blockchain technology bridged it. When the ban on political betting was lifted in 2024, a door was opened. Prediction market platforms walked straight through.

Last month the category minted its first unicorns. Polymarket raised $200 million on a $1 billion valuation. Two days later, Kalshi announced a $185 million round against a whopping $2 billion price tag.

With online trading giants like Robinhood getting into the game, investors see big things in store. Have prediction markets achieved liftoff? And what about the blockchain upstarts who hope to exploit the leaders’ Achilles Heel?

Hedging the future

Prediction markets let anyone try their hand at forecasting the future and get paid when they get it right. Everyone from sports fans to current affairs buffs can channel their inner Nostradamus and take a punt on topics both trivial and weighty – perhaps the likely Rotten Tomatoes score for the next Superman movie, or who will be New York City's next mayor.

For economists and others who try to tap into the wisdom of crowds, the success of prediction market platforms makes perfect sense. They offer a future where every uncertainty can be priced, hedged, and insured against.

They're built on the idea that a large pool of bettors with cash on the line can foresee an outcome better than any one expert, no matter how deep their expertise. The risk of losing money switches on a collective power of premonition, the thinking goes. The right algorithm can bring it to the surface.

For end users, prediction market platforms make participation easy.

• Create an account and log in, select the event they want to bet on, and decide the size of their wager.
• Buy a share of an event contract, which is tied to a particular outcome (e.g., ‘this will be / won't be the hottest July ever').
• Wait for the final outcome. Shares trade for between a penny and a dollar. Once the final outcome is known, those who chose correctly get a payout for each share.

If you buy a share for 48 cents, predicting that this July will be the hottest on record – and the prediction comes true – you make 52 cents on the dollar. The final payout depends on the number of correct-outcome shares you purchased.

A New DeFi Growth Opportunity?

How attractive is that as a proposition? Figures from Dune Analytics show that monthly volume, active users, and new account signups on Polymarket all spiked in the runup to the 2024 US election, but then settled back.

Min Jung, an analyst at Presto Research, believes that prediction markets show promise but suffer from a liquidity problem.

"Even at their peak, attention has been mostly short-lived and limited to certain periods like elections, with only the top three to five markets having enough volume for people to trade in size (e.g., $1000)."

Another issue can be the time it takes for a prediction market event to resolve. In sports, you know an event outcome within a few days. For some prediction events, users don't know exactly when the resolution will happen or if it will occur before the event contract's end date.

Prediction markets will struggle to attract and keep these zero days to expiration (0DTE) users, Jung added, "as they prefer quick resolutions."

And of course, the algorithmic truth-telling they rely on isn't infallible. A 2024 event contract on the likelihood of a US government shutdown caused an uproar when it resolved to a 'yes' (meaning a shutdown will happen) even after the federal budget bill gained assent and ‘no shutdown' was confirmed.

Predicting upstarts

Another issue that could stymie the category's growth is a lack of freedom and flexibility. On Kalshi and Polymarket the platforms gate-keep the events worth predicting. End users shop from a list of pre-set event contracts and select the one that interests them most.

New entrants want to let users decide what's worth a wager.

Prediction market startup XO Market sees an opportunity to leverage dissatisfaction by making its system AI-driven. Instead of relying on humans, XO's oracle uses data analytics and machine learning to quickly calculate outcomes transparently and verifiably.

If a conflict does arise, XO has its own dispute panel, though it's composed of staked XO token holders who are chosen by a randomizing algorithm to vote on a resolution. A quadratic voting mechanism is also applied to minimize the influence of any one large token holder.

XO Market also aims to be permissionless, which would set it apart from Polymarket and Kalshi, where event contracts are approved by human arbitration. XO would allow end users more power to shape new markets by drafting questions for new event contracts and helping define resolution criteria.

The take away

Bookies have long been known to be better predictors of political contests than traditional pollsters, yet mainstream media treat psephologists as oracles while bookies' odds get tacked on to election news as colorful footnotes.

Now that their mathematical reliability is better understood, prediction markets are turning betting into an attractive pastime – very attractive, as recent valuations attest.

From stock markets to fantasy sports leagues, a significant number of people devote loads of time, effort, money, and brainpower to working out what comes next. Prediction markets trade on this very human propensity and democratize it to a wider spectrum.

Decentralized, blockchain-based platforms promise a more trusted and insight-driven betting environment. They can also deliver big money. The infamous Polymarket whale who went all-in on a Trump White House win received an estimated payout of around $85 million. Who would have foreseen that?

Benzinga Disclaimer: This article is from an unpaid external contributor. It does not represent Benzinga’s reporting and has not been edited for content or accuracy.