Regulatory shifts and a decisive court win for Kalshi have put US prediction markets back in play. With event-based contracts edging closer to the mainstream, venues are racing to deliver futures-grade technology capable of handling surging liquidity, real-time pricing, and the scalability demands of high-stakes speculation, writes Sergey Samushin, head of exchange solutions, Devexperts.

Traders often commend financial markets for their ability to price-in future outcomes. Those who view current market prices as having already incorporated all relevant information believe them to represent the aggregate knowledge of all their respective participants.

The wisdom of crowds

Prediction markets, decision markets, or idea futures, as they’re also known, tend to receive a great deal of press in and around high-profile elections. They correctly predicted the 2024 re-election of Donald Trump, as well as the winner of the New York mayoral primary, Zohran Mamdani.

One of the classic demonstrations of the wisdom of crowds comes from Francis Galton. In 1906, at a fair in Plymouth, 800 visitors took part in a challenge to guess the weight of an ox carcass. Galton showed that the median guess of the participants came to within 1% of the actual weight of the carcass.

Despite their storied history and perceived accuracy, they’ve failed to capture the popular imagination due to a regulatory climate that’s historically been hostile to them. This is due to the space they occupy between traditional asset markets and gambling.

A changing regulatory landscape

In 2023, Kalshi, the largest US prediction market provider was denied regulatory approval to list and clear its election contracts. In September 2024, a US judge ruled in favour of Kalshi, which led to the CFTC appealing the ruling. In October, the court again sided with Kalshi, allowing the company to run these markets.

The current administration is expected to be more lenient on financial regulation, as evidenced by its pro crypto stance, and the recent approval of perpetual-style futures for US traders.

Some are now expecting prediction markets to be the next big thing for US speculators as regulatory tailwinds encourage other venues to introduce their own events-based trading products.

Prediction markets at-a-glance

In essence, events-based traders purchase contracts in favour or against a future outcome, such as “the S&P 500 will hit 7000 in 2026,” or “Oleksandr Usyk will retire from boxing with an undefeated record”. Each correct prediction is worth $1 per contract, with the prices for “yes” or “no” outcomes depending on how many other traders have taken that side of the bet.

For example, there’s currently a prediction market on whether the EU will lose a member state by 2030. A “no” bet is currently valued at around 80 cents per contract, whereas a “yes” bet costs around 20 cents.

Those expecting a “yes” outcome may purchase contracts at 20 cents today and sell them to buyers closer to the 2030 deadline should it become evident that the EU is indeed likely to lose a member as the price rises to reflect this. Alternatively, they may wait for the contract to be resolved and earn a dollar for every 20 cent contract purchased.

Underlying technology

The relative simplicity of these markets belies the technological complexity they require. Predictions are exchange-traded, with participants (both individuals and market makers) having access to the relevant order books, rather than the venue acting as principal. This means they have far more in common with futures and options exchanges in the technologies that underpin them.

For this reason, there are technological overheads that have to be met. The news cycle is crucial to traders being able to evaluate the likelihood of an outcome, which means that they have to be able to price-in this information around the clock. The uptime requirement of these venues is similar to crypto exchanges in this regard.

This carries over into the listing, resolution, and settlement of individual contracts, which should be manageable on the fly without the need for technical restarts. In order to maintain perpetual uptime, management of contracts has to be conducted through APIs via administrative interfaces in business logic, rather than requiring changes to be made in code and technical system restarts.

Another important requirement is sufficient headroom when it comes to scalability. Interest in event outcomes tends to peak around certain high-profile events, which means that systems must be able to handle many times the average in terms of activity.

In the interests of both reliability and scalability, replication is advisable as a strategy. For uptime, this calls for multiple instances of key components that can be switched to in the event of technical difficulties. For scalability, horizontal scaling can be achieved by using separate instances of the matching engine for particularly popular events.

Finally, despite the binary nature of outcomes, contracts require flexibility in the way they can be resolved, such as the ability to define when an outcome has come to pass. In the examples above, the S&P 500 may or may not reach the target price on any day of 2026, whereas Usyk will only be able to be defined as undefeated after his official retirement.

Looking to the future

Beyond being a powerful tool for assessing crowd wisdom, events-based trading is set to grow as clarity is achieved on the regulation of these markets. Whether or not it impacts incumbents on the financial and gaming ends of the spectrum remains to be seen.

We can expect on-going innovation in this space, which will give various underlying technologies a chance to compete for the most effective and efficient means to operate prediction markets at scale.