Wisdom Of The Crowd
The Past, Present, And Future Of Prediction Markets 

Prediction markets are booming, offering bets on the next U.S. president, the Super Bowl champion, and when Taylor Swift will get married. We explore how betting markets came to be, along with their implications for markets and the general public.

(Published April 10, 2026)

Prediction markets are all the rage in the media. News articles cite betting market odds to “make sense” of what’s going on in the world. Want to know the odds of the presidential election or whether there will be a military attack somewhere in the world? Check online betting markets like Polymarket or Kalshi — anyone aged 18 or older can bet on anything.¹ You could even bet on when Taylor Swift will get married or the weather in Central Park.

And trading volume on prediction markets has skyrocketed. At the beginning of 2026, the weekly notional market volume of prediction markets topped $6 billion, with 232 million monthly transactions and nearly 3 million unique users since 2024 (see Figure 1).²

Theoretically, betting markets are supposed to aggregate information from millions of users, thereby reflecting the “true” outcome of an event. Since money is at stake, the bets carry more weight (“put your money where your mouth is,” as the saying goes). 

But do betting markets truly reflect the wisdom of the crowd? Or are prediction markets just another example of the world turning into a giant casino? In this article, we explore the past, present, and future of prediction markets and assess their potential for forecasting.

Shared Origin

Sports betting is the oldest form of prediction market, dating back at least to Ancient Greece, during the original Olympic Games, and to Roman gladiatorial contests and chariot races.

Beyond athletic events, the earliest recorded prediction markets were in 16th-century Rome, where sensali, or brokers, circulated slips of paper between bettors and banks on papal election outcomes (Did You Know? Predicting Popes).

Later in 18th-century London, coffeehouses were venues where elites traded on parliamentary scandals and prime ministers. In particular, one of the coffeehouses that facilitated bets, Jonathan’s, would later be renamed the London Stock Exchange.

It wasn’t until 1988 that the first computerized, real-time betting exchange platform emerged in the U.S., created by the University of Iowa professors, the Iowa Electronic Markets (IEM), were designed to analyze the market-based probabilities of election outcomes.

Although participation was limited to university staff due to gambling regulations, the Iowa Electronic Markets were a huge success—IEM odds beat mainstream election polls like Gallup and ABC News 74% of the time from 1984 to 2005. In the 1988 election, IEM even accurately predicted Bush’s margin of victory at exactly 53.2%.⁶ 

Following IEM’s success, Intrade became the first publicly open online prediction market and quickly became a widely referenced “crowd forecast” for U.S. presidential elections. However, Intrade was later sued by the Commodity Futures Trading Commission (CFTC) for operating without legal registration despite offering commodity derivatives, a case that eventually led to its shutdown.

Trial And Error

Regulatory barriers to betting markets are not unfounded. To most people (including regulators), the betting market is pretty much akin to online gambling: extremely speculative and addictive, with little productive value.

After all, traditional betting markets are very similar to sports betting and gambling—they are designed to be “house-based,” where one bookmaker sets the odds and takes bets from traders. If you win, the bookmaker pays you, but if you lose, the bookmaker gets to keep your money. 

The issue with traditional models is that the odds of an event are fixed by the house, which can reflect emotional biases and be deliberately skewed to guarantee the bookmakers’ profits.

But modern prediction markets discovered a better way. In the 2000s, Betfair, in the UK, adopted a peer-to-peer framework, or exchange model, that operates more like a financial exchange than a Vegas-style gambling venue. 

Specifically, in exchange models, users bet against one another, with opposing positions matched in real time and prices determined by supply and demand. And there is no “fixer”—traders can also enter and exit positions at any time, allowing odds to update continuously. The improvement in predictive power is obvious: research comparing 5,478 matches of European football leagues suggests that prediction markets forecast more accurate outcomes than do bookmakers, by self-selecting more skilled and information-savvy bettors (see Figure 2).⁷ 

The two most dominant betting markets today, Kalshi and Polymarket, both operate on an exchange model. Polymarket is the first decentralized, on-chain, peer-to-peer betting market that settles in cryptocurrency (stablecoins) instantly across borders. Kalshi, on the other hand, was the first regulated exchange model. It operates like a peer-to-peer betting exchange, but Kalshi runs the order books and guarantees settlement of orders under CFTC regulation as an event-linked futures instrument. 

As of March 2026, Kalshi accounts for 31% of the outstanding notional volume in major betting markets, and its volume has grown 44-fold since April 2024 (see Figure 1 again). Interestingly, the composition of online bets today continues to largely reflect their origins: most betting volume is concentrated within politics and sports (see Figure 3). 

Track Record 

As financial market economists, we are particularly interested in how modern betting markets perform in economic forecasting. 

The most important bond market forecast is the path of the federal funds rate. So, we compared Kalshi’s market odds of the number of Fed rate cuts in 2024 and 2025 versus the number of rate cuts implied by the Overnight Index Swap (OIS) market—a financial derivative market in which traders take an investment position on when and how many rate cuts or hikes the U.S. central bank will deliver in the next 12 months. 

Both Kalshi and OIS market odds of rate cuts tracked closely throughout the year (see Figure 4). Although Kalshi’s Fed funds market's daily trading volume is one-millionth of the OIS market’s, Kalshi odds respond equally quickly to news, such as new data releases or Fed speeches, suggesting that it attracts sufficient liquidity and expertise to be on par with professional financial forecasts. 

Meanwhile, recent research has also found that Kalshi forecasts for other economic indicators, such as the Consumer Price Index (CPI) and the unemployment rate, are similar to those of the professional survey consensus, such as those from Bloomberg.⁸ 

But Kalshi rewards traders for being right, whereas public professional forecasters get better rewarded for being “popular,” or in line with consensus. Some professional forecasters would rather align with others than publish an outrageous forecast that turns out to be wrong, which then dampens their reputation. On Kalshi, though, everyone is anonymous, and the only thing at risk is the money you trade, which is linked directly to how close you get to the true outcome of the event. In fact, there may even be an incentive to bet against groupthink, as it could win you additional profit.

Valued Odds

Now, one may ask, what’s the point of having prediction markets if they just give you similar insights as existing financial assets?

In short, prediction markets can provide a lot more value.

First, prediction markets are more accessible than financial markets for those who want to express and profit from their views, as they are simpler to execute and easier to understand, with no embedded leverage or multipliers and minimal required investment. 

Second, prediction markets attract interest and, in turn, offer collective information on a much broader array of events than do financial markets. Even within the financial market world, prediction markets offer additional granularity. For example, prediction markets can fill in the gap for high-frequency forecasts of economic indicators such as recession odds, inflation, and the labor market, for which no single financial market asset or derivative exists. Or, rather than betting on the overall stock price, traders can bet on a company’s earnings, planned capital expenditures, or what the CEO says during the earnings call, which again provides the audience with more collective insight than just the stock price.

Third, even for existing financial market derivatives from which we can derive a forecast, prediction markets may offer the “cleanest” odds that most accurately reflect the “true probability” of an event. After all, asset markets serve purposes beyond directional betting; they are also used for hedging and exposure management, whereas prediction markets are completely binary—you only win when you are closer to the actual outcome of the specified event. Research shows that during the night of the Brexit result announcement, Betfair odds converged to actual odds faster than the pound-dollar futures exchange market.⁹ 

Fourth, contract-based prediction markets like Kalshi provide insights that traditional financial derivatives cannot, as they reveal the full probability distribution of outcomes rather than a weighted average measure of all trades. This additional detail matters because recent Fed research has revealed that the mode (the single-highest probability outcome) of Kalshi market odds on Fed rate cuts has been perfectly accurate on the Federal Open Market Committee (FOMC) press conference days, a feat that neither surveys nor the Fed funds futures market has been able to achieve.¹⁰ 

The Odds Ahead

Nonetheless, prediction markets today are far from mature and perfect. Regulatory obstacles remain. Polymarket was barred from operating in the U.S. until late 2025, and Kalshi continues to face numerous active lawsuits.

But we remain optimistic about the future of prediction markets. 

Some critics point to the presence of “whales” (traders with large trading volumes) as the key driver of prediction market odds, rather than to the wisdom of the crowd. Indeed, in Polymarket, whale orders account for only 5% of daily trades but 60% to 70% of total trade volume (see Figure 5).¹¹ 

Perhaps large trade volumes simply reflect more insight. After all, traders with better access to information or expertise in certain fields are much more likely to hold high-conviction views and want to profit from them. This type of informational asymmetry exists everywhere, even in financial markets. But as the co-founders of Kalshi pointed out, the purpose of prediction markets is to incentivize more people to provide more information about a much wider range of events than financial markets.¹² So, in a sense, prediction markets can help address information asymmetry on almost any topic one can think of.

More importantly, prediction markets exhibit a positive feedback loop and network effect: increased volume—hence liquidity—improves forecasts, making them a more useful and well-known forecasting tool, which in turn attracts more users and more liquidity, further improving forecasts. The recent rise of stablecoins and the ability to bet and trade on-chain (Kalshi recently began supporting stablecoin deposits) could further attract liquidity globally.¹³

Betting against the future of betting markets? Start a bet on Kalshi or Polymarket, and we will take the other side!

Endnotes

6.  Berg, J. E., Nelson, F. D., & Rietz, T. A. (2008). Prediction market accuracy in the long run. International Journal of Forecasting, 24(2), 285–300.https://doi.org/10.1016/j.ijforecast.2008.03.007 

7.  Franck, E., Verbeek, E., & Nüesch, S. (2010). Prediction accuracy of different market structures—bookmakers versus a betting exchange. International Journal of Forecasting, 26(3), 448–459.https://doi.org/10.1016/j.ijforecast.2010.01.004

8.  Diercks, A. M., Katz, J. D., & Wright, J. H. (2026). Kalshi and the rise of macro markets (Finance and Economics Discussion Series 2026-010). Board of Governors of the Federal Reserve System. https://doi.org/10.17016/FEDS.2026.010 

9.  Auld, T., & Linton, O. (2018). The behaviour of betting and currency markets on the night of the EU referendum. International Journal of Forecasting, 35(1), 371–389.

10.  Diercks, A. M., Katz, J. D., & Wright, J. H. (2026). 

11.  andy_chelsea. (n.d.). Polymarket whale order observation. Dune.https://dune.com/andy_chelsea/polymarket-whale-order-observation.

12.  Cheeky Pint. (2026, March 17). Creating prediction markets (and suing the CFTC) with Tarek Mansour and Luana Lopes Lara. Substack.https://cheekypint.substack.com/p/creating-prediction-markets-and-suing.

13.  In November 2025, Kalshi partnered with Coinbase and is now supporting USDC stablecoin deposits.