The History of Prediction Markets and What Comes Next

Open a trading app, and you may now find a different kind of trade besides stocks and options. Instead of buying a company or betting on where its share price goes, you can trade on whether inflation comes in above a certain number or the Federal Reserve changes interest rates. Prediction markets are becoming part of the places people already go to invest.

So how did prediction markets get here, and why are major brokers and exchanges joining in? Their story goes back centuries, through early betting markets, university research and the first online platforms.

We'll explore that history to explain what has changed, why the industry is growing and what could come next for you as a trader.

An old idea with a much bigger audience

You think something will happen. Someone else thinks it won't. A prediction market lets you put money behind your view by buying a Yes or No contract, with rules that spell out what counts as a win. You can hold it until the result is confirmed or sell it to another trader beforehand.

Basically, prediction markets put a price on a familiar pub argument: "I reckon I'm right, mate." "Well, put your money where your mouth is." Now there's a betting platform for that.

Some of the best documented early examples involve politics. Economists Paul Rhode and Koleman Strumpf traced political futures markets to 16th-century Italy, 18th-century Britain and 19th-century America. Their research also examined organized US presidential betting markets between 1868 and 1940, before scientific opinion polling became established. Read the historical research. It's fascinating.

Those early markets looked very different from the apps we use today, but the idea was much the same. When people put their own money behind a prediction, what do the odds tell us about how likely it is to happen?

A university gave prediction markets a new purpose

In 1988, researchers at the University of Iowa launched the Iowa Political Stock Market, later known as the Iowa Electronic Markets. Around 100 people from across campus joined the initial experiment, trading contracts linked to the presidential election.

The researchers wanted to see whether trading could produce useful forecasts. That distinction matters. You might strongly support one candidate but think another will win. Once money is involved, what you hope will happen and what you are prepared to back can be very different things.

Iowa also tested markets covering flu outbreaks, hurricanes, initial public offering prices and movie openings. Prediction markets were becoming a way to study how people share information and make forecasts across different subjects. The university explains the experiment here.

Suppose an outcome is priced at a 70% chance and it does not happen. Was the market wrong? One result cannot tell you whether 70% was a sensible estimate. If the odds are consistently well judged, outcomes given a 70% chance should happen roughly seven times out of ten across a large enough set of comparable events. The interesting test is how those forecasts perform over time, rather than whether the favorite wins every time.

This led us to a question we particularly enjoyed digging into at GnosisFi.

Who Writes the Question and Sets the Odds?

Someone first has to turn an interesting question into a contract. On Polymarket, the markets team creates markets, with user suggestions. Kalshi also accepts ideas from users and its own researchers, then reviews the proposed contract and writes the terms, including the source used to confirm the result. The platform decides what people can trade, but that is different from deciding the odds. Polymarket Help Center

For an exchange such as Kalshi or Polymarket, the prices come from orders placed by traders. That includes market makers, traders who regularly offer prices at which they are willing to buy or sell. There is no requirement for every new question to start at 50%, because the two possible answers are not necessarily equally likely. Kalshi Help Center

Take a hypothetical question: "Will the Fed cut rates at its next meeting?" On Polymarket, someone willing to pay 60 cents for Yes can be matched with someone willing to pay 40 cents for No. Together, they provide the $1 that will eventually go to the winning contract. That gives us an opening trade implying roughly a 60% chance of a cut. As more orders arrive, the price changes.

The percentage on the screen is a summary of those prices, rather than a separate forecast calculated by the platform. Polymarket generally displays the midpoint between the best buying and selling prices, using the last traded price when that gap is wider than 10 cents. So the headline probability and the price you can actually buy at may differ.

Why Can Those Prices Be Useful Forecasts?

Think about the people following a Fed decision. One studies inflation, another follows employment data, and someone else closely tracks policymakers' speeches. If they believe the available price understates the chance of a rate cut, they have a reason to buy. Their orders help bring that research into the price, although the crowd can still misjudge the evidence.

There is research behind the interest in these markets. A February 2026 paper published in the Federal Reserve's research series found that Kalshi's forecasts were comparable in accuracy to established benchmarks, with statistically significant improvements for federal funds rate and headline CPI forecasts in the comparisons studied. For core CPI and unemployment, Kalshi's accuracy was statistically similar to the benchmarks rather than better. That is a finding from the paper's authors, rather than an official Fed endorsement of a trading platform. Read the research.

But what does an accurate probability look like? If 100 comparable events are each given a 70% chance, you would expect roughly 70 to happen if those estimates are well judged. The other 30 happening differently would be consistent with the forecast. Researchers call this calibration, and it helps explain why one losing favorite does not tell you whether a market's odds are generally useful.

For you as a trader, 70% means the outcome is more likely to happen than not, but there is still a 30% chance it won't. It also gives you a better question to ask when someone claims a platform is accurate: how well did its prices match the results across many events, and how early were those forecasts made?

The internet made it easier to take the other side

Betfair launched its exchange on June 9, 2000. Customers could back an outcome or take the opposite side, offering prices to one another through a matching platform. If you have placed a limit order on an exchange, the basic idea will sound familiar. Read Betfair's launch history.

Sports helped that exchange model reach a much wider audience. Platforms such as Intrade also offered event contracts, while PredictIt later developed a following around political markets. Their businesses followed different legal routes: the CFTC brought a case against Intrade in 2012, while PredictIt operated under conditional staff relief for an academic research market. CFTC Intrade record and PredictIt relief.

The internet solved part of the access problem. You no longer had to find someone locally who disagreed with you. But platforms still needed a legal basis for offering their contracts, a reliable way to determine results and enough trading interest for people to enter and exit.

You do not always have to wait for the answer to make a trade. Suppose you buy 100 Yes contracts at 40 cents, costing $40. New information arrives and buyers are now offering 60 cents. If you sell all 100 at that price, you receive $60 and make $20 before fees, even though the event has not happened yet. You are trading how expectations change as well as deciding what you think the final result will be.

Financial exchanges began building their own markets

Kalshi received its designation as a US contract market in November 2020. Polymarket, which also began operating in 2020, used blockchain infrastructure. These were separate approaches developing during the same period, not one industry moving from crypto into traditional finance. Kalshi's designation and the CFTC's Polymarket record.

Established derivatives exchanges were involved too. CME Group introduced event contracts in September 2022 covering daily price outcomes in benchmark futures markets, including stock indices, energy and metals. These were daily options on futures, which shows why "event contract" can cover products with different structures and settlement values. CME's launch notice.

ForecastEx started operations on August 1, 2024, with eligible Interactive Brokers clients able to trade contracts on economic releases and climate indicators. Robinhood introduced presidential election contracts through ForecastEx on October 28 that year. Event trading was reaching people through brokerage accounts they already used. ForecastEx launch and Robinhood announcement.

How Crypto Helped Bring in a New Audience

Crypto gave prediction markets another route to reach traders. As mentioned earlier, platforms such as Polymarket used blockchain infrastructure to let people buy and sell positions on real events. Prediction markets already existed, but they were now reaching an audience comfortable with digital wallets and trading online.

We can see why the idea appeals. If you already follow economic announcements, political developments or product launches to help you decide what to buy or sell, a contract on the event itself is a natural extension. Instead of working out how the news might affect Bitcoin or a stock, you can take a position on whether the event happens.

Plenty of people trade both. In S&P Global's August 2026 US survey, more than half of prediction market traders, 55%, said they had also bought or sold crypto in the past year. Read the research.

Prediction Markets Are Coming to the Apps You Already Use

You might be interested in prediction markets without wanting to open and fund another account. Now, more brokers are adding them to their existing apps, letting eligible customers trade on events alongside their other investments. For everyday traders, that makes it easier to give them a try.

In December 2025, Coinbase began rolling out US prediction markets with Kalshi supplying market flow at launch. The company presented them alongside its broader expansion into stocks and derivatives. A customer coming from crypto could now encounter event contracts through the same app. Coinbase's announcement.

Interactive Brokers took another step in May 2026, announcing a unified interface connecting Kalshi, CME Group and ForecastEx. It lets eligible clients compare available liquidity, prices and fees across connected venues, with order routing designed to select the best available net price. Read the IBKR announcement.

That can make a practical difference. Suppose a genuinely equivalent $1 payout contract costs 42 cents on one venue and 45 cents on another. Buying 100 contracts costs $42 or $45 before fees. Comparing them could save you $3, provided the result, deadline and settlement rules really match.

The app you use and the exchange handling your trade may be different companies. You might place a trade through your broker's app, but the contract is listed on a partner exchange. Check which exchange that is, because its rules determine how the result is confirmed and when the contract pays out.

The main players do different jobs

Some companies run the exchange, while others let you trade through their app. This table explains what each of the main names does and how they fit together.

How much has trading grown

Pew Research Center's analysis of Kalshi and Polymarket data, including Polymarket US, shows the scale of the increase.

Combined monthly global volume rose from about $810 million in July 2024 to $53 billion in July 2026.

Source: Pew Research Center's analysis of The Block data. Figures are rounded and show combined trading on Kalshi and Polymarket, including Polymarket US. Volume is counted at each contract's $1 settlement value rather than the price paid for it, so these totals are larger than the money traders actually put in.

Much of that growth came from sports trading. So while the industry's numbers are impressive, they do not mean trading on inflation, interest rates, or business events has grown at the same pace.

Who actually uses prediction markets

So who is using prediction markets? An August 2026 survey by S&P Global asked around 2,500 US adults who use the internet. One in ten said they had traded on prediction markets in the past year.

Among those prediction market traders, 61% were under 35, while 13% were 55 or older. Around two-thirds were men, and 55% had also traded crypto during the year. These findings give us a picture of the US audience, rather than traders worldwide.

That points to a younger, predominantly male audience with considerable crypto overlap. But there is still plenty of room for awareness to grow: 44% of survey respondents had never heard of prediction markets.

We should note that these are survey results, not audited platform account records or a global demographic breakdown. Read S&P Global's full analysis and methodology.

Regulation will shape the next chapter

Prediction markets are growing, but where you can trade still depends on local laws.

Countries classify these products differently, and a platform's US exchange status does not automatically allow it to serve customers elsewhere.

In the United States, event contracts operate within the CFTC framework, but disputes over federal and state authority have continued, particularly around sports. The CFTC filed lawsuits challenging several states' actions in 2026. CFTC litigation announcement.

The courts have since split on the central question. In April 2026 the Third Circuit held that sports event contracts are swaps under federal law, and that federal law overrides state gambling rules. The Ninth Circuit reached the opposite conclusion in August and the Sixth Circuit did the same in September, both unanimously: these contracts are not swaps, and state gambling law still applies. In October 2026 the CFTC published a proposed rule on whether event contracts belong inside the swap definition at all.

So regulators and courts still disagree over who has the authority to oversee these markets, and the answer may differ depending on where you live.

Further reading: Great Britain, France, Australia and Singapore. This is a selected snapshot as of October 2026, not a complete eligibility list.

For you, the useful check is specific: can this platform offer this contract where you live? Age requirements, customer eligibility and available markets depend on the operator and jurisdiction.

What could come next?

The developments below are our assessment of where the industry could go. Some have already started, but their scale and commercial success remain uncertain.

More trading in the apps you already use

Brokerage integration is already happening. We expect more platforms to offer event contracts besides existing products, because customers can research an event, compare ways to trade it and manage positions in one place. That does not turn an event contract into a stock, option or perpetual future; each still has its own terms and risks.

The same expansion is reaching platforms known for perpetual futures. Hyperliquid's HIP 4 introduces fully collateralized outcome contracts alongside its existing infrastructure. Its documentation distinguishes these from leveraged perps: outcome contracts have dated settlement and do not themselves involve leverage or liquidations. This is another example of an established trading platform adding a different product. Read the official documentation.

Better comparisons between venues

IBKR's approach points toward more competition over execution. Over time, more services could help you compare what a trade actually costs across exchanges. The useful improvement would be comparing equivalent contracts after fees, with enough available orders to fill your trade, rather than simply displaying the lowest headline price.

Odds becoming part of everyday market research

In October 2025, ICE, the owner of the New York Stock Exchange, announced its Polymarket investment arrangement and plans to distribute its data to institutional customers. In March 2026, ICE reported a further $600 million direct investment following an initial $1 billion. Established finance is interested in the information these markets produce as well as the trading business. ICE's original announcement and investment update.

We expect to see prediction market odds more often beside economic calendars, company earnings reports and investment research. Even if you never place a bet, they can help you follow what traders expect. After an inflation report, for example, you could check whether the odds of a Fed rate cut have risen or fallen and compare that reaction with your own view.

More specific ways to trade or hedge an event

Consider an investor worried about an unexpectedly high inflation release. A contract paying on that result could offset part of a portfolio loss if the result occurred and their portfolio fell. It would not be a perfect hedge, because stocks can react differently and the contract has its own cost, but it shows why traders might want both products available.

More help with research and automation

AI could help you find the latest economic reports, compare forecasts and track changes in the odds, giving you more time to focus on the trade. But if it predicts an outcome, you should be able to see what information it used, how it reached that estimate and how its previous predictions turned out.

Finally, we at GnosisFi wanted to explore what we think will become the norm in the next few years.

Could Predictions and Perps Become Part of the Same Trade?

Imagine following the next Fed decision on one screen. You can trade a contract on whether rates change, then take a separate position on how stocks, gold or Bitcoin might react. That combination could give traders more ways to act on the same research, all within a familiar platform.

Major trading platforms are already offering prediction markets and perpetual futures side by side.

In May 2026, the CFTC approved Kalshi's Bitcoin perpetual futures contract. Polymarket's international platform also offers perps across assets including stocks, crypto and commodities. Coming from the other direction, Hyperliquid has introduced outcome contracts alongside its existing trading infrastructure. Availability and eligibility differ between platforms. CFTC.

These products answer different questions.

A prediction contract might ask whether the Fed cuts rates at a particular meeting. A stock index perp lets you trade the market's reaction before or after that decision. Having both available could help you choose the trade that best matches your view, rather than using one product for every situation.

Leverage on event contracts could take this a step further. Where permitted, you could hold a larger position by putting up only part of its value as collateral. That could make some trades and hedges possible with less money upfront, but it would also change how you manage the position.

Suppose you buy a leveraged Yes position. You still believe the event will happen, but a news report sends the odds sharply lower. If your collateral falls below the required level, your position could be closed automatically. You might eventually get the prediction right and still lose the trade because you could not hold it through that move.

That is different from buying a fully paid contract and holding it until the result. It is also different from trading an ordinary asset perp on a prediction platform. Kalshi describes its perpetuals margin account as separate from its predictions balance, while Hyperliquid's outcome contracts are fully collateralized and do not themselves involve leverage or liquidations.

How Kalshi separates the two balances.

Some developers are now exploring leveraged trading on the probabilities themselves. Perp.com describes long and short positions on event probabilities with up to 5x leverage. Its website currently labels the product an early access testnet, so it is an example of what is being developed rather than an established live service. It is far too early to know whether it will work.

There are still practical questions to answer. If you're trading whether the Fed cuts rates at its next meeting, that trade eventually has a result. So what happens to a contract advertised as having "no expiry" once the decision is announced? You also need to understand how its price is set, what it costs to keep the position open and when a move against you could trigger liquidation.

The possibilities are interesting. You could trade the Fed's decision directly, use a stock index perp to trade the market's reaction, or potentially use leverage to trade changes in the odds before the meeting. Each gives you a different way to act on your research, depending on what you expect to happen and when.

We're intrigued, but we'd want to see these products handle real money and fast news before expecting them to attract a much wider audience. If they work well, we think larger platforms could introduce similar products. The early developers would then need to compete on price, ease of use and whether traders can reliably get in and out of their positions.

Monitor the next stage for yourself

Choose one event you already follow, perhaps an inflation report or the next Fed meeting, and look at how its odds change as the date approaches. You will start to see why the same information interests a casual reader, an active trader and a brokerage building its next product.

Got a view on the next Fed decision, an election or where Bitcoin is heading? Visit our GnosisFi Prediction Markets page to find related markets on Polymarket and Kalshi. Check the odds, see how they've moved and find events due to settle soon. Start with something you follow and see whether the price looks attractive to you.

Then put the trade through our free prediction tools. Bet Check shows what you'll pay, your potential profit and what you'll lose if you're wrong. Market Picker helps you compare markets, while Odds Gap Finder lets you compare your forecast with market odds and available estimates.