Mentioned in This Article
1. A Hedge That Missed by Design
NEXTPredict 2026, the prediction markets conference held at Convene Hudson Yards in New York on October 22 and 23, brought exchanges, trading firms and policymakers together less than two weeks before the U.S. midterm elections. Its organizers opened with a losing trade: a $12,000 premium on Kalshi for contracts that would have paid roughly $3 million had more than half of scheduled JFK arrivals been canceled the day before the event.
Elections took up much of the program, but executives also used the two days to describe where the industry plans to expand next, from small-business hedging to financial contracts and the regulatory treatment of event trading.
NEXTPredict co-founder Pierre Lindh spoke about the wager on Thursday during a panel titled "Hedging the Unhedgeable." He said the team had lost the trade the previous day, and that he was pleased about it, because losing meant the conference had gone ahead as planned. The contracts were a form of protection against the event being disrupted, and they expired worthless.
The wager also illustrated a gap in ordinary insurance. Lindh explained that standard event policies cover outright cancellation, but force majeure events tend not to cancel a gathering. Instead they damage it, through lost attendance or disrupted travel, and that kind of partial harm is hard to insure through conventional products. A contract tied to a measurable outcome, such as the share of arrivals at a major airport that are canceled, can fill that space.
The structure of the bet is worth spelling out. A buyer pays a small premium up front, here $12,000, and receives a much larger payout, about $3 million, if a defined condition is met. If the condition is not met, the premium is lost and the contract settles at zero. That asymmetry is the same one found in insurance and in options, and it is why traders describe event contracts as a hedging instrument and not only a way to speculate.
2. Who Took the Other Side of the Trade?
Susquehanna, the trading firm, took the opposite position. Joe Grubb, the firm's head of business development, said Susquehanna modeled the risk in-house in a day or two. Pricing unusual risks, he said, is part of the firm's DNA, and the contract was one more problem of that kind.
That exchange matters for how prediction markets work in practice. A contract only trades if someone is willing to quote a price on the other side, and a professional market maker with its own models can do that for a niche event far faster than an insurer could underwrite a bespoke policy.
Nicolas Hull, director of business development at Kalshi, said smaller businesses are already using the exchange to hedge. He gave the example of a New York bar that protected a promotion tied to a Knicks playoff run. In his description, Kalshi can list almost any contract so long as it is certified with the Commodity Futures Trading Commission and has a verifiable data source to settle against.
Those two conditions, regulatory certification and a reliable settlement source, define the boundaries of what the exchange can offer. They also explain why event-risk contracts of the JFK type are possible: airport cancellation data is published and can be checked after the fact.
For a small business, the appeal is practical. A bar planning a promotion around a playoff run, or an organizer with a large event on the calendar, faces risks that no insurer is likely to underwrite at a sensible price. A listed contract with a published settlement source lets such a firm buy protection in minutes and see the price at which others are willing to sell it.
3. Can Prediction Markets Replace Election Polls?
Chuck Todd, the veteran political journalist and former political director at a major television network, was one of the more closely watched speakers. Before the conference he told Lindh he was open-minded about whether markets or pollsters would better call the midterms. By Friday's panel, titled "Who Calls the Midterms," he appeared close to a convert.
Todd said he is bullish on prediction markets for journalism and described them as a kind of fact checker, because money adds an incentive to be right. He added that markets have effectively replaced election-night exit polls as the quick read on how a race is going. He did not accept, however, that polling itself is broken. A good poll, he said, simply costs a lot to produce.
Other panelists urged caution. Rajiv Sethi, a professor at Columbia University, argued that speed should not be mistaken for accuracy, since a market can move quickly and still be wrong. John Phillips, chief executive of PredictIt, said the strength of these markets is the wisdom of crowds rather than the wisdom of whales, a reference to the risk that a few large traders could dominate prices.
The accuracy debate had a fresh example. Ahead of Brazil's presidential runoff on Sunday, Polymarket priced Luiz Inácio Lula da Silva at roughly a 20% chance against Flávio Bolsonaro. Lula won a fourth term. Observers attributed the miss to Brazil's ban on prediction markets, which left the odds largely in the hands of traders outside the country.
The exchange between speakers also showed how differently the two tools work. Polls sample a population at a point in time and publish an estimate with a margin of error. A market aggregates the views of people who have staked money, and its price updates continuously. Each has weaknesses: a poll can be slow and costly, and a thin market can be pushed around by a handful of participants or by traders far from the event.
4. How Far Do the Contracts Extend Beyond Elections?
The companies on stage described a business that is widening quickly. JB Mackenzie, vice president and general manager for futures and prediction markets at Robinhood, said the product began with elections, moved into politics and sports, and is now expanding into economics and financial markets. He spoke in conversation with a financial-television journalist among the several media figures the organizers booked.
Sports remain the largest category by activity, Mackenzie said, but contracts on Bitcoin (BTC), crude oil and gold are the fastest growing. His explanation was that a probability is simpler for a customer to grasp than many conventional derivatives. A contract that pays if oil closes above a stated price asks a plain yes-or-no question, without the vocabulary of options.
Bailey Springer, who handles regulatory strategy outside the United States for Polymarket, said regulators in global financial hubs increasingly accept that prediction markets are not going away and that they have real investment and hedging uses. Polymarket's position is that they should be regulated as derivatives, with the oversight that classification implies, rather than treated as a form of gambling.
The distinction is central to the industry's legal arguments. A derivatives classification brings exchange-style rules on reporting, surveillance and market integrity, while a gambling label would subject the contracts to a patchwork of local licensing regimes.
The broadening of subject matter also reflects who is joining the market. A customer who first came to trade an election result may later look for a contract on crude oil or gold, and a platform that already holds that customer's account can list the additional markets at modest cost. For firms such as Robinhood, which already serve retail investors in stocks and other products, prediction contracts extend an existing relationship rather than require a new one.
5. Why Do Retail Traders Lag Professional Firms?
Ordinary traders usually reach a news event later than professional firms, which means they often arrive after the price has already moved. Farokh Sarmad, president and co-founder of a digital media company, made that point on Thursday's "Race to Zero Latency" panel, arguing that retail users tend to show up too late to profit from breaking news.
His proposed fix is to combine reading and trading in a single place, so that a person can act on a story at the moment it is published. His company is pursuing that idea through the Information Exchange, a product built on Solana (SOL). For the approach to work, Sarmad said, everything has to move on chain, because settlement and data would need to sit on the same infrastructure as the content.
The theme fits the retail push that several speakers described. As exchanges list more contracts on economic data and asset prices, the gap between a professional desk with a dedicated feed and an individual with a phone becomes a design problem for the platforms as much as for the traders.
Speed matters in this setting because a prediction market reprices as soon as new information appears. A trader who reads a headline minutes after a professional firm has acted is, in effect, buying at a price that already contains the news. Any product that shortens that interval, whether through faster data or through a tighter link between a report and an order, changes who captures the gain.
6. What Comes Next for the Industry?
Institutional interest was visible on the second day. Morgan Stanley led sessions at the conference, a sign that large financial firms are studying the market rather than watching it from a distance.
Lindh said the next debate for the sector will compare prediction markets with insurance. The lost JFK hedge offered an early example of that comparison: a small premium bought exposure to a loss that traditional cover would not have paid out, and a trading firm was willing to price it within a day or two.
The near-term test comes with the midterm elections, now less than two weeks away, when the arguments made on stage about speed, accuracy and who participates will meet real prices. Brazil's runoff has already supplied one case for the skeptics, while Todd's shift in view supplied one for the supporters.