Is Trading Prediction Markets Gambling or Investing?

The honest answer: it’s legally contested, and functionally it sits in between. Under federal law, the contracts traded on Kalshi and Polymarket are event contracts — a species of derivative — listed on exchanges regulated by the Commodity Futures Trading Commission, the same agency that oversees oil futures and interest-rate swaps. Under the laws of several states, at least when the underlying event is a ballgame, those same contracts are wagers, and offering them without a gaming license is illegal gambling. As of August 2026, federal appeals courts are split on who’s right, New York is suing Kalshi for at least $36 billion, and New Jersey has been knocking on the Supreme Court’s door. Nobody gets to give you a one-word answer yet, because the courts haven’t produced one.

What can be answered precisely is how the machine works, how it differs from a sportsbook, and what the unsettled legal status means for you in practice.

Why is this even a question?

Because two regulatory regimes claim the same product.

The federal framing: a prediction market contract is a binary derivative — it pays $1 if an event happens, $0 if it doesn’t — and the Commodity Exchange Act gives the CFTC exclusive jurisdiction over derivatives traded on federally designated exchanges. Kalshi has been a CFTC-designated contract market since 2020. Polymarket re-entered the US through the same door: the CFTC issued an amended order of designation in November 2025, and the platform relaunched for US users that December. The CFTC itself has taken Kalshi’s side against the states, filing amicus briefs arguing that event contracts are swaps within its exclusive jurisdiction and that state gambling laws are preempted — a step the agency takes rarely.

The state framing: putting money on the outcome of a sporting event is a bet, whatever instrument it’s wrapped in, and states — not Washington — have always licensed and policed betting. On this view, a “Will the Chiefs win?” contract is a sports wager sold without a gaming license, to customers as young as 18 in states that require bettors to be 21. More than 30 states have filed briefs backing that position in one court or another.

The courts are split. On April 6, 2026, a divided Third Circuit panel ruled for Kalshi in KalshiEX LLC v. Flaherty, holding that sports event contracts traded on a CFTC-designated exchange are swaps under the Commodity Exchange Act and that New Jersey’s gambling laws likely can’t reach them; New Jersey was given until August 4, 2026 to ask the Supreme Court to review that ruling. Three months later a federal judge in Manhattan went the other way, denying Kalshi’s bid to block New York’s gaming regulators and finding Kalshi had not shown its preemption argument was likely to win. On July 31, New York’s attorney general sued Kalshi outright, alleging it runs an illegal gambling operation and seeking at least $36 billion in penalties and disgorgement. Kalshi disputes the allegations and has appealed the Manhattan ruling; the platforms remain live nationwide while the cases proceed. Note the shape of the fight: nearly all of it is about sports contracts. Nobody is suing anyone over Fed-rate markets.

How is a prediction market different from a sportsbook?

Set the legal labels aside and the structural difference is real: a sportsbook is a house, a prediction market is an exchange.

When you bet at a sportsbook, the book is your counterparty. It sets the line, builds its profit margin (the vig) into the odds, wins when you lose, and can limit or close the accounts of customers who win too often. When you trade on Kalshi or Polymarket, your counterparty is another trader who took the opposite side. The exchange sets no line and takes no position — it matches orders and collects a fee either way, so it has no economic reason to care whether you win.

	Sportsbook	Prediction market
Counterparty	The house	Another trader
Prices/odds set by	The book	Supply and demand
Operator earns	Vig built into odds; wins when you lose	Trading fee on each fill, win or lose
Exit before the event	Cash-out at the book’s price, if offered	Sell to the market anytime
Consistent winners	Routinely limited or banned	Welcome — they pay fees
Primary regulator	State gaming commissions	CFTC (contested by states)

The cost difference is measurable. A standard -110 sportsbook line means risking $110 to win $100 on each side — roughly a 4.5% hold for the house. On Kalshi, the taker fee is $0.07 × price × (1 − price) per contract: at 50¢ that’s 1.75¢, or 3.5% of your stake, and it shrinks toward the extremes — 0.63¢ on a 90¢ contract. Polymarket US charges takers up to $1.50 per 100 contracts at 50¢ under its July 2026 schedule and pays makers a rebate. Both figures as of August 2026. Fees are lower than vig, and limit orders cut them further — but both systems are negative-sum for the players in aggregate: every dollar won is a dollar someone else lost, minus the operator’s cut. That part, prediction markets and sportsbooks genuinely share.

A sportsbook profits when you lose; an exchange profits when you trade.

Is it skill or chance?

The legal test in many states asks whether skill or chance is the “dominant factor” in the outcome. Prediction markets sit awkwardly on that spectrum, partly because “prediction markets” is not one thing.

On the chance end: a contract on tonight’s coin-flip-close game, held to resolution, is close to a wager in substance whatever it’s called in law. On the skill end: a trader who prices Fed decisions better than the market, buys at 60¢ what should trade at 70¢, and exits when the price converges is doing something recognizably like trading — the same expected-value arithmetic that drives options desks. The instrument allows both. Three features push it toward the trading end: prices are set by an open order book you can quote into, not by a house you can only accept; you can sell before resolution, so you can profit from being right about the probability without being right about the outcome; and being consistently good gets you volume rebates, not a banned account. None of that guarantees skill wins — most participants in any negative-sum arena lose after fees — but the structure rewards it in a way a sportsbook is built to prevent.

Is it investing?

Not in the sense your index fund is. An investment claims a stream of future cash flows — earnings, coupons, rent — and can compound. An event contract is a fixed-pot transfer: it expires at $1 or $0 on a known date, and the money comes entirely from the trader on the other side. That makes it a derivative — closer to a short-dated option than to a stock. Calling it “investing” oversells it; calling it “gambling” ignores the exchange structure, the open price discovery, and the fact that federal law currently regulates it alongside futures. “Trading” is the accurate word, with everything that implies about risk.

What does this mean practically?

Your money’s max loss is defined. Contracts are fully collateralized: buy at 40¢ and your worst case is losing 40¢ per contract. No margin calls, no losing more than you put in.

Taxes are genuinely unsettled. The IRS hasn’t said how event-contract profits are classified. Practitioners variously treat them as ordinary income, capital gains, or gambling winnings — and the label matters, because starting in 2026 the 2025 tax law caps deductible gambling losses at 90% of winnings, meaning a break-even year could still produce a tax bill under the gambling treatment. Platforms do not issue comprehensive trade-level 1099s for these profits, but the income is taxable regardless. If the sums are meaningful, this is a question for a tax professional, not a blog post.

The legal risk sits mostly on the platforms, not you. The lawsuits, cease-and-desist letters, and the $36 billion demand target the exchanges. No state, as of August 2026, is pursuing individual traders. But an adverse ruling could force a platform to pull specific contract types — sports above all — from specific states, and positions in delisted markets would be unwound on the exchange’s terms.

The consumer-protection layer is different, not absent. A CFTC-regulated exchange must keep customer funds segregated and its markets surveilled for manipulation — but you don’t get state gaming-commission protections like mandated self-exclusion registries, and deposits carry no SIPC or FDIC-style guarantee on trading losses. Whichever label wins in court, the money you trade is money you can lose.

Quick answers

Is Polymarket gambling?
Legally contested. Federally, Polymarket US operates as a CFTC-designated derivatives exchange; several states have sent cease-and-desist letters treating event contracts — especially sports contracts — as unlicensed gambling. As of August 2026, no court has issued a final, nationwide answer.

Is Kalshi legal in my state?
Kalshi operates nationwide under its federal designation and has fought state enforcement in court with mixed results — it won an injunction against New Jersey at the Third Circuit but lost a similar bid in New York, where the attorney general is now suing it. The platforms remain live everywhere while the litigation runs.

What’s the real difference from sports betting?
Structure. A sportsbook sets the odds, takes the other side of your bet, and profits when you lose. A prediction market matches you with another trader, earns a fee either way, lets you sell out before the event resolves, and doesn’t limit winners.

Do I owe gambling taxes on my winnings?
Unclear — the IRS hasn’t issued guidance, and practitioners use ordinary-income, capital-gains, or gambling treatments. The gambling label is the costliest, since 2026 rules cap deductible gambling losses at 90% of winnings. The profits are taxable under any theory.

Can I lose more than I put in?
No. Event contracts are fully collateralized: your maximum loss is the price you paid. There’s no leverage and no margin call on standard contracts.

So is it gambling or investing?
Neither cleanly. It’s trading short-dated derivatives on event outcomes — regulated like finance at the federal level, challenged as gambling by states, structured like an exchange rather than a casino, and negative-sum after fees like any wagering pool. Treat it with the seriousness both labels demand.