What are prediction markets?

Every market has two sides. A YES share pays out if the event happens; a NO share pays out if it does not. When the event resolves, the winning side is worth $1 per share and the losing side is worth nothing.

Because exactly one side must win, YES and NO always sum to $1. If YES is trading at $0.62, NO is trading at $0.38. Buying one YES share at $0.62 risks $0.62 to win $1 — a profit of $0.38 if the event occurs, and a total loss of the $0.62 if it does not.

The $0.62 in that example is not an arbitrary number. It is the price at which enough buyers and sellers were willing to trade, and it therefore encodes the market’s aggregate estimate that the event has roughly a 62% chance of happening.

The logic is straightforward: if traders believed the true probability were 80%, buying YES at $0.62 would be obviously profitable in expectation, so they would buy — pushing the price up until it stopped being a bargain. Prices settle where the expected value is roughly neutral, which is another way of saying they settle at the market’s honest estimate. This is why prediction-market prices are often quoted as forecasts.

It also means the prices update continuously as news arrives, which is the property that makes them interesting to read even for people who never trade them.

Shares trade continuously until the event resolves, so a position can be closed at any point at the prevailing price. A trader who bought YES at $0.30 and watched it rise to $0.70 can sell for a $0.40 gain per share without ever finding out how the event actually ends.

This is what distinguishes a prediction market from a conventional bet. A bet is locked once placed and pays only on the final result. A prediction-market position is a tradable asset whose value moves with the odds, and it can be exited early to take a profit or cut a loss.

A market is only as good as its resolution. Each one specifies in advance what will count as the answer and what source determines it. When the event concludes, that resolution process determines the winning side, and holders of the winning shares redeem them at $1 each.

On CoinXchange Predictions, markets, liquidity and resolution all come from the underlying Polymarket protocol. CoinXchange supplies the trading interface only — it has no role in deciding an outcome and cannot alter a resolution. Trading is non-custodial: positions are held on-chain in the user’s own wallet, so there is no account balance to withdraw because nothing was ever deposited with CoinXchange.

Prediction markets are regulated as financial or gaming products in many countries, and CoinXchange does not offer them everywhere. The product is not available to users in a substantial list of jurisdictions including the United States, United Kingdom, Canada, Germany, France, Italy, Australia, Japan, Singapore, the Netherlands, Poland and Brazil, together with comprehensively sanctioned countries.

The restriction is enforced at the network edge before the application loads, and applies regardless of wallet. Users elsewhere remain responsible for the laws where they live. CoinXchange’s cross-chain swap and other tools are globally available and unaffected.

A losing share is worth exactly zero — there is no partial recovery for being nearly right. Markets can also be thinly traded, which makes it possible to enter a position but hard to exit one at a fair price. And resolution can turn on the precise wording of the question rather than the spirit of it, so reading the resolution criteria before trading matters more than it first appears.

Nothing here is financial advice, and no position has a guaranteed return.