How to read prediction-market odds as a trader
Prediction-market prices are probabilities with costs attached. How to convert them, why the resolution criteria matter more than the number, and where the odds lead or lag price.
THE SHORT ANSWER
A prediction-market contract that settles at $1 if an event happens trades at roughly the market's implied probability: 62 cents is about a 62% chance. Before using that number, read the resolution criteria and check the spread and depth, because a thin book can quote a probability nobody is willing to trade in size.
Prediction markets have become a routine input for event-driven traders because they express a market view in the one unit that other instruments do not: an explicit probability. A contract that pays $1.00 if an event occurs and $0.00 if it does not will trade near the collective estimate of that event happening. At $0.62, the market is pricing roughly a 62% chance.
Converting price to probability, and back
The conversion is direct but the details cost money. Implied probability is the contract price divided by the payout. Fees, the bid-ask spread and — on some venues — the cost of capital locked until resolution all sit between the quoted number and the number you can actually transact at.
| Price | Implied probability | Payout if right | Return if right |
|---|---|---|---|
| $0.10 | 10% | $1.00 | +900% |
| $0.35 | 35% | $1.00 | +186% |
| $0.50 | 50% | $1.00 | +100% |
| $0.75 | 75% | $1.00 | +33% |
| $0.92 | 92% | $1.00 | +8.7% |
The right-hand column explains a persistent feature of these markets. At $0.92 a contract has 8.7% of upside and 100% of downside, so it takes a genuinely high conviction — or a very short time to resolution — to be worth the capital. That asymmetry is part of why the tails are frequently mispriced relative to a well-calibrated forecast, and it is also why a 92% quote is not the same claim as a 92% probability from a statistical model.
Why the odds are useful even if you never trade them
The practical value for a trader in another market is a benchmark. Before a scheduled decision, an equity or crypto position is implicitly a bet on an outcome. A prediction market makes the consensus explicit, which turns a vague sense that something is priced in into a number you can disagree with.
- It quantifies surprise. If a decision is priced at 88% and the alternative lands, the repricing is proportional to the 12% nobody was positioned for.
- It timestamps the shift. Odds often move on a report before the underlying market does, because the contract is a purer expression of the same news.
- It separates event risk from direction. You can believe an outcome is likely and still conclude the asset move is already priced.
- It provides a post-mortem baseline: what did the market believe an hour before the release?
That last use is underrated. Reviewing an event trade against the odds at the time of entry is a far better test of the decision than reviewing it against the outcome, and it belongs in the record described in how to build a trading journal.
Where the number breaks down
A prediction market is only as good as its book. Several failure modes recur and all of them are visible before you rely on a quote.
- Thin liquidity: a mid-price between a $0.40 bid and a $0.70 offer is not a 55% consensus, it is an absence of one.
- Stale quotes: low-volume contracts can sit unchanged through news that clearly changes the answer.
- Longshot bias: very low-probability contracts often trade above a fair estimate because the payout profile attracts buyers.
- Capital cost: on venues without interest on collateral, a contract resolving in nine months is discounted for the tied-up capital, which pulls prices toward the middle.
- Resolution risk: the price includes the chance the criteria resolve in a way you did not expect.
The discipline is the same one that applies to any other quoted market: check the spread, check the depth, and prefer contracts where both sides are being made in size. A 62% quote on $400 of resting depth is a curiosity. The same quote on a deep, actively arbitraged book is information.
Turning odds into an expected move
The most directly useful conversion is from a probability into a rough repricing estimate, because that is the number that decides whether an event is worth trading at all. The arithmetic is a two-branch expectation: if the market is 88% sure of outcome A and an asset is expected to sit at 100 under A and 92 under B, the price today should be near 0.88 x 100 + 0.12 x 92 = 99.04.
Invert it and the trade becomes visible. If A lands, the asset moves from 99.04 to 100 — under 1%. If B lands, it moves to 92, a fall of over 7%. The asymmetry is not an opinion about the outcome; it is a direct consequence of the odds. Heavily-priced events always pay little for being right about the consensus and a great deal for being right about the tail, which is why the interesting question before a scheduled decision is rarely "what will happen".
| Odds on A | Move if A lands | Move if B lands | Ratio |
|---|---|---|---|
| 50% | 4.0% | -4.0% | 1.0x |
| 70% | 2.4% | -5.6% | 2.3x |
| 88% | 1.0% | -7.0% | 7.3x |
| 95% | 0.4% | -7.6% | 19.0x |
The table assumes an 8-point spread between the two outcome prices; the ratio column is what generalises. At 95% confidence the market pays roughly nineteen times more for the surprise than for the consensus, which is the same structural asymmetry described in FOMC day explained and the reason so much event risk is expressed through options rather than direction.
Using odds alongside a scheduled release
The workflow that adds most is comparing three things before a known event: the prediction-market odds, the pricing implied by the instrument you actually trade, and the positioning shown by funding and open interest. When those three disagree, the disagreement is the trade idea — and the invalidation is usually the moment they converge.
The Forecandle console puts event odds beside the scored news feed and live market data so this comparison does not require three tabs and a stopwatch. For the scheduled events themselves, the economic-calendar process covers what to write down before the release.
Frequently asked questions
- Do prediction-market prices equal probabilities?
- Approximately. A binary contract paying $1 trades near the implied probability, so 62 cents is about 62%. Fees, spread, the cost of capital locked until resolution and longshot bias all pull the traded price away from a fair estimate, most noticeably at the extremes.
- Are prediction markets more accurate than analysts?
- They aggregate opinions weighted by willingness to risk money, which tends to produce well-calibrated forecasts on liquid, clearly-defined contracts. On thin or ambiguously-worded contracts they are no better than the handful of participants quoting them, and sometimes worse.
- Why do prediction-market odds move before the underlying market?
- The contract is a direct expression of the event, so a credible report changes it immediately. Other markets have to translate the same news through positioning, hedging and second-order effects, which takes longer and is noisier.
Related reading
- MacroFOMC day explained: the statement, the projections and the press conferenceWhy FOMC days often produce two opposite moves, what the dot plot changes, and how to read a decision that was fully priced beforehand.
- MacroHow to use an economic calendar before a volatile sessionA pre-event process for inflation, labour and central-bank releases: what to write down beforehand, how to read a surprise through the current regime, and when to stand aside.
- News intelligenceA trader's checklist for breaking market newsHow to separate a headline that changes market structure from one that only creates noise: source, delta from expectations, transmission channel and confirmation.
Or browse the full library on the research index, and see the same data live in the console.