Minimalist Trading
  1. Home
  2. Learn
  3. How prediction markets work

How prediction markets work

A prediction market sells yes or no contracts on a future event. Each pays $1 if it comes true and $0 if not, so the price in dollars is read as a probability. Here is the math behind it.

  • 6 min read
  • By a.b
  • Updated November 2026

The contract

A prediction market lists a question with a clear yes or no answer and a set end date, for example whether a data release will be above a stated level. A YES contract pays $1.00 if the answer is yes and $0.00 if no. A NO contract pays the other way. Because the payout is fixed, the only thing that moves is the price you pay, and it always sits between $0.00 and $1.00.

Why a price reads as a probability

If a YES contract costs $0.62, the market is saying the event is about 62% likely. The logic is the break-even point: a buyer who is right 62% of the time on contracts bought at $0.62 gets back, on average, exactly what was paid (0.62 × $1.00 = $0.62). Pay less and the same buyer would profit on average, so buyers push the price up. Pay more and sellers push it down. The price is the crowd's estimate, not a fact, and it can be wrong.

Payoff of one contract (E)

OutcomePayReceiveProfit or loss
Event happens (YES wins)$0.62$1.00+$0.38
Event does not happen$0.62$0.00−$0.62

You risk $0.62 to make $0.38. Selling a YES contract at $0.62 is the same bet in reverse, and equals buying NO at $0.38: you risk $0.38 to make $0.62. With 100 contracts the numbers are $62 at risk and $38 or $100 at stake.

This page is education, not advice, and not an invitation to trade. Every price and fee below is invented (E) to show the arithmetic. Rules on who may trade event contracts differ by country, so check your own regulator.

YES plus NO is usually more than $1

Quoted prices include the market's spread. If YES costs $0.62 and NO costs $0.41, buying both costs $1.03 for a guaranteed $1.00 back, a 3 cent loss, or 2.9% of the outlay. Dividing each price by 1.03 gives probabilities that add to 100%: 0.62 ÷ 1.03 = 60.2% and 0.41 ÷ 1.03 = 39.8%. The same step is used to remove the bookmaker margin from betting odds, so the quoted 62% overstates the market's view by about 2 points here.

Fees raise the break-even

Suppose a fee of $0.02 per contract (an invented figure). A YES contract at $0.62 then costs $0.64 all in, and you need the event to happen more than 64% of the time to come out ahead. Always add the fee to the price before judging an edge. Fee schedules differ and change, so read the live one.

Edge: your belief against the price

Expected value per contract is your probability minus the price. It is only an edge if your probability is better than the market's, and you cannot know that in advance.

Your beliefPrice paidExpected value per contractAs % of stake
70%$0.62+$0.08+12.9%
62%$0.62$0.000%
55%$0.62−$0.07−11.3%

The 70% row only pays on average if 70% is really right. If the true chance is 62%, the same trade has no edge at all. This is the same logic as expectancy in trading: a positive number from a guess is not a proven edge.

Cheap and expensive contracts are not symmetric

A contract at $0.95 pays only 5 cents if right and loses 95 cents if wrong, so one loss erases 19 wins. A high price feels safe because it usually wins, but the break-even hit rate is also 95%.

PriceWinLoseWins needed per loss to break even
$0.50+$0.50−$0.501
$0.80+$0.20−$0.804
$0.95+$0.05−$0.9519

Test a rule yourself: buy 100 contracts at $0.80, one at a time, and 75 of the events happen. You spent $80, received $75 and lost $5, even though three in four were right. Being right often is not the same as being right often enough.

Sizing a binary bet

A contract has one of two outcomes, so the Kelly fraction has a simple form: f = (p − q) ÷ (1 − q), where p is your probability and q the price. With p = 70% and q = $0.62, f = 0.08 ÷ 0.38 = 21%. That assumes your 70% is exactly right. Five losses in a row at 10% of the account each leaves 0.95 = 59% of it. Because the probability is only an estimate, a common approach is a fraction of Kelly or a fixed small risk, as explained in how much to risk per trade. You can test the formula in the Kelly criterion calculator.

What this guide does not cover

It does not list live prices, rank platforms, link to any platform or say whether a given market is a good bet. It explains the arithmetic so that any quoted price can be read with fees and spread removed.

Checklist before reading any price

  1. Read the exact question and the rule that decides the winner, including the source and the end time.
  2. Add the fee and the spread to get the real break-even probability.
  3. Write down your own probability before looking at the price.
  4. Decide the most you can lose, as a share of your account, before you enter.

Quick questions

What does the price of a prediction market contract mean?

It is the cost in dollars of a contract that pays $1 if the event happens. After removing the spread and fees, the price is read as the market's estimated probability, for example $0.62 as about 60 to 62%.

How do you make money on a prediction market?

Only when your probability is better than the price after fees and spread, and you size the trade so a run of losses is survivable. Buying at $0.62 needs the event to happen more than 62% of the time just to break even.

Is a prediction market the same as betting?

The payoff has the same two-outcome shape as a fixed-odds bet, and the same math applies. Legal treatment differs by country and by product, so check your own regulator before trading.

Share it or follow along

If this helped, share it with someone who trades. We post new worked risk examples on LinkedIn every week: follow the Minimalist Trading Page to get the next one.

Keep reading

a.b

Written by a.b. I joined TradingView in December 2015 and publish indicators there. These guides are for education only. Message me on TradingView chat with questions or topics you want covered.

Last updated November 2026. For education only. Not financial advice.