Prediction markets, explained

A market that trades on the future prices truth in real time. Here's how the mechanism actually works — the math, the money, and why every serious market is about to have one.

In an ordinary market, a price tells you what something is worth. In a prediction market, a price tells you how likely something is. A contract that pays $1 if an event happens trades at whatever the crowd believes the odds to be. If it changes hands at 62¢, the market is saying 62% — and it updates the instant new information arrives, with real money behind every revision.

That single property — a continuously-updated, money-weighted probability — is why prediction markets keep outperforming polls, pundits, and committees on questions that have a clean answer.

A price is the most honest poll ever taken — every respondent has money on the answer.

A price is a probability

Consider a binary contract: it settles at $1 if some event resolves YES and $0 if NO. A risk-neutral trader will buy it whenever the price sits below their estimate of the true probability and sell when it sits above. In equilibrium, the price is pinned to the market's collective probability estimate:

$$ \text{price} \;=\; \mathbb{E}[\text{payoff}] \;=\; (1)\cdot p \;+\; (0)\cdot(1-p) \;=\; p $$

So the order book is a probability distribution. Read the price, read the odds. A market trading at $0.62 is a 62% forecast — no translation needed.

Who takes the other side?

Early markets are thin: you want to bet, but there's no one to bet against. The fix is an automated market maker — a formula that will always quote a price. The classic is Hanson's Logarithmic Market Scoring Rule (LMSR), which prices a basket of outcomes from a single cost function:

$$ C(\mathbf{q}) \;=\; b \,\ln\!\left( \sum_{i} e^{\,q_i / b} \right) $$

Here $q_i$ is the number of shares outstanding on outcome $i$, and $b$ is a liquidity parameter — bigger $b$ means deeper books and smaller price moves per trade. The cost to move the market from $\mathbf{q}$ to $\mathbf{q}'$ is simply $C(\mathbf{q}') - C(\mathbf{q})$. Differentiate, and the instantaneous price of each outcome falls out as a softmax:

$$ p_i \;=\; \frac{e^{\,q_i / b}}{\sum_j e^{\,q_j / b}} $$

Two things to notice. The prices always sum to one — $\sum_i p_i = 1$ — so the market is a coherent probability distribution by construction. And the maximum the market maker can ever lose is bounded by $b \ln n$ for $n$ outcomes: liquidity has a known, finite price. That boundedness is what makes a market operable rather than a charity.

The category went from a curiosity to a billion-dollar month

For a decade prediction markets were an academic toy. Then the volume arrived. Monthly trading went from under $100M in early 2024 to more than $13B by the end of 2025 — the bulk of it sports — and the sell side now models the category on a path to $1 trillion a year by 2030.¹

Monthly volume — illustrative. Sources in notes.¹

Why now

The structural unlock was regulatory. In 2020 the U.S. CFTC recognized event contracts as a distinct asset class² — turning a grey-zone product into a licensed one. Kalshi, the first federally-regulated exchange, was just valued at $11B.³ And the incumbents are circling: ICE — owner of the NYSE — committed up to $2B to Polymarket, for the right to distribute its data,⁴ with CME and Cboe entering the category.

What makes them hard

Three problems separate a toy from an exchange. Liquidity — without a market maker, books are empty (the LMSR above is one answer). Resolution — someone must adjudicate the real-world outcome, cleanly and without dispute. And regulation — a market that pays out on real events is, in most jurisdictions, a licensed financial product. The first two are engineering. The third is the moat: the right to operate legally is granted per country, and it does not transfer.

That last problem is the one we spend our time on. The mechanism is well understood; the license is not. Seeker's bet is that every country will regulate prediction markets — and the winner in each is whoever shows up with the compliance infrastructure to operate inside the rules.

For the bigger picture — why these markets are worth building at all — read Why prediction markets matter.

Notes

Category trajectory — sub-$100M/mo (early 2024) → more than $13B/mo by the end of 2025 → on track for ~$1T/yr by 2030. Pew Research Center; The Block; Bernstein.
CFTC recognition of event contracts as an asset class, 2020.
Kalshi raised $1B at an $11B valuation (Series E, December 2 2025, led by Paradigm). Kalshi.
ICE — owner of the NYSE — committed up to $2B to Polymarket (October 2025), structured around becoming the global distributor of its event data. ICE; FinTech Weekly.