Prediction
Markets UK

An independent guide to
an uncertain subject.

Our editorial approach
01 / PARTICIPATION

Who makes money on prediction markets? 76.5%

US research on profit concentration, maker and taker returns, and the risks of interpreting a winning trade as a repeatable strategy.

Updated · Research note
The short reading

The available studies describe strikingly uneven outcomes. They do not show that a novice can reproduce the returns of the best-performing accounts—or that the same patterns apply to a UK betting exchange.

This is research and context, not a market recommendation or a statement of UK product availability.
At a glance76.5%of positive Polymarket profits went to the top 1% of profitable users in one studyView original source
Measure / 01$67bntrading volume studiedPolymarket sample in the Akey and co-authors working paper.View original source
Measure / 02−32%average taker returnApproximate average loss reported in the Kalshi study.View original source
Measure / 03−10%average maker returnApproximate average loss for the comparison group.View original source
01 / Analysis

How profits are split

Akey and co-authors examined a large Polymarket trading sample and found that gains among profitable accounts were highly concentrated. Their 76.5% figure uses the top 1% of users who had positive profit and loss—not the top 1% of everybody with an account. It is a result from a particular platform and period, not an expected return for future participants.

02 / Analysis

Makers, takers and the cost of immediacy

A maker posts an order that can be matched later; a taker accepts an available order. Burgi, Deng and Whelan report that takers in their Kalshi analysis lost almost 32% on average, while makers lost about 10%. Selection, fees and trading against better-informed participants all matter. A maker can still lose, and providing liquidity is not a risk-free income stream.

03 / Analysis

Why cheap contracts can be expensive

A ten-cent contract may look affordable, but it is only fairly priced if the event occurs often enough after costs. Longshots can attract buyers for the excitement of a large payout. If the implied chance is higher than the true chance, repeatedly buying them may lose money despite occasional wins. This is a behavioural and pricing risk, not a formula for finding easy profit.

04 / Analysis

What a UK reader can—and cannot—infer

These figures come from US or globally accessible services. Different operators have different contracts, customer protections, markets and entry rules. Before comparing them with a UK exchange, check the legal operator and product. The safest conclusion is about uncertainty and concentrated outcomes, not a recommended trading tactic.

Figure / maker-taker-loss

Average loss in the Kalshi sample

Average loss in the Kalshi sampleHorizontal bar chart. % average loss; exact figures appear in the data table below. MEASURE % AVERAGE LOSS Takers −32%Makers −10%
Chart data: Average loss in the Kalshi sample
Measure% average loss
Takers−32%
Makers−10%
Takers and makers both had negative average returns in the cited study.

Reading note. Bars encode the magnitude of loss; a taller bar means a worse average result. These are sample averages, not forecast returns.

SourceThe economics of Kalshi prediction markets
Method / How to read this

Methodology.

The headline is from an academic working paper, and the maker/taker comparison is from a separate Kalshi analysis. Samples, periods and profit measures differ. This page does not pool them into a single population. The cited papers disclose their own authorship and funding; no operator paid this site for placement.

Reference / Original record

Sources.

  1. 01
    Who Wins and Who Loses in Prediction Markets? Evidence from Polymarket (opens in a new tab)
    Research papers

    Funder: Academic working paper; check the paper for author disclosures.

    Akey and co-authors analyse Polymarket trading gains across $67bn in volume. Working-paper estimates may be revised.

  2. 02
    The economics of Kalshi prediction markets (opens in a new tab)
    Research papers

    Funder: CEPR/VoxEU publication; see authors’ disclosures.

    Burgi, Deng and Whelan report average losses for maker and taker groups, not guaranteed outcomes for individuals.

Follow the original documents for definitions, scope and subsequent revisions. Browse the full source library ↗

Questions / In context

Questions worth asking

01Do most prediction-market traders make money?

The cited studies show uneven outcomes in their samples, but they do not establish a universal percentage for every product or person. Do not assume market participation is profitable.

02What is the difference between a maker and a taker?

A maker leaves a quote available for others to match; a taker trades against an existing quote. Their costs and selection risks can differ.

03Does the 76.5% result include all users?

No. It describes the share of positive profits captured by the top 1% of users with positive profit and loss in that Polymarket sample.