# Is Polymarket a good side hustle? What the data says on prediction markets

For most people, no. On one count 84% of Polymarket addresses are in the red, and profits are concentrated. The numbers, who wins, and a free way to test.

Source: https://rizon.markets/learn/are-prediction-markets-a-good-side-hustle (updated 2026-11-04)

> **Short answer:** For most people, no. On [one count of 2.5 million addresses](https://sergeenkov.com/polymarket-profitability/), 15.9% of Polymarket traders were in profit and only 2% had made over $1,000. Under 1% cleared $5,000 in any one month. Most people's positions are small, so most lose little, but the profits go to a small group, largely professional and largely automated. It can be a fine hobby on a budget you can lose. As an income, the numbers are against you.

## Who makes money

| Source | Finding |
|---|---|
| [Sergeenkov](https://sergeenkov.com/polymarket-profitability/), April 2026, 2.5M addresses | 15.9% in profit; 2% above $1,000; 0.32% above $10,000 |
| Same | $5,000 in a month: 0.98% in any one month, 0.1% two in a row, 0.015% four in a row |
| [Wall Street Journal analysis, as reported](https://en.cryptonomist.ch/2026/05/04/prediction-markets-traders-lose/), May 2026 | 67% of profits went to 0.1% of accounts |
| [Bloomberg](https://news.bgov.com/crypto/prediction-market-users-suffer-broad-losses-as-bots-reap-gains), April 2026 | Over 100,000 accounts lost at least $1,000, almost twice the number that made that much |
| [Pew](https://www.pewresearch.org/short-reads/2026/07/22/what-we-know-about-the-typical-polymarket-user/), July 2026, 11,989 wallets over six weeks | The typical wallet's net loss was under $2; 58% gained or lost under $100; 7% gained over $1,000, 9% lost that much |

Pew's is the useful corrective. Most users aren't being wiped out: they're trading small and losing small. The Sergeenkov count has its limits too: addresses aren't people, one person can run many, and he counts only realised PnL.

## Why winners win

[A study of 588 million Polymarket trades](https://ideas.repec.org/p/cpr/ceprdp/21615.html) found "successful traders provide liquidity using limit orders ... while unsuccessful traders take liquidity using market orders". So the split is partly about who pays the spread and who collects it. Bots do a lot of the collecting: [researchers measured about $40 million of arbitrage profit](https://arxiv.org/abs/2508.03474) on Polymarket in one year, with the top account making $2 million.

If you're clicking Buy on your phone, you're usually the one paying. On top of the spread, takers pay a fee of [up to 1.75¢ a share](https://rizon.markets/learn/polymarket-fee-calculator).

## Compared with a side hustle

| | Typical side hustle | Prediction markets |
|---|---|---|
| Worst case | You earn nothing | You lose your capital |
| Paid per hour | Yes | No |
| The edge you need | Being reliable | Knowing something the price doesn't |
| Who you compete with | Others in your area | Full-time traders and bots |

There's one case for it. If you really do know a subject better than the market, like a local election or a niche industry, then that knowledge has a price. The strongest places to use it are [near 50¢](https://rizon.markets/learn/why-every-market-on-rizon-is-50-50), where a better forecast is worth the most. Even then, expect most of your income to come from a small number of trades, and count the time you spent.

## How to test yourself cheaply

1. **Forecast on paper for a month.** Write down your probability for 50 questions and compare it with the market's. If you can't beat the market on paper, you won't beat it with money.
2. **Practice in a real book.** [Polymarket has no demo](https://rizon.markets/learn/polymarket-demo-account), but simulators copy its prices, and [Rizon](https://rizon.markets/?ref=learn-are-prediction-markets-a-good-side-hustle) gives you a live order book with paper money and a public leaderboard. Neither pays out.
3. **If you go real, cap it.** A budget you can lose whole, one venue, and a log of every trade with the fee counted. Check that it's available and legal where you live.

## FAQ

### Do people actually make money on Polymarket?

Yes, a minority: 15.9% of addresses on one count, most of them by small amounts. 2% had made more than $1,000.

### How much can you make on Polymarket?

There's no typical number, because the distribution is so skewed. Only 0.98% of addresses made $5,000 in any given month, and very few repeated it.

### What percentage of Polymarket traders are profitable?

It depends on the method: 15.9% in Sergeenkov's April 2026 count and about 30% in [another from December 2025](https://sergeenkov.com/polymarket-profitability/), which treated some flows differently. Studies count addresses, not people, so treat any exact figure with care.

### Why do people lose on Polymarket?

They pay the spread and fee on every trade, they tend to trade against better-informed people and bots, and they often have no edge to begin with. [Fixed windows](https://rizon.markets/learn/fixed-windows-vs-horizon-contracts) add a cost of their own on short markets.

### Is Polymarket legit?

It's a real, large market, and its positions settle by its stated rules. Whether you can use it depends on where you live. Being legit doesn't make it a good way to earn.