Studies of prediction market data consistently show the same pattern: the majority of active traders lose money over time. The most recent on-chain analysis — published April 2026 by independent analyst Andrey Sergeenkov, covering 2.5 million Polymarket wallet addresses — found 84.1% of traders in the red. Why? And what does the profitable minority do differently?
The data
Prediction market trading data is public on-chain, and two independent studies have measured this directly:
- April 2026 (Andrey Sergeenkov, 2.5M wallets): 84.1% of traders unprofitable; the odds of any trader earning $5,000+/month consistently were under 1% in a given month
- December 2025 (DeFi Oasis, 124M trades across 1.7M addresses): 70% of trades unprofitable — an earlier, narrower methodology that the April study superseded by also capturing token splits and merges
- Across both studies, a small fraction of traders — well under 5% — account for a disproportionate share of all profitable volume
Why most traders lose
Overconfidence in their predictions. Prediction markets are efficient. The crowd has already priced in most publicly available information. If you think an event is 70% likely and the market says 65%, you probably don't have an edge — you just feel like you do.
Taking directional risk without compensation. Most traders buy YES or NO and hold until resolution. If the market was efficiently priced when they bought, expected value is zero (minus fees). Fees alone guarantee a negative expected return for the average directional trader.
Overtrading. Each trade has a fee. Frequent trading on thin edges guarantees negative returns even if each individual bet has a small positive expected value in theory.
Recency bias and narrative trading. Traders over-weight recent news. After a dramatic event, they pile into the obvious-seeming side — which is already priced in.
What the winning minority do differently
The consistently profitable traders on prediction markets generally fall into two categories:
Information traders: People with genuine informational edge — insiders, domain experts, or professional researchers who know something the crowd doesn't. This is rare and hard to sustain.
Arbitrageurs: Traders who don't predict outcomes at all. They exploit mispricings — between platforms, within markets, or between related contracts. No prediction required. Pure math.
Arbitrage: the market-neutral alternative
Cross-platform arbitrage between Polymarket and Kalshi is the most accessible form of market-neutral trading in prediction markets. You don't need to predict anything. You just need to be faster than the market at identifying and closing price gaps. Arbitrage Agent does this automatically — join the waitlist for early access.
Related articles
- Arbitrage Polymarket: How to Profit Without Predicting
- Polymarket vs Kalshi: Why the Same Event Has Different Prices
- How Much Can You Make with Prediction Market Arbitrage?
- Best Polymarket Arbitrage Agent: Automated vs Manual Comparison
- Do Prediction Markets Limit Winning Traders? What We Found (2026)
Key terms: Directional Exposure · Cross-Market Arbitrage