Prediction Markets Reward the Few, Not the Crowd

The Crowd Is Not the Signal

Prediction markets are often sold as digital democracy for probabilities, but the latest research points to a harsher reality: the market does not reward the average participant equally. Instead, a small cluster of informed traders appears to capture a disproportionate share of gains while most users pay for the privilege of being early, wrong, or both. That matters because these platforms are increasingly used to price political outcomes, macro events, and even crypto-adjacent narratives. If the edge is concentrated, then the “wisdom of the crowd” is really the wisdom of a select few.

The implication is important for anyone treating prediction markets as pure consensus engines. In practice, the order book is a battlefield between liquidity providers, skilled takers, and a much larger base of casual users. That is not a moral judgment; it is a structural one. Once spreads, fees, and informational asymmetry enter the picture, the market can still produce useful probabilities, but not because everyone is equally informed. It works because the few who know more are able to express that knowledge fast enough to move price.

What the Study Suggests

The research cited in the article found that roughly 3.5% of informed traders, including market makers and skilled takers, captured more than 30% of total profits on prediction platforms, while around 67% of users absorbed the entirety of losses. The authors used a sign-randomization test, repeating each account’s historical trades 10,000 times to estimate profit-and-loss distributions under randomized conditions. That kind of methodology matters because it tries to separate genuine skill from luck, noise, or one-off streaks.

This finding also fits a broader academic pattern. Earlier work on prediction markets has long shown that more experienced traders tend to outperform and that markets often become more efficient as weaker participants exit. More recent research on decentralized prediction markets has similarly argued that skilled traders can profit by exploiting the biases of less skilled participants. The newer NBER work on Kalshi also reinforces that these venues are becoming more important as real-time forecasting tools, especially for macro and policy expectations.

Why This Matters For Crypto Markets

For crypto readers, the lesson is not that prediction markets are broken. It is that they are closer to a microstructure game than to a simple aggregation of public opinion. In a thin, fast-moving market, price discovery depends heavily on who is willing to provide liquidity, who can process information quickly, and who understands the resolution rules better than everyone else. That is why the same market can be informative and extractive at the same time. The number printed on the screen may be accurate, but the path to that price is often uneven.

That distinction matters more as prediction markets expand beyond niche crypto circles. They are now being used to interpret macro prints, election odds, and event outcomes with increasing seriousness. But a market that is useful for forecasting does not need to be fair in the everyday sense. It only needs enough informed capital to push price toward truth. The structural cost is that the most attentive traders can earn persistent edge while the majority subsidizes market efficiency through losses and fees. That is the hidden subsidy inside “crowd wisdom.”

What This Means For Investors (Our Take)

The practical takeaway is straightforward: treat prediction market prices as signals, not as proof of consensus. A market can be directionally right and still be dominated by a narrow elite of traders. For investors, that means probabilities should be cross-checked against liquidity, volume concentration, and how close a market is to resolution. When participation is shallow, the tape can look smarter than it really is. When informed flow is concentrated, the market may be telling the truth for reasons most users do not share.

What to watch next is whether this concentration persists as institutional participation deepens. If more professional market makers and data-driven traders enter, prices may become cleaner even if retail outcomes worsen. Also watch whether platforms disclose more about trader performance, since that would help separate genuine forecasting skill from simply being first in the queue.

Focus: Prediction markets may forecast events well, but the profits appear to belong to the few who understand the game best.

Monica Ramires, Senior Markets Analyst, The Chain Journal