The truth about predicting market liquidity: How 290,000 market data points reveal the virtual and real aspects of the ecosystem

When Polymarket announced the launch of the U.S. real estate prediction market, the market had high expectations for this emerging category. However, the trading volume after launch was disappointing—only a few hundred dollars in daily transactions, far below the buzz on social media. Behind this seemingly contradictory phenomenon lies a core truth about prediction markets: liquidity is not evenly distributed but highly concentrated.

To fully reveal this truth, we conducted an in-depth analysis of 295,000 of Polymarket’s market historical data. This investigation ultimately pointed to six key findings that allow us to re-understand the operational logic of prediction markets.

The Extreme Segmentation Behind the Data: Two Worlds of the Market

Prediction markets appear democratic and open, but data reveals a harsh reality—liquidity is extremely uneven.

Among the 295,000 markets, there are 67,700 ultra-short-term markets with cycles less than 1 day (accounting for 22.9%), and 63.16% of the trading volume in these markets is zero. In other words, over 13,800 markets are “hanging orders waiting” with no takers. This scene is reminiscent of the meme coin frenzy on Solana—large numbers of assets rushed online and ultimately abandoned, piling up without interest.

But this is not the most extreme contrast. When observing over different time spans, a startling divergence appears: the longer the time horizon, the deeper the liquidity.

Markets with cycles longer than 30 days number only 28,700, but their average liquidity reaches $450,000. In comparison, markets with cycles within 1 day have an average liquidity of only about $10,000. This indicates that capital participating in prediction markets shows a clear “long-term sedimentation” tendency—large funds do not compete with retail traders in short-term battles.

Sports and Politics: Two Poles of Liquidity Concentration

In prediction markets, the activity levels vary greatly across different categories, revealing the true needs of market participants.

Ultra-short-term sports predictions have become “high-frequency gambling.” The average trading volume for sports markets with cycles less than 1 day reaches $1.32 million, far exceeding the $44,000 in crypto predictions during the same period. Users seek immediate dopamine feedback—whether a team will win tonight, whether an athlete will break a record. The criteria for judging these events are simple and intuitive, making them ideal for short-term speculation.

In contrast, political predictions have become a safe haven for long-term capital. Markets related to U.S. politics (mainly cycles over 30 days) have an average trading volume of $28.17 million, with an average liquidity of $811,000. The appeal of these predictions does not lie in volatility stimulation but in macro strategic space—such as the 2028 U.S. presidential election outcome or whether certain policies will be implemented. These are the “big events.”

These two extremes form the “head-tail effect” of prediction markets—either providing immediate sensory stimulation or offering deep macro hedging, with other markets caught in an awkward middle ground.

The Dilemma of Crypto Predictions: Short-term Liquidity Crisis

For participants hoping to conduct short-term crypto trading within prediction markets, a brutal truth must be faced: liquidity is too shallow.

The average trading volume for crypto-type short-term predictions is only $44,000, which causes slippage issues for any sizable trading capital. Meanwhile, the mainstream use of crypto prediction markets is evolving—they are gradually becoming a “simple options hedging tool.” Users prefer to predict things like “Will BTC break $150,000 by year-end” or “Will a certain token fall below a specific price in 6 months,” rather than engaging in minute-level short-term volatility.

This signals a market maturity: strategies without sufficient liquidity are doomed to fail due to slippage.

The “Water and Soil” Mismatch in Real Estate Predictions: The Cost of Cold Start

Logically, real estate predictions should be a promising niche—relatively high certainty and longer cycles. But reality has defied this assumption.

After launch, the activity level in the real estate prediction market was far below expectations. There are three reasons: first, such markets require participants to have professional knowledge, unlike “whether a team will win” which has a low barrier; second, the volatility of real estate markets is relatively low, lacking frequent event-driven opportunities; third, professional players have no counterparties, and amateur players dare not enter—creating a vicious cycle of illiquidity.

This phenomenon reveals an important truth about prediction markets: emerging categories face a “cold start dilemma,” and the key to breaking this deadlock is to first attract enough professional participants to provide liquidity depth.

The “Spotlight Effect” of Liquidity: The Inevitable Concentration of Capital

When we segment markets by trading volume, a deeper truth emerges: liquidity is extremely unevenly distributed.

Markets with trading volumes over $10 million number only 505 but account for 47% of all trading volume. Conversely, markets with trading volumes between $10,000 and $1 million (156,000 markets) have a total trading volume of only 7.54%. This means most newly launched prediction contracts face the fate of “going to zero” immediately after listing.

Capital is not like evenly distributed sunlight but is concentrated around a few ultra-important events—like a spotlight. This truth is crucial for participants—only markets with abundant liquidity reveal true value; in illiquid markets, they are traps.

The Rise of Geopolitics: New High-Liquidity Territory

Among all prediction market categories, geopolitics is rapidly rising. This category accounts for 29.7% of activity (854 active markets out of 2,873 historical contracts), making it the most efficient among all sectors.

This reflects a shift in user focus toward real-time issues. Geopolitical events often have high uncertainty and immediacy, while also involving macro global patterns, attracting a broad range of participants from retail to institutional. The frequent abnormal trading in these markets also confirms the capital’s strong interest in geopolitical predictions.

The Evolution of Prediction Markets: From Utopia to Professional Tools

Synthesizing these truths behind the data, we see that prediction markets are undergoing a profound differentiation. They are no longer the “predict everything” utopia but are gradually evolving into highly professional financial tools.

In this ecosystem, successful markets tend to have at least one of two qualities: either providing immediate, frequent feedback mechanisms (like sports predictions) or offering deep macro strategic space (like political predictions). Markets lacking narrative density, with feedback cycles that are too long and low volatility, will gradually be phased out by decentralized order books.

For participants, the key truth to remember is: in prediction markets, understanding the distribution of liquidity is far more important than blindly chasing the next “hundredfold prediction.” Only where liquidity is abundant can you find real opportunities; in illiquid areas, even the best predictions will only lead to losses.