Anyone who has conducted prediction market research has probably encountered the same problem: your carefully designed arbitrage strategies look perfect on paper, but when launched, they leave you dumbfounded. The issue isn't with your mathematical formulas; it's that the market simply lacks liquidity.

This phenomenon has recently become particularly evident with Polymarket. Last year, they launched a US real estate prediction market, which sounds very innovative, right? But in reality, these markets only see a few hundred dollars in daily trading volume. There's a lot of buzz on social media, but the market remains cold and quiet. The gap is so large that it's almost ironic.

What's really going on? We decided to let the data speak.

We collected historical data from 295,000 Polymarket markets, and the results were a bit eye-opening: 67,700 markets, or 22.9%, have cycles shorter than 1 day; 198,000 markets, or 67.7%, have cycles under 7 days. In other words, two-thirds of the markets are ultra-short-term.

Even more astonishing, among these ultra-short-term markets, 21,848 are still active. Sounds like a lot? Check the trading volume: 13,800 markets had zero trading volume in the past 24 hours, accounting for 63.16%. That means roughly 60% of active markets are actually untraded.

It's like low-level battles in PvP games—seems lively, but in reality, there's little liquidity. The liquidity problem in prediction markets might be more severe than we imagined.