Written by: Clow, the Blockchain in Plain Language

Last month, a friend who trades perpetual contracts told me he was done with them.

Not because he was scared of losses—though he indeed lost a significant amount—but because he discovered a "cleaner" way to gamble. Polymarket launched a 5-minute Bitcoin price prediction market, where he can buy a bunch of "up" shares for $10, and 5 minutes later, even if Bitcoin only increases by 1 cent, he can get back $100. If it drops? He loses those $10, gone, clean and simple.

No liquidation, no funding fees, no suffocating feeling of being liquidated by a spike at three in the morning and then having the price pulled back.

He said: "This thing is like a crypto scratch-off ticket, but the odds are right in your face."

Data shows he is not alone. On March 11, Polymarket revealed that in less than a month since the launch of the 5-minute prediction market, the daily trading volume has already surpassed $60 million, accounting for 67% of all directional prediction trade volume on the platform. With 288 settlement windows each day, from morning to night, every 5 minutes, it never stops.

Prediction markets were once used for betting on the U.S. election and the Super Bowl. Now, they have turned into a 24-hour operating slot machine.

The Three Mountains of Perpetual Contracts

Why are retail investors fleeing perpetual contracts?

The answer is simple: perpetual contracts are too unfriendly to retail investors. First, liquidation. If you take a long position with 10x leverage, a 10% price drop can wipe you out, and even if the price goes back up an hour later, it doesn't matter to you; your position has already been devoured by the exchange. Second, funding fees. When longs are crowded, you have to pay a "rent" to shorts every 8 hours, with costs increasing the longer you hold. Third, price spikes. In the early morning when liquidity is thinnest, a price spike of hundreds of dollars can wipe out a slew of stop-losses.

The 5-minute prediction market eliminates all three of these issues.

You buy a $0.1 "up" share, and the worst outcome is losing that $0.1. But if you win, you get back 1—10 times the return. It doesn't matter how Bitcoin fluctuates in between; you only look at the price at the moment the 5 minutes end. No liquidations, no funding fees, no "precision explosions" from the market maker.

Simply put, this is a form of speculation with completely transparent risks. You know beforehand how much you can potentially lose, which is impossible with perpetual contracts—unless you use no leverage, but who trades perpetual contracts without leverage?

For those who once were keen on meme coins and hundred-fold contracts, the 5-minute prediction market is tailor-made: high frequency, exciting, low barrier to entry, immediate results. This is not replacing perpetual contracts; this is stealing the users away from perpetual contracts.

How does Polymarket achieve this?

Technically, Polymarket uses a "Conditional Token Framework" (CTF), with each 5-minute window being an independent binary event: Did Bitcoin go up or not. Settlements run on the Polygon chain, which is low-cost and fast; price feeds use Chainlink Data Streams to ensure that the settlement price isn't determined by the platform itself.

But the truly smart design is in the transaction fees.

With such short cycles of 5 minutes, what’s the biggest fear? Delay arbitrage. Some people can place bets a few tenths of a second before the Chainlink price feed updates, almost guaranteeing a profit. Polymarket's response is clever: dynamic fees. When market probabilities approach 50% (the point of maximum uncertainty), the fee for eating the order is higher, reaching up to 1.56%. When the outcome becomes more certain (probabilities near 0 or 100%), fees drop almost to zero.

This means that if you want to arbitrage in the most "profitable" ambiguous zone, you must first pay a not-so-cheap toll. The opportunity for making money via internet speed is significantly compressed, forcing you to bet based on true judgment.

The collected fees are not wasted—20% is directly returned to market makers, incentivizing them to provide deeper liquidity on the order book. In just one month since launch, the 5-minute market has already developed the depth to accommodate large trades.

AI Robots Have Arrived

With a daily trading volume of $60 million, how much do retail investors contribute? Probably not as much as one might think.

Developers on Reddit are already sharing trading bots for the 5-minute market, claiming win rates of over 80%. These bots use "mechanism marking" technology to automatically determine whether the current environment is a bull market, bear market, or sideways trend, and then switch between different prediction strategies. With 288 settlement windows each day, the backtesting data is extremely rich, allowing AI to accumulate the sample size that traditional markets take years to gather, in just a few days.

More importantly, Polymarket just announced a collaboration: On March 10, the platform partnered with Palantir and TWG AI to use the Vergence AI engine developed by the latter two to monitor trading activities, screen for non-compliant accounts, and detect insider trading.

There’s a subtle contradiction here. On one hand, the platform welcomes AI traders—they provide liquidity and make the market more efficient; on the other hand, it has to guard against AI cheating—using faster data sources to jump ahead, cross-platform arbitrage, or even price manipulation. Polymarket's solution is to use AI to monitor AI.

As for who will win this "AI vs AI" arms race, no one knows yet. But one thing is certain: the proportion of human participation in the trading volume of the 5-minute market will only decrease.

Exchanges Are Stirring

Faced with the encroachment of prediction markets, traditional exchanges are uniformly reacting: if you can't beat them, recruit them.

In early March, Binance launched Opinion (OPN) via Launchpool, a prediction market infrastructure protocol, in an attempt to corner the prediction market from the protocol layer. Coinbase directly integrated Kalshi's contracts at the end of January, allowing U.S. users to trade prediction markets within the Coinbase app. Gemini is even more aggressive, spending five years obtaining a DCM license from the CFTC, and building Gemini Predictions, covering all 50 states in the U.S.

Three paths, three strategies, but the goal is the same: keep those users who are flowing towards prediction markets.

The trend is best illustrated by Kalshi's story. In 2024, Kalshi's annual trading volume was about $300 million, an unremarkable number. Then it integrated NFL event contracts into Robinhood, allowing tens of millions of retail users to buy prediction contracts just like stock purchases. In 2025, Kalshi's annual trading volume soared to $23.8 billion, with annualized volume hitting close to $50 billion at one point.

The jump from $300 million to $23.8 billion was not due to Kalshi becoming stronger, but because it found distribution channels. Prediction markets are no longer a niche tool that users have to seek out actively—they are embedded in brokerage apps, payment software, and even media platforms, becoming a fundamental financial function.

This dimensionality reduction in distribution capability is what exchanges truly fear.

The Sword of Regulation Hangs Overhead

The faster prediction markets grow, the more apparent the contradictions in regulation become.

In the U.S., the CFTC has deemed prediction contracts as "swaps," categorized as federally regulated financial derivatives, and submitted a court opinion in February 2026 to assert exclusive jurisdiction. However, state gaming commissions do not agree— the chairperson of Nevada's Gaming Control Board stated: "In our view, this is just sports betting, plain and simple." Close to 20 states have launched lawsuits against Kalshi or issued stop orders, and 37 states have formed an alliance against the federal government encroaching on their turf.

The result is an absurd situation: legal at the federal level but illegal at the state level. Polymarket has already acquired a CFTC-licensed entity by late 2025 to relaunch a U.S. version in waiting list mode, but as of today, its expansion speed remains constrained by the tug-of-war between federal and state authorities.

In Asia, the situation is more direct. The Singapore Gambling Regulatory Authority directly shut down Polymarket in January 2025, treating all bets, whether on Bitcoin or the Super Bowl, as illegal gambling. The Hong Kong Securities and Futures Commission is slightly more courteous, allowing professional investors to trade virtual asset derivatives, but firmly shutting the door on retail investors.

The EU, on the other hand, is trapped in a classification quagmire: if the prediction contracts are anchored to financial indices, they are financial instruments under MiFID II; if anchored to non-financial events, multiple member states directly treat them under gambling bans. France, Belgium, and Romania have already been at the forefront of blocking them.

The global regulatory attitude toward prediction markets can be summed up in one sentence: everyone wants to regulate, but no one knows how to.

Conclusion

Polymarket's 5-minute market demonstrates one thing: retail investors have never wanted to "hold assets," but rather "gamble on outcomes."

When a platform can take speculation to the extreme with simpler rules, lower barriers to entry, and faster feedback cycles, high-leverage products from traditional exchanges are no longer the only choice. Exchanges are already scrambling to recruit prediction markets, while regulators argue whether it is gambling or finance—but users do not care about these definitions.

Every 5 minutes, 288 times, day after day.

The votes cast by retail investors with their feet are more honest than any regulatory definition.

免责声明：本文章仅代表作者个人观点，不代表本平台的立场和观点。本文章仅供信息分享，不构成对任何人的任何投资建议。用户与作者之间的任何争议，与本平台无关。如网页中刊载的文章或图片涉及侵权，请提供相关的权利证明和身份证明发送邮件到support@aicoin.com，本平台相关工作人员将会进行核查。