Building liquidity amid uncertainty

Author: Prathik Desai Translation: 善欧巴, Golden Finance

Can a company achieve revenue growth without changing (or even while significantly reducing) trading volume? Core economic theory tells us that raising prices will drive away customers. But Polymarket has just completed a price increase: although it lost some customers, revenue nevertheless grew.

Before the beginning of this year, the globally largest prediction market by trading volume had been operating on a zero-revenue model. For most of its development history, users were free to trade, deposit, or withdraw without paying the platform any fees.

And all of that was completely changed within less than a quarter of a year.

At the start of January this year, Polymarket first announced it would charge fees on certain categories of markets. Three months later, on March 30, this prediction market rolled out a full fee policy: it announced fees would be charged on all categories except geopolitics markets, and even increased the fee rates for categories that were already charged.

Within just eight days after the new rules were published, the total fees collected by the platform had already exceeded the sum of the entire previous month.

In this edition of the analysis, I will walk you through Polymarket’s brand-new fee structure—how it reshapes the platform’s liquidity-building logic in an environment full of market uncertainty.

The inverse relationship between trading volume and fees

After Polymarket broadly increased fee rates across various categories, the expected situation appeared: the cost of prediction trading went up, causing trading volume to fall. Did the platform anticipate this outcome? The answer is almost certainly yes—that’s exactly what basic economic laws predict. But the platform doesn’t need to worry too much.

Although higher fee rates reduce users’ demand to trade on Polymarket, the drop in trading volume was not proportional to the increase in prices—which also aligns with basic economic principles. The end result is that Polymarket’s total fee revenue grew.

This is not the first time Polymarket has abandoned the traditional belief that “more trading volume means faster business growth.” Back in January, the platform already led the way by rolling out a fee policy for some market categories.

This shift is directly reflected in Polymarket’s daily fee rate (the proportion of fees as a share of total trading volume). This metric has continued to rise steadily.

From early January to the late March period, this ratio gradually increased—from about 0.001, doubling to 0.002. After March 30, the fee rate increased directly by two times, breaking through 0.007—equivalent to generating about $0.7 in revenue for every $100 of trading volume.

But how exactly does a business achieve revenue growth with trading volume staying flat (or even lower)? The answer lies in how the platform precisely targets the user group that is willing to pay a premium for a high-quality trading market.

And the cleverness of this model is hidden in its fee calculation formula.

Pricing in uncertainty

Unlike traditional fixed commission mechanisms, Polymarket does not use a single fixed fee rate. Instead, it dynamically adjusts fees based on the probability of an event’s outcome. The formula is: fee = position share × fee rate × price × (1−price)

All markets on Polymarket trade using “Yes/No” share formats. Share prices float between 0 and 1 dollar, reflecting traders’ expectations of an event’s likelihood. For example, a price of $0.90 means traders collectively believe the event has a 90% chance of occurring. Polymarket’s fee design rule is: fees are highest when the event probability is close to 50%, and as the result moves toward certainty (i.e., probability approaches 0 or 1), fees gradually fall to the minimum.

This model ensures the platform captures the highest profits in the markets where outcomes are most disputed. This design fully matches economic logic. Imagine if an event’s probability is 90% or 10%: the potential payoff from trading is already thin. If high fees further compress profits, traders will give up on trading—ultimately suppressing trading activity in the margin markets.

Let’s use an example market: “Whether Bitcoin will reach $100k before June 30.” Suppose the price of the “Yes” share is $0.90, representing that traders are 90% confident Bitcoin will rise to $100k by the deadline. If you buy 100 shares at this price, you pay $90. If your prediction is correct, you can earn $100, for a net profit of $10.

With a fixed fee rate of 1.5%, this $90 trade would pay a $1.35 fee, dramatically cutting the maximum potential profit of $10. Polymarket’s probability-banded fee schedule solves this problem. If you buy 100 shares—fees are calculated according to the event’s likelihood.

When the share price is $0.90, the fee is only $0.65 (calculation formula: 100 shares × 0.072 fee rate × $0.90 price × $0.10 (1−price)); and when the probability is split evenly at 50%, the fee reaches the peak of $1.80 (calculation formula: 100 shares × 0.072 fee rate × $0.50 price × $0.50 (1−price)).

This design concentrates fee revenue at the nodes of the probability curve where the value in the trade-offs is highest.

Data source: Polymarket official documentation

The rebate flywheel effect

This bell-curve fee model also has another clever design.

Polymarket not only charges a premium to active traders in exchange for tighter bid-ask spreads and high-quality liquidity, but it also distributes part of its fee income to market makers in the form of rebates. Active traders are users who directly complete trades by clicking market “Buy / Sell.” Market makers, by contrast, build a deep order book by posting limit orders, thereby improving liquidity for the platform.

The rebate mechanism attracts more market makers to join, which in turn improves liquidity and attracts more active traders. The rebate proportions differ across categories: crypto categories at 20%, political and sports categories at 25%, and financial categories at 25%. Only geopolitics markets do not charge active traders’ fees, and they also do not provide rebates to market makers.

Data source: Polymarket official documentation

Polymarket applies a zero-fee, zero-rebate policy to geopolitics and global events markets, reflecting its operational thinking for that category: the platform wants to attract traffic with the characteristics of zero barriers, zero cost, and high exposure. Once traders move onto the platform, it then guides them to trade across categories, capturing fee revenue from other categories.

A bet on liquidity deployment

Polymarket has been operating for more than five years, yet in the past four years it has consistently struggled to achieve precise product-market fit. The 2024 U.S. election became the key turning point that broke the deadlock. Before that, the platform was little known and had few users.

After the election frenzy faded, outsiders questioned whether prediction markets could keep developing, while sports and crypto categories filled the traffic gap.

With the rollout of this fee policy, Polymarket’s revenue received a major short-term boost. But can the platform retain users and prevent them from switching to competitors?

History may hold the answer.

When Uber launched in 2012, it took a 20% commission from each trip fare, with the rest paid to drivers. Over the years, the commission percentage kept being raised, and by 2025, in some scenarios it even exceeded 40%. Drivers organized protests and some chose to leave—but it did not affect operations: new drivers kept flowing in, and riders still relied on the app.

The takeaway from this case is: if a platform gradually raises fee rates, as long as the competitor experience is worse, users will accept the higher prices. Does Uber’s approach comply with moral standards? I hold a more cautious view.

But Polymarket’s move is far more than just a simple price increase.

Raising fixed fees extracts value from users’ unfavorable position of having no better alternative. Polymarket’s bell-curve fee schedule, however, charges the highest fees exactly at the stage of trading where the value in the traders’ game is highest—and it also feeds part of the revenue back to market makers. It is these market makers that make the platform worth keeping for users.

Better rebates attract more market makers. More market makers lead to tighter bid-ask spreads and a deeper order book. This is crucial because Polymarket uses a centralized limit order book model: every trade incurs the cost of the bid-ask spread, and large orders can also produce slippage when sweeping the book. These costs won’t show up on the fee statement, but they are ultimately borne by traders.

On a platform with scarce liquidity, even if trading is free of fees, traders still suffer losses from wide spreads and high slippage. On Polymarket, even though traders pay a small fee, they can obtain better execution thanks to the high-quality liquidity supported by rebates. If this flywheel works properly, Polymarket—through lower total trading costs enabled by high liquidity—will be lower than low-liquidity platforms. In that case, the fee is merely a small cost paid for a better trading experience.