How Polymarket makes money: Burning cash today to mint a token economy tomorrow

Polymarket spent three years facilitating tens of billions of dollars in trading volume while reporting almost no protocol revenue. That changed in 2026. After returning to the U.S. and rolling out trading fees across nearly all of its markets, the platform now collects more than $1 million in fees on its busiest days, an annualized run rate of roughly $338 million by on-chain estimates. In October 2025, the NYSE's parent company Intercontinental Exchange (ICE) invested $2 billion, valuing Polymarket at over $8 billion. By April 2026 the company was in talks to raise fresh capital at a reported $15 billion valuation.

For most of its life, Polymarket optimized for growth over revenue, charging no fees and paying users to supply liquidity. It has now flipped that model, layering fees on top of its volume while still building toward a token-based system meant to make the economics durable. Understanding how the world's largest prediction market went from zero reported revenue to a nine-figure fee run rate shows where information markets are headed.

In this breakdown, we'll unpack how Polymarket now captures value through trading fees and hidden mechanisms, why it paid $112 million just for regulatory clearance, and how the upcoming POLY token is meant to turn soaring volume into lasting profit.

Polymarket was founded in 2020 by Shayne Coplan under the name Union. The platform emerged at a moment when early blockchain prediction markets like Augur had failed to gain traction, hampered by high Ethereum gas fees and poor user experience. Polymarket's mission was straightforward: democratize access to information markets and provide superior forecasting through crowd wisdom.

The platform operates as a decentralized exchange for event-based trading. It doesn’t work with fiat currencies, like USD or EUR. Instead, users deposit USDC stablecoin on the Polygon blockchain and trade binary outcome shares priced between $0 and $1 on real-world events. If you buy a "Yes" share at $0.70 and the event happens, you collect $1. If it doesn't, the share becomes worthless. The price at any moment reflects the market's collective probability.

Behind the scenes, Polymarket uses Automated Market Makers (AMMs) for liquidity and UMA's Optimistic Oracle for outcome resolution. 

For years, what set Polymarket apart was its zero trading fees. While competitors like Kalshi charged around 1.2%, Polymarket took nothing, using free trading and global crypto rails to drive explosive growth. The platform processed $3.3 billion on the 2024 U.S. election alone, with viral Twitter embeds of live odds amplifying its reach. That free-trading era ended in 2026, when Polymarket began charging variable fees across most of its markets.

That growth came with regulatory consequences. In 2022, the CFTC fined Polymarket $1.4 million for operating as an unregistered derivatives exchange and forced it to block U.S. users. The company spent the next three years building toward re-entry. In July 2025, it acquired QCX, a CFTC-registered exchange, for $112 million. That cleared the path back to American customers, and Polymarket relaunched in the U.S. in early 2026, with an iOS app rollout following that spring.

Polymarket spent its first years reporting no protocol revenue, running on what amounted to a negative take rate: it charged no fees and paid users to hold positions, funded entirely by venture capital. In 2026 it switched the meter on. The platform now charges trading fees across nearly all markets and clears more than $1 million on its busiest days, while still leaning on hidden mechanisms and venture funding to subsidize growth.

The sequencing is deliberate. Polymarket built volume first, then began charging for it, and it is holding its biggest monetization lever, the POLY token, in reserve until the U.S. business matures.

Polymarket doesn’t charge trading fees, but its system still generates revenue under the hood. The engine behind this is its Automated Market Maker (AMM)—a pricing model that uses math, not human market makers, to quote prices and guarantee liquidity.

Here’s the key idea: every AMM trade creates a tiny bit of value for the pool. Because prices adjust along a curve, traders always pay slightly more to buy than they would receive if they sold at that exact moment. That built‑in gap—think of it as the AMM’s version of a bid‑ask spread—accumulates over time.

A simple example: If someone buys a YES share at $0.52, the curve might adjust so that the next seller only receives $0.51. That one‑cent difference stays in the pool. On its own, it’s tiny. But across millions of dollars of volume, it becomes meaningful.

Those accumulated spread gains don’t go to external liquidity providers. They flow into Polymarket’s Treasury, which supplies and manages the liquidity. In effect, the platform captures the value that traditional market makers would normally earn.

This stands in sharp contrast to an order‑book exchange. There, human market makers post bids and asks—say, $10.00 to buy and $10.05 to sell. If one trader hits their ask and another hits their bid, the market maker earns the $0.05 spread for providing liquidity. With an AMM, the pool earns the spread instead—and because Polymarket controls the pool, it captures that value directly.

Today, this implicit revenue appears to be Polymarket’s primary source of operational funding. It isn’t reported as income because it’s typically recycled straight back into liquidity incentives or retained in the Treasury. But it’s real, it’s measurable, and without it, zero‑fee trading simply wouldn’t be sustainable.

Polymarket manages a substantial pool of deposited USDC collateral, known as Total Value Locked (TVL). This capital sits on the platform as users hold positions in active markets. Rather than letting it sit idle, Polymarket deploys it into yield-generating strategies.

The platform pays out 4% annual holding rewards to users from Treasury funds. To sustain this subsidy without depleting investor capital, the Treasury must earn more than 4% on deployed collateral. The difference between gross yield and the 4% paid to users is net interest margin—a critical but undisclosed revenue source.

This yield capture functions as subsidized customer acquisition. By paying users to hold positions, Polymarket incentivizes capital deposits and market participation. The 4% reward keeps liquidity sticky and ensures markets remain liquid even during low-activity periods.

The strategy only works if the Treasury can consistently generate returns above 4%. Given that USDC can be deployed into low-risk DeFi protocols or traditional money market instruments, this is achievable. But it requires active management and exposes the platform to interest rate risk. If yields fall below 4%, the subsidy becomes a direct cost.

The POLY token is the linchpin of Polymarket's long-term monetization strategy. The company's CMO has confirmed that a token and an airdrop are coming, introducing utility through governance and staking. It gives Polymarket a way to route fee revenue back to its most active users rather than simply extracting it.

Here's how it works: with POLY live, the trading fees Polymarket already collects can be routed to token stakers or accrued to the protocol treasury. Users who stake POLY earn a share of platform revenue, creating a direct incentive to hold and use the token. And Polymarket, as a significant holder of its own token, benefits both from token appreciation and from earning a share of staking rewards.

This turns raw volume into a measurable revenue stream. Polymarket generated about $31 billion in volume between September 2025 and February 2026 and now clears more than $2 billion in a typical week. Even a 0.5% take rate on that pace would generate hundreds of millions of dollars a year. As volume grows, so does revenue, and staking turns active traders into stakeholders.

The timing is strategic. Polymarket held the token back until its U.S. platform was live and its fee model was proven, minimizing regulatory scrutiny during the sensitive re-entry. Delaying the launch also gives the company time to structure POLY to avoid being classified as a security.

Once live, token-based value accrual is expected to become the core of Polymarket's reportable economics. It is central to justifying a valuation that reached over $8 billion in the ICE round and a reported $15 billion in its 2026 fundraising talks. Fees have proven the model can generate cash. The token is meant to make that cash flow durable and to reward the users who create the volume.

Trading fees have moved from theory to Polymarket's clearest source of real revenue. The platform first charged fees only on crypto and sports contracts. On March 30, 2026, it expanded variable taker fees to nearly all categories, including politics, finance, economics, culture, weather, and tech. Daily fee revenue crossed $1 million on April 1, up from $696,000 two days earlier.

The structure is dynamic rather than flat. Fees are priced off probability, peaking near a 50% chance of an outcome and dropping toward the extremes. Crypto markets carry the steepest rate at 1.80%, while sports sit lowest at 0.75%. Geopolitics and world events stay fee-free. Makers pay nothing and instead collect daily rebates worth 20% to 25% of the fees collected in their category.

At the April 2026 pace, on-chain analysts pegged Polymarket's annualized fee run rate at roughly $338 million. That is a genuine revenue line for a platform that reported none a year earlier, though it still trails Kalshi's reported $1.5 billion annualized run rate. The strategy is to let fees fund operations now, then use the token to capture and redistribute that value over time.

Sustaining that run rate is not guaranteed. Fees can dampen the very volume they tax, and Polymarket's active traders on the global platform slipped to about 643,000 in April 2026 from more than 733,000 a month earlier. The company is balancing fees low enough to keep liquidity deep against a rate high enough to generate meaningful income.

Polymarket's real-time prediction data is valuable beyond the platform. Hedge funds, media companies, and policymakers all want access to crowd-sourced probabilities on events ranging from elections to economic indicators. ICE's $2 billion investment, which valued Polymarket at over $8 billion, explicitly targets this opportunity.

ICE plans to become the global distributor of Polymarket's event-driven data, packaging odds as sentiment factors alongside traditional financial data. This positions Polymarket as information infrastructure, not just a trading venue. The data itself becomes a product.

Data licensing carries near-pure profit margins. Once the data exists, selling it has a low marginal cost. Polymarket could charge for API access, historical data, or custom analytics tools. It could create indices based on prediction market odds or offer derivatives tied to event probabilities.

Polymarket's cost structure is heavily skewed toward regulatory compliance and user subsidies rather than typical tech operations. The company is spending aggressively to achieve market dominance before monetization.

The $1.4 million CFTC fine in 2022 was just the beginning. The real cost of compliance came in July 2025, when Polymarket acquired QCX for $112 million. This wasn't an acquisition of technology or users. It was pure regulatory arbitrage—buying CFTC-registered status to serve American customers legally.

That $112 million is arguably the ultimate Customer Acquisition Cost for the entire U.S. market. It's a one-time fixed expense that unlocks access to the world's largest derivatives market. By incurring this cost, Polymarket mitigates regulatory risk and differentiates itself from unregulated DeFi peers.

Beyond the acquisition, the company faces ongoing compliance costs. It hired former CFTC Chairman J. Christopher Giancarlo as an advisor. It maintains dedicated legal and compliance teams to manage CFTC reporting, market certifications, and geo-blocking technology to prevent U.S. users from accessing the global site.

The 4% annual holding rewards represent millions in annual costs, depending on market positions. This subsidy is funded by the Treasury and functions as heavily subsidized marketing. By paying users to hold positions, Polymarket ensures markets remain liquid even during low-activity periods.

The NHL partnership likely includes licensing fees, adding to the cost of building mainstream credibility. These expenses are strategic. Polymarket is treating liquidity as a moat, using venture capital to build a network effect that competitors can't easily replicate.

This approach mirrors DeFi "liquidity mining" strategies, where platforms pay users to provide capital. The difference is that Polymarket is doing it at scale, with institutional backing, and with a clear path to monetization once the token launches.

Polymarket covers Polygon blockchain gas fees via meta-transactions, absorbing costs that users would otherwise pay. It integrates with UMA's Oracle for market resolution, MoonPay and Stripe for fiat onramps, and Coinbase for crypto deposits. Each integration carries development and maintenance costs.

The platform recently added stock index markets, requiring new development work. Smart contract audits and security measures are ongoing expenses.

Operating out of Manhattan means high overhead. Engineering, compliance, legal, and customer support teams all command premium salaries in one of the world's most expensive cities. Polymarket also brought on Nate Silver as an advisor in 2024, adding credibility but also cost.

Security audits and cyber liability insurance are necessary given the platform's exposure to hacks and exploits. Potential lobbying efforts for favorable regulations add to the expense. M&A integration costs from the QCX acquisition are still being absorbed.

Polymarket dominates crypto-native prediction markets and clears more than $2 billion in a typical week, but it faces intensifying competition from regulated exchanges, traditional betting platforms, and DeFi challengers. The lines between these categories are blurring as crypto platforms become regulated and traditional platforms add prediction markets.

Kalshi is Polymarket's closest competitor, and in the U.S. it is the clear leader, holding roughly 90% of the market. The two platforms crossed a combined $150 billion in lifetime volume in April 2026, and both now clear multiple billions of dollars in weekly volume, but they took different paths to get there.

Kalshi has been CFTC-regulated from inception and works in fiat, which gave it legal certainty in the U.S. while Polymarket was locked out. It built an early lead in sports and now reports annualized trading volume of about $178 billion. Polymarket's strength has been politics and crypto, and its global, crypto-native rails let anyone trade without a U.S. license.

Both now generate real fee revenue, but at very different scales. Kalshi's annualized revenue has topped $1.5 billion, while Polymarket's newer fee model runs at roughly $338 million annualized. The two are converging on price: Polymarket added fees to a formerly free platform, while Kalshi has leaned into sports contracts that ride a federal loophole.

Kalshi raised about $1 billion in 2026 at a $22 billion valuation and has opened informal talks about an IPO. That leaves Polymarket, in talks at a reported $15 billion, valued at a discount to its rival for the first time, partly because investors are unsure how much of its volume is organic versus token-airdrop farming. The winner will be whoever combines regulatory legitimacy with the broadest coverage and the best user experience.

Betfair, the world's largest betting exchange, operates on a 2-5% commission model. DraftKings acquired Railbird, a CFTC-registered prediction market, in late 2025. These platforms bring massive user bases and marketing budgets that dwarf Polymarket's.

Polymarket differentiates through breadth and zero or low fees. Traditional platforms have trust and simplicity advantages for mainstream users. They're household names with established customer acquisition channels.

FanDuel and others may follow DraftKings into prediction markets, intensifying competition. The question is whether traditional platforms can match Polymarket's liquidity and market variety. Polymarket's crypto rails enable global participation, while traditional platforms are constrained by state-by-state licensing.

Augur, the original Ethereum prediction market, failed due to high gas fees and poor UX. Polymarket learned from those mistakes, choosing Polygon for low costs and curating markets for quality. Other players like Limitless Exchange, Zeitgeist, and Manifold remain niche. A newer threat comes from Hyperliquid, whose HIP-4 framework brings prediction markets to a high-volume on-chain exchange, and from Binance's wallet, which has begun testing in-app prediction trading.

SX has accumulated $151.8 million in three-year volume. BetSwirl has $38.4 million. Azuro has $8.7 million. All are dwarfed by Polymarket's tens of billions in volume. The liquidity moat creates a "winner-take-most" dynamic. New entrants face a cold start problem: without liquidity, markets don't function.

Polymarket has shifted from a crypto-native prediction app into a fee-charging, partly regulated event-trading exchange with institutional data ambitions. Its next phase depends on three levers: defending its U.S. re-entry, launching the POLY token, and building out ICE-backed financial data.

U.S. commercialization is now underway. Polymarket relaunched to American users in early 2026 and expanded through an iOS app that spring, with sports as the initial wedge into the world's largest wagering market. The open question is whether it can win share from Kalshi, which still controls roughly 90% of U.S. prediction-market activity.

The POLY token is still the company's biggest unplayed card. Fees have already moved Polymarket off a zero-revenue base, but the token is meant to route that revenue to stakers and the treasury and to reward the traders who create volume. The design emphasizes governance and staking while trying to avoid a security classification.

ICE transforms Polymarket from a betting platform into a financial data supplier. The partnership aims to create event-based indices and market-moving probability feeds for institutional investors. If successful, Polymarket becomes an input into Wall Street—not just a consumer app. Tokenization will connect real-world events to traditional financial rails.

Long term, Polymarket aims to formalize event trading as a new derivatives class and to position itself as a global probability oracle for businesses and governments. Fees have answered the first question of whether it can make money. The next question is whether the POLY token can turn that revenue into a durable, defensible economy while Kalshi pulls ahead in the U.S. If it can, the reported $15 billion valuation looks like a floor. If it can't, Polymarket risks being the higher-volume platform that still lost the market it created.