Fed Contracts Show Repeated Cross-Platform Pricing Gaps
Kalshi and Polymarket have repeatedly priced the same Federal Reserve and political outcomes far enough apart to create apparent cross-platform arbitrage opportunities, highlighting how liquidity remains fragmented even as prediction markets grow into a major trading category.
Data tracked by prediction-market intelligence platform Predictbook shows that the differences were not confined to a single stale quote or short-lived market dislocation.
On Sept. 10, Predictbook recorded multiple spreads involving the Federal Reserve’s September decision. At 10:19 a.m. ET, a 25-basis-point September rate increase could effectively be paired by buying the “Yes” contract on Kalshi for 56 cents and the opposing “No” position on Polymarket for 37 cents.
The combined cost was 93 cents for positions that would ordinarily produce a combined $1 payout if the contracts settled consistently, implying a gross 7% gap before trading fees, slippage and other costs.
The same combination appeared again at 1:26 p.m. ET. By 3:08 p.m., Polymarket’s “No” price had fallen to 35 cents while Kalshi remained at 56 cents, increasing the apparent spread to 9%.
That matters because a pricing disagreement that survives across several observations over multiple hours is more difficult to dismiss as a single bad tick.
The September contract was not the only example.
Predictbook recorded a 6% spread on the December 25-basis-point Fed hike contract at 11:26 a.m. on Sept. 10, with Kalshi “Yes” at 43 cents and Polymarket “No” at 51 cents. That widened to 8% by 2 p.m. and 10% shortly after 3 p.m., when the opposing contracts cost a combined 90 cents.
Two days later, the difference was even larger.
At 3:27 p.m. ET on Sept. 12, Predictbook showed Kalshi’s December hike “Yes” contract at 43 cents while a Polymarket “No” position cost 45 cents. The resulting 88-cent combined price produced an apparent 12% gross spread.
Predictbook also identified a 9% gap on the October Fed contract later that afternoon.
The divergence is particularly notable because expectations around the Federal Reserve’s September meeting have been moving rapidly after stronger US labor and inflation data increased the probability of another rate increase.
By Sept. 13, the most immediate September pricing discrepancy had largely closed. Public market snapshots showed the 25-basis-point hike near 79% on Kalshi and approximately 79.5% on Polymarket, demonstrating that cross-venue prices can eventually converge even after meaningful temporary gaps.
However, the repeated December differences suggest the issue is not simply one platform being permanently more hawkish on the Fed. Different portions of the policy curve can diverge independently as traders, capital and liquidity shift between venues.
Brazil Election Markets Show the Same Pattern
The same phenomenon has appeared outside monetary policy.
Predictbook recorded repeated gaps in contracts asking whether Luiz Inácio Lula da Silva will win Brazil’s 2026 presidential election.
At 10:52 a.m. ET on Sept. 10, Lula “Yes” was priced at 45 cents on Polymarket while “No” cost 46 cents on Kalshi, leaving a 9% gross difference. By 1:43 p.m., the Polymarket price had moved to 44 cents while Kalshi’s opposing contract remained at 46 cents, creating a 10% gap.
The same 10% configuration appeared again later that afternoon.
On Sept. 12, Predictbook continued detecting differences in the Lula market, although the quoted spread had narrowed to around 5% at several observations.
The repeated Brazil signals matter because they suggest the Fed discrepancies are not necessarily specific to macro markets. They may instead reflect a broader structural feature of Polymarket and Kalshi: two separate pools of traders, capital and market makers can maintain different estimates of the same outcome for meaningful periods.
There are important qualifications.
Predictbook labels these observations as arbitrage signals, but its public signal feed carries a 30-minute delay. A quote displayed by the service therefore should not be treated as proof that the same trade was still available when a user saw it.
Displayed prices also do not establish how much size was available at the quoted level. A trader attempting to buy thousands of contracts could consume the best orders and quickly erase the apparent profit.
Trading fees further reduce the headline spread. Kalshi charges transaction fees based partly on contract price and expected earnings, while Polymarket applies taker fees to several market categories.
More importantly, apparently identical contracts do not always have perfectly identical settlement language.
Independent contract-comparison services examining the September Fed markets found that Kalshi and Polymarket would be expected to produce the same result under ordinary circumstances but identified differences in published policies covering unusual situations such as canceled meetings, missing information or revisions.
That means the trade may look mathematically hedged without being legally identical under every possible resolution scenario.
Fragmented Liquidity Is the Bigger Story
The interesting part is not whether someone could have captured exactly 7%, 10% or 12% on one of these trades.
The bigger issue is that two increasingly important markets can maintain substantially different probability curves for the same event.
Prediction markets are supposed to aggregate information into a price. If one platform says an event has roughly a 56% probability while another effectively prices the same outcome closer to 63%, there is no single “market probability.” There are competing probabilities shaped by who happens to trade on each venue.
That is classic market fragmentation.
Traditional financial markets deal with the same problem, but mature assets are surrounded by professional market makers, high-frequency firms and routing systems whose entire business is removing price differences between venues.
Prediction markets have not fully reached that stage.
Kalshi and Polymarket have grown rapidly, but their capital pools remain separated. They have different account structures, settlement infrastructure, regulatory frameworks and user bases. Polymarket’s global platform is blockchain-based and uses USDC, while Kalshi operates as a regulated US derivatives exchange.
Moving capital between those environments is not equivalent to moving an equity order from one stock exchange to another.
That friction allows price differences to survive.
It also creates an opportunity for specialized arbitrage firms. As volumes increase, there is an obvious economic incentive for traders to maintain funded balances on multiple venues, monitor corresponding order books continuously and automatically buy whenever combined opposing positions fall sufficiently below $1.
If that capital arrives, spreads should become smaller and disappear faster.
That would arguably improve the informational quality of prediction-market prices. Research has already suggested that only a small minority of sophisticated participants account for much of Polymarket’s price discovery, while most traders primarily supply liquidity.
Cross-platform arbitrageurs could serve a similar function: they do not need to know whether the Fed will hike or whether Lula will win. They simply need to know that the same event should not be priced materially differently in two places.
But tighter integration would introduce other questions. Regulators are already focusing on market manipulation, information advantages and surveillance as event-contract volumes increase. Faster arbitrage could make markets more efficient while simultaneously concentrating more activity among sophisticated automated traders.
The industry is also attracting increasingly large amounts of capital. Kalshi has raised more than $1 billion as the prediction-market sector expands, while Polymarket has become a significant source of political and economic pricing. That growth makes persistent cross-market discrepancies harder to dismiss as a niche technical curiosity.
The same markets are already being used to track major policy events. Large Polymarket positions have appeared around the Clarity Act, while Fed contracts now react in real time to inflation, payrolls and central-bank commentary.
As those probabilities increasingly become inputs for investors, journalists and financial firms, whether two major venues agree starts to matter.
For now, the Sept. 10-12 data suggests they do not always agree quickly.
And while some gaps disappear, their repeated reappearance across the Fed curve and political markets suggests fragmented liquidity remains a structural feature rather than a one-off pricing error.
The next development to watch is whether professional arbitrage capital closes those gaps as the industry matures, or whether differences in access, regulation and market structure allow separate probability curves to persist even as volumes grow and legal fights over prediction markets continue.
Shane Neagle is a financial markets analyst and digital assets journalist specializing in cryptocurrencies, memecoins, prediction markets, and blockchain-based financial systems. His work focuses on market structure, incentive design, liquidity dynamics, and how speculative behavior emerges across decentralized platforms.
He closely covers emerging crypto narratives, including memecoin ecosystems, on-chain activity, and the role of prediction markets in pricing political, economic, and technological outcomes. His analysis examines how capital flows, trader psychology, and platform design interact to create rapid market cycles across Web3 environments.
Alongside digital assets, Shane follows broader fintech and online trading developments, particularly where traditional financial infrastructure intersects with blockchain technology. His research-driven approach emphasizes understanding why markets behave the way they do, rather than short-term price movements, helping readers navigate fast-evolving crypto and speculative markets with clearer context.