Perpetual futures use funding payments between longs and shorts to keep the contract price anchored to spot. When the same asset trades on multiple venues, funding rates rarely line up exactly — and that gap is the basis of funding rate arbitrage.
The comparison everyone gets wrong
Hyperliquid settles funding hourly. Binance and Bybit settle every 8 hours. Comparing the raw per-period rates side by side without annualizing for each venue’s actual interval produces a spread number that looks meaningful but isn’t — it’s comparing an hourly figure to an 8-hourly one as if they were the same unit. The correct comparison is each rate annualized as rate × (24 / intervalHours) × 365.
Where the data actually comes from
Hyperliquid’s own predictedFundings endpoint already returns the predicted next funding rate for the same coin across Hyperliquid, Binance Perp, and Bybit Perp in a single call — including each venue’s funding interval, which is what makes correct annualization possible without separately querying each exchange.
Why a wide spread isn’t free money
A large annualized spread tells you funding compensation differs meaningfully between venues right now — it does not account for trading fees, slippage on entry and exit, or the capital required to hold both legs of a hedge (long on one venue, short on the other) simultaneously. Funding rates also reset every settlement period; a wide spread today can normalize by the next funding window.
Summary
Funding rate arbitrage is a real strategy, but only if you compare rates on an apples-to-apples annualized basis and price in the real cost of running both legs of the trade. Raw rate comparisons and headline spread numbers alone will mislead you.