Cross-exchange funding-rate arbitrage

Funding Spread · Delta-neutral Execution

This strategy pairs a long position on a lower-rate venue with a short position on a higher-rate venue. It aims to reduce directional exposure and collect the rate difference, but it is not risk-free or perfectly market-neutral.

In this guide
  1. 01Constructing the hedge
  2. 02Settlement-interval pairing
  3. 03The execution threshold
  4. 04The risks that remain
01

Constructing the hedge

A common setup goes long on the lower standardized rate and short on the higher standardized rate, with notionals matched as closely as contract specifications allow.

Equal notionals reduce directional exposure but do not eliminate it. Mark prices, contract multipliers, basis, fill timing, and collateral differ across venues, so the two legs can diverge during a fast market.

02

Settlement-interval pairing

The ranking checks each pair rather than excluding every hourly quote. 1h can pair with 1h / 4h; 4h can pair with 1h / 4h / 8h; 8h can pair with 4h / 8h. Only direct 1h-versus-8h candidates are excluded.

Allowed pairs use a common eight-hour observation window, but shorter intervals can still reset before the window ends. The displayed spread is not locked return.

03

The execution threshold

Opening and closing both legs creates four executions. If each execution is modeled at 0.04% in fees plus 0.02% in slippage, the simplified round-trip cost is 0.24% before basis and funding-transfer effects.

At a gross 0.020% per eight hours, roughly twelve similar windows would be needed just to cover that simplified cost. The current spread is therefore a screening input, not an executable return.

04

The risks that remain

Different mark-price rules, contract multipliers, liquidation systems, and settlement times can make equal notionals drift apart.

Exchange failure, withdrawal suspension, margin imbalance, and a rapid funding reversal can turn a nominal arbitrage into a loss.

Frequently asked questions

QIs the displayed spread guaranteed profit?

No. It is a screening metric before execution costs, basis risk, and funding changes.

QWhy standardize the interval?

Allowed interval pairs need a common observation window, but the result is still not locked return.

QAre all one-hour quotes excluded?

No. Kieran Lab retains 1h-versus-1h / 4h candidates and excludes only direct 1h-versus-8h pairs.

Open the live tool

Compare cross-venue funding spreads

K