Home / Ideas / BTC/ETH Cointegration Pairs
BTC/ETH Cointegration Pairs
market-neutral mean reversion on the ETH/BTC spread
BTC and ETH are strongly cointegrated (~96% in cited tests), so the ETH/BTC ratio mean-reverts around a slow-moving equilibrium. Practitioner and academic backtests report annualized returns of ~15–16% with Sharpe ~1.0 to 2.23 net of ~10 bp costs. Market-neutral construction removes crypto-beta, which is attractive while directional BTC is capped by macro.
Why it's relevant now
With BTC range-bound and no clear trend, a beta-neutral relative-value trade harvests the ETH-vs-BTC oscillation instead of betting on direction.
Universe
ETH-USD and BTC-USD (yfinance daily). Trade the spread/ratio; dollar-neutral legs.
How it works
Compute rolling z-score of the ETH/BTC log-ratio (e.g. 30–60 day window); short the spread at +2σ, long at −2σ, exit near 0σ; equal dollar legs.
Expected performance
Research-derived Sharpe estimate: 0.6–1.2. sources 1.0–2.23 gross; discounted for costs, single-pair concentration, and cointegration-break risk).
Backtest this idea with SignalChain
This is a research lead — not a finished backtest. SignalChain takes an idea like this and runs the whole pipeline inside Claude Code: it researches the concept against academic and practitioner sources, sets benchmarks, writes and lints a VectorBT backtest, runs it, and grades the result PASS/FAIL. One command:
/signalchain Market-neutral pairs trade on ETH-USD vs BTC-USD: z-score the 40-day log-ratio, short at +2 sigma, long at -2 sigma, exit at zero.
Get SignalChain — $49 →
Research & sources
- blog.amberdata.io — blog.amberdata.io
- ijsra.net — ijsra.net
- insights.deribit.com — insights.deribit.com
More crypto and digital-asset strategies ideas
Not financial advice. This page describes a research idea, not a recommendation. Any performance figures are hypothetical, research-derived estimates and are not indicative of future results. SignalChain is a research and educational tool; you are solely responsible for any decisions you make.