Backtesting a Cointegration-Based Crypto Pairs Strategy
Summary
This article presents a multi-year backtest of a market-neutral crypto pairs strategy using ETC and FIL quoted in USDT. The broader method selects a pair based on cointegration and mean-reversion evidence, estimates a hedge ratio, tracks the spread with rolling Z-scores, and opens positions when deviations reach preset thresholds. It describes exits using reversion targets or stop-loss rules, with transaction costs and slippage considered.
The reported test shows positive cumulative and annualized returns, with volatility, drawdown, Sharpe ratio, profit factor, win rate, and average win-to-loss figures used to characterize performance. The narrative emphasizes that a small number of large winning reversions outweighed more frequent losses, and recounts several early stop-outs followed by large winners. It also cites an ADF test and a Hurst exponent below 0.5 as support for stationarity and mean reversion. These are results for one selected pair and historical sample; they do not establish robustness across assets or future periods. The supplied discussion is incomplete near its treatment of trading frictions, and backtest assumptions may affect the reported outcomes.
Key ideas
- The strategy trades deviations in a cointegrated spread between ETC and FIL rather than predicting overall crypto direction.
- ADF testing and a Hurst exponent below 0.5 are presented as evidence supporting stationarity and mean reversion.
- A hedge ratio and rolling spread Z-scores define the market-neutral positions and their entry signals.
- The reported backtest had a 40% win rate, with larger average wins than losses driving performance.
- Results from one historical pair and test period may not generalize, even when costs are included.
Tags
This summary was written by Stratmill's research agent from the original; it is not a copy of the source.