Renko and Kagi H-Statistics for Pairs Selection and Trading
Summary
The article presents a pairs-trading framework that uses Renko- or Kagi-style constructions to identify turning points in a spread. From those points, it derives H-statistics: H-inversion counts directional changes, H-distance summarizes turning-point moves, and H-volatility measures average movement between turns. The proposed selection method forms spreads from log prices, sets a construction threshold using historical spread volatility, and ranks pairs by H-inversion. A contrarian strategy is suggested for mean-reverting processes, while momentum is associated with other H-volatility conditions.
The article reports applications to S&P 500 constituents and cryptocurrency pairs, with periodic pair selection and transaction-cost scenarios. It describes declining performance over time for the equity application and sensitivity to transaction costs and unstable results in crypto. It also notes that liquidity and resulting slippage were not included in the crypto backtest. The claimed theoretical profitability for contrarian trading depends on assumptions about mean reversion, and practical results remain sensitive to threshold choice, spread construction, market structure, and costs.
Key ideas
- Renko and Kagi constructions identify turning points and confirmation points in a tradable spread.
- H-inversion is proposed as a measure for ranking pairs with mean-reverting characteristics.
- The framework uses spread volatility to set a turning-point threshold and selects pairs with high H-inversion.
- The choice between momentum and contrarian trading depends on the process characteristics summarized by H-volatility.
- Reported performance varies across markets and transaction costs, while crypto liquidity and slippage are not modeled.
Tags
This summary was written by Stratmill's research agent from the original; it is not a copy of the source.