Hurst-Filtered Mean Reversion in Stock-Index Spreads: The HBB Model
This working paper proposes the HBB model, a trading signal for a US stock hedged against the S&P 500. The model fits a rolling hedge ratio between the stock and the index, scores the resulting spread in Bollinger-style z-units, and estimates a Hurst exponent from the variance of hedged price changes at lags of 1, 2, 4, and 8 bars. A buy or sell signal fires only when the spread sits two standard deviations from its mean, and the Hurst estimate falls below 0.5, marking a mean-reverting spread. The model runs on fifteen-minute bars to limit bid-ask noise in the Hurst estimate. The paper also proposes a Bayesian extension, which carries the Hurst estimate across trading sessions as a belief rather than recomputing it each bar. It lists the model’s limitations and four testable predictions, and includes the full Pine Script code for TradingView in an appendix. The paper presents a hypothesis only. The author has not tested the model on market data. The idea belongs to the author. Claude, an AI model by Anthropic, wrote the code and drafted the text.
Authors
- Pragnyatejas Balemarthy (ORCID: https://orcid.org/0009-0000-9361-4498)
Publication Details
- Journal
- Zenodo (CERN European Organization for Nuclear Research)
- Published
- 2026-11-12
- DOI
- https://doi.org/10.5281/zenodo.24166559
- Primary Topic
- Financial Markets and Investment Strategies
- Type
- article
- Field-Weighted Citation Impact
- 0.00