Testing Bitcoin Correlation as a Crypto Relative-Strength Signal
Summary
This document explains how Pearson correlation can describe the relationship between cryptocurrency prices and Bitcoin, then explores whether that relationship contains a relative-performance signal. Using four-hour Binance data for 144 currencies listed at the start of 2023, the analysis normalizes price series, compares each currency’s correlation with Bitcoin, and contrasts the average paths of the most and least correlated groups. It reports that, in the period examined, the more Bitcoin-correlated group rose more, while the less correlated group served as a hedge when shorted. A split-period calculation is also presented to address the initial use of future data in estimating correlation.
The discussion links co-movement to market leadership and investor behavior, while suggesting rolling correlations and separate calculations for rising and falling markets as follow-up analyses. The evidence is historical and tied to one market period and a selected universe; it does not establish that the relationship persists or is tradable after costs. The initial grouping method also depends on how the sample and correlation window are chosen, so it should be validated out of sample before use.
Key ideas
- Pearson correlation ranges from negative to positive linear association and can be calculated across crypto price series.
- The study uses four-hour price data for a selected set of currencies during 2023.
- In the examined period, the group more correlated with Bitcoin outperformed the less correlated group.
- The document acknowledges look-ahead bias and describes a split-period check to reduce it.
- Rolling and regime-specific correlations are suggested, while persistence and trading costs remain untested.
Tags
This summary was written by Stratmill's research agent from the original; it is not a copy of the source.