The Hurst exponent tries to quantity momentum, random walk, and mean reversion. Not necessarily a backtest, but more of an analyses on how the Hurst exponent behaves. Based on preliminary tests, in order to match the S&P500 return using the hurst exponent one needs to leverage 3x in order to get the same return. However, the risk adjusted return over long periods seem to compensate. This strategy could likely be replicated buying UPRO or other triple leveraged stocks.
Tried to also look into Marco Lopez Deprados KCA, but not much ground work was reached in this area (reason while the KCA.py is included)