Hello there:
I have taken the course of momentum strategies in Phyton, and I noticed something, in that course, there is a document that actually adresses how to calculate the Hurst exponent using the last 8 returns (9 periods) of a sample stock Data (Quantra by QuantInsti | Courses on Algorithmic and Quantitative Trading),
but some time ago… I posted a question on how to calculate the Hurst exponent using Ibridgepy , and I was directed to install a Husrt library in python for that … now this library uses more periods (I thik 100 or 1000), to calculate this very exponent, is there a method of this library to calculate the Husrt exponent using only 8 periods ??
Another doubt that I have is if this exponent is as useful when calculated with a very volatile stock ?? for example, some months ago ( around 28th Feb 2021), tickers GME KOSS and had more than +100% percent returns… could the Hurst enponent have helped to predict this price movements???
Hi Ghery,
The "compute_Hc" function from the "hurst" library will give an error if the length of the series is less than 100. If you want to calculate the Hurst exponent for a shorter period like 10 you can refer to this article.
The purpose of the Hurst exponent is to tell if a time series is mean reverting or trending. So it could defiantly help you predict the price movement. But you should always other features according to your strategy and backtest it thoroughly.