I have a set of Sharpe Ratios collected from various models and I would like to check whether these are statistically significantly different from each other. Have seen that many research papers apply t-test to compare two or more Sharpe Ratios against a benchmark but I cannot find a plausible way or an explanation how this is carried out in practice.
Any links/references to python/R code is very much appreciated. Thank you.
Have a look at :https://rdrr.io/rforge/PeerPerformance/man/sharpeTesting.html
https://cran.r-project.org/web/packages/SharpeR/SharpeR.pdf
https://arxiv.org/pdf/1905.08042.pdf
Hi Patrick Bezzina
Have a look at this: https://towardsdatascience.com/calculating-sharpe-ratio-with-python-755dcb346805
Regards
Patrick Bezzina !
First, I suggest that you check if your data are parametric or nonparametric. T-test is used only when the data are parametric. Use the shapiro-wilks test to verify this situation. I prefer use R, for this analyses, you may the aproach found here: http://www.sthda.com/english/wiki/paired-samples-t-test-in-r
- Badges
- Science topic
- Similar topics
- Computer Science and Engineering
- Programming Languages
- Python