This package implements a moment-free estimator of the Sharpe (signal-to-noise) ratio. The algorithm does not require the computation of any moments by estimating the Sharpe ratio based on the cumulative sum of (i.i.d.) increments or returns. This algorithm is much more precise (efficient) when increments are heavy-tailed.
The estimator computes the difference between the drawdown and drawup durations. First, the cumulative sum of x (e.g. prices) is computed, then the total drawdown and drawup durations are computed. Precision is improved by averaging the estimator on several random permutations of x.
The data used to fit the spline is saved as spline_data.csv in the data folder.
The current implementation only supports Python 3.
pip install --no-cache-dir pysharperratio
Maintainers: Amir Sani & Damien Challet
Challet, Damien. "Sharper asset ranking from total drawdown durations." arXiv preprint arXiv:1505.01333 (2015).
Preprint available at http://arxiv.org/abs/1505.01333
@article{challet2015sharper, title={Sharper asset ranking from total drawdown durations}, author={Challet, Damien}, journal={arXiv preprint arXiv:1505.01333}, year={2015} }