Noise Fit, Estimation Error and a Sharpe Information Criterion

48 Pages Posted: 21 Feb 2016

See all articles by Dirk Paulsen

Dirk Paulsen

John Street Capital

Jakob Söhl

Delft University of Technology - Delft Institute of Applied Mathematics (DIAM)

Date Written: February 19, 2016

Abstract

When optimizing the Sharpe ratio over a k-dimensional parameter space the thus obtained in-sample Sharpe ratio tends to be higher than what will be captured out-of-sample. For two reasons: the estimated parameter will be skewed towards the noise in the in-sample data (noise fitting) and, second, the estimated parameter will deviate from the optimal parameter (estimation error). This article derives a simple correction for both. Selecting a model with the highest corrected Sharpe selects the model with the highest expected out-of-sample Sharpe in the same way as selection by Akaike Information Criterion does for the log-likelihood as measure of fit.

Keywords: Model Selection, Sharpe Ratio, Akaike Information Criterion, AIC, Back-testing, Noisefit, Overfit, Estimation Error, Sharpe Ratio Information Criterion, SRIC

JEL Classification: C13, G11

Suggested Citation

Paulsen, Dirk and Söhl, Jakob, Noise Fit, Estimation Error and a Sharpe Information Criterion (February 19, 2016). Available at SSRN: https://ssrn.com/abstract=2735087 or http://dx.doi.org/10.2139/ssrn.2735087

Dirk Paulsen (Contact Author)

John Street Capital ( email )

London
United Kingdom

Jakob Söhl

Delft University of Technology - Delft Institute of Applied Mathematics (DIAM) ( email )

Mekelweg 4
Delft, Holland 2628
Netherlands

Register to save articles to
your library

Register

Paper statistics

Downloads
171
Abstract Views
1,316
rank
174,827
PlumX Metrics