Influence Functions for Risk and Performance Estimators

32 Pages Posted: 9 Jul 2019

See all articles by Shengyu Zhang

Shengyu Zhang

HomeStreet Bank

Doug Martin

University of Washington

Anthony-Alexander Christidis

University of British Columbia

Date Written: July 6, 2019

Abstract

We first provide a brief introduction to influence functions and their use for computing standard errors of risk and performance estimators. We then derive the influence functions for a number of risk estimators and a number of performance estimators, and use the formulas to graphically display the shapes of the various influence functions. The latter provide a useful basis for comparing different risk and performance estimators with regard returns data influence on the estimators, including the influence of outliers and the appropriateness of the estimators with regard to how they reflect gains, losses, and performance. We briefly discuss an R package RPEIF, available on CRAN, forcomputing and plotting influence functions of risk and performance estimators.

Keywords: Risk estimator, performance estimator, influence function

JEL Classification: C13, C14, G10

Suggested Citation

Zhang, Shengyu and Martin, R. Douglas and Christidis, Anthony-Alexander, Influence Functions for Risk and Performance Estimators (July 6, 2019). Available at SSRN: https://ssrn.com/abstract=3415903 or http://dx.doi.org/10.2139/ssrn.3415903

Shengyu Zhang

HomeStreet Bank ( email )

601 Union Street
Ste. 2000
Seattle, WA 98101
United States

R. Douglas Martin (Contact Author)

University of Washington ( email )

Applied Mathematics & Statistics
Dept. of Statistics
Seattle, WA 98195
United States

Anthony-Alexander Christidis

University of British Columbia ( email )

2329 West Mall
Vancouver, British Columbia BC V6T 1Z4
Canada

Do you have a job opening that you would like to promote on SSRN?

Paper statistics

Downloads
34
Abstract Views
370
PlumX Metrics