Testing for structural breaks in return-based style regression models

Yunmi Kim, Douglas Stone, Tae Hwan Kim

Research output: Contribution to journalArticlepeer-review

Abstract

It is important for investors to know not only the style of a fund manager in which they are interested, but also whether this style is constant or changing through time. The style of a fund manager can be estimated by the so-called style regression, and a great deal of research has been carried out to investigate the statistical properties of style regression methods. However, there has been no formal and statistically valid method to test for a change in manager style when the two typically imposed restrictions (sum-to-one and non-negativity) are jointly present in style analysis. In this study, we apply and extend the results of Andrews (Econometrica 61:821–856, 1993; Estimation when a parameter is on a boundary: theory and application, Yale University, 1997a; A simple counterexample to the bootstrap, Yale University, 1997b; Econometrica 67:1341–1383, 1999; Econometrica 68:399–405, 2000) to develop a valid testing procedure for the possibility wherein the location of any possible change does not need to be specified and the case of multiple shifts is accommodated. When our proposed test is applied to the Fidelity Magellan Fund, it is revealed that the fund’s style changed at least twice between 1988 and 2017.

Original languageEnglish
Pages (from-to)61-76
Number of pages16
JournalFinancial Markets and Portfolio Management
Volume35
Issue number1
DOIs
Publication statusPublished - 2021 Mar

Bibliographical note

Funding Information:
We thank Markus Schmid (the editor) and two anonymous referees. Yunmi Kim acknowledges that this work was supported by the 2020 Research Fund of the University of Seoul.

Publisher Copyright:
© 2020, Swiss Society for Financial Market Research.

All Science Journal Classification (ASJC) codes

  • Accounting
  • Finance

Fingerprint Dive into the research topics of 'Testing for structural breaks in return-based style regression models'. Together they form a unique fingerprint.

Cite this