The Value of Alpha Forecasts in Portfolio Construction

31 Pages Posted: 3 Aug 2007 Last revised: 19 May 2008

See all articles by Kingsley Y. L. Fong

Kingsley Y. L. Fong

University of New South Wales - School of Banking and Finance

David R. Gallagher

Bond University

Adrian D. Lee

Deakin University - Department of Finance (Property and Real Estate)

Date Written: May 19, 2008

Abstract

This study examines a portfolio strategy which selects stocks using the undisclosed monthly holdings of Australian active fund managers. When considering a large range of strategies incorporating fund portfolio holdings information, the top performing strategies are robust to data-snooping and are both economically and statistically significant. These strategies are short term in nature, with significant performance lasting up to two months. However, when we account for look-ahead bias in the formation of a strategy, we document the absence of statistically significant performance. When we consider a strategy following the best performing strategy holding 20 stocks or more in the previous month, we find statistically significant alpha of at least 6.88 percent per year.

Keywords: Portfolio construction, Fund mimicking strategies

JEL Classification: G11, G23

Suggested Citation

Fong, Kingsley Y. L. and Gallagher, David R. and Lee, Adrian D., The Value of Alpha Forecasts in Portfolio Construction (May 19, 2008). Available at SSRN: https://ssrn.com/abstract=1004434 or http://dx.doi.org/10.2139/ssrn.1004434

Kingsley Y. L. Fong (Contact Author)

University of New South Wales - School of Banking and Finance ( email )

UNSW Business School
High St
Sydney, NSW 2052
Australia

David R. Gallagher

Bond University ( email )

Centre for Data Analytics, Bond Business School
Gold Coast, QLD 4229
Australia

Adrian D. Lee

Deakin University - Department of Finance (Property and Real Estate) ( email )

70 Elgar Road
Melbourne, VIC 3125
Australia

Do you have negative results from your research you’d like to share?

Paper statistics

Downloads
507
Abstract Views
2,398
Rank
103,077
PlumX Metrics