The Reach of the Disposition Effect: Large Sample Evidence Across Investor Classes

Authors


  • *We gratefully acknowledge support for the research from the Australian Stock Exchange. The provision of CHESS data by ASX required that we maintain the confidentiality of these data and required that the data not be made available to other researchers without ASX permission. We have complied with both conditions. This paper is an outgrowth of Nick Chappel's undergraduate honours thesis at The University of Sydney. We acknowledge the helpful comments of commentators and colleagues, especially Phil Dolan, Ken Peasnell, Peter Pope and Peter Swan. We thank Joe Tang for programming assistance. We claim all errors as our own.

Philip Brown
UWA Business School, M250
The University of Western Australia
35 Stirling Highway
Crawley, WA 6009
Australia
philip.brown@uwa.edu.au and philip.brown@unsw.edu.au

ABSTRACT

We examine detailed daily Australian Stock Exchange share registry data for investors in IPO and index stocks between 1995 and 2000 and find that the ‘disposition effect,’ investors' reluctance to crystallize losses and relative eagerness to realize gains, is pervasive across investor classes. However, traders instigating larger investments tend to be affected less by the disposition bias. Our novel findings include that (a) the disposition effect ameliorates over time, being undetectable from around 200 trading days after purchase, (b) the ‘house money’ effect tempers the disposition effect, (c) shareholder loyalty schemes also partially offset investors' relative preference for selling winning stocks, and (d) the reversal of the disposition effect in June (the last month of the Australian tax year) does not occur among investors unable to take advantage of tax shields. In line with earlier research, our results support a tax-related explanation for the June effect rather than window dressing or momentum explanations. Finally, we confirm Odean's finding that the disposition effect is not driven by diversification motives, or by higher transaction costs associated with lower-priced stocks.

Ancillary