Return explanatory ability and predictability of non-linear market models

Hung, D.C.

Return explanatory ability and predictability of non-linear market models

Hung, D.C.

Authors

D.C. Hung

Abstract

Recent literature supports the pricing of higher-order systematic co-moments of returns. This paper provides some support for the quadratic-market model that is consistent with the three-moment CAPM in explaining time-series returns of the winner and the smallest size portfolios. This study further uses three innovative methodologies in analysing the ability of the linear CAPM, the quadratic- and the cubic-market models in predicting one-period ahead returns on individual stocks, equally- and value-weighted portfolios of momentum, size and country sorts. The results are surprising but important that the higher-moment CAPM market models do not outperform the linear CAPM in the return predictability tests.

Citation

Hung, D. (2007). Return explanatory ability and predictability of non-linear market models

Online Publication Date	Aug 1, 2007
Publication Date	Aug 1, 2007
Deposit Date	Mar 9, 2009
Publicly Available Date	Mar 9, 2009
Series Title	Durham University Business School Economics Finance Accounting Working Papers
Keywords	Asset pricing, Non-linearity, Return predictability.
Public URL	https://durham-repository.worktribe.com/output/1168161
Publisher URL	http://www.dur.ac.uk/dbs/faculty/working-papers/