Assessing the usefulness of daily and monthly asset-pricing factors for Australian equities

Philip Gray, Angel Zhong

Research output: Contribution to journalArticleResearchpeer-review

Abstract

This paper considers which factors are worthy of inclusion in asset-pricing models for Australian equity returns at daily and monthly frequencies. We utilise the recently developed methodology of Harvey and Liu (2020) to employ a performance metric specifically designed for model comparison and to conduct statistical inference that accommodates the possibility of data mining. The clear and consistent finding is that the market risk premium is the most important factor. This is the case for both daily and monthly asset-pricing specifications, and irrespective of whether individual stocks or portfolios are used as test assets. There is no compelling evidence that any other factors are useful, implying that a parsimonious model suffices.

Original languageEnglish
Number of pages31
JournalAccounting & Finance
DOIs
Publication statusAccepted/In press - 2021

Keywords

  • Asset pricing
  • Data mining
  • Data snooping
  • Factors
  • Market risk premium
  • Pricing error

Cite this