Independent Sources of Information About UK Expected Returns

Abstract

We investigate the sources of independent information about UK expected stock returns, taking into consideration both firm-level characteristics and firm-specific investor sentiment. Cochrane (2011) argues that US research has produced a ‘veritable zoo’ of characteristics that help explain the cross-section of returns and calls for research on which of the characteristics in this ‘zoo’ are independent sources of information about expected returns. We respond to this call but using UK data. We largely follow the approach of Green, Hand, and Zhang (2017) in the US. We find that only 7 (out of 94) characteristics are common to the two countries in explaining expected stock returns. Unlike the US, there is little evidence that small capitalisation stocks affect the analysis in the UK. In the presence of a comprehensive list of firm characteristics, firm-specific investor sentiment is no longer associated with cross-sectional stock returns. When only focusing on the identified UK significant characteristics to predict expected stock returns, their performance is statistically indistinguishable from using all the potential characteristics; whereas in the US, using all the studied characteristics provides statistically superior predictive ability.   In this context, we discuss the current concern with the replicability of results in the finance literature, whether in the same jurisdiction or across jurisdictions, in justifying our study. Our findings have important implications to both researchers and practitioners who seek an adequate model of returns designed specifically for the UK.

Keywords: firm characteristics, investor sentiment, stock returns, returns prediction, replication