Investor sentiment, momentum, and stock returns: an examination for direct and indirect effects

Abstract

This paper aims to analyse the impact of behavioural biases on asset pricing by hypothesising that sentiment and momentum are relevant risk factors in Pakistan equity market. The paper also examines the influence of sentiment and momentum factors on market risk, size, and value premiums by estimating the interacted asset-pricing model. To carry out the empirical analysis, monthly stock returns of firms listed on Pakistan Stock Exchange are used for the period 2000–2013. The empirical results indicate that both investor sentiment and momentum factors have a significant impact on the required rate of returns. Specifically, it is found that the premium for both factors is positive and statistically significant. Further, the estimated results provide evidence that the inclusion of these two factors in the Fama-French three-factor model considerably increases the prediction power of the model. The results also reveal that the inclusion of sentiment factor in the Carhart four-factor model significantly increases the prediction power of the model. Yet, the estimation results indicate that the prediction power of the model further increases when the interaction terms are added to the model in order to examine the indirect effects of sentiment and momentum. The results of the interacted model provide evidence of a significant impact of investor sentiment and momentum factors on market risk, size, and value premiums. Although investor sentiment negatively affects all the three premiums, the effect of momentum is positive for both market risk and size premium, whereas, it is negative for value premium. The findings are helpful in explaining and understanding the effects of behavioural biases on stock returns in Pakistan. The findings of the indirect effects suggest that investor sentiment and momentum factors significantly increase the chance of mispricing in Pakistan equity market.

Cite this work

No public copy is stored for this record, so only the metadata is cited.

Stored only when both papers are in the corpus — 24 citations exist upstream.

Papers that cite the same references — the relational fallback while embeddings are still backfilling.