Slow Trading And Stock Return Predictability

Matthijs Lof

Aalto University

Matti Suominen

Aalto University School of Business

October 15, 2015

Abstract:

Market returns predict the future abnormal returns on small and illiquid stocks, implying attractive dynamic investment strategies for investors investing in the size premium or in small and illiquid stocks either directly or through exchange traded funds. We provide evidence that this return predictability is due to institutional investors’ trading patterns: When rebalancing their portfolios the institutional investors initially buy (sell) relatively more the large and liquid stocks. In the case of illiquid stocks they split their orders over several days to avoid excessive price impact, thus inducing predictability in stocks returns. We provide evidence that some hedge funds exploit this return predictability.

Slow Trading And Stock Return Predictability – Introduction

Returns on large and liquid stocks predict the returns on small and illiquid stocks (See e.g. Lo and MacKinlay (1990) and Chordia and Swaminathan (2000)). We demonstrate the economic significance of these ‘return spillovers’ by suggesting various attractive trading strategies exploiting the predictability. We provide evidence that hedge funds are positively exposed to these strategies. Finally, we find that the return spillovers originate partly from the trading patterns of large institutional investors: While these institutions execute trades in large liquid stocks swiftly, consistent with Vayanos (2001) they split their trades in small illiquid stocks over several days to avoid excessive price impact.

Returns on the static size premium, i.e. a portfolio that is long in small stocks and short in large stocks, have been rather weak, in particular in recent decades1. In this paper we show that the size premium is however highly profitable when timed correctly. Exploiting the return spillovers from large to small stocks, we propose strategies that are long in the size premium following periods with positive market returns and short in the size premium following periods with negative market returns. With daily, weekly, and even monthly portfolio adjustment these strategies yield positive highly attractive abnormal risk-adjusted returns. Even low-cost versions of such strategies that rely only on long positions in smallcap exchange traded funds (ETFs) yield positive abnormal returns.

While earlier research has argued that the predictability of small-stock returns can be due to informed traders trading in the more liquid stocks and the information therefore being first revealed in the liquid stocks, see e.g. Chordia, Sarkar, and Subrahmanyam (2011), we provide evidence of a supplementary force behind the return predictability. Our results suggest that the return predictability is partly due to institutional investors’ trading patterns: When rebalancing their portfolios the large investors initially buy (sell) relatively more the large stocks, where they can quickly increase or reduce their overall risk exposure. In the case of the small stocks, in turn, they split their orders over several days. When the institutions have large buying or selling needs both in the liquid and illiquid stocks this leads to predictability in stocks returns. This result provides evidence supporting for the model by Vayanos (2001), who finds that investors optimally slow down trades in small illiquid stocks to avoid price impact. Consistent with this theory, we find that the returns spillovers are larger during periods when market liquidity is low.

This paper is organized as follows. Section 2 presents the data. In Section 3 we document the lead-lag relation between large capitalization stock returns and small market capitalization stock returns. In Section 4 we show also that this predictability implies highly attractive trading strategies, some of which can be executed with low cost ETFs. We provide here also evidence that this lead lag relation is stronger in illiquid markets. In Section 5 we investigate the lead-lag relations in trading volume and provide evidence that the observed return predictability can be related to institutional investors’ trading patterns. In section 6 we demonstrate that some hedge funds exploit the return predictability by increasing their exposure to the small stocks after good returns in the stock market. Section 7 concludes the paper.

return predictability

return predictability

See full PDF below.