Do short‐term institutions exploit stock return anomalies?

• Published date: 01 February 2022
• Author: Yinfei Chen,Wei Huang,George J. Jiang
• Date: 01 February 2022
• DOI: http://doi.org/10.1111/fire.12284

DOI: 10.1111/fire.12284

ORIGINAL ARTICLE

Do short-term institutions exploit stock return

anomalies?

Yinfei Chen1Wei Huang2George J. Jiang3

1Department of Finance, School of Business,

CentralConnecticut State University, New

Britain, Connecticut, USA

2College of Saint Benedict and Saint John’s

University, St. Joseph, Minnesota, USA

3Department of Finance and Management

Science, Carson College of Business,

WashingtonState University, Pullman,

Washington,USA

Correspondence

GeorgeJ. Jiang, Department of Finance and

ManagementScience, Carson College of Busi-

ness,Washington State University, Pullman,

WA99164, USA.

Email:george.jiang@wsu.edu

Abstract

The literature documentsthat institutional investors in

aggregate trade in the “wrong” direction of stock return

anomalies. This paper explains the puzzle by isolating active

institutions with a shorter-term investment horizon (“short-

term institutions”). Based on 10 well-known stock return

anomalies, we find that these short-term institutions indeed

exploit anomalies. The results established in the litera-

ture are mostly driven by institutions with a longer-term

investment horizon, who are likely more interested in long-

run financial performance. In addition, given that institu-

tional investors are momentum traders, we examine sep-

arately how institutional investors trade on momentum

and contrarian anomalies. We show that institutions gen-

erally trade in the right direction on momentum anoma-

lies but wrong direction on contrarian anomalies. How-

ever,over relatively short trading windows, short-term insti-

tutions trade in the right direction of both momentum

and contrarian anomalies. Furthermore, we show evidence

that when short-term institutions exploit anomalies, their

trades generate significantly positive abnormal returns. Our

results also imply that trading by short-term institutions

appears to be subject to liquidity provision by long-term

institutions.

KEYWORDS

institutional demand, long-term institutions, short-term institu-

tions, stock return anomaly

Financial Review. 2022;57:69–94. wileyonlinelibrary.com/journal/fire ©2021 The Eastern Finance Association 69

70 CHEN ET AL.

JEL CLASSIFICATION

G12, G14, G23, G32

1INTRODUCTION

Institutional investors are generally perceived to be informed and skilled. However, the literature documents that

institutional investors in aggregate do not seem to take advantage of well-documented stock return anomalies. For

instance, Ali et al. (2000) find that institutional investors fail to profit from the negative relation between the accrual

component of earnings and future stock returns. Lewellen (2010) presents evidence that institutional investors in

aggregate hold market-likeportfolios and do not seem to bet on stock return predictive characteristics such as book-

to-market ratio, asset growth, and profitability. In a more recent study, Edelen et al. (2016) show that institutional

investorsactually trade in the opposite or “wrong” direction of cross-sectional stock return anomalies associated with

valuation, profitability, corporate investment, financing, financial distress, and momentum. They show that institu-

tions increase holdings on stocks in the short leg of the anomaly or stocks with ex post negative abnormal returns

and decrease holdings on stocks in the long leg of the anomaly or stocks with ex post positive abnormal returns. At

the same time, the literature documents evidence that anomalies disappear once they are documented byacademics

(McLean& Pontiff, 2016). Moreover, institutional investorsmake up the vast majority of financial markets (Gompers &

Metrick, 2001). If institutions make up most of financial marketsbut do not exploit anomalies, how can the anomalies

disappear?

In this paper, we explain the puzzle of institutional investors tradingagainst known anomalies. When we isolate

active institutions with a shorter-term investment horizon (“short-term institutions”), we find that these institutions

indeed exploit known cross-sectional stock return anomalies overshort horizons. We show that the main results doc-

umented in the literature are driven by institutions with long-term horizons that presumably do not attempt to profit

off short-term mispricing but instead are more interested in long-term growth and profitability.

Our study is motivated by the following arguments. Short-term institutions are characterized as having high port-

folio turnover with short investment horizons (Bushee, 2001; Yan& Zhang, 2009). They tend to trade more actively

and focus more on gains from active trading. On the other hand, long-term or non-transient institutions provide sta-

ble ownership to firms and trade passively because they are geared toward longer-term dividend income or capital

appreciation. The literature documents evidence that short-term activeinstitutions predict breaks in the sequence of

earnings increases (Ke & Petroni, 2004) and exploitthe post-earnings announcement (PEAD) drift (Chen et al., 2017).

Yan an d Zhan g (2009) and Ke and Ramalingegowda (2005) find that trading byshort-term institutions predicts future

stock returns and earnings surprises, whereas trading by long-term institutions is not associated with future stock

returns or earnings. Motivated by these studies, in this study we explore the extent to which short-term institutions

exploit stock return anomalies.

We include 10 well-known return-predictive variables in our study: total accruals, asset growth, investment-to-

assets ratio, net operationassets, net share issues, book-to-market ratio, O-score, gross profitability, return on assets,

and stock return momentum. As documented in the literature, these variables havestrong predictive powers of future

stock returns.1Consistent with the literature, we construct hedge portfolios based on the return-predictivevariables.

Specifically,the long (short) portfolio consists of the top or bottom three deciles, whichever have higher (lower) future

1Cooper et al. (2008) document a negative relation between asset growth and future stock returns. A number of studies show that value stocks with low

book-to-marketvalue outperform growth stocks with high book-to-market value (Ali et al., 2003). Lyandres et al. (2008), Fama and French (1992), La Porta

et al. (1997), Lakonishok et al. (1994), Loughran (1997) document that investment-to-assets ratiois negatively associated with future stock returns, and

Xing(2008) finds that firms with high investment growth tend to underperform subsequently. Novy-Marx (2013) and Fama and French (2015) document that

grossprofitability has a strong power in predicting cross-sectional stock returns. Specifically, firms with high gross profit-to-assets ratio generate significantly

highersubsequent returns than firms with low gross profit-to-assets ratio. The literature shows that firms with low net share issues significantly outperform

thosewith high net share issues over both long and short horizons (Daniel & Titman, 2006; Fama & French, 2008; Ikenberry et al., 1995; Jiang & Zhang, 2013).