The consequences of non‐trading institutional investors

• Published date: 01 September 2023
• Author: Mohammad (Vahid) Irani,Hugh Hoikwang Kim
• Date: 01 September 2023
• DOI: http://doi.org/10.1111/fima.12418

DOI: 10.1111/fima.12418

ORIGINAL ARTICLE

The consequences of non-trading institutional

investors

Mohammad (Vahid)IraniHugh HoikwangKim

Darla Moore School of Business, University of

South Carolina, Columbia, South Carolina, USA

Correspondence

Mohammad (Vahid)Irani, Darla Moore School

of Business, Universityof South Carolina, 1014

Greene St., Columbia, SC 29208, USA.

Email: Vahid.Irani@moore.sc.edu

Abstract

We document that institutional investors do not trade a

single share, on average, in one of five stocks in their port-

folio for an extended period. Investors with high inaction

are likely to underperform in the future. Our results show a

similar underperformance for stocks with a high non-trading

levelof institutional investors. We investigate several behav-

ioral biases as potential drivers of the non-trades and find

no evidence of distraction, overconfidence, and disposition

effects. Institutional investors’ tendency to sell stocks with

salient price movements and recency bias best explainstheir

inactions. Overall, the non-trading behavior of institutional

investors serves as a unique predictor for their future per-

formance and potential behavioral biases are driving this

predictability.

KEYWORDS

behavioral bias, institutional investors, non-trades, portfolio

management

1INTRODUCTION

Inaction is a widely observed behavior of economic agents. Contrary to the usual assumptions of conventional asset

pricing theories, the empirical literature finds that retail investors often do not change their portfolio positions for

extended periods, which is often called portfolio inertia (e.g., Agnew et al., 2003; Madrian & Shea, 2001). Portfolio

inertia can have a first-order impact on risk premia due to incomplete risk sharing among investors (Chien et al.,

This is an open access article under the terms of the Creative Commons Attribution-NonCommercial License, which permits use,

distribution and reproduction in anymedium, provided the original work is properly cited and is not used for commercial purposes.

© 2023 The Authors. Financial Management published by Wiley Periodicals LLC on behalf of Financial Management Association

International.

Financial Management. 2023;52:433–481. wileyonlinelibrary.com/journal/fima 433

434 IRANI ANDKIM

2012; Gust & López-Salido, 2009). As the influence of institutional investors increases in the stock market, there

is a renewed interest in understanding how institutional investors make their portfolio decisions, which directly

influence asset prices (Basak & Pavlova, 2013; Choi et al., 2017; Dasgupta et al., 2011a; Gompers & Metrick, 2001;

He & Krishnamurthy,2013;Koijen&Yogo,2019; Nofsinger & Sias, 1999; Vayanos& Woolley, 2013). So far, however,

there has been little research investigating to what extent institutional investors engage in the non-trading of their

investment positions. It is also unknown whether such inactions are the result of a strategic choice to minimize the

cost of information processing (Kacperczyk et al., 2016;Sims,2003,2010;Steiner et al., 2017) or is a manifestation of

investors’ behavioralbiases (Gabaix, 2019). More importantly, investigating the non-trading of institutional investors

and the potential drivers for such behaviors would shed additional light on understanding the impact of the increasing

influence of institutional investors on asset prices (Brunnermeier et al., 2021).

In this paper, we first document the degree of non-tradingamong institutional investors and assess whether such

inaction is related to their future performance. Instead of focusing on a specific group of institutional investors (e.g.,

equity mutual funds), we examine the behavior of all types of institutional investors, encompassing banks, insurers,

investment companies, investment advisors, corporate pensions funds, public pensions, and university endowment

funds. The extensive sample also helps us investigate the overallimpact of institutional investors’ portfolio inaction

on stock returns. Importantly, we test several possible explanations driving the portfolio inaction of institutional

investors as well.

Toexplore the extent of institutional investors’ inaction, we examine stock-holding information in their 13F filings,

studying 40 million investor-stock quarterly observations. We consider an institutional investor is inactivein trading

stock if the number of shares held by an institutional investor in the current calendar quarter has not changed from

the number held in the previous calendar quarter.

To preview the analysis result, we find that institutional investorsdo not trade a single share for one out of five

stocks in their portfolio for at least a quarter of the year, on average.There is also great heterogeneity across insti-

tutional investors in their non-trading behavior.Insurance companies, corporate pension funds, public pension funds,

and university endowments are the most inactive institutions, while independent investment advisors are the most

active ones. Institutional investors with small portfolios are less likelyto trade any shares of a stock in their portfolio.

This behavior is more likely to happen when a stock’s portfolio weight is small relative to the overall portfolio value

and the investor has a concentrated portfolio. Regarding stock-level characteristics, we find that non-tradedstocks

are likely to be small and illiquid, suggesting that the transaction cost at least partially contributes to the non-trading

behavior. Size and illiquidity,however, do not fully explain institutional investors’ inactions. Non-Traded stocks also

have lower volatility, lower profitability, and lower institutional ownership. These stocks have low book-to-market

ratios, suggesting that institutional investors do not buy-and-hold value stocks for extended periods to benefit from

the value premium. In addition, we find that security lending is not the main reason for institutional investors’inaction.

In our additional analysis, we evaluate whether institutional investors choose inaction to catch profitable trading

opportunities in the future. We run regressions of a non-tradedstock’s future return on the length of inaction (i.e., the

time elapsed since the initial purchase of a stock until it is traded),controlling for various stock-level and investor-level

characteristics. Our analysis shows that the length of inaction is negatively related to the profits of subsequent returns

of the stock, implying that institutional investors do not choose inaction to time profitable future trades.

Although the average degree of inaction may not sound sizable, we find that the inaction provides useful infor-

mation about institutional investors’ future performance. Our analysis shows that inaction is negatively associated

with the overall future performance of institutional investors. Controlling for various characteristics of institutional

investors, we estimate a predictive regression model of 3-month-ahead risk-adjusted returns on the inaction level of

institutional investors. We find that the coefficient estimate on the inaction level is negative and statistically signifi-

cant. The result still holds up to 12-month-ahead risk-adjusted returns, suggesting that the predictability of non-trade

for future performance is long lasting.

The negative return implication of inaction is not isolated to an individual institutional investor. Weevaluate the

aggregate performance implication of inaction and active trading strategiesforallinstitutional investors. Every quar-

ter,we categorize each institutional investor’s stock trades into non-trade and active tradinggroups. We then compute

IRANI ANDKIM 435

1-month-ahead value-weighted returns on each trading strategy for each institutional investor. Averaging such

returns across all institutional investors, we form a time series of portfolio returns representing non-trade and active

trading strategies. We measure alphas from time-series regressions of the portfolio returns representing non-trades

and active trades based on various asset pricing models. The results show that the alpha for the non-trade is negative

andstatistically significant across all asset pricing models employed. This is not the case for the active trading portfolio.

These results suggest that inaction might undermine institutional investors’ performance at the aggregate level.

To evaluatewhether the inaction of institutional investors is priced into stock returns, we investigate the return

implication of non-tradesat the stock level. We calculate the inaction ratio for each stock as the fraction of non-traded

shares out of the total number of shares held by institutional investors. Every quarter, we sort stocks into quintile

portfolios based on the inaction ratio. The portfolio-sorting analysis shows that the future risk-adjusted returns are

lower for stocks with higher inaction ratios than for those with lower ratios. We also runFama–MacBeth regressions

of the future excessreturns of stocks on the inaction ratio and other well-documented firm characteristics associated

with stock returns, such as size, book-to-market, momentum, volatility,leverage, and profitability. The analysis shows

that there is a negative and significant correlation between the inaction ratios of stocks and future stock returns. In

sum, we find that non-traded stocks are likely to underperform in the future, undermining the overallperformance of

institutional investors.

Additionally, we examine whether the correlation between stock-level inaction and returns can be explainedby

mispricing. This is tested by including a stock’s exposureto the average mispricing score calculated by Stambaugh and

Yuan (2017) in the Fama–MacBeth specification. If the association between stock-level inaction and returns is pri-

marily driven by mispricing, the significance of the result will decrease when the mispricing score is included. Our

results show that the coefficient’s statistical significance decreases and becomes insignificant, indicating that the

underperformance of inaction stocks is mainly due to mispricing.

We further investigate several behavioral biases as potential drivers for the portfolio inaction of institutional

investors.We explore five plausible behavioral channels: distraction, overconfidence, disposition effect, salience trad-

ing,and recency bias. Additional analyses allow us to rule out the distraction, overconfidence, and disposition effects as

main drivers of the non-trade of institutional investors.The analysis result shows that the probability of selling is low-

est at around zero profits, and it subsequently increases as the size of losses or gains increases, a V-shaped pattern on

a horizon from loss to gain regions. This result is consistent with the salience tradinghypothesis, stating that investors

do not substantially change their positions when they do not observe any material price movements of their stocks

(Ben-David & Hirshleifer,2012; Hartzmark, 2015). Additionally, we observe that the stocks closed out after negative

price movements would havehad positive returns, consistent with the recency bias (Chakrabarty et al., 2017).

We also test our main hypothesis using individual mutual fund sample. There are two empirical challenges of using

the 13F institutional investors’ holdings data to investigate the impact of inactions. First, the 13F holdings data are

largely reported quarterly, and the low frequency of the data and the possibility of window dressing behavior may

impact the identification of inaction. Second, the patterns of inaction can be reflected in the fund flows of outside

investors, which is not easily observable for some institutional investors such as banks and insurers. By using the

mutual fund sample, we verify that our main result remains robust even when controlling for unobserved trading

behaviors between reporting periods and fund flows.

In severalrobustness checks, we extend the length of the non-trading period to define institutional investors’ inac-

tion up to 6 months. Our analysis, based on these stringent criteria, still finds that the inaction negatively predicts

the future performance of institutional investors. The analysis result is not adversely biased by potential intraquar-

ter round-trip trading (i.e., buying and selling the same number of shares within a quarter). Puckett and Yan (2011)

report that intraquarter round-trip trading by institutional investors generates higher returns. Because some non-

trades could havebeen short-term round-trip trading with positive returns, our result of negative returns for inactive

investors would be upward biased; the actual returns for inaction would be evenmore negative than we report in our

analysis. Moreover,in a series of untabulated analyses, we find our results are not driven by a specific sample period