Measuring Mutual Fund Flow Pressure as Shock to Stock Returns

• Author: MALCOLM WARDLAW
• DOI: http://doi.org/10.1111/jofi.12962
• Date: 01 December 2020
• Published date: 01 December 2020

THE JOURNAL OF FINANCE •VOL. LXXV, NO. 6 •DECEMBER 2020

Measuring Mutual Fund Flow Pressure as Shock

to Stock Returns

MALCOLM WARDLAW∗

ABSTRACT

A large and rapidly growing literature examines the impact of misvaluation on firm

policies by using mutual fund outflow-induced price pressure to isolate nonfundamen-

tal price variation. I demonstrate that the standard approach to computing outflow-

induced price pressure produces a measure that is inadvertently a direct function of

a stock’s actual realized return during the outflow quarter, raising doubts about its

orthogonality to fundamentals. After removing these direct measurements of return,

outflows generate a fairly negligible quarterly decline in returns, with no subsequent

reversal, and many established results in this literature no longer hold. I provide

suggestions for future analysis.

FINANCIAL ECONOMICS HAS LONG BEEN interested in whether nonfundamen-

tal movement in stock prices impacts corporate decision making. Empirically

identifying such impact is challenging, however, as it requires an independent

shock to stock prices that is both fully observable to the econometrician and

completely orthogonal to firm fundamentals. Over the last 10 years, a rapidly

expanding literature has used the investor flows to and from mutual funds as a

source of exogenous price pressure. The idea behind this approach is that large

investor redemptions may place pressure on mutual funds to sell the stocks

they hold. If the required sales are sufficiently large, the funds’ liquidity needs

may put downward pressure on prices that is unrelated to the fundamental

value of the underlying stocks. If cleanly identified, this price pressure then

creates a laboratory for studying market feedback effects by breaking the en-

dogenous relationship between prices and fundamental firm value.

∗Malcolm Wardlaw is with University of Georgia, Terry School of Business. Special thanks to

Elizabeth Berger, Nicholas Crain, Jonathan Cohn, Alex Edmans,Itay Goldstein, Wei Jiang, Uday

Rajan, Breno Schmidt, Sheridan Titman, Wei Wei, Toni Whited, Steven Xiao, and Alex Young for

comments and suggestions. Thanks also to seminar participants at the University of Michigan,

University of Illinois at Chicago, University of Warwick, University of South Carolina, University

of Georgia, Chinese University of Hong Kong, University of New South Wales, the 2019 Midwest

Finance Association, and the 2019 Western Finance Association annual meetings. The author has

no conflicts of interest as identified by The Journal of Finance’s disclosure policy.

Correspondence: Malcolm Wardlaw, Terry School of Business,University of Georgia, 620 South

Lumpkin Street, B332 Amos Hall, Athens, GA 30602; telephone: (706) 542-0356;

e-mail: malcolm.wardlaw@uga.edu.

DOI: 10.1111/jofi.12962

© 2020 the American Finance Association

3221

3222 The Journal of Finance®

Interest in this approach began with Coval and Stafford (2007), who provide

suggestive evidence of this flow pressure using observed mutual fund sales.

However, measuring sell pressure with this approach is of limited usefulness

in cleanly identifying nonfundamental shocks. Since the buy and sell actions

of mutual funds are measured directly, these actions reflect of the informa-

tion used in that decision. Edmans, Goldstein, and Jiang (2012), EGJ, propose

solving this problem by estimating not the sells themselves, but a measure of

the quarterly outflows to each fund scaled by the proportion of each stock that

makes up the mutual fund’s portfolio. Essentially, this measure estimates what

the total sales of each stock in each fund would be if each stock was sold in pro-

portion to the fund’s initial beginning-of-quarter holding of that stock. Since

the measure abstracts from which stocks are sold in the quarter and the infor-

mation about firm fundamentals that those sales might contain, it potentially

satisfies the exclusion restriction for an instrumental variables approach. Ex-

posure to the measure also appears to result in large quarterly price declines

followed by a reversion over the subsequent two years. This observation sug-

gests that the price declines are solely due to the exposure to flow pressure in

the event quarter and hence are unrelated to fundamentals before or during

the quarter. EGJ use this measure to instrument for nonfundamental declines

in market value and the effect that these declines have on takeovers.

Since the publication of EGJ, a large number of papers published in high-

quality finance and business journals have used this measure and basic

econometric approach to identify declines in corporate equity value that are

unrelated to fundamentals and the impact of these declines on a wide array

of corporate decisions. These papers include Derrien, Kecskés, and Thesmar

(2013), who examine the impact on payout; Phillips and Zhdanov (2013), who

examine the impact on R&D; Norli, Ostergaard, and Schindele (2015), who

examine the impact on shareholder activism; Zuo (2016), who examines the

impact on managerial earnings forecasts; Lee and So (2017), who examine

changes in analyst coverage; Bonaime et al. (2018), who examine the impact

on mergers and acquisitions; Eckbo, Makaew, and Thorburn (2018), who

examine stock-financed takeovers; Lou and Wang (2018) who examine the

impact on corporate investment; and Dessaint et al. (2018), who examine the

intraindustry cross-firm impact on corporate investment.1

In this paper, I demonstrate that this approach misidentifies the primary

source of these declines in stock price. This misidentification occurs not

1Each of these articles is published or forthcoming in a high-quality finance or accounting jour-

nal such as Journal of Finance,Journal of Financial Economics,Review of Financial Studies,

Journal of Finance and Quantitative Analysis,andJournal of Accounting and Economics. A num-

ber of papers such as Acharya et al. (2014), Deng, Hung, and Qiao (2018), and Bilinski et al. (2018)

use this measure as a control in their tests. In addition, a large and growing number of current

working papers also utilize this measure for identification, including but not limited to Agarwal

and Zhao (2016), Badertscher, Shanthikumar, and Teoh (2017), Chang et al. (2017), Dong, Hir-

shleifer, and Teoh (2018), Gredil, Kapadia, and Lee (2018), Henning, Oesch, and Schmid (2015),

Honkanen and Schmidt (2017), and Sun (2017). Recent working papers by Dessaint et al. (2019)

and Gredil, Kapadia, and Lee (2019) have partially responded to the issues raised in an earlier

working version of this paper.