Do investors actually value sustainability indices? Replication, development, and new evidence on CSR visibility

• Date: 01 September 2019
• Published date: 01 September 2019
• Author: Hervé Stolowy,Luc Paugam,Rodolphe Durand
• DOI: http://doi.org/10.1002/smj.3035

RESEARCH ARTICLE

Do investors actually value sustainability indices?

Replication, development, and new evidence

on CSR visibility

Rodolphe Durand | Luc Paugam | Hervé Stolowy

HEC Paris and S&O Center, France

Correspondence

Rodolphe Durand, HEC Paris and S&O

Center, 1 rue de la Libération, Jouy-en-

Josas 78351, France.

Email: durand@hec.fr

Funding information

Society & Organizations Center - HEC Paris

Abstract

Research Summary:In this study, we replicate and expand

Hawn et al.’sstudy(Strategic Management Journal , 2018,

39, 949–976) that used Dow Jones Sustainability World

Index (DJSI) events to measure variations in firms' Corpo-

rate Social Responsibility (CSR)-activism and examined

their effect on a firm's stock price. We use DJSI events to

capture variations in firms' CSR visibility, holding CSR

activism constant by restricting our analyses to CSR-

equivalent firms. First, we find similar results on stock price

(i.e., no impact) and on trading volumes. Second, because

professional market participants paymore attention to CSR-

oriented firms and use visible cues such as DJSI events, we

study and find that additions to DJSI lead to more analysts

following a firm, and that continuations on the DJSI lead to

an increase in equity being held by long-term investors.

Managerial Summary:We replicate and complement a

recent study that showed that the events of a firm's addition,

continuation, or deletion in a major sustainability index

(DJSI) had little impact on stock market reactions (Hawn

et al. Strategic Management Journal, 2018, 39, 949–976).

We redefine the comparison set that was previously used

and compare across observationally equivalent firms in

terms of CSR orientation in order to isolate the visibility

effect of DJSI events. We find that these events do not sig-

nificantly impact stock price and trading volumes. How-

ever, we expand the analys is and find that sustainabil ity

events attract more attention from financial analysts and

Received: 16 July 2018 Revised: 17 March 2019 Accepted: 20 March 2019 Published on: 9 May 2019

DOI: 10.1002/smj.3035

Strat Mgmt J. 2019;40:1471–1490. wileyonlinelibrary.com/journal/smj © 2019 John Wiley & Sons, Ltd. 1471

lead to an increase in the percentage of shares held by long-

term investors indicative of a trend that professional inves-

tors pay more attention to CSR-visible firms over time.

KEYWORDS

analyst coverage, CSR, long-term investors, replication, stock market

returns, sustainability

1|INTRODUCTION

A substantial literature has considered whether markets react to firms' Corporate Social Responsibil-

ity (CSR)-related activities, finding mixed results. A recent study (Hawn et al., 2018) that provided

the most rigorous empirical approach to date by using an event study (instead of correlational associ-

ations) found at most limited stock market reaction to additions, continuation, or deletion from a high

profile sustainability index, the Dow Jones Sustainability World Index (hereafter “DJSI”). Attention

to several elements might be useful in testing the robustness of the study, including assumptions

about relevant comparison groups, use of alternative analytic techniques, and considering possible

outcomes beyond stock market price reactions. Our study offers a quasi-replication (Bettis, Helfat, &

Shaver, 2016, fig. 3) and extends the study in order to examine the robustness of their results and

whether additional insights can be drawn from this particular empirical setting (Ethiraj, Gam-

bardella, & Helfat, 2016).

1

The choice of a universe from which to draw and of criteria to match observationally equivalent

firms (so-called placebos) is not theoretically neutral. Hawn et al. (2018) match firms on three

criteria: industry (four-digit SIC), EBITDA, and total assets in the year of announcement. The ratio-

nale of such a matching strategy is to consider investments across all possible comparable competing

firms in order to examine variations in CSR activism. While some investors (e.g., retail investors)

may rely on these index-related events to assess CSR activism, evidence suggests that other catego-

ries of investors, in particular specialized and institutional investors (e.g., Ioannou & Serafeim, 2015)

may not (see also Appendix A in File S1). In this study, we reason that investors increasingly con-

sider CSR-oriented firms as suitable and distinct investments and that, therefore, DJSI events may

influence such investors' investments as DJSI events affect the visibility of some firms' CSR versus

others, among all CSR-oriented firms. According to Global Sustainable Investment Alliance (2014),

sustainable investment funds managed $21.4 trillion worth of assets in 2014 (30.2% of assets man-

aged globally) whereas in 2012, sustainable funds were managing “only”$13.3 trillion. These fund

managers are constrained to invest on CSR-oriented firms (PIMCO, 2017). As a result, instead of

comparing the effect of DJSI events across the total universe of firms in Compustat to examine varia-

tion in CSR activities, we compare treated firms (i.e., firms that enter, stay in, or exit the DJSI) with

CSR observationally equivalent placebos, that is, firms with equivalent levels of CSR that could have

been included in the DJSI but were not.

2

1

Replications of precursor works can lead to different conclusions—see Zhao and Murrell (2016), replicating Waddock and

Graves (1997) or to corroborate and extend previous conclusions (this paper).

2

Placebo firms are “placebo-event firms”because they are determined for each event that we study: addition, continuation, and

deletion. To mark the difference in the matching universe vis-à-vis the original study, we refer to CSR-equivalent firms for our

replication and to placebo firms when mentioning the original study.

1472 DURAND ET AL.