ESG Didn’t Immunize Stocks Against the Covid-19 Market Crash

Elizabeth Demers is Professor at the University of Waterloo School of Accounting and Finance. This post is based on a recent paper by Professor Demers; Jurian Hendrikse, Master’s student at Tilburg University; Philip Joos, Professor at Tilburg School of Economics and Management; and Baruch Lev, Philip Bardes Professor of Accounting and Finance at the New York University Stern School of Business. Related research from the Program on Corporate Governance includes The Illusory Promise of Stakeholder Governance by Lucian A. Bebchuk and Roberto Tallarita (discussed on the Forum here); Socially Responsible Firms by Alan Ferrell, Hao Liang, and Luc Renneboog (discussed on the Forum here); and For Whom Corporate Leaders Bargain by Lucian A. Bebchuk, Kobi Kastiel, and Roberto Tallarita (discussed on the Forum here).

From its high on February 19th, 2020, the S&P 500 index lost more than one-third of its value in a little over a month as the potential human and economic consequences of the global pandemic began to be incorporated into share prices. In the wake of this COVID-19-induced stock pummeling, there were widespread claims being made by large investors and fund managers (such as Blackrock, Morningstar), purveyors of ESG data (such as MSCI), as well as by the financial press (including Fortune, the Financial Times, and the Wall Street Journal) that companies with higher environmental, social, and governance (“ESG”) performance scores were immunized against the pandemic-induced value destruction.

The notion that ESG activities contribute to stock price resilience during periods of crisis is premised upon the belief that corporate social responsibility activities help to build social capital and trust in the corporation. These bonds, the story goes, will motivate the company’s stakeholders (employees, customers, suppliers, financiers, government, society, etc.) to help the firm weather the challenges imposed by a crisis. Indeed, several studies of the 2008-2009 global financial crisis (“GFC”) purport to provide evidence of such downside risk protection. An alternative view of corporate social responsibility suggests that executives may choose to improve their company’s ESG scores at the expense of shareholders in order to build their own personal reputation. From this agency theory perspective, ESG investments are at best wasteful, and probably even harmful to shareholders (e.g., by increasing the propensity for management entrenchment). To the extent that investments in corporate social responsibility reflect poor management and/or agency problems, higher ESG scores could be an indication of potential hindrances to the firm’s resilience during challenging times. So which is it? Our study undertakes an extensive set of analyses in order to address this question in the context of U.S. equities during the current global pandemic.

Our first set of analyses regress buy-and-hold abnormal stock returns during the “crisis” quarter (i.e., January through March 2020) on the firms’ ESG scores, while controlling for various other factors that have been shown to affect stock performance, such as accounting-based measures of financial performance, liquidity, leverage, intangible asset investments, variables capturing institutional investor interest and shareholder orientation, firm age and market share, and the firm’s industry affiliation, as well as market-based variables that are known to be determinants of returns. As expected, our results show that COVID crisis returns are associated with the firm’s financial flexibility (i.e., leverage and cash positions), as well as with industry indicators and market-based measures of risk. While we also confirm that ESG and stock returns exhibit simple pairwise correlations for the firms in our sample, our multiple regression results provide robust evidence that ESG is not significantly associated with stock market performance during the first quarter of 2020, once other expected determinants of returns have been controlled for. Notably, however, COVID crisis returns are positively associated with the firm’s stock of internally developed intangible assets (such as R&D, IT, and brands), and this association is both statistically and economically significant. These results suggest that, contrary to conventional wisdom, innovation-related assets rather than social capital investments offered greater immunity to unexpected COVID-induced share price declines.

To further substantiate the irrelevance of ESG scores in determining crisis period stock price resilience, we undertake an Owen-Shapley decomposition of the explained variation in stock returns (i.e., the returns regression model’s R2). The results of these analyses indicate that three groups of explanatory variables offer almost all of the model’s explanatory power for crisis period returns: market-based risk variables, industry fixed effects, and accounting-based measures capturing the firm’s performance, liquidity, leverage, and its stock of intangible assets. Other variables contribute very little to the model’s explanatory power, and ESG is responsible for a meagre 1% of the total explained variation in stock returns during the first quarter of 2020. This is a sobering finding considering all the hype.

Our next set of analyses examine whether ESG is a significant explanatory variable in the same fully specified returns regressions using returns from the second quarter of 2020 as the dependent variable. The results from these analyses indicate that during this market rebound, cash and investments in innovation-related assets are once again positively associated with returns to an economically significant degree, but firms with higher ESG scores, in fact, significantly underperformed less socially responsible companies.

We perform two final sets of analyses. First, we investigate the extent to which there are common indicators of share price resilience across the two most recent and extreme market crises – the GFC of 2008-2009 and the global humanitarian crisis in 2020. This is particularly important in light of the increasing expected frequency and severity of major unexpected shocks to global capital markets (World Economic Forum 2019). Using a GFC period based logit model for which the dependent variable is set to one if the firm falls into the top decile of returns performance from August 2008 to March 2009 (i.e., “winners”), and is set to zero if the firm falls into the bottom decile of returns (i.e., “losers”), we regress this indicator on various parsimonious sets of accounting, market, ESG, and other measures of risk in order to obtain a series of within-sample estimated coefficients. We then fit each separate set of coefficients to the out-of-sample data that was available at the end of 2019 in order to derive predicted Q1 2020 COVID crisis period winners and losers. Our analyses show that a parsimonious combination of accounting- and market-based variables yields a model that is successful both within- and out-of-sample in predicting crisis period winners and losers. Our analyses further indicate that not only does an ESG-only prediction model perform little better than a purely random categorization of winners and losers, adding ESG to the combined accounting- and market-based prediction model contributes little to the model’s performance. Given the inherent differences in the nature of the GFC and COVID crises, our success in predicting winners and losers across the two crises is not a trivial feat. Contrary to all the hype, however, ESG plays no meaningful role in this success.

Our final analyses document the economic significance of the preceding out-of-sample prediction models by constructing a series of hedge portfolios that go long (short) in firms that each of the respective models predicts will be winners (losers). These analyses confirm that the combined accounting- and market-based prediction model yields economically significant abnormal returns for the Q1 2020 COVID crisis period. Consistent with all of our previously reported analyses, ESG offers little improvement to the model’s investment success.

Overall, our study provides robust evidence that, contrary to widespread claims, ESG is not an important determinant of crisis period stock returns, nor does it offer meaningful out-of-sample predictive power to help discriminate between crisis period winners and losers. Rather, traditional accounting-based measures of the firm’s financial flexibility, a measure of the firm’s stock of intangible assets, the firm’s industry affiliation, and common market-based measures of risk are all significant in explaining crisis and recovery period returns. Furthermore, our key findings—that firms’ financial flexibility and innovation-related assets are positive indicators of crisis resilience, while ESG scores are not—are generalizable across the two most recent, but characteristically very different, global crises.

While our results don’t speak to the longer-term shareholder value creation of responsible corporate citizenship, an approach to doing business that we generally support and advocate for, they do provide robust evidence that ESG’s touted role as a vaccine against unexpected market shocks is grossly overstated.

The complete paper is available here.

Both comments and trackbacks are currently closed.

One Comment

  1. John
    Posted Sunday, September 20, 2020 at 4:47 pm | Permalink

    This is an interesting paper and raises good points. However, I think that it misses some of the channels through which ESG influenced performance. Traditional funds generally underperformed ESG funds during the covid shock, after all. If all traditional fund managers needed to do was look at metrics like liquidity and R&D to protect themselves, why didn’t they?

    The problem is, this type of analysis seems to assume you know a shock is about to happen. Of course, if I sold all my CCC rated companies in Dec 2019 I would’ve done better than those who didn’t. But if I had sold them in Jan 2016 I’m not so sure how I’d look.

    Investors tend to write off the possibility of a sudden black swan shock. As we’ve seen many times, they tend to put 1 in a million odds on things that are more like 1 in 100. And that causes them to pick companies with shakier balance sheets, etc when things are going alright. Because until the downturn happens, those companies tend to pay you more.

    But perhaps adding an ESG constraint tames this tendency? Perhaps it implicitly pushes investors to take a longer term investment approach (even if they don’t realize it)?