Hit 'Em Where It Hurts, ESG Investing 2.0

Published in Investments & Pensions Europe

“Managers of firms don’t like people who short sell their stock, especially if the short sellers are accusing the firms of fraud and even more especially when the fraud accusations are true.”
‒ Owen Lamont (2004) “Short Sale Constraints and Overpricing”

Responsible investment, otherwise referred to as ESG (Environmental, Social, Governance) investing, has become an important topic for many investors and their asset managers over the last several years. Asset owners are increasingly concerned not only with their fiduciary responsibility to deliver financial results but also with impact on their constituents and the broader global community, making the questions and policies around these topics more central to an investment decision. ESG concerns have often led investors to avoid holding certain controversial stocks, whether they are cluster munitions or coal producers, as well as companies in more mainstream industries that rank poorly on ESG metrics.

However, avoiding investment in a company may be like the proverbial tree falling in the woods that no one hears — did it make a sound? Companies have an interest in attracting shareholders, but they aren’t necessarily aware of those investors who choose not to hold its stock — or why. We suggest a new approach to ESG investing that we believe may be more effective in making negative investor views known to management — while at the same time potentially improving portfolio expected returns.

This approach, which we label 130/30 Sustainable Relaxed Constraint investing, eliminates poor ESG performers from the potential universe of long positions but removes the no-shorting constraint typically applied to equity portfolios and instead allows limited shorting (i.e., total gross position of -30% of the portfolio’s net asset value) of companies, including those with a poor ESG profile. The result is a portfolio which avoids taking a long position in a poorly-ranked ESG name in all circumstances, but will actively short such a name when our overall return-forecasting model sees it as potentially beneficial to the portfolio.

Why allow shorting?

Shorting can be a forceful expression of the ESG considerations driving the restriction in the first place. Investors who do not condone a given firm’s activities would not hold the stock of that firm, but could be open to shorting such a stock. Shorts may send a strong message to corporate management, and are more difficult to ignore than investors who simply abstain from holding shares. Of course, short positions do not give the investor a vote at the shareholder meeting. However, we believe they have a large and important advantage in that corporate management is keenly aware and interested in them. Put simply, corporate management pays close attention to what the short community thinks about its company.

Sometimes this can provoke an aggressive reaction — unfortunately hardly a constructive response, but one that is illustrative of how important this can be to the management. Recently, Cleveland-Cliffs' CEO threatened short sellers on an earnings call. 1 1 Close Wall Street Journal, "Cleveland-Cliffs CEO Lashes Out at Analyst, Short-Sellers," October 19, 2018. Another telling example is the reaction of Dick Fuld, the CEO of Lehman Brothers: Lamont (2012) reports that “Mr. Fuld and other Lehman executives have accused investors who short the firm’s stock — making money by betting the price will fall — of spreading false rumors about the firm’s ability to finance its businesses... ‘I will hurt the shorts, and that is my goal.’” Of course, we believe that it is best to resolve disputes amicably and reach agreement through a dialogue. Unfortunately, this may not be an easy option for investors who are not prepared to hold any stock in the company. Rather than risk being ignored by the management altogether we believe that such investors may agree to sometimes be on the short side of the company and thus “hit ‘em where it hurts.”

Additionally, allowing shorting in general has the potential to improve performance by allowing for a fuller expression of an investor’s views. This is a first-order effect that’s been well documented in prior literature (e.g. “Towards More Information-Efficient Portfolios” [Clarke, de Silva, and Sapra, 2004] and “Understanding Relaxed Constraint Equity Strategies” [Ang, Michalka, Ross, 2017]).

Can ESG information help identify candidates for shorting?

We do not advocate that investors should short all controversial companies, all the time — but shorting may be very useful when aligned with their investment view. In fact ESG-type information might be one input into an investment view — for example, information about a given company’s earnings quality.

Earnings quality reflects how conservative a company is in accounting for its profits, which in turn reflects that company’s governance. Firms with stronger corporate social responsibility scores (and presumably better governance, and perhaps a better overall ESG profile) tend to adopt conservative accounting processes (e.g., Kim et al., 2012).

A simple indicator illustrating this idea measures the importance of cash in a company’s profits: when profits are derived mainly from cash they tend to be more conservative; profits derived mainly from accruals may be less certain.

Accruals are effectively a promise of cash coming in the future; in the best case, the promise is risky and may not materialize; in the worst case, the promise may be based on aggressive accounting choices. Indeed, research shows companies that are subject to SEC enforcement actions tend to have abnormally high accruals prior to such actions (e.g., Richardson, Sloan, Soliman, and Tuna, 2006). Such companies also have a higher likelihood of earnings restatements (e.g., Richardson, Tuna, and Wu, 2002).

Not surprisingly, academic research (e.g., Sloan 1996) suggests that it has paid to tilt toward companies with higher earnings quality. High accruals, and the subjectivity involved in accruals, may lead to temporary accounting distortions that make accounting statements less informative. Investors may not appreciate that mechanism and make systematic errors, temporarily driving up the prices of high-accrual companies; since such companies may on average be overpriced, they may realize negative excess returns going forward.

Impact on investment performance

How large is the performance impact of allowing for shorting, and in particular allowing shorting of companies with a poor ESG profile? To address this question we turn to a hypothetical backtest on a straightforward strategy combining, at equal risk weights, a set of valuation, momentum and defensive indicators that are often used by quantitative managers to assess a stock’s attractiveness.

We compare the performance of the same underlying investment model, first in a purely long-only fashion with ESG-type screens; then allowing for a 130/30 ‘Relaxed Constraint’ implementation (i.e., allowing for some shorting) but keeping the ESG screens on both the longs and the shorts; and finally our preferred 130/30 implementation with ESG screens on the longs, but allowing shorts on screened companies. The ESG screen we incorporate here removes stocks with MSCI ESG scores below 2, capturing roughly 10% of the MSCI World index universe. In short, the hypothetical ESG Sustainable Relaxed Constraint 130/30 had higher performance than either the hypothetical ESG-screened long-only or the screened 130/30 portfolio, which is constrained from holding poor ESG names on the long or short side (see Exhibit 1). Allowing shorting of poorly-ranked ESG names may help performance, while simultaneously sending a strong signal to management.

Our expectations for performance are quite clear: we expect to see a meaningful improvement from allowing any shorting, in line with prior studies on the topic (e.g., the two studies we cited above, Clarke et al., 2004 or Ang et al., 2017). This change leads to a potential first-order improvement in returns.

Relaxing a constraint further should help performance, so it is not surprising to see another potential improvement from allowing shorts on the stocks previously excluded for ESG purposes. In fact, this second improvement seems particularly large. Ex ante, we expect to see higher returns, but they arise only when the investment view happens to be sufficiently negative on one of the restricted stocks. The screen is more than cosmetic, removing over 10% of stocks from the investment universe, but it is still surprising to see that those 10% of stocks have such a meaningful impact on performance. One reason may be that over our sample period, limited by the availability of ESG data, some stocks with particularly poor ESG scores experienced poor performance.

As we have illustrated, a 130/30 Sustainable Relaxed Constraint strategy, which limits the ability of a portfolio to take a long position in a poorly-ranked ESG security but allows for limited shorting of any stock including those that are poor ESG performers, may outperform both long-only portfolios and traditional ESG-screened versions whether they be long only or 130/30. Moreover, taking a short position in a poorly ranked ESG security is more in line with the desired outcome of ESG investors — namely to encourage change at the underlying company to better align with ESG objectives. Company executives of poorly-ranked ESG stocks, who find their shares shorted, may not like seeing ESG investors bet against their company — but that is exactly the point, and why ESG investors might consider improving their portfolio’s expected performance and sending a strong signal — by relaxing the no-short constraint of their ESG implementations.

References:

Ang, I, A. Michalka, A. Ross, 2017, “Understanding Relaxed Constraint Equity Strategies,” AQR Whitepaper
Clarke, R. G., H. de Silva, and S. Sapra (2004) “Toward More Information-Efficient Portfolios,” the Journal of Portfolio Management, Vol. 31:3, pp. 54-63
Kim, Y., M.S. Park, B. Wier (2012) “Is Earnings Quality Associated with Corporate Social Responsibility?” The Accounting Review, Vol. 87:3, pp. 761-796
Richardson, S. A., R. G. Sloan, M. T. Soliman, and I. Tuna (2006) “The implications of accounting distortions and growth for accruals and profitability,” The Accounting Review, Vol. 81:3, pp. 713-743
Richardson, S. A., I. Tuna, M. Wu (2002) “Predicting Earnings Management: The Case of Earnings Restatements,” working paper, University of Pennsylvania
Sloan, R. G. (1996) “Do stock prices fully reflect information in accruals and cash flows about future earnings?,” The Accounting Review, Vol. 71:3, pp. 289-315