In our paper, Sustainable Investing: Evidence from the Field, which was recently made available on SSRN, we survey 509 equity portfolio managers from both traditional and sustainable funds on whether, why, and how they incorporate firms’ environmental and social (“ES”) performance into investment decisions. The answers reveal several interesting results, organized into four groups, that point to a more complex model of investor behavior than currently used in the academic literature.

Beliefs

Our first question asked respondents to rank the importance of actual ES performance (not ratings) for long-term firm value relative to five other value drivers: strategy and competitive position, operational performance, corporate culture, governance, and capital structure. ES performance received the lowest average support, with 73% ranking it fifth or sixth; its average ranking was lowest even among sustainable investors. Notably, ES ranks significantly below governance (“G”), even though ESG factors are often bundled together; below corporate culture, another intangible; and below capital structure, even though the latter is irrelevant in perfect capital markets.

The low ranking of ES performance does not mean that investors view it as irrelevant. Many free text entries emphasized that all value drivers are interlinked, that ES can affect the firm’s competitive position or operational performance, and that deficiencies in any of the value drivers are a concern. Others indicated that poor ES performance can be a signal of other problems, and that specific ES issues are highly material in specific industries. Thus, many portfolio managers have a nuanced view of ES that emphasizes granularity, omitted variables, and interactions with non-ES value drivers.

Moreover, when we asked respondents to rate the financial materiality of ES performance on eight dimensions, 85% (including 78% of traditional investors) rated at least one ES issue as material. Both investor types view ES performance on employee and consumer-related issues as most important, potentially because they are internalized even in the absence of regulation. More sustainable than traditional investors view environmental issues as material.

We next turned from long-term value to long-term returns, to study views on market pricing. 73% of sustainable investors expect good ES performers to deliver positive alpha, and a notable 45% of traditional investors agree. Unexpectedly, by far the most popular reason is that ES performance is correlated with other factors that improve shareholder returns, rather than mattering directly. The second is that ES is directly valuable but the market fails to price it in. Investor short termism and unawareness of the financial materiality of ES are seen as more important than insufficient information. There is even greater similarity in the two investor types’ expectations for poor ES performers, with 61% (67%) of traditional (sustainable) investors predicting negative alpha. This suggests that some traditional investors view ES as a hygiene factor, where poor performance matters more than good.

We then asked investors whether, from a shareholder value perspective, firms over- or underinvest in the eight ES issues. For all eight, the modal response was that they invest at the optimal level, potentially explaining why investors are selective in their support of shareholder proposals and their engagements. Yet, for each issue, more than 40% of investors believe that firms either over- or underinvest; across all issues, 68% of investors believe that companies overinvest in at least one (most commonly greenhouse gas emissions), and 51% that they underinvest in at least one (most commonly ecological impacts). The most supported reasons for overinvestment are pressure from the media, the public, employees or investors; underinvestment is attributed to investor or company short termism.

Objectives and constraints

Our next set of questions asked about investors’ objectives and the constraints they operate under. Only 27% of investors (24% traditional, 30% sustainable) would tolerate companies sacrificing even one basis point of annual return for ES performance; both types explained that fiduciary duty prevents them from doing so. This contrasts models where investors assign significant weight to social value. While plausible for asset owners, it does not describe delegated asset managers constrained by fiduciary duty. Instead of having an objective function that trades off social against financial value, fund managers take ES performance into account largely to improve financial returns or, as the next questions will show, to satisfy constraints.

71% of investors (61% traditional, 84% sustainable) report that ES constraints such as firmwide policies, fund mandates, and client wishes led them to make different stock selection, voting, or engagement decisions than they otherwise would. The most frequent consequences were that investors had to avoid stocks that they believed would improve returns or diversification; for 41-77% of constrained investors (and thus 29-55% of investors overall), these constraints reduced financial returns. Paradoxically, ES constraints sometimes led investors to take actions that reduced their ES impact, such as not investing in ES laggards whose performance they could have improved.

These responses show that the industry does not readily partition into traditional funds with a purely financial objective and sustainable funds with both financial and social objectives, nor into unconstrained traditional and constrained sustainable funds. Instead, both types of funds have a dominant financial objective but also face a range of formal and informal ES constraints. Since these constraints are imposed either by the fund family or its clients, one may wonder why they exist. One explanation, reinforced by our interviewees, is that they are a second-best solution to a principal-agent problem. Writing a contract that requires asset managers to maximize their asset owners’ weighted objective over ES and returns may be infeasible. Instead, asset managers offer a menu of funds with different constraints that they believe reflect clients’ ES concerns. When selecting a fund, clients choose its constraints plus the manager’s ability to maximize returns subject to those constraints, rather than its objective function. A model of sustainable investing with delegated asset management that features such an equilibrium has, to the best of our knowledge, not yet been written.

Actions

Our third set of questions investigates how ES considerations affect investor actions: stock selection, voting, and engagement.

Stock selection: 77% of investors (66% traditional, 91% sustainable) “often” or “very often” incorporate ES performance into stock selection. The reasons, however, differ. For sustainable funds, the fund mandate is most important, followed by alignment with client values, and firmwide policies. These three constraints rank higher than “to improve fund returns” or “to avoid downside risk.” For traditional funds, the two financial reasons are most important, followed by the three constraints.

Despite these differences, financial reasons cause a majority of both investor types to regularly incorporate ES performance into stock selection: 74% of sustainable and 51% of traditional investors do so to avoid downside risk, improve returns, or reduce volatility. Avoiding downside risk is a more common reason than improving expected returns, and much more important than reducing volatility. Incorporation of ES performance for financial reasons is correlated with fund managers’ beliefs on whether ES is a source of alpha: thus, while ES constraints are mainly determined by the type of mandate, ES-based actions taken in pursuit of returns are mainly determined by beliefs. Impact (affecting firms’ cost of capital or rewarding / penalizing companies for ES performance) is less important, mattering for just 20% of traditional investors and a minority (41%) of even sustainable investors.

Voting: Consistent with the importance of fiduciary duty, only 27% of investors (24% traditional, 31% sustainable) have voted for a shareholder proposal that was even slightly negative for shareholder value, even though 78% have supported value-neutral proposals. Such voting, especially for negative-value proposals, is driven more by ES constraints than the proposal’s expected impact on society. The effect on other companies owned by the investor is least important, suggesting limited support for “universal owner” motivations.

Engagement: 76% of investors (64% traditional, 92% sustainable) have engaged with companies to improve their ES performance. Such engagement is motivated primarily by the expected impact on shareholder value, followed by the issue’s importance to clients, the firm, and wider society. Marketing motivations, such as concerns for the fund’s sustainability rating and reputation, are seen as least important. The main reasons why some investors never engage are their small stake and the costs of engagement, consistent with standard cost-benefit analysis. These responses suggest that most asset managers are reluctant to undertake ES engagement that is not in their clients’ financial interest.

Specific ES issues

We finally asked investors whether and why they take into account carbon emissions or board diversity, two ES issues that receive significant attention. Despite their differences (diversity is a social issue whose effects are mainly internalized by the company, emissions are an environmental issue whose effects are mainly externalities), the responses are similar. Neither is considered of high importance, and the most common reason for considering either is its impact on society. Consistent with prior responses, reducing downside risk and complying with fund mandates, firm policies, and client values are more important reasons than improving returns. More investors associate emissions with lower returns and diversity with higher returns than the reverse, in contrast to academic research that shows the opposite or no link. In free-text fields, investors explain that they consider multiple forms of diversity, while most academic research focuses on demographic diversity.

Additional conclusions

We can draw four broader conclusions from the results. First, the asset management industry is unlikely to lead the charge in improving the aggregate ES performance of firms. Most investors do not place significant weight on ES objectives beyond what is required to improve financial returns, nor do they believe that firms are systematically underinvesting in ES. This may explain why academic research generally finds that SI has a limited impact on companies’ ES performance. This need not be because asset managers are greenwashing, but because they are bound by fiduciary duty and believe that most companies are investing in ES optimally.

The second is that differences between traditional and sustainable investors are smaller than commonly believed. Both types recognize fiduciary duty and the priority of financial returns, with similarly low proportions willing to sacrifice financial returns for ES performance or to vote for ES proposals that are even slightly negative for shareholder value. Both face ES-related constraints that affect their portfolio composition. Many of their beliefs are also similar. Both types rank ES performance low relative to other value drivers, yet over three quarters of both view at least some ES issues as financially material. Both often tilt their portfolios based on ES performance to improve risk-adjusted returns, and both engage with companies to improve ES performance. Empirical studies on the effectiveness of SI often focus on funds labelled as “sustainable” or “responsible,” or compare them to funds without such labels. However, the performance of explicitly sustainable investors may not be representative of the performance of sustainable investing. Some funds with sustainable labels do not incorporate ES into stock selection for financial reasons or engage on ES performance; many funds without such labels do both.

Third, there is large heterogeneity of beliefs, and of actions driven by beliefs, which does not polarize neatly across traditional vs. sustainable lines. For example, 45% of traditional investors expect good ES performers to deliver positive alpha, 44% no alpha, and 11% negative alpha. These beliefs affect behavior, with believers in ES outperformance much more likely to use ES performance in stock selection and engage with companies on ES. This contrasts prior research that attributes differences in investor behavior to different preferences, and suggests that it may be fruitful to incorporate heterogeneous beliefs into models of sustainable investing. For asset owners, it is important to understand that whether portfolio managers act as “traditional” or “sustainable” depends more on their investment beliefs and ES constraints than how their fund is labelled.

The final conclusion is the need to better understand how asset managers reflect, or fail to reflect, their asset owners’ preferences. Several prior studies provide evidence that many retail investors care about ES performance. The vast majority of traditional and sustainable asset managers, however, prioritize financial returns. ES performance affects their investment decisions because of ES constraints, or because they view it as a predictor of returns. It is an open question whether self-selection of asset owners with different preferences to managers with different beliefs and constraints achieves asset owners’ ES objectives, particularly given the difficulty of observing beliefs.