Feature Article: Will the New Flows of Assets into ESG Funds Produce Alpha for Investors?


Investing in funds based on the rating of companies on environmental, social, and governance metrics are undoubtedly growing in popularity. What is much less clear is whether ESG metrics provide a signal that leads either to increased relative returns (alpha) or reduced risk of loss.

In this article, I take a Bayesian approach, and start by describing two sets of prior views about the potential effectiveness of ESG information as a source of alpha. The first is deductive, based on investment theory, and the second is inductive, based on my experience as a corporate executive.

I then review recent research on ESG investing, and then weigh this evidence in light of the priors to reach a conclusion about the likely alpha generation potential of ESG-based investing.

Prior View

In general, in order to justify investing in an active strategy, like ESG, instead of a broad asset class index fund, one must believe one of these propositions is true:

(1) The active fund will deliver higher returns than the asset class fund, but with higher risk.
(2) The active fund will deliver lower returns than the asset class fund, but with lower risk.
(3) The active fund will deliver higher returns than the asset class fund, but with lower risk.

The first two propositions are consistent with mainstream investment theory. Whether then third proposition is true is the eternal question about active management. One’s belief about the likelihood that an active manager will deliver better than market returns with lower risk (after expenses) must rest on a set of further assumptions:

(1) The active manager has (legal) access to superior information; and/or,
(2) The active manager has a superior model for deriving investment insights from publicly available information; and,
(3) These advantages will last (at least for a period of time); and,
(4) You can identify these managers either because you have access to superior information about active managers and/or a superior model for making sense of public information about them.

As a starting point, we know from the evidence that the base rate of active managers who deliver higher returns (after expenses) with less risk than broad asset class index funds is historically small, declines as the time horizon lengthens (i.e., superior performance is hard to sustain), and in recent years has declines (e.g., because of the increased use of algorithmic trading strategies, and historically low price volatility in many asset classes).

So the question becomes, is there any reason to expect that an active manager would have access to superior ESG information about potential investments, or a superior model for deriving alpha generating insights from publicly available ESG information about them (e.g., their scores on various ESG metrics)?

My corporate experience helps to answer these questions. At a public company I ran, the goal we sought to maximize was total shareholder return relative to an appropriate comparable benchmark (our sector index). However, our pursuit of this goal was subject to a number of constraints, including environmental stewardship, our legal and ethical behavior, our safety record, employee, customer, host community, and supplier satisfaction, and the like.

In practice, the distinction between the objective we were trying to maximize and the constraints we faced was not cut and dried, because on some of the constraints we could act in ways that would, sometimes through multiple channels, produce a competitive advantage for us, that would drive superior value creation.

For example, a superior environmental and safety record meant that we were put on some customers’ preferred supplier list, which resulted in higher revenues. Another example was our choice to increase employees’ pay as they acquired more certified competences rather than wait for a promotion slot to open up, which reduced both employee turnover and operating costs and improved service quality, all of which drove better returns.

The key point to recognize is that we took actions that would have given us a higher rating on some of today’s ESG metrics (had they existed back then) because we believed that they would strengthen our value proposition and business model, and thus generate higher shareholder returns. Put differently, we undertook these actions because they were a part of our business strategy, not in an attempt to boost our composite score on someone’s ESG metric.

In theory, an active manager who understood all this (and we did discuss it on earnings calls) and had a model which accurately incorporated these initiatives into an earnings forecast could have invested in our stock and generated alpha (we outperformed our index benchmark by 43% while I was CEO). But this insight would have been driven by the analysis of our strategy and implementation capacity, not by our composite score on ESG metrics (which was unlikely to provide much information about whether our shares were over or undervalued).

I’ll now move on from environmental and social actions to the issue of governance. I have no doubt that high quality governance contributes to shareholder value creation. That said, I don’t believe that current governance metrics capture the interacting causes from which this advantage emerges (apart, perhaps, from whether a board has separated the chair and CEO roles). There is no metric that measures the quality of a board’s non-executive directors, the issues on a board’s agendas over the course of a year, the information it receives from management and its interaction with them, and the quality of the processes it uses to carry out its fundamental duties of setting direction, allocating resources, hiring and evaluating the CEO, monitoring strategic risk, and evaluating and fairly reporting results to stakeholders.

On balance, my prior view is that investing in companies that have high composite scores on various (and often conflicting) ESG metrics is unlikely to generate sustained risk and expense-adjusted alpha for an investor. However, there is an important caveat – it may be the case that companies that score highly on ESG indexes are less likely to suffer severe losses due to environmental, safety, and other reputational disasters that, in the worst case, could threaten their right to continue operating.

What Does the Research Say?

Conclusions

The weight of recent research evidence on ESG investing is generally consistent with my prior view, and leads me to these conclusions:

1) Simply investing in a portfolio of companies with high ESG ratings is unlikely to generate risk and expense adjusted alpha relative to a broad asset class index return. In fact, as ESG based index funds accelerate their accumulation of assets under management, it may have just the opposite effect. As Ibbotson and his co-authors note in a new research paper (“Popularity: A Bridge Between Classical and Behavioral Finance”, which is covered in this month’s Evidence File) investors tend to bid up the price of assets that have characteristics to which they are attracted, and in so doing reduce their expected returns.

2) However, this relationship is likely not symmetrical. While assets with very low ESG ratings will have their prices bid down and thus their expected future returns increased, research indicates that these low ratings are signals of significant ESG risks of uncertain magnitude that, historically, have led to realized underperformance versus the asset class benchmark.

3) Consideration of ESG issues is most likely to lead to alpha generation when, as part of an active investment management process, they are analyzed in the broader context of a company’s business environment and value creation strategy. However, as is true of all active management, this type of ESG related alpha will likely be very difficult for investment managers to sustain over longer periods of time.