The Disappointing Reality about ESG Fund Performance

On November 8, 2019, at 11:50pm, this article was corrected. The original version stated that, because ESG/SRI investors choose from a smaller universe of securities than an unconstrained manager, they should expect to underperform a risk-adjusted benchmark. That is not correct. Reducing the size of the investable universe does not reduce the expected returns. The article was edited to correct that error. For more information on that correction, see the comments in APViewpoint.

We’d all like our investments to make a positive contribution to environmental, social and governance (ESG) issues. But, as predicted by economic theory, ESG funds have suffered a performance deficit over a long time horizon.

The question of whether ESG funds have outperformed an appropriate benchmark on a risk-adjusted basis has been studied extensively (see here for a survey of that research). But I was intrigued by a recent study by Wayne Winegarden, an economist and senior fellow with the Pacific Research Institute (PRI). Winegarden looked at 30 ESG funds, 18 of which have a 10-year track record and the remaining 12 of which outperformed the S&P 500 over the recent past.

Only 1 of the 18 outperformed the S&P 500 over a five-year investment horizon, and only 2 beat the S&P 500 over a 10-year horizon. The results were no more encouraging for the 12 recent outperformers.

What is responsible for this underperformance? Let’s look at the sobering data on ESG performance, and then the explanations for it.

The sad reality

Let’s say you have two equally competent portfolio managers, one of whom can select investments from an unconstrained universe, but the other is restricted to a subset of that universe. Our intuition says that the former will outperform the latter. As Winegarden said to me when I interviewed him over the phone, “Options have value. Anytime you restrict your options, you're going to be harming your potential performance.”

ESG investors choose from a subset of available securities. But contrary to our intuition, that should not lead to underperformance. It could lead to increased risk, as measured by standard deviation, but only if the size of the universe was reduced to about 30 securities.