In response to my post about Don Boudreaux’s and my recent op/ed in the Wall Street Journal in which we argued that ESG would get in the way of maximizing shareholder value, frequent commenter (and friend) David Seltzer pointed out that the “annualized return was 0.02% higher for the S&P 500 ESG Index than the S&P 500.” It seemed to him to follow that ESG investing does not hurt shareholder value.

We had a phone call recently in which I explained why it does hurt shareholder value and why the evidence on shareholder returns is not evidence.

Here’s my explanation. Let’s say a bunch of hypothetical firms decide, without warning, that they will go ESG. If I’m right that it creates uncertainty about what steps the company will follow, then the market value of those firms should fall relative to the market value of firms that haven’t made such an announcement but instead have announced that they won’t do ESG.

If that happens, that won’t contradict the findings that David reports above. The reason is, essentially, arbitrage. Once the firms’ values have fallen relative to the values of the other firms, it would be a disequilibrium if their values didn’t rise just as much as those of the other firms from this point on. So someone examining the values of ESG firms will not find a lower rate of return. The rate of return, risk adjusted, will be the same. But the announcement of ESG will have caused a one-time reduction of the value. (Of course, if they get even more “ESGer” in the future than the participants in the market expected at first, the market values of those firms should rise more slowly than the market values of the other firms.)

This, by the way, is why financial economists do event studies. They want to find an event that is a surprise to the market and that is expected to affect the market value of specific firms. So they look at cumulative average residuals of that subset of firms from a few days before the event to a few days after.

I’m going from memory here about what I learned from dozens of financial economics presentations at the University of Rochester’s Graduate School of Management in the mid to late 1970s when I was an assistant professor, and also what I used to complete my Ph.D. dissertation in 1976. If the literature has changed substantially, I’m open to hearing about it. But here’s what ChatGPT told me:

The event window is the period of time over which the effects of the event are expected to be reflected in the stock prices of the affected companies. The event window typically starts a few days before the event and ends a few days after the event.