ESG investing proponents believe that ESG has alpha. Not only can you expect to make more money in the ESG version of an asset, but at the same time, you will be a force of good in the world. ESG opponents argue that ESG investing has no alpha and that fund managers who pursue ESG investing act against their shareholders’ best interests by focusing on social outcomes rather than maximizing sharehodler wealth - and this may be unlawful depending on who you ask.

Regardless of whether ESG outperforms the market or not, an interesting question is what utility ESG has for the average investor. Does it preserve value in a recession? Can it be used to diversify? How about ESG data, is it worth paying attention to? Of course you can look for answers to these questions in news articles and studies, but then you get this:

followed by studies like this:

Like anything else in markets, ESG seems to have gone through a narrative cycle and right now the sentiment is strongly against it.

It went from this:

to this:

To illustrate, WSJ reported in January that executives are purposely avoiding using the term. But does this shift in sentiment mean that it would be profitable to fade it?

The Data

I happened to have a dataset of Refinitiv’s ESG scores for S&P 500 stocks from Jan 2010 to Oct 2022. It’s a monthly dataset and I couldn’t update it to 2024, but I thought it would still offer some insights. I also downloaded daily price data for every iteration of the S&P 500 during this period from Yahoo! Finance. I removed any tickers whose ESG scores I didn’t have, and the resulting balance of missing and available tickers looked like this:

I thought it was representative enough, but there is a small caveat to the results because the data wasn’t full bar across the study period.

ESG Characteristics of S&P 500 Stocks

In January 2010, about 40% of the stocks had high ESG and roughly 58% had low ESG. Towards the end of the period, circa Oct 2022, it was more balanced. Apart from mid-2015 to 2017 when the percentage of high ESG dropped to roughly 20%, the number of stocks with high ESG has been on a steady ascent. Mark that period for later.

Average ESG scores in each sector rose steadily over the timespan. Besides communication services, the financial services and energy sectors have the lowest ESG scores because energy directly affects the environment(Environment factor) and financial services are linked to many fraud cases (Governance factor). The median chart looks similar so it’s not because of outliers.

There is no real difference between the risk-reward characteristics of high and low ESG stocks. I highlighted the outliers. RIG (Transocean) is an offshore drilling company. DexCom (DXCM) is a pharmaceutical company that produces diabetes monitoring stuff. Empire Petroleum (EP) is an oil company. And GameStop.

Nothing rather remarkable about High ESG stocks in regards to their risk-reward properties.

A Brief Primer on ESG and It’s Importance in Financial Markets

Skip this section if you don’t want to know how ESG became a thing that people cared about. 

In it’s earliest form, sustainable investing had elements of religion. Its earliest roots can be traced to religious discourse such as a sermon in 1872 by John Wesley dubbed ‘The Use of Money’, whose main takeaway was that we should ‘gain all we can without hurting our neighbor’. Then, as corporations grew and became widespread in the early 1900s, their responsibilities to outside parties came under debate. An example is the famous Dodge v Ford case in 1919. John F. and E. Dodge both owned Dodge Bros. Motor Car plant which owned 10% of Ford Motor Co. They sued Henry Ford because he used the company’s assets to grow the business instead of paying dividends. The judge, Chief Justice Ostander , wrote on behalf of the Supreme Court of Michigan: “A corporation is organized and carried on primarily for the profit of the stockholders. The powers of the directors are employed to that end.” The ruling fully endorsed stockholder theory, which meant that corporations didn’t have to give a damn about being ‘socially responsible’, unless that is, it maximized sharehodler value.

Still around those times, corporations were viewed as impersonal legal entities and the public generally distrusted them. Even though they didn’t need to care about social issues, there was pressure on managers to win the hearts of the people. Labor unions at the time were focused on improved working conditions so corporations begun to pay attention to these issues in their PR efforts, giving birth to Corporate Social Responsibility (CSR). The main emphasis back then was on a corporation’s responsibilities to its employees; not unlike how the main emphasis of modern day sustainable investing is on a corporation’s responsibilities to the environment. Academics of the time churned up stakeholder theory as a direct challenge to stockholder theory , arguing that the right decision-making framework for managers involved honoring the corporations responsibilities to different stakeholders i.e. employees, shareholders, society etc. Many academic papers in the 1930s and 1940s sought to define the various responsibilities. By the 1950s, CSR was a formal thing and remained conceptually the same until the globalization phenomenon made it clear that CSR was only focused on a select few stakeholders, and this wasn’t enough for global companies as they could do good at home, but misbehave elsewhere.

Then, in 1999 at the World Economic Forum, the then UN Secretary General Kofi Annan urged business leaders to align their business goals with UN’s goals. In 2004, he sent a letter to 50 CEOs of major financial institutions to collaborate with the UN to integrate environmental, social, and governance factors to capital markets and later, an initiative was formed called the PRI which presently has some 7000 signatories and $121 trillion AUM. Hence ESG became a thing, and it spanned ESG funds, green bonds, rating agencies and more.

A hot issue between ESG proponents and opponents is whether ESG investing has alpha or not. ESG opponents argue that a fund manager should be looking to maximize shareholder wealth, anything else be dashed. There is therefore no need for special ESG anything because if ESG has alpha, then the market will reward high ESG assets whether they are labeled as such or not. A smaller faction of ESG opponents argue that since the higher the risk, the higher the reward, then it should pay more to buy into low ESG stocks. ESG proponents argue that sustainable companies ought to outperform their less sustainable peers so ESG investing should maximize shareholder wealth. This was a major selling point for ESG funds, but Bloomberg reported that ESG studies that fueled Investing boom were flawed. 

The debate has potential legal ramifications depending on which side regulators land in when it is settled. For instance if it is the latter, you could sue a fund manager who lost your money by choosing sustainable investment alternatives over others if you didn’t sign up for it.

ESG as a Risk Factor

According to iShares:

This approach hails from Fama and French who in 1992 showed that asset returns could be explained by common risk factors like value, size and exposure to market risk. Since then, academics have sought other risk factors to add to the list which so far also includes momentum and profitability among others. There is an ongoing discussion in academic literature about whether sustainability should be included too. Fama and French warned about including just any old risk factor, but the search for risk factors has been led by pragmatists who are looking for what works and what doesn’t, without caring for why. That’s because the relevance of these factors changes. A 2020 study found that the original factors had negative returns from 2010-2019. So it would be interesting to see if ESG qualifies to be a risk factor.

To this end, I replicated the Fama-French factors, and then constructed similar momentum and ESG factors. The formulas are in the appendix. SMB (small-minus-big) is the size factor which professes that small-caps should outperform large-caps. HML (high-minus-low) is the value factor which professes that high (book-to-market) value stocks should outperform low ones. MOM is momentum-high momentum stocks should outperform low momentum ones, and ESG professes that high ESG stocks should outperform low ESG ones. The charts below show the value of $100 invested in each factor. Momentum ruined the y-axis so I put it in a different chart with the other factors that went up.

Replicated size factor (Replicated-SMB) tanked. Original value factor (HML) also tanked but then recovered some during the covid rally. Original size (SMB) was unch, and ESG mostly tanked but then recovered in the covid rally before taking a nose dive after Russia invaded Ukraine. So sustainability(ESG) came out on top among size (SMB) and value (HML) even though it didn’t really have noteworthy returns. These findings are generally in line with the aforementioned study that reported fama-french factors having negative returns.

Replicated value (Replicated HML) and momentum soared. That market premium (Mkt-Rf) soared is trivial. I highlighted the start of Covid and the Russian-Ukraine war so it would be easier to see how the risk factors performed during crisis periods. ESG performed well during covid, but the chart below this paragraph shows that everything except bonds (TLT and BND) did well in covid, then dropped when Russia invaded Ukraine. DSI is the iShares social ETF made up of 400 stocks which is supposed to omit companies in controversial industries like gambling etc. I wanted to contrast it to the ESG factor, and it looks like DSI is just tracking SPY. During crisis times correlations go to 1 so ESG’s performance from 2020 could be down to how everything was highly correlated at that time. Also during the mid-2015 to 2017 period when a low portion of the stocks had high ESG, the ESG factor rose, but the same time so did SPY. So it looks like positive returns in both DSI and ESG factor both depend on what SPY is doing.

I did some regressions to check if the stock returns were sensitive to these risk factors. I grouped the models depending on whether the factors were obtained from Kenneth French’s website (Original), or replicated, and whether ESG and MOM were added to the SMB, HML and Market Premium (Mk-Rf) model. The chart shows box plots of the p-values for each factor plus the adjusted R-squared of the model.

About 40% of the stocks had an adjusted R-squared above 50 which can be interpreted to mean the model’s had some explanatory power on the returns. Market premium is almost always highly significant except for a few outliers. Notice that even though momentum outperformed everything else, few stocks are sensitive to that risk factor. The next chart is another way of visualizing the same information. It shows the percentage of stocks whose p-values were significant. Only about 20% of the stocks had a significant response to momentum, and between 20-30% to ESG.

So overall, ESG as a risk factor has negligible returns at best and few stocks respond to sustainability risk.

ESG Data as an Indicator

I attempted to determine whether you should care about ESG data. Say you own a stock, and it’s ESG rating goes down, it is time to sell? What if it’s ESG score went up instead? Do you buy more? Or sell? A 2019 study that found that the stock market reacts a little bit (-0.1%) to negative ESG news, but has no reaction to positive news.

I backtested if it would have been profitable to make a directional bet based on a change in ESG score. For entry signals, I normalized the changes in ESG score by previous score to give percentages, then I varied the holding period and how big a change would trigger an entry long or short. The chart below is a heatmap of the total returns of the different strategies.

Every version of the strategy had positive returns but nothing spectacular. The best performing strategy was to take a directional bet based on a large percentage change in ESG score (20% or more), then hold for two years (504 days). Taking a directional bet on a 5-10% change in ESG score and holding for 3 months returned about 20%. Here are more stats for the best performing strategy. The benchmark return is a buy-and-hold strategy.

I also checked how the same strategies performed on stocks that had a statistically meaningful sensitivity to ESG. These had a higher total return of about 50% if you took directional bets on a 10-15% change in ESG score and held for a year. Holding for two years was suboptimal but also profitable.

There were about 10 stocks that were extremely sensitive to ESG. The results show no consistency.

Conclusions

The findings show three things:

• High ESG stocks have the same risk-return properties as low ESG stocks.

• ESG makes a lousy risk factor that is correlated to the market in crisis times. So no diversification benefits as well.

• Taking directional bets on ESG scores has negligible profits and depends on the size of the change in ESG score, the holding period, and whether or not the stock is already sensitive to ESG risk.

Keep in mind that the data had some missing tickers for various iterations of the S&P 500, and that this analysis used Refinitiv’s ESG scores, and focused exclusively on the S&P 500. The results may vary if you use non-index stocks, stocks from other indices, or other countries. The results may also vary if you use a different ESG ratings provider. There are no established standards for determining ESG ratings leading to a wide divergence in scores among rating agencies. There is a lot of subjectivity involved as the quoted article highlights:

Epilogue

In lieu of inundating you with charts about the ESG properties of the stocks, I had decided to make a dashboard that I was to present in this section but couldn’t finish it in time. Once it is done, I will update this page and let you know. Every post will come with supplementary stuff like code, tradingview indicators, data, supplementary results and more.

Appendix

The Fama-French factors are created as follows:

First,  split the stocks based on the size (Small-S, Big-B) and the book-to-market (BE/ME) ratios (HIgh-H, Medium-M, Low -L) to get six portfolios.

- For the size factor, make the split using the median of market equity (the grand total of market capitalization).

 - For the value factor, make the split using the 30th and 70th percentiles of the book-to-market ratios.

 - For the momentum factor, split the previous month's returns at the 30th and 70th percentiles.

 Then, construct the portfolios as follows: