Explaining ESG equity index fund performance

By Jan-Carl Plagge, Global Head of Active-Passive Portfolio Research, Investment Strategy Group, Vanguard Europe

• Many ESG equity index funds show substantial and persistent deviations in industry allocations relative to the market – although the trend is declining.
• The impact of these deviations on performance, however, has been less persistent: depending on the time period, it can be positive or negative, accounting for the majority of return differences between an ESG fund and the market in some periods and the minority in others.
• Treating industry returns as a possible factor in explaining ESG fund returns, we find that only a few have had a statistically significant impact on fund returns, both before and after controlling for style factors.
• What’s more, we find substantial variations in industry allocations among individual ESG funds, so investors are best served by assessing them on a fund-by-fund basis.

As environmental, social and governance (ESG) investment products continue to grow in popularity, many investors will be curious to know what drives the performance of ESG funds relative to the performance of the broader market.

Can the growing number of investors considering these products find any discernible patterns in performance?

To help answer this question, we investigated whether over- and underweights of industries in ESG equity index funds relative to the broader equity market drive returns. For our analysis, we looked at equity index funds and ETFs with a US investment focus that indicated the use of ESG factors in their investment process, according to Morningstar, from 1 January 2006 to 31 December 2020.

Our previous research based on a similar range of ESG funds found that the performance of many funds deviated from the broader market and that part of the difference in performance could be explained by exposures to style factors¹. What we did not explore in that study was the role of industry allocations.

Many ESG funds under- or overweight companies based on their business activities, such as the production of tobacco or the extraction or distribution of oil and gas. ESG funds that apply an exclusionary screening approach explicitly (and often solely) focus on such business activities. As a result, these funds should have industry allocations that systematically differ from those of the broader market.

Funds that apply “inclusionary approaches”, which often select companies by assessing their exposure to ESG risks, also tend to have different industry allocations compared to their non-ESG peers because ESG risks are often industry-specific.

Substantial deviations have declined over time

Firstly, we composed an ‘ESG market portfolio’ as an asset-weighted aggregate of all funds in our sample. For this ESG market portfolio, we found substantial and persistent deviations in industry allocations across our review period. Industries that were the most heav­ily over- or underweighted were quite stable through time. Financial services and technology were the two most heavily overweighted industries, while energy and industrials were the most heavily underweighted in absolute terms.

It was noticeable, however, that these deviations fell markedly over time. The underweighting of energy, for example, dropped from 4.7 percentage points on average in January 2006 to just 0.9 percentage points in December 2020. It seems likely that this was driven, at least in part, by the overall decline in the weight of this industry in the US market over that time.

Next, we looked at the impact of deviations in industry allocation between our ‘ESG market portfolio’ and the US equity market on the total return of the ESG market portfolio. For this, we decomposed return differences into two elements: the ‘industry allocation’ effect and the ‘selection/interaction’ effect. As its name indicates, the industry allocation effect tells us which part of the performance difference is due to differences in industry allocations, while the second effect simply captures the remainder (which could be driven by differences in the way individual stocks are weighted within industries). The left-hand side of the below chart shows both effects on an annual basis².

The effect of industry allocations varied quite significantly from year to year: depending on the time-period, the effect was positive or negative, accounting for the majority of return differences between the ESG market portfolio and the market in some periods and the minority in others.

But we noticed variability not only for the entire industry allocation effect but also for the contribution of individual industries (see right-hand side of the below chart). The 2008 global financial crisis and its policy response showed particularly hefty gyrations. As one might expect, the overweight in financials had a negative impact on relative performance in 2008. This overweight then contributed positively in the following years as markets rebounded. The opposite was true for energy. The underweight in energy had a positive impact on relative performance in 2008 and a negative impact in the following years.

Return attribution by industry.

Decomposition of industry allocation effect.

Past performance is not a reliable indicator of future results.

NOTES: Left side: The figure displays the difference in returns between the industry-adjusted benchmark portfolio and the US equity market (the ”industry allocation effect”) and the difference in returns between the ESG market portfolio and the ESG market portfolio specific industry-adjusted benchmark (the “selection/interaction effect”). Right side: The figure displays the breakdown of the “indus­try allocation effect” into its single-industry specific elements.

Source: Vanguard analysis based on US equity index funds and ETFs that indicated the use of ESG factors in their investment process, according to Morningstar, from 1 January 2006 to 31 December 2020. Analysis based on monthly gross returns in USD. Gross returns used to better isolate ESG risk/return factors.

US equity market was proxied by the Morningstar US Market Index, which measures the performance of US securities and targets a coverage of 97% of the investable market. Industry returns were based on 11 Morningstar industry indices, which, if weighted according to each industry’s portion of the US market’s capitalisation, represent the Morningstar US Market Index.

Industries as a factor

Next, we explored the extent to which the performance of our ESG market portfolio can be explained by variables such as the market factor, style factors such as size, value, profitability and investment, as well as market-neutral industry returns. After controlling for the market and style factors, we found that only a handful of our industry ‘factors’ played a statistically significant role in explaining the ESG market portfolio’s performance.

Among these, energy was one of the most persistent. The relationship between the performance of the energy sector and the performance of our ESG market portfolio was negative. Hence, when the energy sector outperformed the broader market, the ESG market portfolio underperformed and vice versa. This makes intuitive sense given that the energy sector was underweighted in the ESG market portfolio relative to the market.

Variability in individual funds

Having analysed the characteristics of the ESG market portfolio, we next looked at the underlying individual funds. Given the pretty much limitless ways to construct ESG funds, we expected to find quite some dispersion in our findings, and we were not disappointed.

For some industries, the level of ‘disagreement’ about their representation was higher than for others. In line with observations made at the aggregate level, the majority of ESG funds underweighted energy and overweighted technology. For other industries, however, it was less clear. With financials, for example, we found a roughly 50% split in funds under- or overweighting the industry.

In terms of performance, we found similarly diverse results across the various ESG funds. When some funds underweight an industry in a given year and others overweight the same industry in that year, it is unsurprising that these positions have the opposite effect on performance.

That said, and when looking across all industries, the allocations of the median ESG fund tended to converge to that of the market over time, which chimed with the findings at the aggregate level.

Understand the specifics of each fund

So what are investors to take from our findings? Our results revealed that deviations in industry allocations between ESG funds and the broader market did have an impact on relative ESG fund performance. However, the impact is very time-dependent and less systematic than many investors may expect.

If the observed trend of declining industry deviations should persist, it seems reasonable to expect that the magnitude of the impact of industry deviations on relative returns will become smaller.

That said, we did see quite a significant variability at the level of single funds– and trends based on aggregates or medians reveal little when it comes to a specific fund.  Therefore, investors should always assess the investment implications of industry allocations on a fund-by-fund basis.

¹ Research originally published in the Journal of Portfolio Management: Plagge, J.-C. and D. M. Grim. 2020. 'Have Investors Paid a Performance Price? Examining the Behavior of ESG Equity Funds.' JPM Vol 46 Issue 3 Ethical Investing: 123–140. Performance of US index and active equity mutual funds and ETFs that indicate the use of ESG factors, as determined by Morningstar. Time period observed: 1 January 2004 - 31 December 2018. The original research was conducted in 2019 and was then updated with data to the end of 2021.

² We conducted a performance attribution based on Brinson, G., R. Hood, and G. Beebower. 1986. “Determinants of Portfolio Performance.” Financial Analysts Journal 42.