» Lire en français: French
Using environmental, social, and governance (ESG) scores of firms belonging to the MSCI World universe, we measure the impact of score-based exclusion on both passive investment and smart beta strategies. We find that exclusion leads to improved scores of otherwise standard portfolios without deterioration of their risk-adjusted performance. Smart beta strategies exhibit a similar pattern, often in a more pronounced way. Moreover, our results demonstrate that exclusion also implies regional and sectoral tilts as well as (possibly undesirable) risk exposures of the portfolios.
5 min read
is a professor of Finance. He works in the area of quantitative investments and sustainable finance.
is a professor of Finance. His main research interests are financial econometrics, the modelling of asset prices, portfolio allocation, and systemic risk.Socially responsible investing (SRI) is an investment approach that incorporates environmental, social, and governance (ESG) factors This approach is increasingly important in the asset management world. In 2018 according to the Global Sustainable Investment Alliance, some $30.7 trillion was invested on the basis of a sustainable investment strategy. That was an increase of 34% in just two years, with Europe leading the way.
At the same time as SRI approaches have been on the increase, traditional, actively managed funds, while still popular, have lost some ground to alternative strategies: on the one hand, passively managed investment consists in investing in firms according their market capitalization; on the other hand, smart beta strategies allow portfolios to be exposed to some investment styles, while avoiding weights to be related to the market capitalization. According to the Boston Consulting Group, of the $79.2 trillion assets under management in 2017 some $16 trillion was passively managed. And while smart beta strategies only attracted some $430 billion, this figure has grown by 30% annually since 2012.
The growth in SRI alongside the increasing use of passive and smart beta strategies raises an important issue: whether integrating ESG factors into an investment approach aimed at improving the ESG score of the portfolio, has an adverse impact on the performance of these strategies. In particular, with passive strategies, whether straying from a weighting focused on market capitalization increases the level of tracking errors with respect to the benchmark. Or if the integration of an ESG filter with a smart beta strategy reduces the efficiency of whatever factor is being targeted with the smart beta strategy.
This is the issue investigated by Fabio Alessandrini and Eric Jondeau in their paper ESG Investing: From Sin Stocks to Smart Beta. To test the effect of integrating ESG strategies Alessandrini and Jondeau used ESG scores for a large set of firms that feature in both the MSCI ESG database, and also the MSCI All Countries World Index (ACWI). This index includes about 2,500 large- and mid-cap stocks across 23 developed and 24 emerging markets and the research covered the period from January 2007 to December 2018. In addition to the world market, they also considered four regional zones – the United States, Europe, Pacific, and emerging countries, – and used the corresponding MSCI index as a regional benchmark portfolio.
Two methods of screening for ESG were investigated. A popular method is to exclude from a portfolio a percentage of firms with the lowest E, S, and G scores (negative filter). A drawback of this negative exclusion approach is that it tends to lead to an underweighting of certain sectors, especially the so-called sin industries. In doing so, it may exclude some firms within the less favored industries that are making good ESG progress. Consequently, a second screening method was investigated which adopted a positive best-in-class approach and used the industry-adjusted average ESG score. This approach avoids part of the sector bias described above.
The authors looked at the effect of applying these two types of ESG filter on the overall ESG score of the portfolio as well as geographical and industry weightings. They also looked at whether applying these ESG filters exposed a portfolio to undesirable risk factor exposures.
ESG filtering produced a substantial improvement in the industry-adjusted average ESG score, consistently across regions and did so without worsening the risk-adjusted performance. Notably, a more aggressive ESG screening (i.e., when 50% of firms with the lowest ESG scores are excluded) resulted in an increase by approximately 30% in the average ESG score of the portfolio. In addition, the risk-adjusted performance (Sharpe ratio) increases on average in all regions.
There was, however, an impact on the regional make-up of the world portfolio. In particular, the ESG filter led to an overweighting in firms from Europe and the Pacific and an underweighting in firms from the US and emerging countries. Looking at the effect on industry weightings, ESG portfolios have a relatively lower weight in some sectors: for instance, the world portfolio holds close to 2.5% of financial firms below the benchmark (16% instead of 18.5%). In contrast, some other sectors benefit from overweighting, such in information technology, with on average a weight 3.6% over benchmark (15% instead of 11.4%).
The results also show that the screening criterion deployed (negative exclusion approach or best-in-class approach) makes a difference in terms of performance. Excluding firms with low scores for the environmental criterion improves the portfolio performance in most regions, but has a limited impact for the social and governance criteria.
The authors also looked at the exposure of portfolios based on ESG screening to standard risk factors, such as the sensitivity to the market, size, value, profitability and momentum, as featured in the well-known Fama and French asset pricing models. The evidence showed that ESG screening might lead to style biases – in particular, exposure to large, relatively expansive, and profitable companies. However, once the exposure to these factors is taken into account, there was no evidence that this led to a negative return relative to the benchmarked index.
As well as looking at the impact of ESG screening on various aspects of a portfolio adopting a passive strategy, the authors also considered the impact of ESG screening with smart beta investing. Smart beta is a strategy where an investor builds a portfolio that passively tracks an index weighted not on market capitalization, but some other factor. A portfolio might, for example, be constructed with a focus on size, low beta, high dividend, value, quality, or momentum. These were the strategies that the authors investigated.
Research showed that findings for passive strategies generally hold and are sometimes even accentuated for smart beta strategies. ESG scores are improved quite considerably and for all regions. Nor do Increases in average ESG scores adversely affect risk-adjusted performance, as measured by the Sharpe ratio. Applying the high dividend yield strategy in the US, for example, leads to an increase in the annualized return from 7.7% with no exclusion to 8.5% with 50% exclusion, with volatility decreasing at the same time, and the Sharpe ratio increasing by 20%. Indeed, Sharpe ratios are either unchanged or higher in the vast majority of strategies and regions, with the exception of some strategies in emerging countries.
Overall the authors show that applying SRI strategies to increasingly popular investment approaches, such as passive investing and smart beta strategies, was possible without adversely affecting either the performance relative to the benchmark or the fundamental philosophy of the investment approach over the period covered. Indeed, using a more general and systematic ESG screening method it was possible to substantially improve the ESG profile of passive and smart beta portfolios without reducing risk-adjusted returns.
With passive investment, this may come at the cost of overweighting in particular regions, in favor of Europe and against the US and emerging countries, with multicounty portfolios. It may also mean making significant sectoral bets for information technology stocks and against financial stocks. While with smart beta, even aggressive filtering out of lower ESG scores still leaves targeted factors in place, albeit with some possible reduction in exposure to those factors.
Related research paper: ESG Investing: From Sin Stocks to Smart Beta, Fabio Alessandrini and Eric Jondeau, .
Featured image by ID 122406398 © Thatsaphon Saengnarongrat | Dreamstime.com