ESG risks in times of Covid-19

ABSTRACT

This study analyzes whether investors take risks related to environmental, social and governance (ESG) factors into account when making portfolio decisions. We exploit the new Morningstar’s ESG risk indicators – introduced at the end of 2019 – to estimate the effect of ESG risk perception on investment fund flows. Our exercise, related to the early phase of the Covid-19 crisis when uncertainty skyrocketed, shows that investors have preferred low-ESG-risk funds, with environmental risks remaining a top concern.

KEYWORDS:

• ESG risks
• sustainability
• Covid-19
• crisis
• mutual funds

JEL CLASSIFICATION:

• G11
• G23

Disclosure statement

No potential conflict of interest was reported by the authors.

Notes

¹ Contemporary to our paper, Pastor and Vorsatz (2020) and Dottling and Kim (2020) also analyse the dynamics of mutual funds during the Covid-19 pandemic. The first one investigates the performance and flows of US actively-managed equity funds finding, among other things, a superior performance of sustainable funds; the second one finds a differentiated investor attitude towards ESG risk, as institutional investors increased inflows to low-ESG-risk assets while retail investors decreased them.

² For example, Hartzmark and Sussman (2019) do not find evidence that sustainable funds performed better than low-sustainability funds in the months that followed the first publication of Morningstar globes in 2016.

³ Globe classes (from 1 to 5) were originally named by Morningstar as Low, Below Average, Average, Above Average and High sustainability. Morningstar switched from ESG scores to ESG risk scores but did not change the name of its globe classes, so the latter are still labelled as low-to-high (sustainability) but now indicate high-to-low ESG risk.

⁴ The advent of ESG risk scores is linked to the debate on the materiality of ESG issues for firm’s performance pioneered by Khan, Serafeim, and Yoon (2016). For details on how Sustainalytics ESG risk scores and Morningstar globes are constructed, see Sustainalytics (2019) and Morningstar (2019), respectively.

⁵ Other fund-specific controls are: time fixed effects, lagged net flows, lagged returns, and lagged cumulated returns since the start of the sample (weekly frequency); the exposure share across 11 industrial sectors and geographical macro-regions, size in terms of total assets, the share of cash and equivalents over total assets in the fund’s portfolio (monthly frequency); fund age; a dummy for exchange-traded funds (ETFs), a dummy for funds restricted to retail investors, a dummy for funds domiciled in offshore centres (fixed effects). The exact definition of each variable is omitted to save space but is available from the authors upon request.

⁶ Standardized coefficients report the effect of one-standard deviation increase in the regressor on the dependent variable.

⁷ In general, ESG scores are expected to remain mostly unchanged during crises as ESG risk is not supposed to be pro- or counter-cyclical: this is confirmed by the fact that between February and March 2020, ESG globes changed for only 5% of the funds under investigation.

⁸ In our estimates the LCD control is generally not significant: one possible explanation is that, since the advent of the new ESG risk scores investors caring of climate risks have started to mainly look at the latter – available at higher frequency – for this purpose.