Results from Europe

Page 90 of 146

Taken together, our results from the multiple regression analyses and Owen-Shapley decomposition suggest that sector fixed effects and classic market-based determinants of returns are the biggest contributors in explaining excess returns during the specified period. By contrast, ESG, E, S and G does not significantly contribute to the explanation of returns, meaning that we fail to reject the null hypothesis for the aggregated ESG score and all the isolated pillars. Secondly, and as shown by the Owen Shapely decomposition, we can conclude that ESG, E, S and G performance in Oceania does not meaningfully contribute to the explanation of returns during the pandemic crisis.

Page 91 of 146

most of traditional accounting-based measures of the company’s financial performance are significant determinants of a company’s share price resilience during the partly exogenous shock.

It is worth mentioning that sectors like Consumer Discretionary, Energy, Health Care, Industrials, Real Estate, and Utilities present statistically significant coefficients on acceptable levels. We observe that Consumer Discretionary, Energy, Industrials and Real Estate is negatively associated with returns while, not surprisingly, Healthcare and Utilities present positive coefficients, indicating that companies in these sectors performed relatively better during the crisis period.

Similar to that of Oceania, the aggregated ESG score in column (4) is not statistically significant in explaining market crisis returns in the first quarter of 2020. As such, we again fail to reject our null hypothesis.

In Appendix (24), we analyze the isolated ESG pillars for the European region. For the Environmental pillar, we find a positive but insignificant coefficient, meaning that we also fail to reject the null hypothesis. Additionally, and similar to the aggregated ESG score analysis, we also find that MKT and DE is significant at a 0.1% level. Notably, the company’s profitability (ROA) and its market capitalization (size) are positive and significant determinants of returns during the crisis period.

We observe significantly different result for the Social and Governance specific analyses.

Specifically, we observe that the Social pillar is positive and statistically significant at the 5% level.

As such, we reject our null hypothesis and conclude that the Social factor is an important and positive indicator of a company’s share price resilience during the outbreak period. In line with Forbes (2020), we hypothesize that the significant result for the social pillar is caused by COVID-19 related social issues which became more material. These issues include but is not limited to employee health and safety, labor practices, product quality and safety and access and affordability.

Page 92 of 146

Table 20: COVID-19 January to March outbreak period within sample regressions for Europe Table 20 show the results from regressing the buy-and-hold excess return (BHARQ1) on our independent variables.

In Column (1) we regress BHARQ1 on Refinitiv Eikons ESG Score and Sector. In Column (2) we add market factors, risk, and return related variables. In Column (3) we add Size and in Column (4) we add company-specific accounting variables. All variables except BHARQ1 and ESG are winsorized at the 2% and 98% levels. All variables

are defined in Appendix (18).

We find opposite results for the Governance specific analysis. In Appendix (24), we find that the G pillar is negatively related to BHARQ1 and statistically significant at the 5% level. As such, we reject the null hypothesis and conclude that the Governance pillar is a significant but negative indicator of a company’s share price resilience during the outbreak period. We present the results from the Owen-Shapley decomposition of R-square for the European region in Figure 14.

Page 93 of 146

Figure 14: Owen-Shapley R² Decomposition analysis outbreak period for Europe

Figure 14 represents the Owen-Shapley R² decomposition analysis during the outbreak period for the European region. ESG consist of: Aggregated ESG score. Stocks risk, return and factor loadings consist of: MKT, HML, SMB, MOM, Momentum and IdioRisk. Sector consists of: All 11 sectors. Company financial consist of: Size, ROA,

DE, DPR, Debt, Earnings and FCF.

Table 20 reports that our most complete regression model (4) explains 38.74% of the cross-sectional variation in the COVID-19 pandemic period returns for the observations in our sample.

Figure 14 presents a pie chart illustrating the proportion of the 38.74% that is explained by each group of variables. Similar to Oceania, we observe that Sector contributes the most to the overall R², with about 44.4% of the explained variation being credited to this variable. Stock’s risk, return and factor loadings are now a close second, accounting for 40.6% of the explained variation.

Company financials account for 14.6% of the explained variation in stock returns, while notably, the ESG group is the least important category, contributing just 0.39% of the overall explained variation in returns during the COVID-19 downturn period in Q1 of 2020.

For the isolated pillar specific analyses, we find similar results despite of S and G being statistically significant at a 5% level (Figure 16). Specifically, we find that E contributes 0.29%, S contributes 1.62% and G contributes 1.32%. Taken together, our multiple regression analyses and Owen-Shapley decomposition again suggest that sector fixed effects and classic market-based determinants of returns are the biggest contributors in explaining excess returns during the specified period. Firstly, due to the insignificant ESG and E coefficients, we fail to reject the claim that ESG or E is a significant share price resilience factor during the first quarter of COVID-19 pandemic. Secondly, we observe that S is positively associated with returns during the outbreak period while G is negatively associated. We conclude that both findings are of economical and

Page 94 of 146

statistically significant relevance and thus reject the null hypothesis for both pillars. Finally, and in line with that of Oceania, the Owen Shapely decomposition provides robust evidence that ESG, E, S and G performance in Europe does not meaningfully contribute to the explanation of returns during the pandemic crisis.