Are institutional investors more concerned about climate change than retail investors?

This paper examines the role of institutional investors in responding to climate change concerns and whether they are more reactive than retail investors. By analyzing S&P 500 data from 2010 to 2021, the study explores the relationship between stock returns, emissions intensity, and unexpected climate concern captured through media analysis. The research categorizes firms into green, brown, and neutral portfolios and distinguishes the impact based on ownership levels. Results indicate that institutional investors were more likely to penalize high-emission firms, particularly from 2010 to 2018, though this trend weakened after 2019. The findings contribute to the debate on how investor behavior drives climate-related financial outcomes and suggest further avenues for research.

Reasoning behind the paper

This research was inspired by a gap in the literature regarding the role of institutional investors in driving the reduction of greenhouse gas (GHG) emissions. Understanding how these investors perceive and respond to climate risks in their investment strategies is essential for lawmakers and regulators in their efforts to combat climate change. Remarkably, existing studies present conflicting evidence on whether institutional or retail investors are more concerned about climate-related risks. This research aims to address this gap by providing a definitive answer to the question: "Are institutional investors more concerned about climate change than retail investors?"

The research of Ardia, Bluteau, Boudt, and Inghelbrecht (2022) suggests that green assets are likely to outperform brown assets during periods of heightened climate concern. The aim of our paper is to explore whether these findings vary depending on the type of investor.

Data

The study utilized data from S&P 500 companies spanning the period from 2010 to 2021. Data was gathered on the stock’s total return index, their emissions intensity, and levels of institutional ownership. To capture the unexpected component of climate change concerns, the Media Climate Change Concerns (MCCC) index developed by Ardia, Bluteau, Boudt, and Inghelbrecht (2022) was used. This index, based on textual analysis, aggregates climate concern from news articles over time. An autoregressive time series regression model was estimated on the MCCC index, with the prediction error serving as a proxy for unexpected media climate change concerns (UMC). The Fama and French five-factor model factors were used as control variables to ensure robustness. The analysis was also extended by incorporating industry-specific data and by comparing concerns based on transition risk and physical risk separately.

Used methods

The first step was to categorise stocks on a daily basis into Green, Brown, or Neutral based on emission intensity. Firms below the 25th percentile were labelled Green, those above the 75th percentile as Brown, and the remaining as Neutral. An equally weighted Green-Minus-Brown (GMB) portfolio was created by taking long positions in green firms and short positions in brown ones. Additionally, a distinction was made by institutional ownership levels of each firm, creating High GMB (high ownership) and Low GMB (low ownership) portfolios. To capture unexpected climate change concerns the prediction error of an autoregressive model on the MCCC index was used as the UMC (unexpected media concern)  variable.

A conditional means analysis was then used to observe how GMB portfolio returns relate to UMC, and High and Low GMB portfolios were compared to identify differences based on institutional ownership. A multivariate linear regression framework was also used to analyse returns of the different portfolios over the full period and sub-periods (2010-2018 and 2019-2021). The main  analysis employs a firm fixed-effect panel regression model on the level of individual stocks. The main focus is on the interaction terms to evaluate how ownership influences the impact of unexpected climate concerns on stock returns. The main analysis is also further expanded by comparing firms on an industry level and by comparing reactions to physical and transition related risks.

Results

The conditional means analysis shows that the Green-Minus-Brown (GMB) portfolio with the highest institutional ownership outperforms the one with the lowest ownership during periods of elevated climate concern, indicating that institutional investors respond more strongly to unexpected climate concerns than retail investors.

The multivariate factor analysis reveals that from 2010 to 2018, institutional investors were more concerned about climate change, with high-ownership GMB portfolios performing better under heightened climate concerns. Meaning that they penalize brown firms more than retail investors. However, from 2019 to 2021, this trend reversed, with higher climate concerns correlating with lower returns across both high and low GMB portfolios, suggesting a general decline in investor responsiveness to climate change, possibly due to increased transparency or a genuine decrease in concern.

At the individual stock level, the analysis confirms that climate concern remains a factor for investors, though its impact weakens after 2018. The strong influence of institutional investors, particularly during 2010 to 2018, supports our hypothesis that institutional investors are more concerned about climate change than retail investors. The most crucial finding is the triple interaction term (IGHG*UMC*ownership), which is both negative and statistically significant. This indicates that for firms with higher institutional ownership, the negative impact of high emissions on stock returns is intensified when unexpected climate concerns arise. In other words, firms with high emissions face greater penalties in stock returns during periods of heightened climate concern, particularly when they have a higher level of institutional ownership. The introduction of the triple interaction term also results in a positive shift in the lGHG*UMC interaction term from negative to positive. This suggests that the effects observed by Ardia, Bluteau, Boudt, and Inghelbrecht (2022) are largely driven by the behaviour of institutional investors.

Conclusions & further research

This study empirically demonstrates that institutional investors exhibit greater responsiveness to climate change news compared to retail investors. The results indicate that from 2010 to 2018, institutional investors were penalising high-emission firms more significantly. However, this concern appears to have diminished from 2019 to 2021.

This paper is meant to lay the groundwork for further research on this topic, therefore several suggestions based on existing literature are proposed that could be interesting avenues for future research. First, incorporating an economic condition factor, based on long-term P/E ratios, to control for the impact of economic conditions on stock returns. This is based on insights from literature showing different performance patterns for SRI stocks during varying economic conditions. Second, adding an emission reduction target factor could help assess whether institutional investors push for more substantial climate actions. Third, including a factor for geographical retail ownership to better understand how local environmental changes influence retail investor behaviour. Lastly, expanding the analysis beyond US firms to a global dataset would test the consistency of these findings. It is also important to note that this research specifically focuses on the unexpected component of climate change concern and does not address other aspects of climate concern.