“Responsible investing: The ESG-efficient frontier” Pedersen, H. Lasse, Shaun Fitzgibbons, and Lukasz Pomorski Flashcards
What is the main idea?
Develop an extension of CAPM for ESG and test it empirically.
Propose a theory in which each stock’s environmental, social, and governance (ESG) score plays two roles:
(1) Providing information about firm fundamentals and
(2) Affecting investor preferences
What is the motivation of this paper?
Asset owners and portfolio managers seek to incorporate ESG factors into their decisions.
There is no theoretical framework on how to do that.
Thus, this paper tries to determine theoretically and test empirically if ESG improves or hurts performance.
What is the methodology used for the theoretical part?
Construct an ESG-SR frontier: maximum Sharpe ratio (SR) at each level of ESG
Consider 3 types of investors:
o Type-U – ESG unaware (seek to max their return per unit of risk)
o Type-A – ESG-aware (have same goal, but use assets’ ESG scores to re-evaluate risk and return)
o Type-M – ESG-motivated (ready to sacrifice performance for high ESG, but still want maximum Sharpe ratio at each ESG level)
What are the choices of those 3 types of investors?
◦ U – tangency portfolio ignoring ESG information
◦ A – tangency portfolio using ESG information
◦ M – a portfolio on ESG-efficient frontier
How does ESG predict future returns when each type of investor is dominant?
◦ When type-U investors prevail and ESG predicts high future profits, high ESG delivers high expected returns (high-ESG stocks are undervalued because type-U ignore useful ESG information)
◦ When type-A prevail, prices of high-ESG stock are bid up, eliminating the premium
◦ When type-M prevail, high-ESG stocks deliver low expected returns (type-M accept lower return for more ESG; high-ESG stocks are overpriced)
When is ESG a positive return predictor?
ESG is a positive return predictor if ESG is a positive predictor of future firm profits and the value of ESG is not fully priced in the market.
(Important finding, bridges the gap between papers arguing that ESG hurts performance and those finding the opposite result).
What is the methodology used for the empirical part?
- Pick measures for ESG and its components
- Construct ESG-efficient frontier for E and G (impossible for S as it is binary, and ESG is similar to E, so they skip it)
- Make some important conclusions
How do authors determine the measures for ESG and its components?
E – carbon intensity
S – sin stock indicator
G – we compute how (un)aggressive a company is in its accounting choices based on the accruals in the financial statements
ESG – aggregate ESG score produced by MSCI, a leading provider of ESG ratings
They then estimate the ESG-SR frontier with those proxies
What are the conclusions about the significance of the measures for explaining returns?
E doesn’t have stat. sign. link to returns
S predicts returns negatively
G predicts future profitability and future returns positively (strongest predictor)
ESG doesn’t have stat. sign. link to returns
Screening (filtering out good ESG stocks and ignoring the rest) hurts SR because it is then impossible to short low-ESG stocks
Through which channels can ESG impact returns?
◦ If ESG correlates with future profitability, then ESG can be used to pick ”better” stocks that will have higher return; this is a positive effect (return premium)
◦ If ESG correlates with investor demand, then high-ESG stocks are sought by many investors, which pushes their valuation up and expected return down; this is a negative effect (return discount)
What do portfolios on the ESG-SR frontier consist of?
These portfolios are combinations of the risk-free asset, the tangency portfolio, the minimum-variance portfolio, and the so-called ESG- tangency portfolio (4-fund separation)