A review Sren Hvidkjr Head of Department, Professor of Finance, CBS - PowerPoint PPT Presentation

ESG Investing: A review Søren Hvidkjær Head of Department, Professor of Finance, CBS

Focus of the Report - and a roadmap How do ESG strategies affect investor’s risk - adjusted return? – Theoretical foundation – Classification of CSR/ESG studies – Evidence: • Sin stocks: Negative sector screening • ESG ratings: Positive/negative screening • Event studies • Active ownership • A meta study – Conclusion and implication of the results 2 Søren Hvidkjær – Nordic SIF 19th September 2017

Theoretical considerations for the ESG investor • For an equity investor who – does not possess insider information and – does not engage in active ownership, • the central performance question is not whether ESG initiatives creates value in the firm, but rather whether the stock price accurately reflects the value of these initiatives. Søren Hvidkjær - Nordic SIF 19th September 2017 3

Abnormal Returns to ESG Strategies – why? • Potential drivers of positive return to ESG 1. Underreaction to ESG information • Evidence of underreaction in other cases • Evidence of underreaction for intangible information 2. ESG strategies have become more common • Potential drivers of negative return to ESG 1. Demand effect: If a stock is ignored by many investors, the price falls which implies higher future expected return 2. Companies e.g. in the tobacco industry have incentives to practice conservative accounting Søren Hvidkjær - Nordic SIF 19th September 4 2017

Abnormal Returns to ESG Strategies? • Demand effect depends on – How widespread are ESG strategies – Slope of demand curve for stocks • Depends on arbitrage activity • ESG strategies can lead to a lack of diversification – More relevant for sector exclusions than single- stock exclusions • ESG strategies can be costly – Especially relevant for passive investors Søren Hvidkjær - Nordic SIF 19th September 5 2017

Corporate Governance and Active Ownership • Potential principal-agent problems – Underinvestment in ESG initiatives: • Management has short horizon (myopic) e.g. because of career considerations. – Overinvestment in ESG initiatives: • Management’s social preferences are paid by the shareholders • Management uses stakeholder arguments for their own gain. • Corporate governance initiatives and active ownership can help resolve P-A problems Søren Hvidkjær - Nordic SIF 19th September 6 2017

Studies of Financial Effects of CSR/ESG Classification: 1. Risk-adjusted abnormal returns to ESG portfolios 2. Event studies of market reaction to ESG events 3. Relationship between ESG/CSR and accounting- based performance measures e.g. ROA 4. Relationship between ESG/CSR and the ex ante cost of capital to firms Søren Hvidkjær - Nordic SIF 19th September 7 2017

Evidence of Abnormal Returns for sin stocks: Selected studies Study Period Market Method Alpha p.a. Hong and 1926- USA Equal weighted, FF 1-4 3-4% Kacperczyk (2009, 2006 factor model relative to JFE, ABS 4* F) similar stocks Fabozzi, Ma and 1970- 21 Equal weighted, 1-factor 8-12% Oliphant (2008, 2007 countries JPM, ABS 2 F) Trinks and 1991- 94 Value weighted, FF global 8-13% Scholtens (2017, 2012 countries 4-factor (misspecification) JBusEth, ABS 3 NF) Hoepner and 2001- Stocks Value- and equal weighted, 3-4% Schopohl (2016, 2015 excl. in FF 1 and 4-factor JBusEth, ABS 3 NF) GPFG/AP ABS rating explained p. 10 in the report. High to Low rank: 4*, 4, 3, 2, 1 Søren Hvidkjær - Nordic SIF 19th September 8 2017

The Portfolio: Equal or Value Weighted? Adamsson and Hoepner (2015) criticize Hong and Kacperczyk for equal weighting: – HK regress an equal- weighted portfolio of sin stocks on a value-weighted market benchmark. This implies that the outperformance could be driven by a small cap performance bias rather than sin stocks characteristics […]. This argument is founded on the empirical observation that small stocks outperform large stocks […]. The exceptionally good performance could hence be due to an over-weighting of small cap stocks and underweighting of large cap stocks. Søren Hvidkjær - Nordic SIF 19th September 9 2017

Hoepner-Schopohl Hoepner and Schopohl’s presentation of results: – We conduct a time-series analysis of the performance implications of the exclusion decisions of two leading Nordic investors, Norway’s Government Pension Fund -Global (GPFG) and Sweden’s AP -funds. We find that their portfolios of excluded companies do not generate an abnormal return relative to the funds’ benchmark index. Søren Hvidkjær - Nordic SIF 19th September 10 2017

Hoepner-Schopohl table 3 AP7 AP1 – 4 GPFG Equal Value Equal Value Equal Value (1) (2) (3) (4) (5) (6) Panel A: CAPM model 1-factor 0.450*** 0.286* 0.373 0.346 0.425* 0.364 Alpha (2.63) (1.83) (1.19) (1.13) (1.65) (1.62) Obs. 166 166 109 109 162 162 Adj. R 2 0.83 0.77 0.73 0.39 0.67 0.59 Panel B: four -factor model 4-factor 0.361** 0.204 0.519* 0.345 0.373 0.292 Alpha (2.04) (1.23) (1.71) (1.19) (1.28) (1.17) Adj. R2 0.85 0.78 0.76 0.47 0.67 0.60 � Monthly returns. *, ** and *** indicates significance at 10, 5, and 1% level. Søren Hvidkjær - Nordic SIF 19th September 11 2017

Hoepner-Schopohl Portfolios 70 GPFG 60 AP1 – 4 AP7 50 No of Stocks 40 30 20 10 0 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 Søren Hvidkjær - Nordic SIF 19th September 12 2017

ESG rating and Stock Returns Borgers et al. (2013, JEF, ABS 3 F): Stocks with high ESG rating outperformed stocks with low rating, but not in 2004-9. Halbritter and Dorfleitner (2015, RFE, ABS 1 F/NF) confirm this and find no return differences when extending to 2012, while Larsen (2016, F/I, ABS 0) finds a difference from 2012-2016 in raw returns. Søren Hvidkjær - Nordic SIF 19th September 13 2017

Did the Market Underreact to ESG? • Borgers et al show high returns for stocks with high ESG rating around earnings announcements in 1992-2004, but not in 2004-09 • Consistent with underreaction in the first period followed by correct pricing in the second period. Søren Hvidkjær - Nordic SIF 19th September 14 2017

E, S or G? Environmental Evidence • Limited environmental evidence: – Studies based on ratings from Innovest use very short period – Studies based on KLD show inconsistent results Søren Hvidkjær - Nordic SIF 19th September 15 2017

E, S or G? Social: Employee Satisfaction Study Period Market Data/Method Alpha p.a. Edmans 1984-2009 U.S. large cap Value weighted 2,1-3,5% (2011, JFE “100 best returns subtracted ABS 4* F) companies to from comparable Return work for in stocks (size and difference up America” market/book equity); to 4 years after Then FF 4-factor formation • Underreaction to intangible information: – Realized profit higher than analysts long term estimates – However not abnormal returns around earnings announcement • Kempf-Osthoff (2005) and Statman-Glushkov (2009) report evidence based on KLD scores consistent with Edmans Søren Hvidkjær - Nordic SIF 19th September 16 2017

E, S or G? Governance Evidence • What is corporate governance – shareholder or stakeholder perspective Study Period Market Data/Method Alpha p.a. Bebchuk, Cohen 1990-99 USA G- and E-index Significant: and Wang (2013, equal- and value weighted 6-14% JFE, ABS 4* F) FF 1-4 factor 2000-08 Insignificant Auer (2016, 2004-12 Europa Sustainalytics ratings (<2%) JBusEth, ABS 3 NF) STOXX 600 Sharpe ratios Søren Hvidkjær - Nordic SIF 19th September 17 2017

Did the Market Underreact to G? • Bebchuk-Cohen-Wang show high returns to highly ESG rated stocks around earnings announcements in 1990-99, but not in 2000-08. – Consistent with underreaction in the first period and correct pricing in the second. Søren Hvidkjær - Nordic SIF 19th September 18 2017

Event studies • How does the market react to ESG initiatives/incidents? Study Sample Event return • Fisher-Vanden and 117 events in 1993-2008: EPA CL: -1% event return – • Thorburn (2011, EPA Climate leaders equivalent to $3bn on average • • JEEM, ABS 3 NF) Ceres Ceres: 0% event return • Krüger (2015, JFE, 2,116 events in 2001-07 Neg. events have neg. return • ABS 4* F) from KLD Pos. events have -/0 return • Flammer (2013, 273 WSJ articles in 1980- Eco-harmful: Neg. return • AMJ, ABS 4* NF) 2009 classified as eco- Eco-friendly: Pos. return harmful or eco-friendly Søren Hvidkjær - Nordic SIF 19th September 19 2017

Event Example in Flammer (2013) • WSJ 14/2 1991: Exxon Appears to Be Close to Settling Valdez Suits for Less Than $1 Billion – "It is our opinion that a $650 million settlement is a severe undervaluation of the natural resources damage caused by the oil spill, and that if this deal goes through, the state and federal governments have been severely outdealed by Exxon," said Mr. Miller. • Eco-friendly or eco-harmful news? Søren Hvidkjær - Nordic SIF 19th September 20 2017