Week 7 - Measurement difficulties with ESG data, Greenwashing Flashcards
Issue 1: What are the “right” metrics to use?
For traditional, ‘shareholder-centric’ investing, we have a single metric ($$).
What is the dimensionality problem when it comes to sustainable/ESG investing?
eg. Tesla, Amazon
- ESG comprises many factors
eg. carbon emissions, labour practices, customer-centric behaviour, good governance etc. - (Investors?) care about many things and are unsure how to WEIGH/TRADE them
- diff. ppl have diff. weights & diff. aspects can be hard to make direct comparison
eg. Tesla is well-known for producing green cars (very high scores in these aspects) but also known for BAD LABOUR PRACTICES like workplace safety issues, wage staff, underpayment of employee (very low scores in other aspects)
eg. Amazon is known for 2 things
- high burnout rate
- paying workers much more to compensate
Rating Amazon will require a decision on which aspect to care about more and put more weight on
Whose preferences - Consideration about single-issue investors as ‘ESG’ investors vs Paying a minimum amount of attention to other factors?
Should you invest in single-issue funds or Green funds that consider a range of issues eg. labour practices, supply chain?
- diff. ppl have diff. DEFINITIONS of ESG
- depends on individual PREFERENCES
Do we care more about a company
(i) doing good things, or
(ii) not doing bad things?
Essentially, how much should we focus on PROSOCIAL FACTORS
eg. well-paid employees, generous corporate charitable support
vs A LACK OF ANTISOCIAL FACTORS
eg. labour violations, illegal pollution
-
Trade-off: ‘doing good things’ covers more dimensions & can LEARN MORE about the co, but HARD TO MEASURE and hard to see all the data!
eg. don’t know how much of the company’s pay(?) is going to tax or paying employees minimum wage - ‘not doing bad things’ is EASIER TO MEASURE but don’t learn as much about the co.
- In practice, most ESG measures capture “doing no harm” rather than “doing good”
2 methods to measure how firms treat diff. stakeholders “S”
- PROXIES for STAKEHOLDER-CENTRIC BEHAVIOUR
eg. product safety violations, gender/racial pay gaps
- narrow but easier to measure - PROXIES for ‘corporate culture’ more generally
- broader but harder to measure
- a common approach similar to this 2nd method is ESG scores
ESG scores, where do commercial vendors get their data from?
- Calculated by various third parties to rank/rate companies on their ESG performance
- heavily used by investors to screen companies on social responsibility grounds - Ratings agencies employ research analysts to collect data points about the co’s behaviour & mapped into an overall score»_space; using ALL AVAILABLE DATA FROM ALL PLACES, no direct interaction with co.
eg. from annual reports, co. website, NGO websites, stock exchange filings, CSR reports, news sources
Pro & 3 cons of ESG scores
eg. Boohoo
Pro
1. ESG scores are mapped from a lot of DATA that would be impractical for most investors (esp. individuals) to collect about every single potential investment
- thus, a STARTING POINT for those interested in understanding co.s’ social responsibility / sustainability
Cons
1. Placing an enormous amt of FAITH into RATING AGENCIES. whose models are a black box & whose ratings don’t correlate well
» more variation in ESG scores among diff. agencies compared to variation in Credit ratings
eg. Boohoo was included in most UK ESG funds and given a high ESG score bc it had a lot of domestic production -> ESG scores tend to look at the first level but nobody checks what really goes on!
- ESG scores seem to fixate on QUANTITY of disclosure but not quality
-
DIMENSIONALITY PROBLEM
- unlike credit scores which measure a single quantity and everyone can agree on the meaning…
» …we don’t know what ratings agencies are EXACTLY MEASURING for ESG scores & how we QUANTIFY the measures (if at all possible) (& diff. agencies place diff. weights on diff. issues)
» so, hard to hold agencies ACCOUNTABLE
- what metric do we use to assess the QUALITY of ratings?
Issue 2: ESG data quality
ESTIMATED emissions data - implications
Comparison to financial reporting,
(i) subject to mandatory disclosure (no self-selection)
(ii) checked by auditors and regulators
The primary metric of the ‘E’ measure is CO2 emissions but a potential issue is that…
- most carbon emissions data is ESTIMATED by VENDORS, rather than actually disclosed by firms
- This data is estimated by an implicit ASSUMPTION of WITHIN-INDUSTRY HOMOGENEITY {bc the private vendors cannot see actual differences in eg. production processes}
- the data are generally CORRELATED to co.s’ FINANCIAL fundamentals, eg. net sales, PPE, financial ratios - Several prominent studies claim that carbon emissions are priced in the stock market.
- If estimated emissions are just DERIVED from FINANCIAL RATIOS, stock market performance doesn’t actually show whether investors care about/respond to EMISSIONS
- the data wouldn’t be helpful bc they might just be reflective of the link between financial fundamentals and stock market performance
Issue 3: Does MANDATORY DISCLOSURE get us the data required?
eg. for ‘S’ disclosure, the UK gender pay gap mandate
eg. EU CSR Directive, EU Sustainable Finance Disclosures Regulation (SFDR)
- potential drawback: still leaves a lot of DISCRETION to companies about what METRICS TO DISCLOSE
- regarding the UK gender pay gap disclosure law, proponents argue that the law & “regulation by shaming” could bring about real change…
- …but SUCCESS of a law depends on its enforcement
- the UK’s Equality & Human Rights Commission (EHRC) was criticised for WEAK ENFORCEMENT -> admitted that they do NOT FACT-CHECK the data & care more about whether firms report on time rather than whether firms’ figures are accurate - light-touch enforcement has led to firms MISREPORTING their gender pay gaps
Verification of Mandatory disclosures
3rd party certification can help depending on…
- whether sth that is certified is actually QUANTIFIABLE
- who picks what’s being certified & who the certifiers are
eg. Morningstar’s Low Carbon designation, Great Places to Work awards, Fair Trade
Morningstar’s Low Carbon designation
- given to investment portfolios w/ low carbon risk & empirically, the low carbon funds do actually contain stocks w/ lower emissions
» good compared to ESG scores bc we actually know what we’re measuring
Great Places to Work awards
1. co.s have to SPEND MONEY to APPLY to GPWI to send out the employee surveys for measuring EMPLOYEE SATISFACTION, in order to be scored (SELF-SELECTION & ‘satisfaction’ can be difficult to measure)
- but firms that don’t feel the need to pay for the survey might still have happy employees
2. while GPWI members seem to earn higher stock returns, is this correlation or causation?
Fair Trade certification
- products (coffee, chocolate, tea etc.) are fair trade if farmers are FAIRLY COMPENSATED
- Pro: the QUANTITY being certified is CLEAR, ie. more likely that these FT-certified products is actually sourced from fairly-paid farmers
» simple & understandable for investors
- Con: firms must PAY TO GET CERTIFIED
> which firms can afford to pay? SELECTION BIAS could turn the cert. into a marketing gimmick
» also might constrain investment set bc limited set of products can be FT certified
3 issues with ESG/CSR audits
- VOLUNTARY (set to change in the EU under the CSRD)
- lead to a SELECTION ISSUE; firms can PICK what gets audited - MINIMAL REGULATION
- in contrast to fin. stt. audits which are heavily regulated & do meaningfully affect acct. quality - SKILLSET MISMATCH by auditors
- many ESG/CSR audits are overseen by auditors whose backgrounds are in FINANCIAL AUDIT instead
2 requirements to overcome the DIMENSIONALITY problem of ESG measurement
- Well-defined, UNDERSTANDABLE measures, even if sometimes narrow in scope
eg. carbon emissions, gender pay gap - VERIFICATION of these measures
Greenwashing
= Claiming to engage in ‘green’ or ‘sustainable’ practices while not actually doing so
» claims that are not outright false but presented in a MISLEADING way
eg. McDonald’s straws were made of recycled materials which might have been misleading that the straws can be recycled
eg. Tesco’s biodegradable teabags are strictly speaking biodegradable in an industrial facility, not as regular consumers would assume
Investors can
1. fall victim to firms’ claims of greenwashing and/or
2. engage in greenwashing of their own
3 groups that firms greenwash for
- CONSUMERS who may want to buy ‘green’ products
- INVESTORS who may want to (or at least give appearance of making) make ‘green’ investments
- REGULATORS/NGOs
- seen as watchdogs
- firms want to pre-empt the pressure of regulation to make changes
eg. firms show smaller pay gap to avoid being called out and forced to reduce
4 ways that firms greenwash
1. Vague terminology or pledges - 2 ways
- Vague, non-specific to improve FUTURE behaviour
» targeting REGULATORS & INVESTORS
eg. Business Roundtable (a large US lobbying group) pledged to serve ALL stakeholders rather than just shareholders, but no commitments to SPECIFIC actions given - Vague promises about the PRESENT firm performance
» more often targeting CONSUMERS
eg. Quorn Foods’ Thai Wonder Grains claims that the product “will help us reduce our carbon footprint” but Quorn was actually referring about its OWN footprint rather than the society’s, as most ppl would naturally imply
» the UK’s ASA banned the advertisement for being too vague
What is the downside of setting net zero by 2050 pledges?
What are the drawbacks of setting Interim targets?
Net zero pledges: MORAL HAZARD
- the target is more than 25 years from now; current management might be tempted to sign up and THEN DO NOTHING since they won’t be in place in 2050
Interim targets
1. setting AMBITIOUS long-term targets that will not be the mgt’s problem anymore in the future
2. vs setting CONSERVATIVE, overly precise short-term targets might slow down own progress of going green (bc don’t want to miss target)
3. impose punishments for not meeting targets or not?
» if don’t punish firms, firms have NO INCENTIVE to meet them
» if do punish, firms might set very cautious & conservative targets that don’t lead to meaningful reduction in CO2 emissions
