What this covers

Patrick interviews Katherine Collins, who works on sustainable investing at Putnam, about what ESG data actually is and where it falls short. She treats ESG as an umbrella that catches nearly everything, with bundled vendor data from providers such as Sustainalytics forming only a small slice of the information available. The conversation starts from a three-part account of investor edge (time horizon, insight, analytics) and keeps returning to where edge might remain in sustainability.

The empirical anchor is a George Serafeim study using SASB categories. Equal-weighting all factors produced small, statistically insignificant alpha, while materiality weighting produced potential alpha above 700 basis points in the initial study. Patrick raises the market-cap bias in vendor scores, which may penalize small solutions-focused companies. Collins offers methane as a leverage-point example: it is potent, comes from identifiable sources, can be spotted by infrared imaging, and mostly leaks for economic reasons. People issues get the longest treatment. CEOs name people as their top concern while disclosure lags. A West Coast bank's loan growth came from long-tenured staff, which analysts chalked up to luck. Early Glassdoor edge was quickly exploited, so she suggests splitting responses by seniority and question type.

The last stretch turns philosophical and contentious. Collins argues that early economists set aside wellbeing and climate through ceteris paribus and never came back, that a money-in, money-out utility function lets investors decouple returns from harm, and that Chicago-school finance is treated almost as a religion. She also draws on Scott Page for why cognitive diversity matters most when conditions keep shifting.

Sharpest takeaway

Katherine Collins argues that ESG and impact investing work best when they focus on the few material, relevant issues for each company and are integrated into the core investment process, not run as a side compliance function, and that the field's data remains early and incomplete.

  • Materiality-weighted ESG factors produced large alpha in the Serafeim study, while equal-weighted factors did not
  • Third-party ESG data is structured but incomplete and mismatched to the real questions, so granular fundamental work is needed
  • ESG teams must be embedded in investment teams, since siloed ESG work generates little value

The argument · threads10 threads · 80 claims
0.85

Weighting only industry-material ESG issues beats equal-weighting all factors for returns.

7 pointscentrality 5/5
  • For governance, the most important issue is whether executive incentive pay is aligned with the underlying structure of the business and what drives good long-term performance, which is more forward-looking than board composition metrics.

    “incentive pay and whether incentive pay is aligned with the underlying structure of the business and what makes for good business over time”

  • In the materiality-weighted version of the study, companies that managed their industry's most relevant sustainability issues best, measured against their own peers, substantially outperformed laggards.

    “companies that did the best job managing the relevant issues for their companies massively outpaced the laggards”

  • Among a checklist of around 300 potential ESG factors, only a small handful are truly material for any given company at a given time, where the company has large impact on the world and stewardship strongly affects its long-term potential.

    “if you have a huge checklist of 300 factors that you think are kind of interesting in the world, again context specific for any given company at any given time, there's a small handful of those that are really, really important”

  • A plain-vanilla ESG approach that equal-weights all sustainability factors shows only marginal outperformance that is not robust enough for an investment process.

    “companies that did better in terms of their sustainability performance, marginally better returns, but again, not enough to really hang your hat on over time”

  • The SASB framework maps which sustainability issues are material for each sector, such as safety and emissions for extractive companies and supply chain and wage policy for retailers, and took about ten years to develop with input from investors, executives, and accountants.

    “the whole point of SASB is to help investors focus on the relevant, material issues for any given sector”

  • Putnam's premise is that tending to relevant, material sustainability issues at companies gives a chance for higher returns, and this can produce meaningful long-term outperformance rather than a trade-off against returns.

    “our premise in my work at Putnam is that actually tending to relevant, sustainability-oriented issues at companies gives you a chance for higher returns”

  • In Serafeim's study using SASB materiality, equal-weighting all ESG factors produced only a small, statistically insignificant alpha, while weighting the industry-material issues more heavily produced a potential alpha over 700 basis points in the initial study.

    “When you just simply weighted the issues that everyone had already agreed were more important more heavily in the model, this huge potential alpha jumped out over 700 basis points in his initial study”

0.71

People-related issues are the largest gap between performance drivers and available information.

7 pointscentrality 4/5
  • People-related issues are the biggest gap between what determines long-term performance and the information available to assess it, because CEOs almost always name people as their top concern yet disclosure and analyst questions focus elsewhere.

    “it's people-related, almost 100% of the time it's people-related. Yet you look at the amount of disclosure we have around people”

  • Early Glassdoor rating data showed short-term information value, especially for companies with rapidly growing review sets, but investors quickly exploited that advantage, so the richer longer-term questions are where edge remains.

    “There's some evidence that particularly for companies that are earlier in their presence on Glassdoor who have a rapidly growing data set that there's actually information value there. Pretty short-term in nature and a lot of investors got on that very quickly”

  • Subdividing Glassdoor responses by employee seniority, tenure, location relative to headquarters, and question type, such as CEO respect versus dress code, may reveal longer-term effects that a simple rating does not.

    “does it matter if the responses are from mid-level employees, from brand new employees, or from senior employees?”

  • People-related factors such as culture and talent retention can drive business outcomes like loan growth that analysts miss because they attribute results to luck rather than to how the outcomes occurred.

    “almost all of it is coming from people who have been with them more than 10 years. So if they didn't have this great culture that was also keeping people in their seats, they wouldn't have had this loan growth.”

  • Companies should track simple longitudinal indicators of employee wellbeing, such as surveys and retention, so they can see how changes like new sales incentives disrupt satisfaction and adjust accordingly.

    “every time we change sales incentives, there's a big disruption in our satisfaction, but this is normal”

  • Recruiting data, especially who applies without being solicited, is among the easiest indicators of employee and company health to monitor because it reflects external attraction.

    “I think recruiting is one of the easiest things to monitor because you're getting that sort of front end, who's attracted to your company from the outside.”

  • The S dimension, meaning people and holistic long-term treatment of stakeholders, ultimately determines outcomes on E and G, because a systemic view of a business's impacts makes the whole picture come together.

    “I think in the end, S actually is what ends up determining the E and the G as well.”

0.67

Standard third-party ESG data is narrow, so practice-level evidence must complement it.

6 pointscentrality 4/5
  • Financial data is a narrow stream of information, and no investor believes it alone tells them everything they need to know about what they are investing in, so ESG data adds a more complete view.

    “There's no one who thinks financial data tells you everything you want to know about what you're investing in”

  • When people talk about ESG data, they usually mean bundled third-party data from vendors such as MSCI, Sustainalytics, Thomson Reuters, Carbon Disclosure Project, and Bloomberg, which is only a small piece of the total available information.

    “all that they're really referring to is bundled up third-party data that's available from vendors like MCI or Sustainalytics”

  • Third-party data can show whether a company has fair employment or anti-corruption policies and whether they report to the board, but those only describe the box, while how the practice lives day to day is what matters.

    “all they do is tell you what the box is around these issues. What you really want to see is how the practice is living and breathing every day.”

  • A useful test is whether the data a company reports aligns with how management describes running the business; firms that use the same management system for employee development as for operations reveal that they treat people issues as real business issues rather than box-checking.

    “The fact that they're taking all of their internal S issues... and using the exact same framework... shows you they think this is a real business issue”

  • The biggest opportunity for quant ESG research lies in linking different datasets and combining factors in new ways, because the field is still narrow and mostly determined by a small set of existing metrics.

    “sometimes the insight or the value is in the combination of different factors that you're looking at. This field is just wide, wide open for that.”

  • Principles are more adaptable than metrics, but the best approach combines principles with relevant metrics that evolve over time, because fixed metrics can lock in attention on a small piece of the story.

    “they're inherently more adaptable, they're inherently better designed to begin with”

0.53

Climate solutions investing needs granular fundamental work, since solutions-company data lags.

6 pointscentrality 3/5
  • Katherine observes that solutions-oriented climate investing resonates more with American investors, who react enthusiastically, while European investors respond with more reserved interest.

    “When I talk about this idea to investors in Europe, they're like, 'Oh, interesting.' When I talk about it to Americans, they're like, 'Yeah! Let's fix it.'”

  • Apple's pledge to use only recycled materials by around 2030 is a world-changing commitment that analysts underreacted to, and it may drive changes in recycling incentives and services strategy.

    “Apple two years ago now announced that they not only wanted to make all their products recyclable but that they want to only use recycled materials, no new materials”

  • Data on solutions-oriented companies is younger and harder to analyze, requiring granular fundamental work, such as understanding how a valve or water quality technology actually performs, to determine real impact.

    “the data for solutions-oriented companies really is that granular, fundamental level analysis”

  • Solutions-oriented companies that reduce climate harm for others may score poorly on third-party ESG metrics because they are smaller and lack disclosure teams, creating tension between the data and their real impact.

    “some of those solutions-oriented companies might not look very good on those third-party metrics. They're smaller companies, they don't have a whole team managing disclosures”

  • There are two theories of change for climate in investing: aligning with established global protocols and mapping to companies that follow them, and backing solutions-oriented companies that help others reduce emissions.

    “There are two schools of thought about... They really are like theories of change”

  • Methane is roughly 80 times as damaging to the environment as CO2, comes from a small identifiable set of sources, can be detected by infrared imaging, and has economic incentives since most emissions come from leaks, making it a high-leverage target.

    “methane is somewhere around 80 times as damaging to the environment as CO2 emissions and it's also from a very identifiable, small subset of sources”

0.53

Investing should weigh purposes and harms beyond money-in, money-out financial returns.

6 pointscentrality 3/5
  • Socially responsible investing historically focused on exclusion, asking what investors were against and seceding from engagement with certain companies for moral reasons, unlike impact investing.

    “socially responsible investing is exactly what the name implies. It's, 'I don't want to be involved with certain types of companies for moral or ethical reasons.'”

  • Impact investing pursues explicit financial returns and explicit social or environmental goals side by side, but these can involve a trade-off and sometimes an either-or with financial return.

    “Pretty often, they are two, there's a little bit of a trade off, right?”

  • A core biomimicry principle is to define the function being pursued with great care, and in any real setting that function usually has several layers beyond financial return.

    “always defining the function that you're after with great care. If your function really is just money in money out, you can go down a certain path”

  • Investing with only a money-in, money-out utility function lets investors decouple their returns from the harms their holdings cause, such as destroying a forest or harming children, which people would reject if stated directly.

    “if you said to someone, 'You can get a 12% rate of return but it will destroy the forest behind your own home,' hardly anyone would take that”

  • Arguing that ESG is only about values with no link to the world is misguided, because values and judgments have always affected the world, and the debate usually reduces to bridging values to ongoing business issues.

    “When in the history of time have values and judgments not ended up affecting our world? It's all the same.”

  • Nearly 100 percent of people name an outdoor place as their favorite, yet over 90 percent of time in the US is now spent indoors, an interesting backdrop for thinking about nature.

    “pretty close to 100% will name a place that's outside, and if you look at how we spend our time, it's over 90% now in the US is indoors”

0.45

ESG must be integrated with investment teams and applied across all allocations.

4 pointscentrality 3/5
  • Within the spectrum from socially responsible to impact investing, ESG integration is the large middle category where investors use quantitative and fundamental methods, including hybrids, to apply newer data to financial decisions.

    “The big, huge middle is where it's really exciting, and that's where I'm living all day every day now. This is the ESG integration category.”

  • ESG work generates investment value only when the ESG team and the investment team are integrated; separate ESG teams that meet occasionally rarely produce value.

    “a team in the corner called the ESG team that often does really, really excellent work, but then the investment team is over in the other corner”

  • Signing the UN PRI and similar pledges is largely a fiduciary and governance exercise that is separate from generating alpha, so firms can endorse the PRI without incorporating ESG into investment decisions.

    “That's very separate from the investment-oriented arguments that we just were making.”

  • Treating ESG as a separate allocation or exception to the main portfolio is inconsistent with its potential return value; if it has return potential, it should be considered across every allocation decision as one of many factors.

    “If you actually think this has good return potential over time, you should be considering it across every decision that you make as one of many factors.”

0.45

Diversity must be functionally real and engineered, not measured by counting headcount.

4 pointscentrality 3/5
  • Common ESG metrics such as counting women on boards or in senior executive roles are a first step but are not close to answering the real question of whether a company's teams are functionally diverse and inclusive.

    “those aren't even close to answering the real question that I have over here. They're sort of the first step on a very long path.”

  • Cognitive diversity helps little in repetitive tasks, where less is better, moderately helps in slowly shifting environments, and is essential when the ground is constantly shifting, as in a company navigating continuous change.

    “if you're doing a repetitive task, like a factory, you really don't need cognitive diversity. In fact, the less the better”

  • Scott Page's newer work shows that demographic diversity, including age, income, race, religion, and academic training, is linked to cognitive diversity, though the two were once thought to be possibly unrelated.

    “His newer work actually shows that there is indeed a link between demographic diversity and cognitive diversity.”

  • A defense contractor with a notably diverse board achieved it by treating diversity as an engineering project with defined goals, a process, and regular data reviews, rather than claiming it could not find qualified candidates.

    “They had none of those complaints and the way they described it was completely as an engineering project.”

0.31

Markets fail to price long-term sustainability value because the tools and work are hard.

4 pointscentrality 2/5
  • Early neoclassical economics set aside broader questions with the 'ceteris paribus' assumption, and its early authors intended to return to them, but the field ran narrowly ahead without doing so.

    “we said, 'Ceteris paribus,' and then we shoved all these really important things over in the corner. And we knew we were doing it.”

  • The market fails to price long-term sustainability value because investors trained in rigorous Chicago-school finance treat its tools as a religion, and because the work is hard, depends on context, and yields results that cannot be measured precisely yet.

    “they've taken it to be a religion instead of a set of tools”

  • Water pricing can create economic value for conservation even when the resource is plentiful, because a London water bill of comparable scale to a Boston mortgage versus a $6 Boston bill shows that price, not physical scarcity, drives the economics of water-related businesses.

    “It was as much as my mortgage had been back in Boston and I was just one person like taking an occasional shower.”

  • About 70 to 80 percent of end investors and wealth owners want more action on ESG, but only about 15 percent of intermediaries feel equipped to act on their behalf, creating a large gap.

    “About 70 to 80% of end investors, of wealth owners, are interested in at least more and taking some form of action in this arena and only about 15% of intermediaries are feeling equipped”

0.22

Sustainability is not a zero-sum tradeoff against returns for healthy businesses.

2 pointscentrality 2/5
  • Sustainability-driven edge could appear in both fundamental business growth and equity re-rating, creating an economic flywheel where the same factor re-rates securities over time and re-accelerates underlying businesses.

    “So it's that same economic flywheel that would rerate your securities over time and also re-accelerate the underlying business.”

  • Investor interest in ESG has risen over the last five years across nearly every demographic, especially women and millennials, and the underlying reason is a shift from viewing ESG as a costly zero-sum tradeoff to a systems-thinking view that healthy businesses take care of these issues.

    “it's going from a very zero-sum mindset to a little bit more of a systems thinking mindset”

0.09

Preferred investments should not depend on regulation, though thoughtful policy can help.

2 pointscentrality 1/5
  • Katherine prefers investments that do not depend on regulation or top-down action, though she accepts that some priorities such as human flourishing can be advanced by thoughtful policy when it arrives.

    “I always both as a person and as an investor want to be betting on things that don't rely on regulation”

  • Department of Labor guidance in the US is easily read as highly cautionary and has kept many US institutions hesitant about ESG, while in Europe ESG consideration is almost a requirement.

    “The Department of Labor has given some advice in the US that is easy to interpret as highly cautionary”