Can machine learning bring ESG investing back on course?

ESG investing appears to be straying from its path but can machine learning help set it back on its course?

The terms ‘pivotal moment’ and ‘turning point’ are overused in the sustainable investment world. Last week, they were earned. Two events forced the industry to do some serious and sobering self-reflection after a year of self-congratulation.

First, Danone CEO Emmanuel Faber, a global leader in corporate purpose and responsibility, was ousted by activist shareholders—themselves PRI signatories.

Then, ex-CIO of Sustainable Investing for BlackRock, Tariq Fancy, penned a no-punches-pulled op-ed deriding Wall Street’s embrace of ESG as “greenwashing the economic system.”

Sustainable investing, he argues, “boils down to marketing hype, PR spin and disingenuous promises.”

To evidence his case, Fancy points to the many instances where existing mutual funds are cynically rebranded as ‘green’, “with no discernible change to the fund itself or its underlying strategies.” In other cases, he adds, ESG products contain fossil fuel companies and other large polluters, such as fast fashion manufacturers.

ESG, he concludes, only matters to investors to the extent to which it’s profitable.

Fancy’s solution to the world’s greatest environmental and social issues lies instead with politicians and regulators, on whom he calls to take the helm. A free market can’t fix the big issues—or itself.

To a point, we agree. Yes, investors are driven by profit. And yes, regulators have an important role to play in building the economic framework of the future.

But investor activity has not detracted from the debate. The snowballing pressure on companies to ‘do better’ has driven change at a company level and shaped the conversation among regulators.

We also believe that, for real change to happen, everyone—from consumers to asset owners to buy side to sell side to regulators—needs to take a seat at the table.

The question is, what do investors need to level up?

Fancy is absolutely right about greenwashing. At the beginning of this year, pundits were quick to broadcast the fact that flows into European sustainable funds had hit €1.1trn in 2020. They were quieter about the broader context. It wasn’t just down to new flows: demand for ESG products prompted asset managers to change the strategy or investment profile of portfolios. A total of 253 funds rebranded in 2020, relative to 505 new fund launches.

Did the underlying holdings change? In many cases, no. With little consensus about what ESG should include, it’s anyone’s guess as to what a sustainable fund constitutes in 2021.

Fancy is also right about the popularity of ostensibly irresponsible holdings among responsible funds. It’s not just oil and gas. Take Vanguard’s FTSE Social Index Fund Institutional Shares (VFTNX), which is 25% FAANGS and Tesla. Many so-called sustainable ETFs and indexes are essentially overweight Silicon Valley. Big Tech is not immune to, and has come under fire for, dubious data privacy, ethics and human labour issues.

But there’s another, spherical-shaped problem with this rudimentary approach. The trend towards ‘ESG’, and corresponding piling into tech and renewable stocks, is arguably creating a bubble. As per Investors Chronicle, shares in Tesla are up fivefold in a year. It's trading on a “mind-boggling” valuation of 109x 2022 earnings. The 50 most popular ESG stocks trade at a 33% premium to the market.

In short, the ESG market is effectively crude over-simplification at a multi-trillion dollar scale.

What does that say about investing?

Not, perhaps, that investors are duplicitous, but that they don’t have the right tools. Current assessments aren’t fit for purpose. Right now, the best-practice approach boils down to ‘tech and renewables good, oil and gas bad’, based in no small part to public perceptions and tick-box exercises.

It doesn’t have to be that way.

If extra-financial analysis is to have the same rigour and sit alongside financial analysis, the information simply needs to be better. Like financial data, it must be uniform, transparent and objective, universally applicable.

The rewards are immense. And the technology is here.

To date, the messy, nuanced, tangled universe of causality has remained unexplored because the means of assessing it simply didn’t exist. But technology, particularly big data subjected to machine learning, can begin to make sense of it all.

Analytics like Util open a door to investors trying to evaluate a company, fund or sector as a sum of all of its parts, or looking to measure all performance. It gives consumers the insight to understand the impact of their capital. And it gives regulators, who are, as Fancy says, critical to this process, the information they need to build the framework of the future.

There was another interesting bit of news this week.

The first plans for life on Mars have just been set into motion. According to projections, by 2054, work on Mars’ first inhabitable city will begin. By 2100, it will welcome humans.

That coincides almost perfectly with two dates with which climate-minded investors will be familiar: achieving net-zero targets by 2050 and keeping global warming below 1.5 degrees above pre-industrial levels by 2100.

Perhaps if we reframed the debate as what investors can achieve with the right horsepower, rather than what they can’t with the tools at hand, we could tackle ESG issues with as much fervour as scientists trying to launch civilisation far from the only planet on which we can rely.

Sign up
Sign up to our Week in Impact newsletter to learn more about what's going on in the world of impact and stay up to date with our latest views, research, and feature releases.