Thread by @jeuasommenulle

Long overdue, the thread I promised about ESG investing – mostly personal experience and thoughts.

Remember I mostly invest in fins – but not only.

For many years, the conversation was going like this:

Allocator: do you guys do ESG investing?

Me: Dude, banks have paid 500bn$ in fines since the GFC, so yeah, we kind of look at all the shit they do!

Also me: some banks have gone bust because of dodgy accounting (Alexandria..), because of money laundering, because of bribery, because they lent tens of M€ to the head of their audit committee, etc. So if you don’t do ESG investing with banks, you might as well pick another job

For years, this was enough, and it made perfect sense.
S&G were a crucial part of the investment process because you didn’t have a choice.

But it was also a “flexible” process – I mean by that no hard, algorithmic, rules.

But over the last two years, this world has disappeared, and we’ve entered the world of box ticking.

Probably because of the tsunami of ESG investments and regulations, the process has become rigid and systematic.

Let me give you an example: recently, an allocator told me they couldn’t invest in my “financials only” fund because the prospectus didn’t explicitly prohibit investments in companies making >10% of sales in the porn business.

Okay, sure. Too bad for PornBank.

There are three main problems here.

The first is the massive data requirement.

The set of rules to apply is absurdly complex and no one can seriously pretend they are applied correctly.

I mean, if you have a system where you can click on an Isin code and it shows you the share of the value chain of that company dedicated to tobacco and the expertise assessment of every board member on a 1-100 scale and the share of staff using recycling bins, please let me know.

The second is the personal judgement involved with those rules.

What’s a board tenure that’s too long?

What’s a controversy?

What’s an acceptable political opinion?

And the most important decision of all: how do you WEIGH all the criteria?

Applying weights to qualitative stuff is not even remotely robust, anyone can do anything, and there is no way of choosing the right approach.

In theory, of course, you could use some machine learning algo to optimize the weights, but unlike credit ratings, for which you want to learn how to predict default probabilities, there is no quantitative metric you try to forecast with ESG ratings: no way to train the computer!

The third is the quantitative obsession.

My investment process used to involve the brain.

I digest tons of data, some quant, some qualitative, look at prices and make a decision.

But now, allocators and regulators want me to put a score on everything!

Is this company’s HR policy 17.5% better than the other one’s, or is it 23.4%?

It’s obviously crucially important, because your fund will need to have a score which is 20% better than the average of score of the 80% best issuers of the universe…

So you need a SCORE for everything.

And, as a quant, let me tell you the first rule of quants: don’t trust numbers.

In 2008 I was sitting next to the guy who was managing 40bn of AAA CDO exposure at a big bank, and he sorely regrets trusting scores.

Seriously, this is bananas.

Not to mention that “choosing the investment universe” is to ESG what “choosing the start date” is to performance marketing.

So we’ve reached a point where most people outsource their problems to others, who have a FAR WORSE knowledge & understanding of the companies and businesses!

But they can do A SCORE.

They have DATA.

They have LISTS.

And obviously, most people will just pick the score and data provider that will make their fund look good – and then will be constrained by that choice, which is bad for future returns, ofc.

Let me take a recent example to illustrate this: a few days ago, the Spanish competition authority said it suspected malpractice in banks’ granting of ICO guaranteed Covid loans.

Old ESG investing: Hmm, this is horrible PR, taking advantage of Covid. Let’s dig into it. [hours later] It's only 20 claims out of 900k loans, the refinancing vs restructuring rule was always unclear, probably not material financially. No price reaction, let’s stay away for now.

New ESG v1:
A new controversy just hit the screen!
This is controversy #20 for Santander, they just hit the threshold: EXCLUSION! [but 5 days later, because it needs to be incorporated in the front to back tool, of course]

New ESG v2: sorry, I can’t do anything, I need to wait until Sustain Analytics updates its ESG score because of this. I’m sure they will have a very detailed analysis of the likely fines. (LOL.)

I think another very good way to illustrate the problem we’re facing, is to compare with credit risk analysis. If you are running a credit fund, surely that is still the most important thing to monitor. Right? Right?

But what's crazy is that in due diligences, you’ll get 10 ESG questions for 1 credit risk question. No one will even look at your credit risk process. Regulators don’t even care! They won’t ask you to put an average rating on your portfolio or to compare it to the benchmark.

If you can’t put a credit rating on a company, nothing will happen. If you can’t put an ESG rating, you’re directly going to hell.
And, of course, ESG ratings are not even reliable! Even regulators know that esma.europa.eu/sites/default/…

But still, we all pretend this is not a problem & we MUST have a score. And that’s why our mailboxes are now flooded with “Your alternative data needs for ESG investing” and “What can AI do for your ESG investing” and “Early bird offer – our ESG investing forum”.

If this isn’t a red flag that something is wrong, I don’t know what is…

Yeah, you've noticed I didn't say anything on climate. It's because I believe it's actually a totally different topic, that thread is too long already and I have a lot to say on climate ! So it’s for a future one!