First and foremost, any manager for any investment process needs to really articulate a philosophical argument for integration of ESG.

They need to know their investment thesis, how ESG fits into their investment process, does ESG change for active assumptions and are you going to build new models to accommodate for ESG, or shoehorn it into existing models are all issues that have to be discussed when having a conversation within a firm about ESG.

After this, the next concern becomes the data itself. There are a couple of things that come to mind: breadth is one consideration.

Part of the beauty of a quantitative investment process is that we have a way of systematically analysing many thousands of stocks. If we are going to go down the path of ESG integration, we want to get ESG information for us much of that selection universe as possible.

Depth is another consideration as the more data points we have for any given company, the better. If we just have one thing to observe per company, it isn’t going to allow us to perform the kind of robust work that we want to do.

The idea of comparability is also important as we want to be able to compare stocks on an apples to apples basis and that means having the same data for all stocks available or as much as possible.

There are some important exceptions to this for our own process, but generally we want to be able to make a base line comparable assessment of ESG for all of the stocks that we are looking at.

The next thing for us would be the idea of history, and for quants this is especially important because we want to try and test how this data behaves, not just in the current environment, but also how it might interact with our portfolio ideas in different environments.

This is a big challenge with ESG data as some of the governance data goes back quite a way, but most robust ESG data really is from the early or mid- 2000s, so we don’t have this great time series to work with.

Finally, we are also interested in seeing good data hygiene and by this I mean that when we look at ESG data, we still see problems with things like identifier mapping, not catching corporate actions, time stamping i.e. when did this piece of information become
publicly available?.

These are examples of what we take for granted in traditional financial statement data which is, frankly, less than perfect within ESG data sets.

This has improved but it is a big challenge when we are working with large chunks of data for thousands of stocks.

When it comes to traditional ESG data, the quantity, coverage and depth of data has increased dramatically over the last five years.

When we look at our global universes now, we see really high coverage everywhere, except emerging markets small cap, which lags behind.

For developed, emerging and certainly for large and mid-cap companies, the coverage has been very good.

For some of the concepts that we are interested in today, we would love to look at them going back in time, but again, some history is just not there and this makes for a challenge.

The good news is that we have it today and for our own purposes we have chosen to think about what research and analysis we can carry out that will be cross sectional in nature.

In addition to improvements in traditional ESG data, we observe a massive amount of alternative ESG data that has come to market.

By this I mean news flow analysis, trending ESG sentiment, natural language process derived data, etc.

We also see data that is coming at us from the physical sciences, geophysical data and physical risk data that is often based on satellite imagery and the like.

These are often giant data sets and the challenge that some vendors have chosen to bite off is to map some of this information to companies and map them to identifiers that we can use as investors, which is a herculean task.

What this has done is given investors such as us access to these trending news flow ideas or these physical risk ideas which are all very exciting but are unproven from an investment perspective and typically, we don’t have a lot of time to work with on
these as well.

We need to strike a balance between being excited, interested and optimistic, but also continuing to be skeptical about how these things may or may not be applicable for the task at hand.

When it comes to working with alternative data or ESG data more generally, we have to be very honest with ourselves on what type of information is actionable within our own investment process.

Any information that has a very short horizon is not going to be appropriate for us as longer-term investors.

With some of this information, whilst it can be very compelling from a story line perspective, it may not be appropriate for our investment style, asset class and models so we need to be honest with ourselves about this.

 

Please Sign In or Register to leave a Comment.