Have you ever wondered what underpins the trillions of ESG assets? Like most things in finance, it boils down to math and methodologies, or what we refer to as “ESG models.” Our research at Truvalue Labs, a FactSet company, has found that most of the conventional ESG models do not capture the multi-dimensional (a.k.a., multi-factor) nature of ESG. When we dug further into this, the common reason was simply that most conventional ESG models are one-size-fits-all and were created in a previous era, namely the Third Industrial Revolution.
ESG Models in the Fourth Industrial Revolution (4IR)
Our groundbreaking new research found empirical evidence that industry, country, and company size are key ESG Materiality factors in our new whitepaper. These factors, along with idiosyncratic behaviors by a company, contribute to a unique company-specific Materiality Signature. Think of a Materiality Signature as the aggregated digital exhaust that lives in unstructured documents about corporations from an outside-in stakeholder perspective. Company Materiality Signatures change over time via what we deem Dynamic Materiality, which has made real-time ESG data a necessity in the 4IR.
The History of Static ESG Models
In portfolio analytics and risk management, multi-factor models are ubiquitous. It is the recognition that we live in a complex and interconnected world, where many simultaneous factors can affect an asset price. Unfortunately, the same is not true when it comes to modeling ESG currently, in fact, most conventional ESG models are static through time.
What does this mean in practice? The below table displays an example of how a conventional ESG model commonly works. A typical model would treat a company in Australia the same as it would treat a company in China, just as it would treat a small-cap the same as a large-cap. Industry models remain relatively static through time and also fail to capture any company-specific materiality characteristics (i.e., they apply the same static KPI and weighting scheme to every company in an industry uniformly).
Illustrative Example of a Conventional ESG Model
|Global Industry Model|
|Greenhouse Gas Emissions||40%|
|Employee Health & Safety||20%|
Why was this done? It wasn’t done with mal-intent, rather it was done this way because these models were developed in the Third Industrial Revolution before Artificial Intelligence and other Big Data technologies existed to scale ESG modeling. Quite simply, there was not enough human analyst and data available at the time to measure the complex multi-dimensionality of ESG in real time. Unlike traditional portfolio analytics which mainly rely on structured data from exchanges, OTC markets, and company financial fundamental information, ESG information often lives in vast amounts of unstructured text documents written in many languages.
Conventional ESG Materiality Models are One-Dimensional
As they say, a picture is worth a thousand words, the below graphic illustrates how much information loss happens when you collapse a multi-dimensional mathematical problem into a single dot on a one-dimensional space. Just think about this for a second, the implication here is that the trillions of assets flowing into ESG are backed by a data model that is conceptually one dot in a multi-dimensional space.
A Better Way Forward
The ability to now process the vast sea of unstructured information in the 4IR is what unlocks the revolutionary jump from static one-size-fits-all ESG models to real-time, granular, company-specific ESG models. The amplification of “S” (i.e., “Social” factors) during the COVID-19 crisis has accelerated the adoption of Dynamic Materiality technology in the market. It is my hope the state of the art concepts and techniques proposed in our new whitepaper will become axiomatic in the sustainability industry one day in order to address the deficiencies in existing ESG models.
Early in my career, a wise person once shared advice with me about fixing a fundamental software architecture design issue, “You can either change the oil now, or fix the transmission later.” Perhaps the ESG version of this philosophy is: “You can either change the electric battery now, or fix the low carbon emissions motor later.” Let’s change the electric battery now in ESG, otherwise we’ll have a much bigger problem later on to confront. Except this time, it’s literally the future of the world at stake if we get this wrong as a society.
For more information, please download the ESG Materiality Factors in the Fourth Industrial Revolution Whitepaper now.
ABOUT THE AUTHOR
Adam Salvatori is a seasoned industry expert with experience across the financial, technology and startup industries. He previously served as Global Head of Commodities Systematic Trading Strats at Goldman Sachs, and prior to that Adam was the Co-Founder of the Financial Products division at Kensho. He was one of the first industry practitioners to leverage AI to create indices and co-invented the 4IR new economies sector classification system. He started his career as a mainframe cryptographic security engineer at IBM where he was awarded a patent for his work.
He holds a dual B.S. degree in both Computer Science and Business Administration, with a minor in Mathematics from the University at Albany.