Suggesting there is anything wrong with the principles of ESG is to walk a dangerous line, and there is no doubting the good intentions, unfortunately the mechanics of ESG data is in many instances not living up to the hype. As ESG investing philosophies establish solid roots there is the danger that the old adage of ‘Rubbish In, Rubbish Out’ introduces a blight.

This means identifying the problems early and sorting them out, however, some are easier to deal with others. Fortunately, there are three readily identifiable areas of concern, focused on:

1.Poor Data Sourcing: Inconsistent sourcing from places which are not designed to be sources in the first place, using extraction methods of dubious integrity

2.Low Quality Methodologies: Poor data methodologies creating sub-standard products, i.e. ESG Scores/Ratings

3.Vapourware companies with Vapourware Products and services: Not real companies, often employing unstructured business models, with products that prove to be smoke and mirrors.

Given the complexity of many elements existing within the ESG data world, especially Supply Chains, the ESG data consumer faces the challenge of dealing with incomplete, inaccurate, out of date information to input into their investment decision making chain. Thus, introducing the element of unquantifiable risk.

The biggest threat is the lack of standards, as ESG employ and protect their own proprietary methodologies the ESG data consumer cannot compare like for like, how does a rating from Systainalytics compare to one from MSCI? The problem is there can be two right answers and two wrong answers, or perhaps the investor ought to go to FTSE Russell?

Usually the financial industry gravitates towards a natural standard, for instance Bloomberg’s VWAP, but this does not seem to be happening within the ESG Data World, at least not yet.


The ESG data bandwagon has developed due to the haste of asset managers to show their own investors they have green fingers. Due diligence in many cases went out of the window because of the need to show ESG credentials and the fact that a lot of the ESG data available is cheap and accessible often coming from start ups lacking experience, who scrape the data from public websites using AI based tech. Equally the increasing need to benchmark businesses lends to the development of such products as scorings and ratings which unlike indices rely on subjectivity that requires:

1.Proven methodologies,

2.Managed by experienced professionals, and

3.Who are themselves accountable.

Sadly, many start up ESG data providers and topically ESG ratings services have fallen down on all three requirements.


This is driven by inconsistent sourcing from places which are not designed to be sources in the first place, using extraction methods of dubious integrity because they have developed their own proprietary sourcing process and research methodologies, plus:

1.They have not studied existing best practices and most importantly, nor,

2.Do not understand the nature of the data they are trying to source

The last concept is vital in the ESG context. All data is not the same, each point has its own attributes which then projects a different set of values to the data consumer depending upon what that consumer needs and wants to do with it. Many ESG data providers scrape the data from public websites then apply algo driven rules (AI/Big Data) to associate the data into their environment.

As public websites are not kept up to date, change format, can be inaccurate, plus the difficulty in getting 100% coverage, there is a built-in failover even before the analytics begin to start.

Equally the signals generated by the applied algos can be interpreted differently by individual ESG Data suppliers, introducing new levels of opaqueness.


When the ESG data consumer selects a specific ESG data supplier, that consumer is buying into that suppliers view and interpretation of the ESG world, without necessarily understanding what parameters and judgements the ESG has applied to the raw data. In context perhaps a mathematical approach demonstrates the principle and problem:

‘The accuracy of the raw data multiplied by the quality of the methodology equals the value of the output’

So if the accuracy of the raw data is 90%, and the quality of the methodology is 50%, then the data consumer has a value of 45% for the data.

Poor data methodologies create inconsistencies leading to sub-standard products, and this has been most notable with ESG Scores/Ratings, and this is instructive when considering the ESG data universe.

There are three factors at work:

1.ESG Data suppliers cannot provide the depth, breadth, coverage, and accuracy that data consumers demand,

2.There is a lack of standards in defining the parameters to establish a framework in quantifying ESG data for analysis, which means it is difficult for users to compare ESG data suppliers, and,

3.The amount of subjectivity that goes into producing outputs, for instance, a key area constantly referred to by experts is how to treat missing data.


This is really a single issue subject highlighting the often opaque nature of a surprisingly large number of ESG data suppliers. There are many so-called ESG suppliers which do not have their own products and services but are really white labelling other companies. This would not be a major issue except for certain critical aspects of their business:

1.Is often unaware that they are dealing with a white label operation, meaning,

2.The data consumer is even further away from understanding the data from the products and services subscribed to with all the associated problems highlighted in Topics 1 and 2 above.

In addition there are some ESG data suppliers which scrape data and do not even try to employ any form of analytics or methodologies to that data, just suppling the outputs without providing meaningful parameters or guidelines on what the data is.

These companies are fly-by-nights, the type seen jumping onto the crypto bandwagon amongst others. The good news is they do not or will not last long though before they disappear will take in clients not doing their due diligence or doing ESG on the cheap.


The great thing is none of the issues identified are unsolvable, they will cause pain for data consumers that do not complete their proper diligence, and lessons will be learnt. ESG data providers not meeting the right standards will simply lose business and disappear. It may take time, but it will happen. In November 2022 the Japan Financial Services Agency proposed the world’s first code of conduct for ESG data suppliers.

On the flip side those remaining are unlikely to stay independent. ESG as a topic area is just too large, diverse, complex, and expensive to maintain appropriate standards while not providing the margins necessary to support a standalone business, unless it is niche.  Cash rich global data vendors, especially Deutsche Bourse, MSCI, Moody’s, and Refinitiv have for some time actively sought and acquired quality ESG assets. It is noteworthy these companies are key supporters of Japan FSA’s initiative.

The expected implosion will be amongst the ESG data providers, not the data itself, however, it will come at a price, consolidation of ESG data providers as product units within multi-national information vendors. At the same time E, S, & G will become unbundled into more coherent and naturally associating datasets and by implication providing more relevant services to the data consumer.

For the ESG purist this may be seen as the ‘Lunatics running the Asylum’.

Overall, though, ESG data has a very bright future.


Keiren Harris 07 December 2022

Please contact for a pdf copy of the article

For information on our consulting services please email