As ESG data sources grow, so do the number of pitfalls – RAM AI
ESG: a growing wealth of data
Over the last decade, awareness surrounding ESG issues has been growing rapidly, with investors increasingly incorporating non-financial information into their analysis. Companies and third party entities are responding to this demand by providing an ever-expanding range of data, both in terms of their coverage and diversity of fields. We view this new availability of data as a potential source of alpha, bringing a complementary profile to the existing information-set already captured by our factors.
The difficulty was that until recently it has not been possible to access extensive ESG data on most companies. In 2011, fewer than 20% of S&P500 companies disclosed their ESG data. Whereas by 2016, the number of companies issuing sustainability or integrated reports has increased to over 80%.
According to KPMG, circa 75% of the N100 companies (defined as the top 100 companies by revenue in 49 countries) released annual sustainability reports. Below we can see the growth in the number of companies which provide CDP (formerly the Carbon Disclosure Project) with Environmental data, with a 33% increase since 2013.
The most common pitfalls of ESG investing
The struggle for many investors is incorporating those ESG factors which can enhance a portfolio’s risk-adjusted performance. Countless academic papers which study the relationship between corporate & social responsibility and a stocks’ performance reach contradictory conclusions on this point. Below we examine the common pitfalls that befall investors in this space, below we’ve outlined the primary reasons that these can occur:
1. Unreliable data
ESG performance is not reported in a universal format, thus a lack of robustness, comparability, reliability and timeliness can be prevalent. According to the CFA Institute, this remains the most restrictive factor for investors in effectively evaluating non-financial information. Various interest groups such as the Sustainability Accounting Standards Board (SASB) and the Investor Network on Climate Risk (INCR) are helping to drive consistency by standardising the disclosure on specific ESG topics.
This problem can often be augmented when ratings agencies rely on this inconsistent data to calculate the metrics for their differing methodologies. Consequently, the same company can display disparity for the same metric emanating from two different agencies. The same problem can also occur via changes in methodology by these same agencies, making it tricky to interpret time series
2. Data mining & Reverse causality
Repeating the same thing often enough will occasionally yield successful results. Take a single ESG factor in isolation; sufficient research on a given ESG factor could unintentionally reveal an attractive correlation in relation to stock performance. With hundreds of ESG fields and a relatively short data history, the risk of data mining is high and researchers can often uncover spurious relationships between ESG factors and stock performance.
The frequency of data updates (often annual ratings) brings another risk of modelling; if causality is assumed when a correlation is observed. The danger for investors here, is using such a low-frequency measure to analyse the relationship between returns and trading strategies. The fundamental question here is; has the company performed well because they do good, or they do good, because they have performed well
3. Unintentional factor exposures
This pitfall can be two-fold: the inadvertent capturing of a factor via a sub-optimal approach, or an accidental exposure to a factor which the investor has no comprehension. ESG factors considered on a standalone basis may identify characteristics that could be better captured with other fundamental factors.
For example; the risk-adjusted performance of a ESG-tilted portfolio might exhibit an attractive profile, but in reality, it is actually exposed to Quality in a sub-optimal way (i.e. there are better ways to capture this Quality premium). Additionally, a naïve ESG exposure could present an unwanted bias to factors such as Volatility, Market Capitalisation or Sector, with the investor potentially and unwittingly exposed to these risks.
Nicolas Jamet is senior quantitative analyst at RAM Active Investments.