The main credit rating agencies have been improving ESG criteria in their fixed income scoring systems in recent years on the back of investor pressure, but critics claim shortfalls remain regarding ESG credit ratings because of poor disclosure.
The World Bank and Japan’s huge government pension fund have commissioned research to determine deeper methods of ESG integration into fixed income – a key difficulty has been the integration of ESG factors within credit risk timelines.
“Time horizons vary and depend on factors including investment objectives and the type of rated entities or instruments,” the UN’s Principles for Responsible Investment (PRI) says in a recent report.
“ESG factors can have a material impact on an issuer’s long-term financial performance, depending on the financial strength of the entity that issues a fixed income instrument, the type of issuer – whether its sovereign or corporate – and the maturity and structure of a bond,” the PRI report says.
New research paper
A new research paper published by Christoph Klein, managing partner of German asset manager ESG Portfolio Management, seeks to address some of the ESG shortfalls in fixed income.
The paper – Quantitative Credit Rating Models including ESG factors – assessed the impact of ESG factors on credit risk assessments through a quantitative model.
ESG factors in quantitative credit rating models are better suited to assess the “predictive power” of bankruptcy forecasting models than those without these factors, Klein claims.
The statistical model calculates an equation – a so-called “discriminant function” – that seeks to determine which factors of a given set are most relevant in order to classify a company as either solvent, with a good credit quality, or insolvent, with a bad credit quality.
Under Klein’s methodology, good corporate issuers are rated AAA to BBB+ and bad issuers are ranked from BBB to B. However, BBB-rated companies may still be reasonably solvent.
Factors in Klein’s model include credit ratios, such as leverage, liquidity, profitability and retained earnings, as well as standard ESG factors available in Bloomberg and MSCI ESG databases.
Additional ESG factors could be included in quantitative credit rating models when the data quality and quantity increases in the future, Klein says.
Sorting solvent from insolvent
The statistical model automatically selects factors that are significant to identify a solvent company from an insolvent one. In the best-case scenario, it is able to find an equation with factors that have a 100% hit ratio, Klein adds, correctly separating companies into solvent and insolvent.
“The analysis runs many different combinations of factors and ratios, to come up with the final function,” Klein said. The final function in the research is able to correctly classify a company’s credit quality with a hit ratio of 84.6%. This hit ratio was 0.4% better than the equation without an ESG factor, Klein says.
He explained that one reason that the ESG equation hit ratio was below 100% is because of a “grey area” – companies that have neither a good or bad credit quality.
The ESG factor in the final function, which consists of the four most decisive factors, is based on the MSCI Greenhouse Gas Mitigation Strategy Score, Klein adds.
“The complex ESG factor focusing on greenhouse gas emissions, their dynamics over time and transparency of the reporting seems to be a good indicator for the complex and broad ESG risk,” Klein writes in the paper.
The study selected originally 565 companies in certain sub-sectors of the industrial sector, based in AAA or AA rated countries and from the Bloomberg database, while the function was developed based on a training set of 285 corporations.
Klein says the quantum models are only a “starting point” and dialogue with the management is needed. “You should not ignore the other factors, including all other ESG factors or the impact of the UN Sustainable Development Goals,” he says.
Further progress needed
Gianfrate Gianfranco, professor of finance at EDHEC Business School in Nice, says the paper was able to show the association of ESG factors with creditworthiness but lacked proof of “causality”.
“Causality means that variable A unequivocally determines variable B. In this case, we cannot say that good ESG determines unequivocally good creditworthiness.
“It may easily be the other way around: maybe it is just that more financially solid companies have more financial resources to invest in ESG.
“And maybe, when all the variables that express financial solidity are taken into account, the predictive role of ESG just disappears,” Gianfranco says.
Edward Altman, a professor of finance at the Stern School of Business in New York, whose work contributed to Klein’s research, pointed to the importance of governance factors.
“The accuracy of our default risk models improves by about 20 to 25% when we add both governance and macro factors to the financial data of our credit models for small and medium sized firms,” he says.
