Academics use AI to make link between ESG and bond risk

Evidence for poor responsible investment factors among companies, and a link to a higher risk of default in their corporate bonds, has been strengthened, academics claim.

Researchers from Wharton Business School at the University of Pennsylvania say they made the link after evaluating 342 companies on their environmental, social and governance (ESG) performance over eight years to 2017.

Analysis showed that cases of companies with relatively weak ESG performance at a moment in time subsequently experienced high-profile negative ESG events leading to measurable increases in credit risk, the researchers said. Cases included Volkswagen’s emissions scandal and Wells Fargo’s sales practices.

No figures were produced in the abstract to the research – which is called ‘ESG, material credit events, and credit risk’ – but the researchers said there was “clear evidence” that higher-performing companies based on ESG criteria show a “lower incidence of adverse material events”, while companies with lower ESG performance relative to their peers in their industry experienced a higher rate of incidence.

The abstract from the study, which has been published by the Journal of Applied Finance, said despite a growing consensus and consistent evidence that ESG performance is correlated with credit risk, no empirical evidence has yet linked ESG performance to cost or expense variances or revenue shortfalls that could explain these correlations.

The ESG scores were provided by Truvalue Labs. The San Francisco-based firm says it is the first to apply artificial intelligence to uncover “timely and material ESG information at scale”.

Fund manager Oddo BHF Asset Management recently stated applying ESG criteria to its European Convertibles Moderate fund, limiting exposures to companies with low ESG scores.

Olivier Becker, head of convertible bonds at the firm, said: “It is no coincidence that the latest defaults seen in the convertible bonds market – Folli-Follie or Nyrstar, for example – concerned companies with significant governance issues.”