Corporate social performance and financial risk: Further empirical evidence using higher frequency data

Using a unique dataset of corporate social responsibility rating – available on a monthly basis – we shed new light on the relationship between corporate social performance (CSP) and firm risk. Where previous studies use annual (at best) measures of CSP, assuming that a change in CSP leads a change in risk, we formally test the direction of the relationship using Granger causality. Looking at large UK companies over 2002–2018 (for a total number of 19,832 firm-months), we reject any causality (either way) between CSP and financial risk (both systematic and idiosyncratic risk). This shows that the CSP-risk relationship is not an endogenous one, contrary to what previous evidence has found. Given the structure of our panel data (long T and short N), we apply GLS based estimator to correct for serial correlation in our panel regressions. We find strong evidence that CSP has a negative impact on idiosyncratic risk; however, the effect of CSP on systematic risk is not statistically significant. The existence of a contemporaneous, rather than lagged relationship doesn't fare well with established CSP theories. Overall, our original approach has opened a new door to further the study of the link between CSP, financial performance and financial risk.