Boom, Bust and Betrayal- Business cycles and securities fraud in the US between 1996-2019

Abstract

We study the relationship between macroeconomic factors and the number of settled securities fraud cases through the proxy of SEC class action lawsuits. We perform an empirical study of all SEC class action convictions between 1996-2019 and their relationship with business cycles as proxied by a housing price index and GDP. Furthermore, our dynamic model looks at the effect of market volatility and unemployment as well as the persistence in the time series of lawsuits. Alongside this, we study the extreme circumstances of the dot com bubble of 1999. We use non-parametric models such as GAM and MARS to better account for business cycle fluctuations. We find that the most significant effects seem to be coming from the dot com bubble, which was the only extreme outlier in our dataset, and the one month lag of the number of fraud cases. There is also an indication that both market volatility and business cycles may be correlated with the prevalence of securities fraud in our MARS model. However, the robustness checks did not find these significant for GAM.