Abstract

We use inventory write-downs to differentiate opportunistic and non-opportunistic overproduction measures. We posit that non-opportunistic overproduction is positively associated with future write-downs because overproduction generally leads to excess inventory, while opportunistic overproduction (to inflate earnings) is negatively associated with write-downs because write-downs decrease earnings. We find that change-based proxies (deviations from past behaviour) are positively associated with the likelihood of future write-downs, whereas residual-based proxies (deviations from industry norms) are negatively associated with this likelihood, suggesting that the former (latter) primarily capture non-opportunistic (opportunistic) overproduction. Our study highlights the importance of using appropriate overproduction measures for each research setting.