Has Idiosyncratic Volatility Increased? Not in Recent Times

This study successfully replicates the key findings of Campbell et al. (2001). We document that aggregate idiosyncratic volatility increases over their sample period from 1962 to 1997. In out-of-sample analysis from 1926 to 1962 and 1998 to 2017, we find that idiosyncratic volatility (IV) decreases, suggesting that their finding is sample-specific. We compare their measure of IV with those obtained from models such as the Fama and French (1993) three-factor model and find that they are very similar. The Campbell et al. (2001) volatility measures can only be estimated at the aggregate level. An advantage of asset pricing model-based IVs is that they can be estimated at the stock level. Employing these stock-level IV measures, we examine trends in a variety of IV series and how IV relates to commonly analyzed firm characteristics. In doing so, we provide further insight into IV and its time-series trends.