This study addresses the limitations of traditional earnings risk measures, which often rely on moments such as variance, skewness, and kurtosis. For heavy-tailed distributions, these moments may not exist, challenging such analyses. We propose robust conditional Pareto exponents as novel measures of earnings risk, with accompanying estimation and inference methods. Using UK NESPD and US PSID data, we find (1) moments often fail to exist; (2) tail risk rises over the life cycle; (3) job stayers face higher tail risk; and (4) these patterns persist in both the 2007–2008 recession and the 2015–2016 growth period.
Sarpietro, S., Sasaki, Y., Wang, Y. (2025). Nonexistent Moments of Earnings Growth. JOURNAL OF APPLIED ECONOMETRICS, –, –-– [10.1002/jae.70017].
Nonexistent Moments of Earnings Growth
Sarpietro, Silvia;
2025
Abstract
This study addresses the limitations of traditional earnings risk measures, which often rely on moments such as variance, skewness, and kurtosis. For heavy-tailed distributions, these moments may not exist, challenging such analyses. We propose robust conditional Pareto exponents as novel measures of earnings risk, with accompanying estimation and inference methods. Using UK NESPD and US PSID data, we find (1) moments often fail to exist; (2) tail risk rises over the life cycle; (3) job stayers face higher tail risk; and (4) these patterns persist in both the 2007–2008 recession and the 2015–2016 growth period.I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
