Reassessing growth vulnerability

Dooyeon Cho,Seunghwa Rho

DOI: https://doi.org/10.1002/jae.3005

IF: 2.46

2023-09-26

Journal of Applied Econometrics

Abstract:Summary This paper replicates the results of Adrian et al. (American Economic Review, 2019) that GDP growth volatility is mainly driven by the lower quantiles of the distribution which is predicted by the financial condition. It extends their study by estimating the model with the IVX‐QR estimator of Lee (Journal of Econometrics, 2016) and double weighted estimator of Cai et al. (Journal of Econometrics, 2022) considering that the financial condition index is highly serially correlated. Both models are estimated with the smoothed estimating equation approach of Kaplan and Sun (Econometric Theory, 2017). The results show that the findings of Adrian et al. (American Economic Review, 2019) are robust to possible bias due to the existence of persistent predictors. The out‐of‐sample forecasting exercises suggest that methods that are robust to the existence of persistent predictors can improve forecasting performance at the lower quantiles of the GDP growth distribution.

economics,social sciences, mathematical methods