A0704
Title: Local projection based inference under general conditions
Authors: Ke-Li Xu - indiana university (United States) [presenting]
Abstract: The uniform asymptotic theory is provided for local projection (LP) regression when the true lag order of the model is unknown, possibly infinity. The theory allows for various persistence levels of the data, growing response horizons, and general conditionally heteroskedastic shocks. Based on the theory, two contributions are made. First, it is shown that LPs are semiparametrically efficient under classical assumptions on data and horizons if the controlled lag order diverges. Thus, the commonly perceived efficiency loss of running LPs is asymptotically negligible with many controls. Second, LP-based inferences are proposed for (individual and cumulated) impulse responses with robustness properties not shared by other existing methods. Inference methods using two standard errors are considered, and neither involves HAR-type correction. The uniform validity for the first method depends on a zero fourth moment condition on shocks, while the validity for the second holds more generally for martingale-difference heteroskedastic shocks.
