High-Dimensional Mixed-Frequency IV Regression

This article introduces a high-dimensional linear IV regression for the data sampled at mixed frequencies. We show that slope parameter of high-frequency covariate can be identified and accurately estimated leveraging on low-frequency instrumental variable. The distinguishing feature model is it allows handing datasets without imposing approximate sparsity restrictions. propose Tikhonov-regularized estimator study its large sample properties time series data. has closed-form expression easy to compute demonstrates excellent performance in our Monte Carlo experiments. also provide confidence bands incorporate exogenous covariates via double/debiased machine learning approach. In empirical illustration, we estimate real-time price elasticity supply Australian electricity spot market. Our estimates suggest relatively inelastic throughout day.