2019/03/19 by Matteo Mogliani, Mogliani, Matteo, Anna E. C. Simoni +1 · 1 citation
Decision Sciences · Economics, Econometrics and Finance · Social Sciences · #Econometrics (econ.EM) #FOS: Economics and business #Forecasting Techniques and Applications #Insurance, Mortality, Demography, Risk Management #Monetary Policy and Economic Impact
paper · pdf · doi:10.48550/arxiv.1903.08025
openalex publication_date 2019/03/19 · openalex created_date 2025/10/10 · openalex updated_date 2026/07/28
We propose a new approach to mixed-frequency regressions in a\nhigh-dimensional environment that resorts to Group Lasso penalization and\nBayesian techniques for estimation and inference. In particular, to improve the\nprediction properties of the model and its sparse recovery ability, we consider\na Group Lasso with a spike-and-slab prior. Penalty hyper-parameters governing\nthe model shrinkage are automatically tuned via an adaptive MCMC algorithm. We\nestablish good frequentist asymptotic properties of the posterior of the\nin-sample and out-of-sample prediction error, we recover the optimal posterior\ncontraction rate, and we show optimality of the posterior predictive density.\nSimulations show that the proposed models have good selection and forecasting\nperformance in small samples, even when the design matrix presents\ncross-correlation. When applied to forecasting U.S. GDP, our penalized\nregressions can outperform many strong competitors. Results suggest that\nfinancial variables may have some, although very limited, short-term predictive\ncontent.\n