📢 The iCEBDA Econometric seminar

Join us on April 20, 2026 at 12:00 noon (Moscow time, online via Zoom) for a research seminar by

🎓 Valentina Corradi (NYU Abu Dhabi)

Talk: "Sparsity Tests for High Dimensional Time Series Regression"

This seminar introduces a new Hausman-type test for exact sparsity in high-dimensional linear time series models — where the number of predictors may exceed the sample size. The null hypothesis assumes at most k₀ regressors have non-zero coefficients. The procedure requires no covariance matrix estimation, works under general dependence, and uses a sequential testing approach to determine the sparsity bound. The speaker establishes the validity of block bootstrap critical values and shows via simulations that the test performs well in finite samples. An empirical application to U.S. macroeconomic and financial data reveals that very sparse models often suffice, even with a large set of candidate predictors.

🔗 Register here: https://economics.hse.ru/bigdata/seminars