Our Seminars & Workshops
Seminars
Workshops
Econometrics Seminar - Shin Kanaya (Essex)
Title: Universal inference with time series data (by Shin Kanaya and Yuya Sasaki)
Abstract: In this paper, we develop a likelihood-based inference method for time-series data, building on the universal inference framework of Wasserman, Ramdas, and Balakrishnan (2020, PNAS), whose original work focused primarily on the i.i.d. setting. We consider parametric Markov and hidden Markov models, which encompass a wide range of models used in macroeconomics and finance, including autoregressive, (G)ARCH, and state-space models. Our method retains the original spirit of universal inference: the resulting confidence sets and hypothesis tests are valid without relying on conventional regularity conditions and enjoy finite-sample guarantees. A novel use of the Markov property plays a crucial role in our construction, yielding inference procedures that remain uniformly valid over a class of data-generating processes without imposing either ergodicity or stationarity. We illustrate the proposed method through Monte Carlo simulations and empirical applications.