Extensions to IVX Methods of Inference for Return Predictability

ABSTRACT

The contribution of this paper is threefold. First, we demonstrate that, provided either a suitable bootstrap implementation is employed or heteroskedasticity-consistent standard errors are used, the IVX-based predictability tests of Kostakis et al. (2015) retain asymptotically valid inference under the null hypothesis under considerably weaker assumptions on the innovations than are required by Kostakis et al. (2015). Second, under the same assumptions, we develop asymptotically valid bootstrap implementations of the IVX tests. Monte Carlo simulations show that the bootstrap tests deliver considerably more accurate finite sample inference than the asymptotic implementations of the tests under certain problematic parameter constellations, most notably for one-sided testing, and where multiple predictors are included. Third, we show how sub-sample implementations of the IVX approach can be used to develop asymptotically valid one-sided and two-sided tests for the presence of temporary windows of predictability.

WITH

Matei Demetrescu, Department of Statistics, TU Dortmund University, Germany
Iliyan Georgiev, Department of Economics, University of Bologna, Italy, Institut d’Anàlisi Econòmica, CSIC, Spain
A.M. Robert Taylor, Essex Business School, University of Essex, United Kingdom

link to this paper

About Authors

Paulo Rodrigues

Nova Faculty Affiliate

Paulo M. M. Rodrigues obtained his PhD in Econometrics from the School of Economic Studies of the University of Manchester. He is a professor of economics at Nova School of Business and Economics, a senior research economist at the Bank of Portugal, and a fellow of both the Clive Granger Centre for Time Series Analysis of the University of Nottingham and the University of Essex´s Centre for Financial Econometrics.