Contrastación empírica del Efecto Fisher en la Unión Europea mediante técnicas de cointegración con datos de panel

Abstract

Most empirical studies suggest that the Fisher effect, analysed through a cointegrating relationship between inflation and nominal interest rates, does not hold. Westerlund (2008) argues that this fact can be attributed in part to the low power of cointegration tests for individual countries, and that the use of panel data can generate more powerful tests. In the present study, we use, among others, two new panel cointegration tests proposed by that author that, unlike conventional panel cointegration tests, do not assume independence among crosssectional units (countries) and present better size and power properties. These tests are applied to a panel of quarterly data covering fifteen countries of the European Union between 1983:1 and 2009:1. Our results show that there is evidence for a cointegrating relationship between inflation and nominal interest rates for this panel, and that the two variables move one-to-one, as postulated by the full Fisher effect.