COINTEGRATION AND WAVELETS: AN EMPIRICAL ANALYSIS OF THE RELATIONSHIP BETWEEN MONEY AND OUTPUT IN PERU Erick Lahura Department of Economics, Pontificia Universidad Católica del Perú. elahura@pucp.edu.pe Economic Research Department, Banco Central de Reserva del Perú elahura@bcrp.gob.pe Abstract This paper analyses empirically the relationship between money and output in Peru, based on an orthogonal decomposition of series by timescales obtained using wavelets, following Ramsey and Lampart (1998). Specifically, we propose the application of wavelet filtering to analyze cointegrating relationships. No evidence of cointegration between money, real output and prices is found. However, there is evidence of cointegration between non-stationary components of the series that includes different timescale details obtained using wavelets; this result could be considered as an alternative way to represent the existence of hidden co- integration. In this context, it is found that (1) the link between money and real output is not unique, and (2) the direction of causality and exogenity depends on both the timescale and the monetary aggregate considered. Keywords: wavelets, timescale, money, output, cointegration, hidden cointegration, Granger causality, exogenity. JEL Classification: C22, C49.