Template-Type: ReDIF-Paper 1.0 Series: Tinbergen Institute Discussion Papers Creation-Date: 2005-11-05 Number: 04-119/4 Author-Name: H. Peter Boswijk Author-Email: h.p.boswijk@uva.nl Author-Workplace-Name: Faculty of Economics and Econometrics, Universiteit van Amsterdam Author-Name: Franc Klaassen Author-Email: f.j.g.m.klaassen@uva.nl Author-Workplace-Name: Faculty of Economics and Econometrics, Universiteit van Amsterdam Title: Why Frequency Matters for Unit Root Testing Abstract: It is generally believed that for the power of unit root tests, only the time span and not the observation frequency matters. In this paper we show that the observation frequency does matter when the high-frequency data display fat tails and volatility clustering, as is typically the case for financial time series such as exchange rate returns. Our claim builds on recent work on unit root and cointegration testing based non-Gaussian likelihood functions. The essential idea is that such methods will yield power gains in the presence of fat tails and persistent volatility clustering, and the strength of these features (and hence the power gains) increases with the observation frequency. This is illustrated using both Monte Carlo simulations and empirical applications to real exchange rates. See also the article in Journal of Business & Economic Statistics. Classification-JEL: C12; C22; F31 Keywords: Fat tails; GARCH; mean reversion; observation frequency; purchasing-power parity; unit roots File-Url: https://papers.tinbergen.nl/04119.pdf File-Format: application/pdf File-Size: 525823 bytes Handle: RePEc:tin:wpaper:20040119