Mean Reversion in G - 10 Nominal Exchange Rates *

Received wisdom: industrial-country floating exchange rates contain unit roots. In three types of tests, however, the data support nominal-rate mean reversion. First, SUR tests on panels of Group-of-Ten nominal rates frequently reject the null of unit roots in favor of mean reversion for various samples over the current float, the first such results in the literature. Second, in out-of-sample forecasts, mean-reversion models tend to beat random walks. Third, asset-pricing mo del tests support the joint hypothesis of mean reversion and exchange-market efficiency. Much empirical work in international finance assumes nominal rates contain unit roots and thus must be rethought. One example is literature on the “forecast error” or the actual spot less the lagged forward rate; related are tests and models of cointegration of spot and forward rates, and cointegration test of nominal rates. In a second example, asset pricing models that explain currency appreciation rates assume unit-root exchange rates. A third example: nominal rates’ equilibrating role in real-rate adjustment is neglected. Fourth, international portfolio allocation models assume unit-root nominal rates, as do risk management models. Mean reversion is a powerful result: it places strong restrictions on international models. For example, U.S. and German monetary policies must have been consistent with a stable long-run nominal rate over the sample period. Similarly, mean reversion must be countryand period-specific; consistent with this restriction, panels with high-inflation countries relative to the G-10 support unit roots. *For helpful discussion, thanks are due to Jim Bodurtha, Bob Cumby, Behzad Diba, Allan Eberhart, Bardia Kamrad, Bruce Lehman, Jim Lothian, Keith Ord, Akhtar Siddique, Boo Sjöö, Stephen Taylor and Clas Wihlborg. Participants in finance seminar series at The McDonough School of Business and The Gothenburg School of Economics, in a session at the Conference on Pacific Basin Finance, Economics and Accounting, and in a session at the FMA Europe meetings provided helpful comments. Gary Wang provided excellent research assistance. The McDonough School of Business provided summer support and Georgetown University’s Capital Markets Research Center provided research assistance. Part of this paper was written at Göteborgs Universitet, Sweden. Mean Reversion in G-10 Nominal Exchange Rates Abstract: Received wisdom: industrial-country floating exchange rates contain unit roots. In three types of tests, however, the data support nominal-rate mean reversion. First, SUR tests on panels of Group-of-Ten nominal rates frequently reject the null of unit roots in favor of mean reversion for various samples over the current float, the first such results in the literature. Second, in out-of-sample forecasts, mean-reversion models tend to beat random walks. Third, asset-pricing model tests support the joint hypothesis of mean reversion and exchange-market efficiency. Much empirical work in international finance assumes nominal rates contain unit roots and thus must be rethought. One example is literature on the “forecast error” or the actual spot less the lagged forward rate; related are tests and models of cointegration of spot and forward rates, and cointegration test of nominal rates. In a second example, asset pricing models that explain currency appreciation rates assume unit-root exchange rates. A third example: nominal rates’ equilibrating role in real-rate adjustment is neglected. Fourth, international portfolio allocation models assume unit-root nominal rates, as do risk management models. Mean reversion is a powerful result: it places strong restrictions on international models. For example, U.S. and German monetary policies must have been consistent with a stable long-run nominal rate over the sample period. Similarly, mean reversion must be countryand period-specific; consistent with this restriction, panels with high-inflation countries relative to the G-10 support unit roots. Received wisdom: industrial-country floating exchange rates contain unit roots. In three types of tests, however, the data support nominal-rate mean reversion. First, SUR tests on panels of Group-of-Ten nominal rates frequently reject the null of unit roots in favor of mean reversion for various samples over the current float, the first such results in the literature. Second, in out-of-sample forecasts, mean-reversion models tend to beat random walks. Third, asset-pricing model tests support the joint hypothesis of mean reversion and exchange-market efficiency. Much empirical work in international finance assumes nominal rates contain unit roots and thus must be rethought. One example is literature on the “forecast error” or the actual spot less the lagged forward rate; related are tests and models of cointegration of spot and forward rates, and cointegration test of nominal rates. In a second example, asset pricing models that explain currency appreciation rates assume unit-root exchange rates. A third example: nominal rates’ equilibrating role in real-rate adjustment is neglected. Fourth, international portfolio allocation models assume unit-root nominal rates, as do risk management models. Mean reversion is a powerful result: it places strong restrictions on international models. For example, U.S. and German monetary policies must have been consistent with a stable long-run nominal rate over the sample period. Similarly, mean reversion must be countryand period-specific; consistent with this restriction, panels with high-inflation countries relative to the G-10 support unit roots.