This paper investigates whether oil prices have a reliable and stable out-of-sample relationship with the Canadian/U.S dollar nominal exchange rate. Despite state-of-the-art methodologies, we find little systematic relation between oil prices and the exchange rate at the monthly and quarterly frequencies. In contrast, the main contribution is to show the existence of a very short-term relationship at the daily frequency, which is rather robust and holds no matter whether we use contemporaneous (realized) or lagged oil prices in our regression. However, in the latter case the predictive ability is ephemeral, mostly appearing after instabilities have been appropriately taken into account.
In this paper, we explore the relationship between real exchange rates and real interest rate differentials in the United States, Germany, Japan, and the United Kingdom. Contrary to theories based on the joint hypothesis that domestic prices are sticky and monetary disturbances are predominant, we find little evidence of a stable relationship between real interest rates and real exchange rates. We consider both in-sample and out-of-sample tests. One hypothesis that is consistent with our findings is that real disturbances (such as productivity shocks) may be a major source of exchange rate volatility.
Conventional explanations of the near random walk behavior of real exchange rates rely on near random walk behavior in the underlying fundamentals (e.g.. tastes and technology). The present paper offers an alternative rationale, based on a fixed-factor neoclassical model with traded and non-traded goods. The basic idea is that with open capital markets, agents can smooth their consumption of tradeables in the face of transitory traded goods productivity shocks. Agents cannot smooth non-traded goods productivity shocks, but if these are relatively small (as is often argued to be the case) then traded goods consumption smoothing will lead to smoothing of the intra-temporal price of traded and non-traded goods. The (near) random walk implications of the model for the real exchange rate are in stark contrast to the empirical predictions of the classic Balassa-Samuelson model. The paper applies the model to the yen/dollar exchange rate over the floating rate period.
This paper reviews the large and growing literature which tests PPP and other models of the long-run real exchange rate. We distinguish three different stages of PPP testing and focus on what has been learned from each. The most important overall lesson has been that the real exchange rate appears stationary over sufficiently long horizons. Simple, univariate random walk specifications can be rejected in favor of stationary alternatives. However, we argue that multivariate tests, which ask whether any linear combination of prices and exchange rates are stationary, have not necessarily provided meaningful rejections of nonstationarity. We also review a number of other theories of the long run real exchange rate -- including the Balassa-Samuelson hypothesis -- as well as the evidence supporting them. We argue that the persistence of real exchange rate movements can be generated by a number of sensible models and that Balassa- Samuelson effects seem important, but mainly for countries with widely disparate levels of income of growth. Finally, this paper presents new evidence testing the law of one price on 200 years of historical commodity price data for England and France, and uses a century of data from Argentina to test the possibility of sample-selection bias in tests of long-run PPP.
As the recent empirical studies surveyed here illustrate, it is very difficult to demonstrate that the exchange rate risk premium depends (through a portfolio balance channel) on the currency composition of outside assets. The existence of a 'portfolio balance effect' is a necessary condition for sterilized intervention to be a genuinely independent tool of monetary policy. This paper studies U.S./Canada data, and attempts to improve on earlier studies by using higher frequency (weekly) data and by implementing an appropriate instrumental variables technique (2S2SLS). However, we still fail to detect evidence of a portfolio balance effect.