Exchange-rate return predictability and the adaptive markets hypothesis: Evidence from major foreign exchange rates

This study examines return predictability of major foreign exchange rates by testing for martingale difference hypothesis (MDH) using daily and weekly nominal exchange rates from 1975 to 2009. We use three alternative tests for the MDH, which include the wild bootstrap automatic variance ratio test, generalized spectral test, and Dominguez–Lobato consistent tests. We evaluate time-varying return predictability by applying these tests with fixed-length moving sub-sample windows. While exchange rate returns are found to be unpredictable most of times, we do observe a number of episodes of statistically significant return predictability. They are mostly associated with the major events such as coordinated central bank interventions and financial crises. This finding suggests that return predictability of foreign exchange rates occurs from time to time depending on changing market conditions, consistent with the implications of the adaptive markets hypothesis.     

A Variance-Ratio Test of Random Walks in Foreign Exchange Rates

The separate variance-ratio tests under homoscedasticity and heteroscedasticity both provide evidence rejecting the random walk hypothesis, using five pairs of weekly nominal exchange rate series over the period from August 7, 1974 to March 29, 1989. The rejections cast doubt on the random walk hypothesis in exchange rates, which has received support in the existing literature. Furthermore, since the rejections are robust to heteroscedasticity, they suggest autocorelations of weekly increments in the nominal exchange rate series, which may be consistent with the exchange rate overshooting or undershooting phenomenon.     

Nonlinear trends in real exchange rates: A panel unit root test approach

We analyze the possibility of nonlinear trend stationarity as the alternative to unit roots in 23 OECD real exchange rates, 1974–1998, by adding nonlinear time terms to the CIPS panel unit root test of Pesaran (2007). We follow a thorough bootstrapping approach and propose a technique to adjust statistical significance for the use of multiple tests over several time trend orders. The unit root null that all real exchange rates have unit roots is rejected at better than the 0.05 level. Bootstrapped results from a procedure of Chortareas and Kapetanios (2009) suggest that the hypothesis that all are stationary is reasonable. We argue that nonlinear trend stationarity is the most likely alternative hypothesis for at least some of the real exchange rates because: (1) the strongest CIPS rejection occurs when quadratic trends are specified; (2) nonlinear time terms are statistically significant at the 0.10 level; (3) the actual CIPS statistics are more consistent with CIPS sampling distributions from bootstrapped nonlinear trend stationary processes than from linear trend or mean stationary processes.     

Searching for evidence of long-run PPP from a post-Bretton Woods panel: separating the wheat from the chaff

This paper reexamines the issue of long-run PPP using multiple panel tests in the framework of confirmatory analysis. Application of six panel tests under competing null hypotheses to the real exchange rates of 21 industrial countries yields seemingly contradictory evidence on the parity during the post-Bretton Woods period. Regardless of numeraire currency, four I(1) panel tests unanimously reject the null hypothesis in favor of long-run PPP, whereas two I(0) panel tests lend little support to the parity at conventional significance levels. Confirmatory analysis suggests that this puzzling result can be explained either by nonlinear dynamics of the real exchange rates or by a mixture of I(0) and I(1) series in the panel. Monte Carlo experiments indicate that potential mix of I(0) and I(1) series is more relevant to the empirical finding. The use of a sequential classification method sorts out six real exchange rates which exhibit most persistent deviations from long-run equilibrium. Systematic behavior of these series can be characterized better by country specific factors than by observable macroeconomic variables.     

Common Stochastic Trends in a System of Exchange Rates

Univariate tests reveal strong evidence for the presence of a unit root in the univariate time-series representation for seven daily spot and forward exchange rate series. Furthermore, all seven spot and forward rates appear to be cointegrated; that is, the forward premiums are stationary, and one common unit root, or stochastic trend, is detectable in the multivariate time-series models for the seven spot and forward rates, respectively. This is consistent with the hypothesis that the seven exchange rates possess one long-run relationship and that the disequilibrium error around that relationship partly accounts for subsequent movements in the exchange rates.     

Bivariate relative city price convergence in the United States: 1918-1997

As in international tests of purchasing power parity, panel unit root tests have been successful in rejecting a unit root process in U.S. city relative prices over the period 1918-1997. However, there is an empirical question of what the rejection of a ‘panel unit root’, particularly with respect to real exchange rates, means. This paper employs a variety of univariate unit root and cointegration tests which have recently come to the fore. These tests improve the power and reduce size distortion found in standard unit root and cointegration tests such as the Dickey-Fuller and Phillips-Perron tests. I find considerable evidence for rejecting a unit root process in the majority of U.S. city relative prices over the entire sample period and two subperiods. Less successful are stationarity tests conducted on regions of the U.S.     

An extensive analysis on the Japanese markets via S. Taylor's model

Applying S. Taylor's approach (1986), we make an extensive analysis on the Japanese stock market, foreign exchange market and the Japanese Government Bond Futures market. The purpose of this paper is to empirically reveal the structure of the Japanese markets via Taylor's model rather than to propose a new model. For this reason, we include a variety of analyzed data particularly for the Japanese stock market and the foreign exchange market because the results can be used in a different manner. The paper consists of three parts. But each part can be read separately. Part 1: Overshooting hypothesis for Japanese stock prices Part 2: A trend movement in daily/weekly Yen-Dollar exchange rates Part 3: Price variations of Japanese Government Futures. In the first part, the stock prices are shown to over-respond to new information, which is different from the behaviors of stock prices in other markets. In Part 2, a trend movement is revealed in Yen-Dollar exchange rates. In Part 3, a strategy in the Japanese Government Bond futures markets is shown to perform better than a buy and hold strategy.     

RESPONSE OF FINANCIAL MARKETS TO ANNOUNCEMENTS OF THE AUSTRALIAN CURRENT ACCOUNT BALANCE

This paper documents the response of exchange rates, interest rates and stock prices to monthly announcements of the Australian current account balance. Survey data on market participants' expectations and forecasts generated from ARIMA time series models are used to identify the unexpected component of the announcements. The study also controls for day-of-the-week effects that have been documented in the Australian equity market The results support the efficient market hypothesis and show a significant depreciation of the Australian dollar in foreign exchange markets and a significant rise in both short- and long-term interest rates to announcements of larger than expected current account deficits. The study was, however, unable to find evidence of a significant stock price response to the current account announcements.     

Temporal aggregation and unit roots in nominal foreign exchange rates

Extant empirical evidence on the logarithm of the daily and weekly spot exchange rates indicates the presence of unit roots. At a lower frequency, however, there is evidence that monthly spot rates do not have a unit root, although the autocorrelation coefficient may be close to one in absolute value. Unlike earlier studies, the present article applies various tests to data with different frequencies and reports evidence suggesting that exchange rates may not have unit roots. Absence of nonstationarity in the observations implies that econometric models may not have to apply any differencing.The authors gratefully acknowledge Hashem Dezhbakhsh for helpful suggestions. Comments are welcome.     

Do Interest Rates Follow Unit-Root Processes? Evidence from Cross-Maturity Treasury Bill Yields

Abstract. It is widely reported in the literature that interest rates follow integrated processes. Many empirical studies have, in fact, taken this result as a maintained hypothesis. This article demonstrates that the failure to reject the hypothesis that interest rates contain a unit root may be due to the severe power problem of standard test procedures in small samples. We analyze a panel of cross-maturity Treasury-bill yield series by employing a panel-based test. This test exploits cross-maturity variations of the data to improve estimation efficiency and is more powerful than standard tests for unit roots. The critical values of the test statistics are computed by Monte Carlo simulations tailored to our samples. It is found that the null hypothesis that each yield series contains a unit root can be decisively rejected. Our findings cast some doubt on previous studies that rely on the nonstationarity assumption of interest rates.