Forecasting exchange rates using panel model and model averaging

We propose to produce accurate point and interval forecasts of exchange rates by combining a number of well known fundamental based panel models. Combination of each model utilizes a set of weights computed using a linear mixture of experts's framework, where weights are determined by log scores assigned to each model's predictive performance. As well as model uncertainty, we take potential structural break in the parameters of the models into consideration. In our application, to quarterly data for ten currencies (including the Euro) for the period 1990q1–2008q4, we show that the forecasts from ensemble models produce mean and interval forecasts that outperform equal weight, and to a lesser extent random walk benchmark models. The gain from combining forecasts is particularly pronounced for longer-horizon forecasts for central forecasts, but much less so for interval forecasts. Calculations of the probability of the exchange rate rising or falling using the combined or ensemble model show a good correspondence with known events and potentially provide a useful measure for uncertainty of whether the exchange rate is likely to rise or fall.     

Forecasting exchange rates: The multi-state Markov-switching model with smoothing

This paper presents an exchange rate forecasting model which combines the multi-state Markov-switching model with smoothing techniques. The model outperforms a random walk at short horizons and its superior forecastability appears to be robust over different sample spans. Our finding hinges on the fact that exchange rates tend to follow highly persistent trends and accordingly, the key to beating the random walk is to identify these trends. An attempt to link the trends in exchange rates to the underlying macroeconomic determinants further reveals that fundamentals-based linear models generally fail to capture the persistence in exchange rates and thus are incapable of outforecasting the random walk.     

Parity reversion in the Asian real exchange rates: new evidence from the local-persistent model

This article investigates the time-series properties of 13 Asian real exchange rates (RERs) vis-à-vis the US dollar. The half-life point estimates drawn from the local-persistent model are all less than 2 years, with a finite upper bound. There is no evidence to indicate that the Asian financial crisis has altered the speed of the purchasing power parity (PPP) adjustments. We find that the persistence of RERs over the last three decades remains unchanged in majority of the cases. Given the fairly rapid speed of adjustments and their corresponding confidence intervals, we conclude that the PPP puzzle does not exist in these countries.     

A prototype model of the foreign exchange market of Canada: Forecasting capital flows and exchange rates

A balance-of-payments structural model of the foreign exchange market of Canada, endogenizing capital flows, the spot and forward exchange rates and the entities of the monetary sector, is developed using quarterly data for 1971–81. The capital flows have been disaggregated into ten categories and the exchange rates of the Canadian dollar have been analysed against five major currencies. While the model does not adhere strictly to purchasing power or interest rate parity, it does recognize them and it also incorporates other economic fundamentals, expectations and risk. Government interventions, although generated endogenously, are quantified implicitly and globally. The model tracks the post-Bretton Woods in-sample experience and generates ex post predictions reasonably well.     

Real exchange rates and target interest rates in a simple VAR model

This paper proposes a simple vector autoregressions (VAR) model with (real) output and exchange rate shocks on interest rates. Rather than assuming non-recursive identification schemes, we test the identifying assumption of the error term decompositions. Applying the model to quarterly data on major currencies against the U.S. dollar (USD) from 1974 to 1997, interest rate shocks explain - after 3 years - 16% of Canadian dollar/USD (CAD) real exchange rate variations and less than 2% for the mark/USD and yen/USD. Positive innovations of interest rates bring about (transitory) CAD real appreciations in differences and (permanent) appreciations in levels. Canadian real output is more explained by domestic interest rate shocks (19%) than Germanys (5%) or Japans (0.2%). Canada is smaller than the other economies and CAD has been shown to suffer from fear of floating. Our findings support the proposition that domestic shocks dominate output variance under fixed exchange rates. They are also consistent with structural interpretations of the VAR.     

Dynamic stability in a model of dual exchange rates

The dynamic stability of a small open economy operating under a regime of dual exchange rates is shown to depend on residents' net foreign asset position. This result is in contrast to the economy's dynamic behavior under unified fixed-rate and flexible-rate regimes where stability is unrelated to net foreign asset position.     

Dynamic linkages between Asian stock prices and exchange rates: new evidence from causality in quantiles

This article applies the Granger causality test in quantiles to investigate causal relations between stock returns and exchange rate changes for nine Asian markets over the period 1 January 1997 to 16 August 2010. Our empirical results indicate that the quantile causal relations vary across different quantiles and different periods. Although the causal effects of exchange rate changes on stock returns (or stock returns on exchange rate changes) are heterogeneous across quantiles, the overall evidence suggests that most stock and foreign exchange markets are negatively correlated. The result shows that there are more bidirectional causal relations in accordance with this method than the conventional least square (LS) estimation. The symmetry of these quantile causal effects (the ‘averaging effect’) helps to explain why conventional LS method usually obtains an insignificant result of causality.     

The inadequacy of linear autoregressive model for real exchange rates: empirical evidence from Asian economies

Utilizing the formal linearity test of Luukkonen, Saikkonen and Teräsvirta (Biometrika, 75, 491-499, 1998) as diagnostic tool, the empirical finding suggests that the linear autoregressive (AR) model is inadequate in describing the real exchange rates behaviour of 11 Asian economies. It is noted that the conventional battery of diagnostic tests is capable of identifying the inadequacy of the linear model in only three of these series. Moreover, the linearity nature of this behaviour has been formally rejected in favour of the non-linear smooth transition autoregressive (STAR) model. The finding of non-linearity in the data generating process of these real exchange rates warrants that the use of linear framework in empirical modelling and statistical testing procedures in the field of exchange rates may lead to an inappropriate policy conclusions.     

Exchange rates,exchange risk,and Asian export revenue

While depreciation may raise export revenue, associated exchange risk could offset any positive effect. The present paper investigates this net effect for eight Asian countries using a bivariate GARCH-M model that simultaneously estimates time varying risk. The fundamental result is that export markets react differently to exchange rates and associated risk. High degrees of risk apparently stimulate efforts to avoid its impact. Exchange risk has a dominating negative impact for the appreciating Japanese yen. Depreciation has no impact in Malaysia and Singapore, and exchange risk has a negative effect in Singapore. For the other five countries, depreciation stimulates export revenue but risk leads to a negative net effect in Taiwan.     

Monetary policy and exchange rates: Further evidence using a new method for implementing sign restrictions

This paper estimates SVARs for four small and three large economies. Sign restrictions are used to identify all the shocks in the SVARs, while being agnostic about the sign of the response of the real exchange rate to a relative monetary policy shock. The large number of sets of impulse responses to be judged by sign restrictions for either retention or rejection is generated by a newly proposed method which utilizes instrumental variable estimation. The responses show an absence of an exchange rate puzzle in each economy. The peak appreciation following a contractionary monetary policy shock occurs with at most a one quarter delay in the small countries and, for the United States, on impact. For the Euro region and Japan, the peak appreciation is in the long run. There is considerable model uncertainty in the responses.     

A fractionally integrated exponential STAR model applied to the US real effective exchange rate

The aim of this paper is to study the dynamics of the US real effective exchange rate by capturing non-linearity and long-memory features. In this context, we use the family of fractionally integrated STAR (FISTAR) models proposed by van Dijk et al. (van Dijk, D., Franses, P.H., and Paap, R., 2002. A non-linear longmemory model with an application to US unemployment. Journal of Econometrics 110, 135–165.) in the case when the transition function is an exponential function and we develop an estimation procedure. Indeed, these models can take into account processes characterized by several distinct dynamic regimes and persistence phenomena.     

The comovement between exchange rates and stock prices in the Asian emerging markets

This study investigates the comovement between exchange rates and stock prices in the Asian emerging markets. The sample covers major institutional changes, such as market liberalization and financial crises, so as to examine how the short-term and long-term relations change after such events. The autoregressive distributed lag (ARDL) model proposed by Pesaran et al. (2001) is adopted, which allows us to deal with structural breaks easily, and to handle data that have integrals of different orders. Interest rates and foreign reserves are also included in the analysis to reduce potential omitted variable bias. My empirical results suggest that the comovement between exchange rates and stock prices becomes stronger during crisis periods, consistent with contagion or spillover between asset prices, when compared with tranquil periods. Furthermore, most of the spillovers during crisis periods can be attributed to the channel running from stock price shocks to the exchange rate, suggesting that governments should stimulate economic growth and stock markets to attract capital inflow, thereby preventing a currency crisis. However, the industry causality analysis shows the comovement is not stronger for export-oriented industries for all periods, such as industrials and technology industries, thus implying that comovement between exchange rates and stock prices in the Asian emerging markets is generally driven by capital account balance rather than that of trade.     

Asymmetry dynamics in real exchange rates: New results on East Asian currencies

This paper provides new evidence on the purchasing power parity (PPP) hypothesis in six East Asian countries. Based on nonlinear unit root tests, we discovered that the results are broadly consistent with the fact that real exchange rates (RERs) follow a nonlinear mean reversion process. We presented new evidence that the adjustment towards the PPP parity is asymmetric (LSTAR process) above and below the equilibrium value in all but one case — the Malaysian ringgit (MYR). The empirical results suggest that it is important that the conventional tests of PPP be amended to take account of asymmetries in the adjustment process in RERs.     

Forecasting technology diffusion: a new generalisation of the logistic model

The present research proposes a new generalisation of the logistic model aiming at technology diffusion forecasting. Regarding criticisms and failures reported in the literature to apply logistic function for long-term forecasting, in our work we focused on short-term accuracy of forecast. To formulate the model, based on mathematical approximation, at first the differential equation governing the diffusion process is found and then by solving the derived differential equation, the forecast function is obtained. In all steps, mathematical tools from numerical analysis are used. We compared the New Generalized Logistic Model with eight of the most renowned models in the literature. The model led to more accurate fits and forecasts than those obtained from other models we applied for comparison.     

Forecasting exchange rates in transition economies: A comparison of multivariate time series models

This article compares the accuracy of vector autoregressive (VAR), restricted vector autoregressive (RVAR), Bayesian vector autoregressive (BVAR), vector error correction (VEC) and Bayesian vector error correction (BVEC) models in forecasting the exchange rates for five Central and Eastern European currencies (Czech Koruna, Hungarian Forint, Polish Zloty, Slovak Koruna and Slovenian Tolar) against the Euro and the US dollar. Among the specifications composing this battery of multivariate time series models, those with the smallest prediction error still fail to reject the test of equality of forecasting accuracy against the random walk model in short-term predictions, with the exception of the Slovenian Tolar/Euro exchange rate.First version received: October 2002/Final version received: September 2003The authors are grateful to two anonymous referees and the participants in the workshop Monetary and Exchange Rate Strategies Related to the Current European Unions Enlargement Processes, held in Leuven in September 2000, for very helpful comments.     

Nonlinearity in medical expenditures: a new semiparametric approach

Research in empirical health economics has found that the relationship between medical expenditures and age, income and other variables can be highly nonlinear. Moreover, men and women can have quite different medical expenditure patterns due to their differences in life expectancy. Thus it may be difficult to find an appropriate parametric model to capture the highly complicated nonlinearity in medical expenditures, and introducing a simple gender dummy variable in a parametric model may not reveal all the medical expenditure differences between men and women. This study takes a semiparametric approach. In particular, an additive partially linear specification is employed to study the relationship between medical expenditures and age, income, gender and other individual characteristics. Using data from the National Medical Expenditure Survey, the results indicate that the semiparametric approach taken in this study is very promising. They confirm that medical expenditures are nonlinear in income and age, and that men and women have quite different medical expenditure patterns.     

Dynamic linkages between exchange rates and stock prices: Evidence from East Asian markets

This study examines dynamic linkages between exchange rates and stock prices for seven East Asian countries, including Hong Kong, Japan, Korea, Malaysia, Singapore, Taiwan, and Thailand, for the period January 1988 to October 1998. Our empirical results show a significant causal relation from exchange rates to stock prices for Hong Kong, Japan, Malaysia, and Thailand before the 1997 Asian financial crisis. We also find a causal relation from the equity market to the foreign exchange market for Hong Kong, Korea, and Singapore. Further, while no country shows a significant causality from stock prices to exchange rates during the Asian crisis, a causal relation from exchange rates to stock prices is found for all countries except Malaysia. Our findings are robust with respect to various testing methods used, including Granger causality tests, a variance decomposition analysis, and an impulse response analysis. Our findings also indicate that the linkages vary across economies with respect to exchange rate regimes, the trade size, the degree of capital control, and the size of equity market.     

A simple unit root test against asymmetric STAR nonlinearity with an application to real exchange rates in Nordic countries

Existing tests of the unit root hypothesis against the alternative hypothesis of exponential smooth transition autoregressive (ESTAR) nonlinearity implicitly assume symmetry under the alternative. This paper proposes a simple unit root test against the alternative of symmetric or asymmetric ESTAR nonlinearity. In the event that the unit root hypothesis is rejected, a simple test of symmetric versus asymmetric ESTAR nonlinearity is also proposed. The asymptotic distributions of the test statistics are straightforward to establish and finite-sample performance is studied with Monte Carlo simulations. An empirical application involving the real exchange rates of four Nordic countries against the U.S. dollar illustrates the usefulness of the new tests.     

A new multiscale decomposition ensemble approach for forecasting exchange rates

Due to the high complexity and strong nonlinearity nature of foreign exchange rates, how to forecast foreign exchange rate accurately is regarded as a challenging research topic. Therefore, developing highly accurate forecasting method is of great significance to investors and policy makers. A new multiscale decomposition ensemble approach to forecast foreign exchange rates is proposed in this paper. In the approach, the variational mode decomposition (VMD) method is utilized to divide foreign exchange rates into a finite number of subcomponents; the support vector neural network (SVNN) technique is used to model and forecast each subcomponent respectively; another SVNN technique is utilized to integrate the forecasting results of each subcomponent to generate the final forecast results. To verify the superiority of the proposed approach, four major exchange rates were chosen for model comparison and evaluation. The experimental results indicate that our proposed VMD-SVNN-SVNN multiscale decomposition ensemble approach outperforms some other benchmarks in terms of forecasting accuracy and statistical tests. This demonstrates that our proposed VMD-SVNN-SVNN multiscale decomposition ensemble approach is promising for forecasting foreign exchange rates.     

A macroeconomic model for open economies with pegged exchange rates

Most traditional analyses of the internal-external adjustment problem has been based upon a model with only two assets, money and bonds. In this paper real capital is explicitly introduced. As a logical consequence also the investment function is assumed to be different from standard specifications. With this model the effects of monetary and fiscal policy are then shown to be substantially different from traditional results.