Markov-switching Vector Autoregressive Model Research Articles

MODELING THE INTERACTIONS OF STOCK PRICE AND EXCHANGE RATE IN MALAYSIA

After the East Asian crisis in 1997, the issue of whether stock prices and exchange rates are related or not have received much attention. This is due to realization that during the crisis the countries affected saw turmoil in both their currencies and stock markets. This paper studies the non-linear interactions between stock price and exchange rate in Malaysia using a two regimes multivariate Markov switching vector autoregression (MS-VAR) model with regime shifts in both the mean and the variance. In the study, the Kuala Lumpur Composite Index (KLCI) and the exchange rates of Malaysia ringgit against four other countries namely the Singapore dollar, the Japanese yen, the British pound sterling and the Australian dollar between 1990 and 2005 are used. The empirical results show that all the series are not cointegrated but the MS-VAR model with two regimes manage to detect common regime shifts behavior in all the series. The estimated MS-VAR model reveals that as the stock price index falls the exchange rates depreciate and when the stock price index gains the exchange rates appreciate. In addition, the MS-VAR model fitted the data better than the linear vector autoregressive model (VAR).

Markov‐Switching Estimation of Spatial Maize Price Transmission Processes between Tanzania and Kenya

We discuss the Markov-switching vector au toregressive (MS-VAR) class of nonlinear time series models that can be used to analyze recurring discrete structural changes in time series. Hamilton's (1989) seminal Markov switching (MS) model of the U.S. business cy cle triggered considerable interest in the MS approach in economics. Most empirical appli cations to date have focused on the business cy cle and financial markets, but we see potential for MS-VAR models in agricultural economics, for example, in price transmission analysis. In the following, we first provide an overview of the MS-VAR framework. We then present an illustrative application to maize price trans mission between Tanzania and Kenya. The article closes with a discussion of strengths, weaknesses, and potential uses of the MS-VAR approach in price transmission analysis.

The Carry Trade and The Adjustment of the Japanese Yen: Evidence from a Markov-Switching Vector Autoregression Model

Carry trades are speculative activities which involve simultaneously going short a low-rate currency and long a high-rate currency. They are profitable as long as the gains from interest rate differentials are not offset by exchange rate movements. In this paper I investigate the dynamic relationships amongst exchange rate changes, interest rate spreads and carry trades by means of a Markov-switching vector autoregression model. I use regime-dependent impulse response functions to assess (1) how and to what extent shocks to the interest rate differential and the bilateral exchange rate affect the yen carry trade; (2) the consequences of the unwinding of the yen carry trade on the dollar-yen exchange rate. Empirical evidence indicates the presence of a so-called “carry trade” regime, whose timing is consistent with the yen carry trade episodes identified in the literature. Moreover, only when the system is in the “carry trade” regime a shock in the carry-to-risk ratio has a positive and significant effect on the net short positions in Japanese yen and the depreciation of the yen against the dollar is strong and persistent. Finally, a rising carry-to-risk ratio, which in turn reveals an increase in the attractiveness of carry trades, leads to a significant depreciation of the yen only when the system is in the “carry-trade” regime.

Incorporating a leading indicator into the trading rule through the Markov-switching vector autoregression model

This article examines the profitability of trading rules based on the smoothed probability of Markov-switching models and executes two models in Taiwan's case. The results present that both proposed models can earn excess returns over the buy-and-hold strategy and support that both can be used to trade. However, the univariate Markov-switching model, which only uses daily returns series does not successfully capture the trend in the stock market, especially during a bull market. This implies that high-frequency returns series contain lots of noises. In order to overcome this problem, the Markov-switching vector autoregression model that combines a leading indicator and returns is performed in this study. The results indicate a better trading pattern. We conclude that the leading indicator chosen from open interest in the future market increases useful information and reduces noises to improve model estimation, which can well identify the position of bull and bear markets.

Investigating the Structural Stability of the Phillips Curve Relationship

The reduced-form correlation between inflation and measures of real activity has changed substantially for the main developed economies over the post-WWII period. In this paper we attempt to describe the observed inflation dynamics in the United Kingdom, the United States and the euro area with a sequence of New Keynesian Phillips Curve (NKPC) equations that are log-linearised around different, non-zero, steady-state inflation levels. In doing this, we follow a two-step estimation strategy. First, we model the time variation in the relationship between inflation and a real cost-based measure of activity through a Markov-switching vector autoregressive model. We then impose the cross-equation restrictions of a Calvo pricing-based NKPC under non-zero steady-state inflation and estimate the structural parameters by minimising for each inflation state the distance between the restricted and unrestricted vector autoregressive parameters. The structural estimation results indicate that for all the economies there is evidence for a structurally invariant NKPC, albeit with a significant backward-looking component.

Identifying US turning points revisited: the panel model with the regime switching approach

This article proposes a panel model with a regime switching mechanism to analyse the feature of US business cycles. This Markov Switching Panel model is simple and can easily be estimated using Hamilton's (1989) method. We test the ability of the Markov Switching Panel model to identify US turning points using the US coincident indicator data. The empirical evidence shows that this model is highly capable of identifying US recessionary dates. It also has a better forecast performance than the Markov Switching vector autoregressive model.

Revisiting East Asian exchange rates: the same spirit under a different sky

The transmission of shocks among East Asian currencies following the 1997 crisis has been a widely investigated topic using different methodologies. Some studies have utilized linear vector autoregression (VAR) and its tools, such as impulse responses and forecast error variance decompositions. A few on the other hand, focusing on the nonlinearities in exchange rates, employed Markov-switching VAR (MS-VAR) framework, thus attempted to capture asymmetries linked with different regimes. A major problem of typical MS-VAR models, however, lies in the lack of economic intuition unless these are converted into a structurally identifiable form. This article extends such studies by using a different apparatus: first, it combines Markov switching and structural identifying restrictions in a vector autoregression (MS-VAR) framework, thus providing regime-dependent impulse response functions to currency shocks. Second, it also provides impulse responses to shocks associated with regime changes. Empirical findings sh...

The Carry Trade and the Adjustment of the Japanese Yen: Evidence from a Markov-Switching Vector Autoregression Model

The Carry Trade and the Adjustment of the Japanese Yen: Evidence from a Markov-Switching Vector Autoregression Model

Do Leading Indicators Lead Peaks More Than Troughs?

We develop a novel Markov switching vector autoregressive model to investigate the possibility that leading indicators have different lead times at business cycle peaks and at troughs. In this model, coincident and leading indicators share a common Markov state process, but their cycles are nonsynchronous, with the nonsynchronicity varying across regimes. An application shows that on average the Conference Board’s Composite Leading Index leads the Composite Coincident Index by nearly 1 year at peaks but by only 1 quarter at troughs. Allowing for asymmetric lead times yields improved real-time dating and forecasting of business cycle turning points.

Measuring business cycle turning points in Japan with the Markov Switching Panel model

This paper employs a Markov Switching Panel model to measure business cycle turning points in Japan. This Markov Switching Panel model is simple and can easily be estimated following Hamilton's [J.D. Hamilton, A new approach to the economic analysis of nonstationary time series and the business cycle, Econometrica 57 (1989) 357–384] method. We find that this model is highly capable of identifying Japanese recessionary dates, and it also has a forecast performance that is equal to that of the Markov Switching Vector Autoregressive model. The implication that emerges here is that governments, their agencies and other business leaders in Japan and elsewhere should also employ the Markov Switching Panel model to secure complementary data.

A sneeze in the U.S., a cough in Japan, but pneumonia in Taiwan? An application of the Markov-Switching vector autoregressive model

This paper examines the lead–lag relationships among the output of Taiwan, Japan and the U.S. Three testing methods are employed: the traditional linear Granger causality test, Hiemstra and Jones' [Hiemstra, C., Jones, J.D., 1994. Testing for linear and nonlinear Granger causality in the stock price-volume relation. Journal of Finance 49, 1639–1664] nonlinear Granger causality test and Warne's [Warne, A., 2000. Causality and regime inference in a Markov-S switching VAR, Working Paper no. 118, Sveriges Riksbank, Stockholm.] Granger causality test under the Markov-Switching model. We find that the causal ordering is unclear and depends on the model we used. Because Markov-Switching model imposes few restrictions in estimation, we tend to use its estimated results but bear in mind that the evidence is sensitive. First, the common shock hypothesis is found that most probably exists between Taiwan and the U.S. Next, we conclude that Japan tends to lead Taiwan's output, to a certain extent. Last, there is no causal ordering between the U.S. and Japan economies.

Monetary and fiscal policy interactions in Brazil: an application of the fiscal theory of the price level

The aim of the present paper is to verify the predominance of a monetary or fiscal dominance regime in Brazil in the post-Real period. The analysis is based on a model proposed by Canzoneri, Cumby and Diba (2000). This model proposes that there is a relationship between the public debt/GDP and primary surplus/GDP series by using the vector autoregression (VAR) framework and analyzing the impulse response functions. Another aim is the extension of the article written by Muscatelli et al. (2002) about the interactions between monetary and fiscal policies using the Markov-switching vector autoregressive model (MS-VAR) introduced by Krolzig (1997), since the relationship between these policies may not be constant over time. In conclusion, the macroeconomic coordination between monetary and fiscal policies in Brazil was virtually a substitute policy throughout the study period, with a predominantly monetary regime, in opposition to the non-Ricardian policies of the Fiscal Theory of The Price Level.

Stock returns and inflation in Greece: A Markov switching approach

The paper studies the dynamic relationship between real stock returns and expected and unexpected inflation utilizing a Markov Switching vector autoregressive model (MS-VAR). The MS-VAR model has the advantage that it is able to capture the dependence structure of the series both in terms of mean and variance. Univariate and multivariate innovation decompositions are employed to separate inflation into two components, the expected and unexpected. The empirical evidence suggests that real stock returns are not related to expected and unexpected inflation and this result is independent of the method used to separate inflation into the two components. Rather, the results suggest that stock market movements are regime dependent, implying that stock market performance is not predictable.

MODELLING THE US/UK REAL EXCHANGE RATE-REAL INTEREST RATE DIFFERENTIAL RELATION: A MULTIVARIATE REGIME SWITCHING APPROACH*

This paper investigates the empirical link between the US/UK real exchange rate and real interest differential for the period 1959-2002, using a bivariate Markov switching vector autoregression model. We find strong evidence of volatility regime switching in the US/UK real exchange rate-real interest differential relation. There is also evidence of a regime-dependent dynamic link. These findings indicate that allowing for multivariate regime switching can reconcile theoretical models of exchange rates which predict a strong link between real exchange rates and the real interest differential, and previous empirical evidence which failed to uncover such a link. © Blackwell Publishing Ltd and The University of Manchester, 2005.

Classical and modern business cycle measurement: The European case

This paper intends to harmonize two different approaches employed in the analysis of business cycles and, in doing so, it retrieves the stylized facts of the business cycle in Europe. We start with the ‘classical’ approach proposed in Burns and Mitchell (1946) of dating and analyzing the business cycle. The stylized facts retrieved are commented and compared to those obtained by Harding and Pagan (2002) for the U.S.. Two conclusions can be extracted from the results: a) though the turning points obtained for individual countries seem to cluster and would suggest the idea of a common cycle, there are relevant differences in the stylized facts characterizing the business cycle in the individual European economies under analysis; b) moreover, we find relevant differences in the business cycle stylized facts of the European countries and the U.S., mostly in terms of the duration, the amplitude of the cycle and the shape of the recovery. We then adopt the ‘modern’ alternative: the Markov-switching vector autoregression (MS-VAR). The model’s regime probabilities provide an optimal statistical inference of the turning point of the European business cycle. For assessing the capacity of the parametric approach to generate the stylized facts of the classical cycle in Europe, the stylized facts of the original data are compared to those of simulated data. Contrary to the results reported by Harding and Pagan (2002) , we show that the MS-VAR model is a good candidate to be used as an statistical instrument to improve the understanding of the business cycle.

The European business cycle

This paper deals with the existence and identification of a common European growth cycle. Univariate Markov switching autoregressions are used for individual countries in order to detect changes in the mean growth rate of industrial production. A Markov switching vector autoregression model is then used to identify a common cycle in Europe. Three important results are obtained: we find a common unobserved component governing European business cycle dynamics, suggesting the existence of a common business cycle; we propose a dating of the business cycle, both for an index of industrial protection and GDP, and both chronologies appear to be consistent; and finally we retrieve an important set of stylized facts and relate these with those reported for the US. Finally two further issues are investigated: first, the contribution of the European business cycle to the individual country cycles; and second, we undertake an impulse response analysis to investigate the response of each individual country to European expansions and recessions.

Monetary policy rules and regime shifts

A growing body of empirical literature has established interest rate rules as a convenient way to model and interpret monetary policy. However, as pointed out by Rudebusch (1998), vector autoregression (VAR) models used to recover the central banks' reaction functions generally rely on the dubious assumptions of linearity and time invariance. This paper proposes an empirical framework which allows the parameters of an interest rate rule to vary over time allowing for multiple regime shifts. Employing Markov-switching VAR models this study is able to identify significant and persistent shifts which affects the dynamics of the central banks' instrument interest rates. The shifts are mainly driven by discrete changes in inflation targets.

Monetary Policy Rules and Regime Shifts

A growing body of empirical literature has established interest rate rules as a convenient way to model and interpret monetary policy. However, as pointed out by Rudebusch (1998), vector autoregression (VAR) models used to recover the central banks' reaction functions generally rely on the dubious assumptions of linearity and time invariance. This paper proposes an empirical framework which allows the parameters of an interest rate rule to vary over time allowing for multiple regime shifts. Employing Markov-switching VAR models we are able to identify significant and persistent shifts which affects the dynamics of the central banks' instrument interest rates. The shifts are mainly driven by discrete changes in inflation targets.

Regime-Dependent Impulse Response Functions in A Markov-Switching Vector Autoregression Model

In this paper we introduce identifying restrictions into a Markov-switching vector autoregression model. We define a separate set of impulse responses for each Markov regime to show how fundamental disturbances affect the variables in the model dependent on the regime. We go to illustrate the use of these regime-dependent impulse response functions in a model of the U.S. economy. The regimes we identify come close to the old and new regimes found in recent research. We provide evidence that oil price shocks are much less contractionary and inflationary than they used to be. We show furthermore that the decoupling of the US economic performance from oil price shocks cannot be explained by good luck alone, but that structural changes within the US economy have taken place.

Business cycle measurement in the presence of structural change: international evidence

By generalizing Hamilton’s model of the US business cycle to a three-regime Markov-switching vector autoregressive model, this paper analyzes regime shifts in the stochastic process of economic growth in the US, Japan and Europe over the last four decades. Empirical evidence is established for the presence of a structural break in the expansionary GDP growth for the US and Japan based on an output-employment MS vector equilibrium correction model, and a structural break in the context of a common European business cycle. For the United States the long expansions of recent years signify basic changes in the business cycles pattern. In the case of Japan we identify long episodes of rapid economic expansions (existing until the mid 1970s) and long economic recessions (as in the 1990s). In Europe we find after an episode of catching-up in the 1970s, convergence in the business cycle pattern which suggests the notion of a European business cycle. The multi-regime Markov-switching VARs proposed are profoundly checked for their economic content and statistical congruency, and are found to provide a sound statistical framework for a comprehensive analysis of the business cycle.