Home About us Contact
|
Home About us Contact | ||
|
|
||
Cointegration
Terms modified by Cointegration Selected AbstractsCOINTEGRATION OF STOCK MARKETS BETWEEN NEW ZEALAND, AUSTRALIA AND THE G7 ECONOMIES: SEARCHING FOR CO-MOVEMENT UNDER STRUCTURAL CHANGEAUSTRALIAN ECONOMIC PAPERS, Issue 3 2005PARESH KUMAR NARAYAN This paper examines whether the New Zealand equity market is integrated with the equity markets of Australia and the G7 economies by applying both the Johansen (1988) and Gregory and Hansen (1996) approaches to cointegration. The Johansen (1988) test suggests that there is no long-run relationship between the New Zealand stock market and any of the other stock markets considered in the study. The Gregory and Hansen (1996) test finds that the New Zealand and United States stock market is cointegrated, but the New Zealand stock market is not cointegrated with the other stock markets in the study. This suggests that in order to avoid some of the risk through international portfolio diversification there is potential for investors to purchase shares in the New Zealand market and either the Australian market or most of the world's leading equity markets. [source] The Information Signaling Hypothesis of Dividends: Evidence from Cointegration and Causality TestsJOURNAL OF BUSINESS FINANCE & ACCOUNTING, Issue 3-4 2003Mbodja Mougoué This paper uses cointegration and causality tests to study the temporal behavior of dividends and earnings at the individual firm level. We find that, for a sample of 143 non-utility firms, approximately one-fifth of the firms exhibits a temporal relationship between dividends and earnings that is consistent with the information signaling hypothesis of dividends. In the case of 72 utilities, about a third exhibit dividend policies that are consistent with the signaling notion of dividends. Further examination of firm characteristic differences between signaling and non-signaling firms shows that, in the case of non-utility firms, signaling firms tend to be smaller, have a lower growth rate of total assets, and have a higher leverage ratio. In the case of utilities, we find no major differences in firm characteristics between signaling and non-signaling firms. [source] Cointegration, Efficiency and Forecasting in the Currency MarketJOURNAL OF BUSINESS FINANCE & ACCOUNTING, Issue 1-2 2001Wilson H. S. Tong Existing literature on using the cointegration approach to examine the efficiency of the foreign exchange market gives mixed results. Arguments typically focus on econometric testing techniques, with fractional cointegration being the most current one. This paper tries to look at the issue from an economic perspective. It shows that the cointegrating relationship, whether cointegrated or fractionally cointegrated, is found mainly among the currencies of the European Monetary System which are set to fluctuate within a given range. Hence, there is no inconsistency with the notion of market efficiency. Yet, exploiting such a cointegrating relationship is helpful in currency forecasting. There is some evidence that restricting the forecasting model to consist of only cointegrated currencies improves forecasting efficiency. [source] New Improved Tests for Cointegration with Structural BreaksJOURNAL OF TIME SERIES ANALYSIS, Issue 2 2007Joakim Westerlund C12; C32; C33 Abstract., This article proposes Lagrange multiplier-based tests for the null hypothesis of no cointegration. The tests are general enough to allow for heteroskedastic and serially correlated errors, deterministic trends, and a structural break of unknown timing in both the intercept and slope. The limiting distributions of the test statistics are derived, and are found to be invariant not only with respect to the trend and structural break, but also with respect to the regressors. A small Monte Carlo study is also conducted to investigate the small-sample properties of the tests. The results reveal that the tests have small size distortions and good power relative to other tests. [source] The Effect of Linear Time Trends on the KPSS Test for CointegrationJOURNAL OF TIME SERIES ANALYSIS, Issue 3 2001Uwe Hassler The so-called KPSS test for the null hypothesis of cointegration builds on residuals from single equation regressions. Critical values have been provided for regressions with and without detrending. Here it is shown that the latter are not appropriate if the series display linear trends, although this does not mean that detrending is required. In this paper adequate percentiles are suggested for series that follow linear time trends, and tests are based on regressions without detrending. These percentiles are readily available from the literature. [source] Miller and Modigliani, Predictive Return Regressions and Cointegration,OXFORD BULLETIN OF ECONOMICS & STATISTICS, Issue 2 2008Piergiorgio Alessandri Abstract This paper investigates the use of alternative measures of dividend yields to predict US aggregate stock returns. Following Miller and Modigliani [Journal of Business (1961), Vol. 34, pp. 411,433] we construct a cashflow yield that includes both dividend and non-dividend cashflows to shareholders. Using a data set covering the course of the 20th century, we show in a cointegrating vector autoregression framework that this measure has strong and stable predictive power for returns. The weak predictive power of standard measures of the dividend yield is explained by the strong rejection of the implied cointegrating and causality restrictions on the impact of non-dividend cashflows. [source] Cointegration and the Monetary Exchange Rate Model Revisited¶¶OXFORD BULLETIN OF ECONOMICS & STATISTICS, Issue 4 2002Jan J. J. Groen First page of article [source] Permanent-transitory Decomposition in Var Models With Cointegration and Common CyclesOXFORD BULLETIN OF ECONOMICS & STATISTICS, Issue 4 2000Alain Hecq In this paper we derive permanent-transitory decompositions of non-stationary multiple times series generated by (r)nite order Gaussian VAR(p) models with both cointegration and serial correlation common features. We extend existing analyses to the two classes of reduced rank structures discussed in Hecq, Palm and Urbain (1998). Using the corresponding state space representation of cointegrated VAR models in vector error correction form we show how decomposition can be obtained even in the case where the number of common feature and cointegration vectors are not equal to the number of variables. As empirical analysis of US business fluctuations shows the practical relevance of the approach we propose. [source] Random Walks and the Cointegration of the ACLI and NCREIFREAL ESTATE ECONOMICS, Issue 3 2000Leon Shilton Do NCREIF returns influence commercial mortgage underwriters when they adjust capitalization rates? Are the ACLI capitalization series and the NCREIF return series cointegrated at the national and the smaller geographic sub-division levels? This research uses a two-step procedure to test for cointegration. First, the Phillips,Perron unit-root procedure must show that each series is a unit-root random walk. Previous research usually has assumed that these series are random walks, with the implication that the commercial mortgage market is efficient. Second, the Phillips,Ouliaris test of the residuals of a function of the two series determines the possibility of cointegration. At the national level and for the Northeast and Pacific regions the two series are random walks and cointegrated. In other geographic sub-divisions, neither or only one series is a random walk and therefore the data does not support a relationship. The lack of functional relationships in four of the six smaller geographic regions suggests that underwriters are not obtaining the NCREIF information or are ignoring it. The lack of random walks with the implication about capital-market efficiency invites further research. [source] Cointegration, Government Spending and Private Consumption: Evidence from JapanTHE JAPANESE ECONOMIC REVIEW, Issue 2 2004Tsung-Wu Ho Assuming a CRRA preference, this paper shows that there is a cointegration restriction implied by the intra-temporal first-order condition in the consumption function. This restriction predicts a cointegrated system of government consumption, private consumption, and their relative price. Our analysis indicates that, first, Johansen's VECM confirms the theoretical prediction that is supported by the data of Japan; moreover, Bierens' (1997) nonparametric estimator severely contradicts with the theoretical model and fits the data poorly; second, Japanese people have increasing willingness to rearrange their consumption over time. Besides, the intratemporal relationship between private and government consumption remains relatively stable over time. [source] Cointegration Theory, Equilibrium and Disequilibrium EconomicsTHE MANCHESTER SCHOOL, Issue 1 2004Karim Maher Abadir Two variables are said to be cointegrated when they move closely together over time, after proper scaling. Cointegration was taken to be the statistical expression of the notion of equilibrium in economics. But is it still possible to talk of cointegration when ,disequilibrium' economics prevail? We argue that it is, and that the duality is strongest between cointegration theory and economic theories of non-clearing markets. By setting up a simple generic non-parametric model, it is shown that Clower's dual decision hypothesis is a more direct and natural expression of the notion of cointegration than long-run equilibrium is. With sticky prices, quantities (e.g. consumption and income) move together more closely than they would otherwise. As a by-product, the model gives rise to (and justifies from an economics standpoint) a recent statistical approach to modelling economic time series. An observational equivalence between two econometric models is also presented. [source] Efficiency, Cointegration and Contagion in Equity Markets: Evidence from China, Japan and South Korea,ASIAN ECONOMIC JOURNAL, Issue 1 2009A.S.M. Sohel Azad C14; C32; G14; G15 This paper empirically examines whether three East Asian stock markets, namely, those of China, Japan and South Korea, are individually and/or jointly efficient, and whether contagion exists between the cointegrated markets. While individual market efficiency is examined through testing for the random walk hypothesis, joint market efficiency is examined through testing for cointegration and contagion. The present study finds that the hypothesis of individual market efficiency is strongly rejected for the Chinese stock market, but not for the Japanese and the South Korean stock markets. However, when testing for cointegration, market efficiency is strongly rejected for all these markets. We take a simple case of contagion and find that although there is a long-term relationship among the three markets, the contagion hypothesis cannot be rejected only between Japanese and South Korean stock markets, indicating short-run portfolio diversification benefits from these two markets. [source] Determinants of Female Fertility in Taiwan, 1966,2001: Empirical Evidence from Cointegration and Variance Decomposition Analysis,ASIAN ECONOMIC JOURNAL, Issue 4 2006Paresh Kumar Narayan J13; C22; C52 This study investigates the determinants of the fertility rate in Taiwan over the period 1966,2001. Consistent with theory, the key explanatory variables in Taiwan's fertility model are real income, infant mortality rate, female education and female labor force participation rate. The test for cointegration is based on the recently developed bounds testing procedure while the long-run and short-run elasticities are based on the autoregressive distributed lag model. Among our key results, female education and female labor force participation rate are found to be the key determinants of fertility in Taiwan in the long run. The variance decom-position analysis indicates that in the long run approximately 45percent of the variation in fertility is explained by the combined impact of female labor force participation, mortality and income, implying that socioeconomic development played an important role in the fertility transition in Taiwan. This result is consistent with the traditional structural hypothesis. [source] Vector Autoregression (Var) , An Approach to Dynamic analysis of Geographic ProcessesGEOGRAFISKA ANNALER SERIES B: HUMAN GEOGRAPHY, Issue 2 2001Max Lu Vector autoregression (VAR) is a widely used econometric technique for multivariate time series modelling. This paper shows that with several very attractive features, VAR may also provide a valuable tool for analysing the dynamics among geographic processes and for spatial autoregressive modelling. After a brief discussion of the VAR approach, a VAR model for the dynamics of the US population between 1910 and 1990 is estimated and interpreted to illustrate the techniques. The VAR makes it possible to view the interactions among the four variables used in the model (total population, birth rate, immigration and per capita GNP) more adequately. The paper then discusses recent developments in the VAR methodology such as Bayesian vector autoregression (BVAR), spatial prior for regional modelling and cointegration, as well as the limitations and problems that arise from the application of VARs. [source] Long-Run Links among Money, Prices and Output: Worldwide EvidenceGERMAN ECONOMIC REVIEW, Issue 1 2006Helmut Herwartz Quantity theory of money; P-star; panel cointegration analysis. Abstract. Starting from the quantity theory of money we analyse the dynamic relationships between money, real output and prices for an unbalanced panel of 110 economies. Complementary to trivariate analyses we also adopt a P-star model explaining inflation via an equilibrium price level (P-star), which in turn depends on potential output and money. A key issue of the paper is the cross-sectional stability of estimation and inference results. We find cointegration among the considered variables. Particularly for high inflation countries homogeneity between prices and money cannot be rejected. Given homogeneity we find evidence for an error-correction mechanism linking current price changes and the lagged price gap. Parameter estimates indicating the adjustment towards the price equilibrium are larger in absolute value for high inflation countries. The latter results indicate that central banks, even in high inflation countries, can improve price stability by controlling monetary growth. [source] Measuring Monetary Policy in Germany: A Structural Vector Error Correction ApproachGERMAN ECONOMIC REVIEW, Issue 3 2003Imke Brüggemann Monetary policy; cointegration; structural VAR analysis Abstract. A structural vector error correction (SVEC) model is used to investigate several monetary policy issues. While being data-oriented the SVEC framework allows structural modeling of the short-run and long-run properties of the data. The statistical model is estimated with monthly German data for 1975,98 where a structural break is detected in 1984. After splitting the sample, three stable long-run relations are found in each subsample which can be interpreted in terms of a money-demand equation, a policy rule and a relation for real output, respectively. Since the cointegration restrictions imply a particular shape of the long-run covariance matrix this information can be used to distinguish between permanent and transitory innovations in the estimated system. Additional restrictions are introduced to identify a monetary policy shock. [source] The role of permanent and transitory shocks in explaining international health expendituresHEALTH ECONOMICS, Issue 10 2008Paresh Kumar Narayan Abstract While there is a growing literature that examines the issue of cointegration (co-movement over the long run) among health expenditures, there are no studies that examine the issue of common cycles (co-movement over the short run) among health expenditures. We undertake a multivariate variance decomposition analysis of per capita health expenditures of the USA, the UK, Japan, Canada, and Switzerland based on a common-trend,common-cycle restriction framework, to examine the relative importance of permanent and transitory innovations in explaining variations in per capita health expenditures in each of the five countries. Our main finding is that transitory shocks are more important in explaining per capita health expenditures in the UK, Japan, and Switzerland, while permanent shocks dominate variations in per capita health expenditures in the USA and Canada over short horizons. Copyright © 2007 John Wiley & Sons, Ltd. [source] The real exchange rate,real interest rate relation: evidence from tests for symmetric and asymmetric threshold cointegrationINTERNATIONAL JOURNAL OF FINANCE & ECONOMICS, Issue 2 2006Robert Sollis Abstract This paper investigates the existence of threshold cointegration between real exchange rates and real interest rate differentials. Unlike previous work, which generally fails to find evidence of a long-run relationship employing linear models, we employ tests of the null hypothesis of no cointegration derived from nonlinear bivariate models that allow for threshold cointegration under the alternative hypothesis. For six of the countries in our sample our analysis reveals some evidence of a nonlinear long-run relationship between real exchange rates and real interest rate differentials. Asymmetric mean reversion of the equilibrium error is found to be driven by the asymmetric short-run adjustment of the real exchange rate to dis-equilibrium. When threshold cointegration is found to exist, we find stronger mean reversion when the equilibrium error is negative relative to when it is positive. Copyright © 2006 John Wiley & Sons, Ltd. [source] Indexing, cointegration and equity market regimesINTERNATIONAL JOURNAL OF FINANCE & ECONOMICS, Issue 3 2005Carol Alexander Abstract This paper examines, from a market efficiency perspective, the performance of a simple dynamic equity indexing strategy based on cointegration. A consistent ,abnormal' return in excess of the benchmark is demonstrated over different time horizons and in different real world and simulated stock markets. A measure of stock price dispersion is shown to be a leading indicator for the abnormal return and their relationship is modelled as a Markov switching process of two market regimes. We find that the entire abnormal return is associated with the high volatility regime as the indexing model implicitly adopts a strategic position that pays off during market crashes, whilst effectively tracking the benchmark in normal market circumstances. Therefore we find no evidence of market inefficiency. Nevertheless our results have implications for equity fund managers: we show how, without any stock selection, solely through a smart optimization that has an implicit element of market timing, the benchmark performance can be significantly enhanced. Copyright © 2005 John Wiley & Sons, Ltd. [source] Market based debt reduction agreements: a case study on Mexican and Polish Brady bondsINTERNATIONAL JOURNAL OF FINANCE & ECONOMICS, Issue 2 2001Luca Barbone F34; G14 Abstract This paper analyzes some aspects of the workings of the Brady bond (restructured Less Developed Countries debt) market. It concentrates on the effects of the December 1994 Mexican crisis on the risk assessment (as measured by the stripped spread) of Poland, another Brady country. The main findings are: (i) over the sample period, the unit root hypothesis on the risk premium (measured by the stripped spread) of Mexico and Poland cannot be rejected; this is consistent with the idea that the risk premium reflects new information accruing to the market; (ii) comovements in stripped spreads between Mexico and Poland were stronger during the period of the Peso crisis: we do not reject the null of cointegration for the year that includes the crisis (July 1994,July 1995), but we do reject the null for the year starting 6 months after the crisis (July 1995,July 1996); (iii) the crisis has had a strong permanent effect on the risk assessment of Mexico with respect to the one of Poland (550 basis points circa). Copyright © 2001 John Wiley & Sons, Ltd. [source] Tourism demand modelling: some issues regarding unit roots, co-integration and diagnostic testsINTERNATIONAL JOURNAL OF TOURISM RESEARCH, Issue 5 2003Paresh Kumar Narayan Abstract This paper investigates the all important issue of diagnostic tests, including unit roots and cointegration, in the tourism demand modelling literature. The origins of this study lie in the apparent lack in the tourism economics literature of detail concerning the diagnostic test aspect. Study of this deficiency has suggested that previous literature on tourism demand modelling may be divided into two categories: the pre-1995 and post-1995 studies. It was found that the pre-1995 and some post-1995 studies have ignored unit root tests and co-integration and, hence, are vulnerable to the so-called ,spurious regression' problem. In highlighting the key diagnostic tests reported by post-1995 studies, this paper contends that there is no need to report the autoregressive conditional heteroskedasticity (ARCH) test, which is applicable only to financial market analysis where the dependent variable is return on an asset. More generally, heteroskedasticity is not seen as a problem in time-series data. However, the reporting of a greater than necessary range of diagnostic tests,,,some of which do not have any theoretical justification with regard to tourism demand analysis,,,does not diminish the precision of the results or the model. This paper should appeal to scholars involved in tourism demand modelling. Copyright © 2003 John Wiley & Sons, Ltd. [source] Evaluating Commodity Market Efficiency: Are Cointegration Tests Appropriate?JOURNAL OF AGRICULTURAL ECONOMICS, Issue 3 2002Neil Kellard This paper investigates the claim that the finding of cointegration between commodity spot and lagged futures rates reflects the existence of commodity arbitrage and not, as is generally accepted, long-run market efficiency. The methodology of Kellard et al. (1999) is employed to match spot and lagged futures rates correctly for the UK wheat futures contract traded at LIFFE. Bi-variate analysis shows that spot and lagged futures rates are cointegrated with the vector (1, -1), a necessary condition for market efficiency. However, at variance with asymptotic theory, in a tri-variate VECM estimation, the spot rate, lagged futures rate and lagged domestic interest rate are shown to be cointegrated with the vector (1, ,1, 1). The "cointegration" paradox is explained by investigating the relative magnitudes of the forecast error and the domestic interest rate. The small sample results demonstrate that it is impossible to distinguish between the influence of commodity arbitrage and the existence of market efficiency using cointegration-based tests. In summary, this work implies that such tests are not wholly appropriate for evaluating commodity market efficiency. [source] Generalized long memory processes, failure of cointegration tests and exchange rate dynamicsJOURNAL OF APPLIED ECONOMETRICS, Issue 4 2006Aaron D. Smallwood This paper presents evidence that the equilibrium relationship in a system of nominal exchange rates is best described as a stationary GARMA process. The implementation of the GARMA methodology helps explain conflicting and puzzling results from the use of linear cointegration and fractional cointegration methods. Furthermore, we use Monte Carlo analysis to document problems with standard cointegration tests when the attraction process is distributed as a long memory GARMA process. Copyright © 2006 John Wiley & Sons, Ltd. [source] Modelling the trend and seasonals within an AIDS model of the demand for alcoholic beverages in the United KingdomJOURNAL OF APPLIED ECONOMETRICS, Issue 2 2002I. A. Moosa The argument that is put forward in this paper is that failure to represent stochastic trend and stochastic seasonality in an AIDS model leads to a misspecified and possibly structurally unstable model. This proposition is verified by estimating an AIDS model of the demand for alcoholic beverages in the United Kingdom. Three versions of the model are estimated, and it is demonstrated that the version allowing for stochastic trend and stochastic seasonality performs better than the other two versions of the model in terms of the diagnostics tests and goodness of fit measures. The best estimated model turns out to possess the properties of having common components and being homogenous. Further empirical testing reveals the presence of stochastic trends and cointegration between the budget shares of beer and wine. The results clearly indicate that there has been a shift away from the consumption of beer towards wine. Copyright © 2002 John Wiley & Sons, Ltd. [source] The Information Signaling Hypothesis of Dividends: Evidence from Cointegration and Causality TestsJOURNAL OF BUSINESS FINANCE & ACCOUNTING, Issue 3-4 2003Mbodja Mougoué This paper uses cointegration and causality tests to study the temporal behavior of dividends and earnings at the individual firm level. We find that, for a sample of 143 non-utility firms, approximately one-fifth of the firms exhibits a temporal relationship between dividends and earnings that is consistent with the information signaling hypothesis of dividends. In the case of 72 utilities, about a third exhibit dividend policies that are consistent with the signaling notion of dividends. Further examination of firm characteristic differences between signaling and non-signaling firms shows that, in the case of non-utility firms, signaling firms tend to be smaller, have a lower growth rate of total assets, and have a higher leverage ratio. In the case of utilities, we find no major differences in firm characteristics between signaling and non-signaling firms. [source] Cointegration, Efficiency and Forecasting in the Currency MarketJOURNAL OF BUSINESS FINANCE & ACCOUNTING, Issue 1-2 2001Wilson H. S. Tong Existing literature on using the cointegration approach to examine the efficiency of the foreign exchange market gives mixed results. Arguments typically focus on econometric testing techniques, with fractional cointegration being the most current one. This paper tries to look at the issue from an economic perspective. It shows that the cointegrating relationship, whether cointegrated or fractionally cointegrated, is found mainly among the currencies of the European Monetary System which are set to fluctuate within a given range. Hence, there is no inconsistency with the notion of market efficiency. Yet, exploiting such a cointegrating relationship is helpful in currency forecasting. There is some evidence that restricting the forecasting model to consist of only cointegrated currencies improves forecasting efficiency. [source] Pricing of Forward and Futures ContractsJOURNAL OF ECONOMIC SURVEYS, Issue 2 2000Ying-Foon Chow There has long been substantial interest in understanding the relative pricing of forward and futures contracts. This has led to the development of two standard theories of forward and futures pricing, namely, the Cost-of-Carry and the Risk Premium (or Unbiased Expectations) hypotheses. These studies have modelled the relationship between spot and forward/futures prices either through a no-arbitrage condition or a general equilibrium setting. Relatively few studies in this area have considered the impact of stochastic trends in the data. With the emergence of non-stationarity and cointegration in recent years, more sophisticated models of futures/forward prices have been specified. This paper surveys the significant contributions made to the literature on the pricing of forward/futures contracts, and examines recent empirical studies pertaining to the estimation and testing of univariate and systems models of futures pricing. [source] Vertical price leadership: A cointegration analysisAGRIBUSINESS : AN INTERNATIONAL JOURNAL, Issue 3 2002W. Erno Kuiper Here we detail a method to test whether or not retailers allow suppliers to set the wholesale price not only on the basis of the costs faced by the suppliers but also on the basis of consumer demand. Using standard theory, long-run price relationships between the stages in the channel are derived. Next, these static price relationships are imposed on a dynamic model to be tested for cointegration and long-run noncausality, embedding the hypotheses on vertical price leadership. To derive the testable implications of these hypotheses, we show that the common stochastic trend and long-run equilibrium error must explicitly be assigned to variables in the channel model. The model is particularly relevant for industries characterized by a low degree of product differentiation. An empirical application to two Dutch marketing channels for food products gives comprehensible results. [EconLit citations: C32, L12, Q11] © 2002 Wiley Periodicals, Inc. [source] Price relationships in the Queensland barley marketAGRIBUSINESS : AN INTERNATIONAL JOURNAL, Issue 2 2002V. Jyothi Gali Barley can be differentiated into feed and malting barley based on its end-use markets. Substitutability both in supply and in demand complicates analysis of price information in the barley market. This article examines the price linkages between feed and malting barley in the Queensland barley market using cointegration and error correction models. Malting barley prices respond to restore equilibrium relationships with corresponding feed barley prices in the long run, but not vice versa. Thus feed barley prices appear to be a leading indicator of malting barley prices. [JEL codes: L100, C22, N57.] © 2002 Wiley Periodicals, Inc. [source] Estimation and forecasting in first-order vector autoregressions with near to unit roots and conditional heteroscedasticityJOURNAL OF FORECASTING, Issue 7 2009Theologos Pantelidis Abstract This paper investigates the effects of imposing invalid cointegration restrictions or ignoring valid ones on the estimation, testing and forecasting properties of the bivariate, first-order, vector autoregressive (VAR(1)) model. We first consider nearly cointegrated VARs, that is, stable systems whose largest root, lmax, lies in the neighborhood of unity, while the other root, lmin, is safely smaller than unity. In this context, we define the ,forecast cost of type I' to be the deterioration in the forecasting accuracy of the VAR model due to the imposition of invalid cointegration restrictions. However, there are cases where misspecification arises for the opposite reasons, namely from ignoring cointegration when the true process is, in fact, cointegrated. Such cases can arise when lmax equals unity and lmin is less than but near to unity. The effects of this type of misspecification on forecasting will be referred to as ,forecast cost of type II'. By means of Monte Carlo simulations, we measure both types of forecast cost in actual situations, where the researcher is led (or misled) by the usual unit root tests in choosing the unit root structure of the system. We consider VAR(1) processes driven by i.i.d. Gaussian or GARCH innovations. To distinguish between the effects of nonlinear dependence and those of leptokurtosis, we also consider processes driven by i.i.d. t(2) innovations. The simulation results reveal that the forecast cost of imposing invalid cointegration restrictions is substantial, especially for small samples. On the other hand, the forecast cost of ignoring valid cointegration restrictions is small but not negligible. In all the cases considered, both types of forecast cost increase with the intensity of GARCH effects. Copyright © 2009 John Wiley & Sons, Ltd. [source] | |||||||||||||||||||||||||||||||