The properties of cointegration tests in models with structural change (original) (raw)
Related papers
Testing for Multiple Structural Changes in Cointegrated Regression Models
Journal of Business & Economic Statistics, 2010
This paper considers issues related to testing for multiple structural changes in cointegrated systems. We derive the limiting distribution of the Sup-Wald test under mild conditions on the errors and regressors for a variety of testing problems. We show that even if the coefficients of the integrated regressors are held fixed but the intercept is allowed to change, the limit distributions are not the same as would prevail in a stationary framework. Including stationary regressors whose coefficients are not allowed to change does not affect the limiting distribution of the tests under the null hypothesis. We also propose a procedure that allows one to test the null hypothesis of, say, k changes, versus the alternative hypothesis of k + 1 changes. This sequential procedure is useful in that it permits consistent estimation of the number of breaks present. We show via simulations that our tests maintain the correct size in finite samples and are much more powerful than the commonly used LM tests, which suffer from important problems of non-monotonic power in the presence of serial correlation in the errors.
A simple method of testing for cointegration subject to multiple regime changes
Economics Letters, 2002
In this paper we propose a simple method of testing for cointegration in models that allow for multiple shifts in the long-run relationship. The procedure consists of carrying out conventional residual-based tests with standardized residuals from an appropriate Markov switching model. Our Monte Carlo results show that standard tests work well, even though their asymptotic validity can be questioned because they are not based on least-squares residuals. An empirical application to the present-value model of stock prices is also discussed.
A simple method for testing cointegration subject to regime changes
2001
In this paper, we propose a simple method for testing cointegration in models that allow for multiple shifts in the long run relationship. The procedure consists of computing conventional residual-based tests with standardized residuals from Markov switching estimation. No new critical values are needed. An empirical application to the present value model of stock prices is presented, complemented by a small Monte Carlo experiment.
Residual-based tests for cointegration and multiple regime shifts
2001
In this paper we examine the properties of several cointegration tests when long run parameters are subject to multiple shifts, resorting to Monte Carlo methods. We assume that the changes in cointegration regimes are governed by a unobserved Markov chain process. This speci…cation has the considerable advantage of allowing for an unspeci…ed number of stochastic breaks, unlike previous works that consider a single, deterministic break. Our Monte Carlo analysis reveals that testing cointegration with the usual procedures is a quite unreliable task, since the performance of the tests is poor for a number of plausible regime shifts parameterizations. JEL Classi…cation: C12; C22; C52 changes may have on econometric models. Indeed, failure to detect and account for parameter shifts is a serious form of misspeci…cation, therefore a¤ecting inference and leading to poor forecasting performances (see Clements and Hendry, 1999, for example). This is especially relevant for cointegration analysis, since it normally involves long time spans of data, which, consequently, are likely to display structural breaks. Several papers deal with this possibility in a number of empirical applications, such as money demand, term structure of interest rates, purchasing power parity, among others (see references below). Therefore, it is natural to ask what is the impact of possible multiple parameter changes on the …nite-sample power and size properties of several cointegration tests. In this paper, we investigate this issue in a single-equation framework, resorting to Monte Carlo methods.
2009
This paper investigates whether or not multivariate cointegrated process with structural change can describe the Brazilian term structure of interest rate data from 1995 to 2006. In this work the break point and the number of cointegrated vector are assumed to be known. The estimated model has four regimes. Only three of them are statistically different. The first starts at the beginning of the sample and goes until September of 1997. The second starts at October of 1997 until December of 1998. The third starts at January of 1999 and goes until the end of the sample. It is used monthly data. Models that allows for some similarities across the regimes are also estimated and tested. The models are estimated using the Generalized Reduced-Rank Regressions developed by .
THE EFFECT OF STRUCTURAL BREAKS ON THE ENGLE-GRANGER TEST FOR COINTEGRATION
2012
Abstract: This paper extends Gonzalo and Lee's (1998) results by studying the asymptotic and finite sample behavior of the Engle-Granger test for cointegration, under misspecification of the trend function in the form of neglected structural breaks. We allow breaks in level and slope of trend in both dependent and explanatory variables. We also allow these processes to interact with I (1) processes without breaks.
Testing the Null of Cointegration with Structural Breaks
2005
In this paper we propose an LM-Type statistic to test the null hypothesis of cointegration allowing for the possibility of a structural break, in both the deterministic and the cointegration vector. Our proposal focuses on the presence of endogenous regressors and analyses which estimation method provides better results. The test has been designed to be used as a complement to the usual non-cointegration tests in order to obtain stronger evidence of cointegration. We consider the cases of known and unknown break date. In the latter case, we show that minimizing the SSR results in a super-consistent estimator of the break fraction. Finally, the behaviour of the tests is studied through Monte Carlo experiments.
Detrending procedures and cointegration testing: ECM tests under structural breaks
1998
It is well known that all the test for unit roots and cointegration depend on the deterministic elements that are in mean of the variables; constant, trend, breaks, outliers, segmented trends, etc. This is a serious inconvenient for empirical work. In this paper we analyze if those problems could be solved by forming the cointegration tests on extended models, on the components of the series obtained from trend cycle decompositions. We do that by Monte Carlo Simulations allowing for several structural breaks in the data generating process.
Tests for cointegration with structural breaks based on subsamples
Computational Statistics & Data Analysis, 2010
This paper considers tests for cointegration with allowance for structural breaks, using the extrema of residual-based tests over subsamples of the data. One motivation for the approach is to formalize the practice of data snooping by practitioners, who may examine subsamples after failing to find a predicted cointegrating relationship. Valid critical values for such multiple testing situations may be useful. The methods also have the advantage of not imposing a form for the alternative hypothesis, in particular slope vs. intercept shifts and single versus multiple breaks, and being comparatively easy to compute. A range of alternative subsampling procedures, including sample splits, incremental and rolling samples are tabulated and compared experimentally. Shiller's annual stock prices and dividends series provide an illustration.