Report

ECON 6002 Econometrics Memorial University of Newfoundland Nonstationary Time Series Data and Cointegration Adapted from Vera Tabakova’s notes 12.1 Stationary and Nonstationary Variables 12.2 Spurious Regressions 12.3 Unit Root Tests for Stationarity 12.4 Cointegration 12.5 Regression When There is No Cointegration Principles of Econometrics, 3rd Edition Slide 12-2 Fluctuates about a rising trend Yt-Y t-1 On the right hand side Fluctuates about a zero mean “Differenced series” Figure 12.1(a) US economic time series Principles of Econometrics, 3rd Edition Slide 12-3 Yt-Y t-1 On the right hand side “Differenced series” Figure 12.1(b) US economic time series Principles of Econometrics, 3rd Edition Slide 12-4 Stationary if: E yt (12.1a) var yt 2 (12.1b) cov yt , yt s cov yt , yt s s Principles of Econometrics, 3rd Edition (12.1c) Slide 12-5 Principles of Econometrics, 3rd Edition Slide 12-6 yt yt 1 vt , 1 (12.2a) Each realization of the process has a proportion rho of the previous one plus an error drawn from a distribution with mean zero and variance sigma squared It can be generalised to a higher autocorrelation order We just show AR(1) Principles of Econometrics, 3rd Edition Slide 12-7 yt yt 1 vt , 1 (12.2a) We can show that the constant mean of this series is zero y1 y0 v1 y2 y1 v2 (y0 v1 ) v2 2 y0 v1 v2 yt vt vt 1 2vt 2 ..... t y0 E[ yt ] E[vt vt 1 2vt 2 .....] 0 Principles of Econometrics, 3rd Edition Slide 12-8 We can also allow for a non-zero mean, by replacing yt with yt-mu ( yt ) ( yt 1 ) vt Which boils down to (using alpha = mu(1-rho)) yt yt 1 vt , 1 (12.2b) E ( yt ) / (1 ) 1/ (1 0.7) 3.33 Principles of Econometrics, 3rd Edition Slide 12-9 Or we can allow for a AR(1) with a fluctuation around a linear trend (mu+delta times t) The “de-trended” model , which is now stationary, behaves like an autoregressive model: ( yt t ) ( yt 1 (t 1)) vt , yt yt 1 t vt 1 (12.2c) With alpha =(mu(1-rho)+rho times delta) And lambda = delta(1-rho) Principles of Econometrics, 3rd Edition Slide 12-10 Figure 12.2 (a) Time Series Models Principles of Econometrics, 3rd Edition Slide 12-11 Figure 12.2 (b) Time Series Models Principles of Econometrics, 3rd Edition Slide 12-12 Figure 12.2 (c) Time Series Models Principles of Econometrics, 3rd Edition Slide 12-13 yt yt 1 vt (12.3a) y1 y0 v1 2 y2 y1 v2 ( y0 v1 ) v2 y0 vs s 1 t yt yt 1 vt y0 vs s 1 Principles of Econometrics, 3rd Edition The first component is usually just zero, since it is so far in the past that it has a negligible effect now The second one is the stochastic trend Slide 12-14 A random walk is non-stationary, although the mean is constant: E ( yt ) y0 E (v1 v2 ... vt ) y0 var( yt ) var(v1 v2 ... vt ) tv2 Principles of Econometrics, 3rd Edition Slide 12-15 yt yt 1 vt (12.3b) A random walk with a drift both wanders and trends: y1 y0 v1 2 y2 y1 v2 ( y0 v1 ) v2 2 y0 vs s 1 t yt yt 1 vt t y0 vs s 1 Principles of Econometrics, 3rd Edition Slide 12-16 E ( yt ) t y0 E (v1 v2 v3 ... vt ) t y0 var( yt ) var(v1 v2 v3 ... vt ) tv2 Principles of Econometrics, 3rd Edition Slide 12-17 rw1 : yt yt 1 v1t rw2 : xt xt 1 v2t Both independent and artificially generated, but… rw1t 17.818 0.842 rw2t , (t ) Principles of Econometrics, 3rd Edition R 2 .70 (40.837) Slide 12-18 Figure 12.3 (a) Time Series of Two Random Walk Variables Principles of Econometrics, 3rd Edition Slide 12-19 Figure 12.3 (b) Scatter Plot of Two Random Walk Variables Principles of Econometrics, 3rd Edition Slide 12-20 Dickey-Fuller Test 1 (no constant and no trend) yt yt 1 vt (12.4) yt yt 1 yt 1 yt 1 vt yt 1 yt 1 vt (12.5a) yt 1 vt Principles of Econometrics, 3rd Edition Slide 12-21 Dickey-Fuller Test 1 (no constant and no trend) H0 : 1 H0 : 0 H1 : 1 H1 : 0 Easier way to test the hypothesis about rho Remember that the null is a unit root: nonstationarity! Principles of Econometrics, 3rd Edition Slide 12-22 Dickey-Fuller Test 2 (with constant but no trend) yt yt 1 vt Principles of Econometrics, 3rd Edition (12.5b) Slide 12-23 Dickey-Fuller Test 3 (with constant and with trend) yt yt 1 t vt Principles of Econometrics, 3rd Edition (12.5c) Slide 12-24 First step: plot the time series of the original observations on the variable. If the series appears to be wandering or fluctuating around a sample average of zero, use Version 1 If the series appears to be wandering or fluctuating around a sample average which is non-zero, use Version 2 If the series appears to be wandering or fluctuating around a linear trend, use Version 3 Principles of Econometrics, 3rd Edition Slide 12-25 Principles of Econometrics, 3rd Edition Slide 12-26 An important extension of the Dickey-Fuller test allows for the possibility that the error term is autocorrelated. m yt yt 1 as yt s vt (12.6) s 1 yt 1 yt 1 yt 2 , yt 2 yt 2 yt 3 , The unit root tests based on (12.6) and its variants (intercept excluded or trend included) are referred to as augmented Dickey-Fuller tests. Principles of Econometrics, 3rd Edition Slide 12-27 F = US Federal funds interest rate Ft 0.178 0.037 Ft 1 0.672Ft 1 (tau ) ( 2.090) B = 3-year bonds interest rate Bt 0.285 0.056 Bt 1 0.315Bt 1 (tau ) Principles of Econometrics, 3rd Edition ( 1.976) Slide 12-28 In STATA: use usa, clear gen date = q(1985q1) + _n - 1 format %tq date tsset date TESTING UNIT ROOTS “BY HAND”: * Augmented Dickey Fuller Regressions regress D.F L1.F L1.D.F regress D.B L1.B L1.D.B Principles of Econometrics, 3rd Edition Slide 12-29 In STATA: . regress D.F L1.F ROOTS L1.D.F TESTING UNIT “BY HAND”: Source SS Fuller dfRegressions MS * Augmented Dickey 7.99989546 regressModel D.F L1.F L1.D.F 2 3.99994773 Residual 9.54348876 76 .12557222 regress D.B L1.B L1.D.B Total 17.5433842 78 Number of obs F( 2, 76) Prob > F R-squared Adj R-squared Root MSE .224915182 t P>|t| 79 31.85 0.0000 0.4560 0.4417 .35436 D.F Coef. F L1. LD. -.0370668 .6724777 .0177327 .0853664 -2.09 7.88 0.040 0.000 -.0723847 .5024559 -.001749 .8424996 _cons .1778617 .1007511 1.77 0.082 -.0228016 .378525 Principles of Econometrics, 3rd Edition Std. Err. = = = = = = [95% Conf. Interval] Slide 12-30 In STATA: Augmented Dickey Fuller Regressions with built in functions dfuller F, regress lags(1) dfuller B, regress lags(1) Choice of lags if we want to allow For more than a AR(1) order Principles of Econometrics, 3rd Edition Slide 12-31 In STATA: Augmented Dickey Fuller Regressions with built in functions dfuller F, regress lags(1) . dfuller F, regress lags(1) . dfuller F, regress lags(1) Augmented Dickey-Fuller test for unit root Augmented Dickey-Fuller test for unit root TestTest Statistic Statistic Z(t) Z(t) Number of obs Number of obs = = 79 79 Interpolated Dickey-Fuller Interpolated Dickey-Fuller 1% 1% Critical 5% 10%Critical Critical Critical 5% Critical Critical 10% Value Value Value Value Value Value -2.090 -2.090 -3.539 -3.539 -2.907 -2.907 -2.588 -2.588 MacKinnon approximate p-value Z(t)= =0.2484 0.2484 MacKinnon approximate p-value forfor Z(t) D.F D.F Coef. Std. Std. Err. Coef. Err. tt P>|t| P>|t| [95% [95% Conf. Conf.Interval] Interval] F F L1. -.0370668 .0177327 L1. LD.-.0370668 .6724777 .0177327 .0853664 LD. .6724777 .0853664 _cons _cons .1778617 .1778617 Principles of Econometrics, 3rd Edition .1007511 .1007511 -2.09 7.88 7.88 -2.09 0.040 0.000 0.000 0.040 -.0723847 -.0723847 .5024559 1.77 0.082 -.0228016 1.77 0.082 .5024559 -.0228016 -.001749 -.001749 .8424996 .8424996 .378525 .378525 Slide 12-32 In STATA: Augmented Dickey Fuller Regressions with built in functions dfuller F, regress lags(1) Alternative: pperron uses Newey-West standard errors to account for serial correlation, whereas the augmented Dickey-Fuller test implemented in dfuller uses additional lags of the first-difference variable. Also consider now using DFGLS (Elliot Rothenberg and Stock, 1996) to counteract problems of lack of power in small samples. It also has in STATA a lag selection procedure based on a sequential t test suggested by Ng and Perron (1995) Principles of Econometrics, 3rd Edition Slide 12-33 In STATA: Augmented Dickey Fuller Regressions with built in functions dfuller F, regress lags(1) Alternatives: use tests with stationarity as the null KPSS (Kwiatowski, Phillips, Schmidt and Shin. 1992) which also has an automatic bandwidth selection tool or the Leybourne & McCabe test . Principles of Econometrics, 3rd Edition Slide 12-34 yt yt yt 1 vt The first difference of the random walk is stationary It is an example of a I(1) series (“integrated of order 1” First-differencing it would turn it into I(0) (stationary) In general, the order of integration is the minimum number of times a series must be differenced to make it stationarity Principles of Econometrics, 3rd Edition Slide 12-35 yt yt yt 1 vt F t 0.340 F t 1 (tau ) ( 4.007) So now we reject the Unit root after differencing once: We have a I(1) series B t 0.679 B t 1 (tau ) Principles of Econometrics, 3rd Edition ( 6.415) Slide 12-36 In STATA: ADF on differences dfuller D.F, noconstant lags(0) dfuller D.B, noconstant lags(0) . dfuller F, regress lags(1) Augmented Dickey-Fuller test for unit root Number of obs = 79 Interpolated Dickey-Fuller 5% Critical 10% Critical Value Value Number of obs = 79 -3.539 -2.907 -2.588 Interpolated Dickey-Fuller Test 1% Critical 5% Critical 10% Critical MacKinnon approximate p-value for Z(t) = 0.2484 Statistic Value Value Value . dfuller D.F, noconstantTest lags(0) Statistic Dickey-Fuller test for unit root Z(t) -2.090 Z(t) -4.007 D.F Coef. 1% Critical Value -2.608 Std. Err. t -1.950 P>|t| -1.610 [95% Conf. Interval] F L1. LD. -.0370668 .6724777 .0177327 .0853664 -2.09 7.88 0.040 0.000 -.0723847 .5024559 -.001749 .8424996 _cons .1778617 .1007511 1.77 0.082 -.0228016 .378525 Slide 12-37 eˆt eˆt 1 vt (12.7) Case 1: eˆt yt bxt (12.8a) Case 2 : eˆt yt b2 xt b1 (12.8b) Case 3: eˆt yt b2 xt b1 ˆ t (12.8c) Principles of Econometrics, 3rd Edition Slide 12-38 If you have unit roots in the time series in your model, you risk the problem of spurious regressions However, spuriousness will not arise if those series are cointegrated, so that determining whether cointegration exists is also key The series are cointegrated if they follow the same stochastic trend or share an underlying common factor In that case you can find a linear combination of your nonstationary variables that is itself stationary Principles of Econometrics, 3rd Edition Slide 12-39 You must make sure that you have a balanced (potentially) cointegrating regression, so you want to find out the level of integration of your series (usually they are all I(1)) The coefficients in that linear combination form the cointegrating vector, which should have one of its elements normalized to one, because the cointegrating vector is only defined up to a factor of proportionality The cointegrating vector may include a constant, in order to allow for unequal means of the two series Principles of Econometrics, 3rd Edition Slide 12-40 The estimator from a cointegrating regression is superconsistent Principles of Econometrics, 3rd Edition Slide 12-41 Two main approaches can be used to check if there is cointegration: The residual approach The error correction approach Principles of Econometrics, 3rd Edition Slide 12-42 Two main approaches can be used to check if there is cointegration: The residual approach. The classic Engle-Granger approach, based on testing whether the error of the (potentially) cointengrating regression is itself stationary The error correction approach, which test whether the error correction term is significant Principles of Econometrics, 3rd Edition Slide 12-43 Not the same as for dfuller, since the residuals are estimated errors no actual Ones (also no constant!) Note: unfortunately STATA dfuller as such will not notice and give you erroneous critical values! They would lead to an overoptimistic conclusion Principles of Econometrics, 3rd Edition Slide 12-44 Bˆt 1.644 0.832Ft , R 2 0.881 (12.9) (t ) (8.437) (24.147) eˆt 0.314eˆt 1 0.315eˆt 1 (tau ) (4.543) Check: These are wrong! . dfuller ehat, noconstant lags(1) Augmented Dickey-Fuller test for unit root Test Statistic Z(t) -4.543 Number of obs = 79 Interpolated Dickey-Fuller 1% Critical 5% Critical 10% Critical Value Value Value -2.608 -1.950 -1.610 Slide 12-45 Using egranger with option regress Engle-Granger test regression D._egresid Coef. _egresid L1. LD. -.3143204 .3147476 Std. Err. .0691906 .1021565 t -4.54 3.08 P>|t| 0.000 0.003 [95% Conf. Interval] -.4520965 .1113281 -.1765443 .5181671 eˆt 0.314eˆt 1 0.315eˆt 1 (tau ) (4.543) Now these are right! . egranger B F, regress lags(1) Augmented Engle-Granger test for cointegration Replacing variable _egresid... Number of lags = 1 Augmented Engle-Granger test for cointegration Number of lags = 1 Test1% Critical StatisticValue Test Statistic Z(t) Z(t) -4.543 -4.543 -3.515 Critical values from MacKinnon (1990, 2010) N (1st step) N (test) 1%5% Critical Critical Value Value = = 81 79 N (1st step) N (test) 5% Critical Value 10% Critical Value -2.898 -3.515 -2.586 -2.898 = = 81 79 10% Critical Value -2.586 Critical values from MacKinnon (1990, 2010) Slide 12-46 The null and alternative hypotheses in the test for cointegration are: H 0 : the series are not cointegrated residuals are nonstationary H1 : the series are cointegrated residuals are stationary Principles of Econometrics, 3rd Edition Slide 12-47 Let us consider the simple form of a dynamic model: Here the SR and LR effects are measured respectively by: Rearranging terms, we obtain the usual ECM: Slide 12-48 Where the LR effect will be given by: And is a partial correction term for the extent to which Yt-1 deviated from its Equilibrium value associated with Xt-1 Slide 12-49 This representation assumes that any short-run shock to Y that pushes it off the long-run equilibrium growth rate will gradually be corrected, and an equilibrium rate will be restored is the residual of the long-run equilibrium relationship between X and Y and its Coefficient can be seen as the “speed of adjustment” Slide 12-50 Tthis representation assumes that any short-run shock to Y that pushes it off the long-run equilibrium growth rate will gradually be corrected, and an equilibrium rate will be restored Usually So the SR effect is weaker than the LR effect Slide 12-51 . egranger B F, ecm Replacing variable _egresid... Engle-Granger 2-step ECM estimation N (1st step) N (2nd step) = = 81 80 Engle-Granger 2-step ECM D.B Coef. Std. Err. t P>|t| [95% Conf. Interval] _egresid L1. -.2357887 .0723679 -3.26 0.002 -.3798917 -.0916858 F D1. .7529687 .1080218 6.97 0.000 .5378699 .9680675 _cons -.0315012 .0477661 -0.66 0.512 -.1266156 .0636133 Slide 12-52 If you have cointegration, you can run an Error Correction Model, so you can estimate both the long run and the short run relationship between the relevant variables The integration of the variables suggests that we should not use them in a regression, but rather only their differences. We may obtain inconsistent estimates (the spurious regression problem) However, the fact that they are cointegrated (a weighted average of the variables is stationary, I(0)) means that you can include linear combinations of the variables in regressions of their differences in and Error Correction Model (ECM) Slide 12-53 By having already concluding that the variables are cointegrated, we have implicitly decided that there is a long-run causal relation between them. Then the causality being tested for in a VECM is sometimes called “short-run Granger causality” Slide 12-54 The ECM analysis can show (by the magnitude and significance of the EC terms) that when values of the relevant variables move away from the equilibrium relationship implied by the contegrating vector, there was a strong tendency for the variable(s) to change so that the equilibrium would be restored The ECM analysis under cointegration allows us not to throw away the information on the LR effect behind the relationship Slide 12-55 The EC term will be significant if there is a cointegrating relationship Therefore, you can test the existence of cointegration by looking at the significance of that coefficient Slide 12-56 12.5.1 First Difference Stationary yt yt 1 vt yt yt yt 1 vt The variable yt is said to be a first difference stationary series. Then we revert to the techniques we saw in Ch. 9 Principles of Econometrics, 3rd Edition Slide 12-57 Manipulating this one you can construct and Error Correction Model to investigate the SR dynamics of the relationship between y and x yt yt 1 0xt 1xt 1 et (12.10a) yt yt 1 vt yt vt yt yt 1 0xt 1xt 1 et Principles of Econometrics, 3rd Edition (12.10b) Slide 12-58 yt t vt yt t vt yt yt1 0 xt 1xt1 et (12.11) yt t yt 1 0 xt 1xt 1 et where 1 (1 1 ) 2 (0 1 ) 11 12 and 1 (1 1 ) 2 (0 1 ) Principles of Econometrics, 3rd Edition Slide 12-59 To summarize: If variables are stationary, or I(1) and cointegrated, we can estimate a regression relationship between the levels of those variables without fear of encountering a spurious regression. Then we can use the lagged residuals from the cointegrating regression in an ECM model This is the best case scenario, since if we had to first-differentiate the variables, we would be throwing away the long-run variation Additionally, the cointegrated regression yields a “superconsistent” estimator in large samples Principles of Econometrics, 3rd Edition Slide 12-60 To summarize: If the variables are I(1) and not cointegrated, we need to estimate a relationship in first differences, with or without the constant term. If they are trend stationary, we can either de-trend the series first and then perform regression analysis with the stationary (de-trended) variables or, alternatively, estimate a regression relationship that includes a trend variable. The latter alternative is typically applied. Principles of Econometrics, 3rd Edition Slide 12-61 . Principles of Econometrics, 3rd Edition Slide 12-62 Augmented Dickey-Fuller test Autoregressive process Cointegration Dickey-Fuller tests Mean reversion Order of integration Random walk process Random walk with drift Spurious regressions Stationary and nonstationary Stochastic process Stochastic trend Tau statistic Trend and difference stationary Unit root tests Principles of Econometrics, 3rd Edition Slide 12-63 Further issues Kit Baum has really good notes on these topics that can be used to learn also about extra STATA commands to handle the analysis: http://fmwww.bc.edu/EC-C/S2003/821/EC821.sect05.nn1.pdf http://fmwww.bc.edu/EC-C/S2003/821/EC821.sect06.nn1.pdf For example, some of you should look at (quarterly) seasonal unit root analysis (command hegy4 in STATA implements the test suggested by Hylleberg et al. 1990) Panel unit roots would be here http://fmwww.bc.edu/EC-C/S2003/821/EC821.sect09.nn1.pdf Principles of Econometrics, 3rd Edition Slide 12-64 Further issues: more powerful tests A host of new tests have been developed to try and overcome the shortcomings of the first Dickey-Fuller ones Alternative: pperron uses Newey-West standard errors to account for serial correlation, whereas the augmented Dickey-Fuller test implemented in dfuller uses additional lags of the first-difference variable. Also consider now using DF-GLS (Elliot Rothenberg and Stock, 1996) to counteract problems of lack of power in small samples. It also has in STATA a lag selection procedure based on a sequential t test suggested by Ng and Perron (1995) that uses a Modified AIC (command dfgls) Principles of Econometrics, 3rd Edition Slide 12-65 Further issues: reverting the null Some tests use stationarity as the null hypothesis: See kpss which implements the test suggested by Kwiatowski etal. (1992) Kwiatowski, D., Phillips, P. C. B., Schmidt, P. and Shin, Y. (1992). `Testing the Null Hypothesis of Stationarity against the Alternative of a Unit Root', Journal of Econometrics, 54, 91-115. See also Leybourne, S.J. and B.P.M. McCabe. A consistent test for a unit root. Journal of Business and Economic Statistics, 12, 1994, 157-166. This type of test is complementary to the dfuller-type ones, so if they do not give you consistent results, you might be facing fractional unit roots or long range dependence [See lomodrs in Stata to learn more and Baum et al. 1999] Principles of Econometrics, 3rd Edition Slide 12-66 In STATA: Principles of Econometrics, 3rd Edition Slide 12-67 Further issues: unit root tests and structural breaks You may want to consider unit root tests that allow for structural Breaks, otherwise with a more basic test you might think you are detecting a unit root, while all you have is a structural break See Perron, Pierre. 1989. The Great Crash, The Oil Price Shock and the Unit Root Hypothesis. Econometrica, 57, 1361–1401. Perron, P. 1990. Testing for a unit root in a time series with a changing mean, Journal of Business and Economic Statistics, 8:2, 153-162. Perron, Pierre. 1997. Further Evidence on Breaking Trend Functions in Macroeconomic Variables. Journal of Econometrics, 80, 355–385. Perron, P. and T. Vogelsang. 1992. Nonstationarity and level shifts with an application to purchasing power parity, Journal of Business and Economic Statistics, 10:3, 301-320. You can also take a look at the literature review in this working paper: http://ideas.repec.org/p/wpa/wuwpot/0410002.html Principles of Econometrics, 3rd Edition Slide 12-68 Further issues: unit root tests and structural breaks You may want to consider unit root tests that allow for structural Breaks: Stata has zandrews Zivot, E. & Andrews, W. K. Further Evidence on the Great Crash, the Oil Price Shock, and the Unit-Root Hypothesis Journal of Business and Economic Statistics, 1992, 10, 251-270 And Cleamo1 Cleamao2 Clemio1 Clemio2 Clemente, J., Montañes, A. and M. Reyes. 1998. Testing for a unit root in variables with a double change in the mean, Economics Letters, 59, 175-182 Principles of Econometrics, 3rd Edition Slide 12-69 Further issues Apart from the fact that in your cointegration relationship you must choose one variable to be the regressand (giving it a coefficient of one) When you deal with more than 2 regressors you should consider the Johansen’s method to examine the cointegration relationships This is because when there are more than 2 variables involved, there can be multiple cointegrating relationships!!! In this case, you we exploit the notion of Vector Autoregression (VAR) Models that involve a structural view of the dynamics of several variables The generalization of these VAR techniques in this case resulted in the Vector Error Correction Models (VECM) Principles of Econometrics, 3rd Edition Slide 12-70 Further issues Apart from the fact that in your cointegration relationship you must choose one variable to be the regressand (giving it a coefficient of one) When you deal with more than 2 regressors you should consider the Johansen’s method to examine the cointegration relationships You can use vecrank in Stata to run this test Johansen, S. and K. Juselius. 1990. Maximum likelihood estimation and inference on cointegration with applications to the demand for money. Oxford Bulletin of Economics and Statistics, 522, 169–210. Principles of Econometrics, 3rd Edition Slide 12-71 Further issues: unit root tests for panels Since: • many interesting relations involve relatively short time–series and • unit root tests are infamous when applied to a single time series for their low power there may be hope from tests that can be used on short series but available across a cross–section of countries, regions, firms, or industries Principles of Econometrics, 3rd Edition Slide 12-72 Further issues: unit root tests for panels We need to logically extend the unit roots testing machinery for univariate time series to the panel setting • We can choose the null • We need to consider how stationary (or nonstationary) a panel has to be for us to deem it all stationary (or nonstationary) • We can use a logic of pooling the series and finding one indicator or averaging the indicators we find in each series instead • We can use the residual approach or the ECM approach Principles of Econometrics, 3rd Edition Slide 12-73 Further issues: unit root tests for panels One key issue with panel unit root tests is that they should try and consider cross-sectional dependence Only the second-generation tests can account for it, the first-generation tests assume cross-sectional independence Principles of Econometrics, 3rd Edition Slide 12-74 Further issues: unit root tests for panels STATA offers: • MADFULLER for MADF test, which is an extension of the ADF test (not good for longitudinal panels) • The test's null hypothesis should be carefully considered will be violated if even only one of the series in the panel is stationary • A rejection should thus not be taken to indicate that each of the series is stationary Sarno, L. and M. Taylor, 1998. Real exchange rates under the current float: Unequivocal evidence of mean reversion. Economics Letters 60, 131–137. Taylor, M. and L. Sarno, 1998. The behavior of real exchange rates during the post–Bretton Woods period. Journal of 9 International Economics, 46, 281–312. Principles of Econometrics, 3rd Edition Slide 12-75 Further issues: unit root tests for panels STATA offers: Levin Lin Chu (old levinlin now xtunitroot llc) One of the first unit root tests for panel data, originally circulated in working paper form in 1992 and 1993, published, with Chu as a coauthor, in 2002 This model allows for two–way fixed effects and unit–specific time trends This test is a pooled Dickey–Fuller (or ADF) test, potentially with differing lag lengths across the units of the panel Unlike the MADF test, it works with short wide panels Assumes that the autoregressive parameter rho is identical for all cross section units (homogeneous alternatives) Principles of Econometrics, 3rd Edition Slide 12-76 Further issues: unit root tests for panels STATA offers: ipshin now xtunitroot ips The Im–Pesaran–Shin test extends the LLC to allow for heterogeneity in the value of rho (heterogeneous alternatives) Under the null, all series nonstationary; under the alternative, a fraction of the series are assumed to be stationary in contrast to the LLC test, which presumes that all series are stationary under the alternative hypothesis IPS use a group–mean Lagrange multiplier statistic to test the null hypothesis. The ADF regressions (which may be of differing lag lengths) are calculated for each series, and a standardized statistic computed as the average of the LM tests for each equation Im, K., Pesaran, M., and Y. Shin, 1997. Testing for unit roots in heterogeneous panels. Mimeo, Department of Applied Economics, University of Cambridge. Principles of Econometrics, 3rd Edition Slide 12-77 Further issues: unit root tests for panels STATA offers: hadrilm now xtunitroot hadri Hadri et al. LM test whose null hypothesis is that all series in the panel are stationary, just like the KPSS test differs from that of Dickey–Fuller style tests in assuming stationarity rather that nonstationarity Hadri, K., 2000. Testing for stationarity in heterogeneous panel data. Econometrics Journal, 3, 148–161. Principles of Econometrics, 3rd Edition Slide 12-78 Further issues: unit root tests for panels STATA offers: hadrilm now xtunitroot hadri Hadri et al. LM test whose null hypothesis is that all series in the panel are stationary, just like the KPSS test differs from that of Dickey–Fuller style tests in assuming stationarity rather that nonstationarity Hadri, K., 2000. Testing for stationarity in heterogeneous panel data. Econometrics Journal, 3, 148–161. Principles of Econometrics, 3rd Edition Slide 12-79 Further issues: unit root tests for panels STATA offers: nharvey The Nyblom–Harvey Test of Common Stochastic Trends Nyblom, J. and A. Harvey. Tests of common stochastic trends. Econometric Theory, 16, 2000, 176-199. Nyblom, J. and A. Harvey. Testing against smooth stochastic trends. Journal of Applied Econometrics, 16, 415–429. Nyblom, J. and T. Makelainen, 1983. Comparison of tests for the presence of random walk components in a simple linear model. Journal of the American Statistical Association, 78, 856–864. Principles of Econometrics, 3rd Edition Slide 12-80 Further issues: unit root tests for panels STATA offers: Breitung test Breitung, J. 2000. The local power of some unit root tests for panel data. In Advances in Econometrics, Volume 15: Nonstationary Panels, Panel Cointegration, and Dynamic Panels, ed. B. H. Baltagi, 161-178. Amsterdam: JAI Press. Breitung, J., and S. Das. 2005. Panel unit root tests under cross-sectional dependence. Statistica Neerlandica 59: 414-433. Harris-Tzavalis test Harris, D. and Inder, B. (1994). `ATest of the Null Hypothesis of Cointegration', in Non-Stationary Time Series Analysis and Cointegration, ed. C. Hargreaves, Oxford University Press, New York. Harris, R. D. F. and Tzavalis, E. (1999). `Inference for Unit Roots in Dynamic Panels where the Time Dimension is Fixed', Journal of Econometrics, 91, 201-226 Fisher-type tests (combining p-values) Maddala, G.S. andWu, S. (1999), A Comparative Study of Unit Root Tests with Panel Data and a new simple test, Oxford Bulletin of Economics and Statistics, 61, 631-652. Principles of Econometrics, 3rd Edition Slide 12-81 Further issues: unit root tests for panels Other tests Pedroni, P.L., 1999. Critical values for cointegration tests in heterogeneous panels with multiple regressors. Oxford Bulletin of Economics and Statistics 61 (4), 653–670. Pedroni, P.L., 2004. Panel cointegration; asymptotic and finite sample properties of pooled time series tests with an application to the purchasing power parity hypothesis. Econometric Theory 20 (3), 597–625. Principles of Econometrics, 3rd Edition Slide 12-82 Further issues: unit root tests for panels STATA offers: Pesaran’s pescadf (Lewandowski, 2007) and multipurt -- Running 1st and 2nd generation panel unit root tests for multiple variables and lags To detect cross-sectional dependence: xtcds and xtcd Pesaran, M. Hashem (2004) General Diagnostic Tests for Cross Section Dependence in Panels' IZA Discussion Paper No. 1240. Sarafidis, V. & De Hoyos, R. E. On Testing for Cross Sectional Dependence in Panel Data Models The Stata Journal, 2006, 6, StataCorp LP, vol. 6(4), pages 482-496 Principles of Econometrics, 3rd Edition Slide 12-83 Further issues: cointegration tests for panels We also need tests for cointegration in panels Both residual-based and ECM-based (see Breitung and Pesaran, 2005, for a review) Most residual-based cointegration tests, both in time series and in panels, require that the long-run parameters for the variables in their levels are equal to the short-run parameters for the variables in their differences The failure to meet this common-factor restriction can lead to a significant loss of power for residual-based cointegration tests Principles of Econometrics, 3rd Edition Slide 12-84 Further issues: unit root tests for panels, another issue STATA offers: xtwest a ECM-based cointegration test Persyn and Westerlund (2008) Westerlund, J. 2007. Testing for error correction in panel data. Oxford Bulletin of Economics and Statistics 69: 709–748. Principles of Econometrics, 3rd Edition Slide 12-85 Further issues: unit root tests for panels, more issues • We need to worry now also about the heterogeneity of panels (e.g. see xtpmg) • Blackburne, E. F. & Frank, M. W. Estimation of Nonstationary Heterogeneous Panels The Stata Journal, 2001, 7, 197-208 • Also, it is also now necessary to consider the possibility of cointegration between the variables across the groups (cross section cointegration) as well as within group cointegration Principles of Econometrics, 3rd Edition Slide 12-86