{smcl} {* 19jul2007}{...} {hline} help for {hi:urcovar} {hline} {title:Perform Elliott-Jansson test for unit roots with stationary covariates} {title:Syntax} {p 8 14}{cmd:urcovar} {it:depvar} {it:varlist} [{cmd:if} {it:exp}] [{cmd:in} {it:range}] {bind:[{cmd:,}} {cmd:maxlag(}{it:#}{cmd:)} {cmd:case(}{it:#}{cmd:)} {cmd:firstobs} ] {p}{cmd:urcovar} is for use with time-series data. You must {cmd:tsset} your data before using {cmd:urcovar}; see {help tsset}. You may apply {cmd:urcovar} to a single time series of a panel dataset. {cmd:varlist} may contain time-series operators; see {cmd:help varlist}. {title:Description} {p}{cmd:urcovar} performs the test for a unit root in {it:depvar} in the context of one or more stationary covariates (listed in {it:varlist}) which was proposed by Elliott and Jansson (2003) as a generalization of the CADF test of Hansen (1995). The Elliott-Jansson test constructs a VAR in the model of the stationary covariates and the quasi-difference of {it:depvar}. As in the Dickey-Fuller ({cmd:dfuller}) or Elliott-Rothenberg-Stock DF-GLS ({cmd:dfgls}) unit root tests, the model can include no deterministic terms, constants or constants and time trends. The deterministic model is specified by the {cmd:case} option. {title:Options} {p 0 4}{it:varlist} specifies the set of stationary regressors. At least one must be provided. {p 0 4}{cmd:maxlag(}{it:#}{cmd:)} specifies the number of lags to be used in computing the VAR. If not provided, one lag is assumed. {p 0 4}{cmd:case(}{it:#}{cmd:)} specifies the deterministic features of the model. Case 1 specifies no constant nor trend in the representation of {it:depvar} and {it:varlist}. Case 2 allows a constant in the representation of {it:depvar}. Case 3 allows for constants in both {it:depvar} and {it:varlist}. Case 4 also allows for a trend in the representation of {it:depvar}. Case 5 imposes no restrictions, allowing for constants and trends in both {it:depvar} and {it:varlist}. If not specified, case 1 is tested. {p 0 4}{cmd:firstobs} specifies that the first observation of {it:depvar} should be used to define the first quasi-difference. If the option is not specified, that quasi-difference is set to zero. {title:Saved results} {p}{cmd:urcovar} returns the names of {it:depvar} and {it:varlist} and the value of {cmd:case} and {cmd:maxlag} in the return list. As well, the test statistic {cmd:lambda}, interpolated 5% critical value ({cmd:crit5}) and estimated R-squared ({cmd:r2}) are returned. See {cmd:return list}. {title:Examples} {p 8 12}{stata "use http://fmwww.bc.edu/ec-p/data/macro/blanquah" : . use http://fmwww.bc.edu/ec-p/data/macro/blanquah }{p_end} {p 8 12}{stata "urcovar PINCOME UNRATE, case(3) maxlag(8)" : . urcovar PINCOME UNRATE, case(3) maxlag(8)} {p 8 12}{stata "urcovar PINCOME UNRATE, case(5) maxlag(8)" : . urcovar PINCOME UNRATE, case(5) maxlag(8)} {title:Acknowledgements} I am grateful to Graham Elliott for clarifications of the logic of this routine. Remaining errors are my own. {title:References} {p 0 4}Elliott, G. and Jansson, M., 2003. Tests for Unit Roots with Stationary Covariates. Journal of Econometrics, Vol. 115, pp. 75-89. {p 0 4}Elliott, G. and Pesavento, E., 2006. On the Failure of Purchasing Power Parity for Bilateral Exchange Rates after 1973. Journal of Money, Credit and Banking, Vol. 38, pp. 1405-1430. {p 0 4}Hansen, B.E., 1995. Rethinking the univariate approach to unit root testing: using covariates to increase power. Econometric Theory, Vol. 11, 1148-1172. {title:Citation} {p}{cmd:urcovar} is not an official Stata command. It is a free contribution to the research community, like a paper. Please cite it as such: {p_end} {phang}Baum, C.F., 2007. urcovar: Stata module to perform Elliott-Jansson test for unit roots with stationary covariates. {browse "http://ideas.repec.org/c/boc/bocode/s456863.html":http://ideas.repec.org/c/boc/bocode/s456863.html}{p_end} {title:Author} {p 0 4}Christopher F Baum, Boston College, USA{p_end} {p 0 4}baum@bc.edu{p_end} {title:Also see} {p 0 19}On-line: help for {help dfgls}{p_end}