{smcl}
{* 19jul2007}{...}
{hline}
help for {hi:qll}
{hline}

{title:Perform Elliott-M{c u:}ller efficient test for general persistent time variation in regression coefficients}

{title:Syntax}

{p 8 14}{cmd:qll}
{it:depvar}
{it:varlist}
[ {cmd:(}{it:zvarlist}{cmd:)} ]
[{cmd:if} {it:exp}]
[{cmd:in} {it:range}] 
[{cmd:,} {cmd:rlag(}{it:#}{cmd:)}
]

{p}{cmd:qll}  is for use with time-series data.  You must {cmd:tsset} your data before 
using {cmd:qll}; see {help tsset}. You may apply {cmd:qll} to a single time series 
of a panel dataset. {cmd:varlist} may contain time-series operators; see {cmd:help varlist}.

{title:Description}

{p}{cmd:qll} performs the qLL efficient test for general persistence in time variation
in regression coefficients proposed by Elliott and M{c u:}ller (2006). The test contrasts a stable
regression model y = X beta + Z gamma + epsilon from the unstable alternative y = X beta(t) + Z gamma
+ epsilon. This very general specification nests many of the 'structural break' and 
'time varying parameter' models in the literature, allowing for almost any pattern of variation in the coefficients of the X 
variables, with good power and size even in a heteroskedastic context.

{p}{it:zvarlist} specifies the set of Z regressors, whose coefficients
are assumed to be fixed over time.

{title:Options}

{p 0 4}{cmd:rlag(}{it:#}{cmd:)} specifies the number of lags to be used in computing the 
long-run variance matrix. If a positive value is given, that number of lags is used, just
as in {cmd:newey}. If a negative value is given, the optimal lag order is chosen by the BIC
criterion.

{title:Saved results}

{p}{cmd:qll} saves the value of the test statistic, the number of observations in the 
estimation sample, and a three-element matrix ({it:r(cvmat)}) containing the 10%, 5% and 1% critical values
of the test for up to 10 regressors in X. If the number of lags is chosen by the BIC criterion, 
it is also saved as {it:r(biclags)}.
See {cmd:return list}.


{title:Examples}

{p 8 12}{stata "use http://www.stata-press.com/data/r9/lutkepohl.dta" : . use http://www.stata-press.com/data/r9/lutkepohl.dta }{p_end}

{p 8 12}(Quarterly SA West German macro data, Bil DM, from Lutkepohl 1993 Table E.1)

{p 8 12}{stata "qll linvestment lincome (lconsumption)" : . qll linvestment lincome (lconsumption)}

{p 8 12}{stata "qll linvestment lincome (lconsumption), rlag(4)" : . qll linvestment lincome (lconsumption), rlag(4)}

{p 8 12}{stata "qll linvestment L(1/4).lincome (L(1/4).lconsumption) " : . qll linvestment L(1/4).lincome (L(1/4).lconsumption)}

{p 8 12}{stata "qll linvestment lincome (lconsumption) if tin(30,85)" : . qll linvestment lincome (lconsumption) if tin(30,85)}


{title:Acknowledgements}

I am grateful to Graham Elliott for his clear presentation of the underlying theory when 
visiting Boston College in Spring 2007, and for his MATLAB code from which this routine was
constructed. Remaining errors are my own.

{title:References}

{p 0 4}Elliott, G. and M{c u:}ller, U.K., 2006. Efficient Tests for General Persistent Time Variation in 
Regression Coefficients.  Review of Economic Studies, Vol. 73, pp. 907-940.

{title:Citation}

{p}{cmd:qll} 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.
qll: Stata module to perform Elliott-M{c u:}ller efficient test for general persistent time variation in regression coefficients.
{browse "http://ideas.repec.org/c/boc/bocode/s456862.html":http://ideas.repec.org/c/boc/bocode/s456862.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 newey}
{p_end}