*
* Example 19.4 from page 575
*
open data tablef5-1[1].txt
calendar(q) 1950
data(format=prn,org=columns) 1950:1 2000:4 year qtr realgdp realcons realinvs realgovt realdpi $
   cpi_u m1 tbilrate unemp pop infl realint
*
set logc = log(realcons)
set logy = log(realgdp)
*
* This creates a dummy variable for quarter 4. This series with leads 0 to 3
* (that is regressors seas{0 to -3}) gives you a full set of dummies.
*
seasonal seas
*
linreg(define=ardl) logc
# logc{1 to 3} logy{0 to 3} seas{0 to -3}
*
* Extract the polynomials for the autoregressive part (which will be
* 1-a(1)x-a(2)x**2-a(3)x**3), and the distributed lag part.
*
compute arpoly=%eqnlagpoly(0,logc)
compute dlpoly=%eqnlagpoly(0,logy)
*
* Compute the long run effect
*
compute arlags=%polyvalue(arpoly,1)
compute ardllt=%polyvalue(dlpoly,1)/arlags
summarize(title="Long-Term Effects-ARDL Estimates",value=ardllt,$
   vector=||ardllt/arlags,ardllt/arlags,ardllt/arlags,1.0/arlags,1.0/arlags,1.0/arlags,1.0/arlags,0,0,0,0||)
*
* Compute the implied dynamic lag coefficients. This is done by dividing the DL
* polynomial by the AR polynomial. Since that's, in fact, infinitely long, we
* need to set the number of lags required.
*
compute ardlest=%polydiv(dlpoly,arpoly,7)
*
linreg logc
# logy{0 to 7} seas{0 to -3}
summarize(title="Long Term Effects-OLS Estimates")
# logy{0 to 7}
*
compute [vector] ardlest=ardlest
compute [vector] olsest =%xsubvec(%beta,1,8)
report(action=define)
report(atrow=1,atcol=1) "Lag" %seq(0,7)
report(atrow=2,atcol=1) "ARDL" ardlest
report(atrow=3,atcol=1) "Unrestricted" olsest
report(action=format,atrow=2,picture="*.###")
report(action=show,window="Table 19.3 Lag Coefficients in a Rational Lag Model")