* * @EndersSiklos( options) u start end * does various types of the unit root regressions with threshold breaks * on the residuals from an Engle-Granger cointegrating regression * (assumed to be done already---the input u are the residuals). * * The regression run is * * du = (rho1 D1 x u{1} + rho2 D2 x u{1}) + lags of du * * where D1 is the dummy for threshold series < tau and D2 is * 1-D1. * * It produces three test statistics: * * t-max is the larger of the two t-statistics on rho1=0 and rho2=0. * phi is the joint F-test for rho1=rho2=0 * equal is the test for rho1=rho2. * * Parameters: * u = series of residuals from Engle-Granger regression * start end = range to use [defined range of u] * * You can either input the threshold value or request a search. * * Options: * LAGS=# of lags on the differences [1] * THRESHOLD=series which determines the threshold * * TAU=threshold value [not used] * PI=fraction of high and low empirical values for the threshold series * which are omitted in the search. * * If you provide a value of TAU, the regressions are done with that * fixed value for tau. If you don't, the values of the threshold series * (excluding the PI fraction at each end) are searched for the one that * minimizes the sum of squared residuals * * Reference: Enders and Siklos(2001), "Cointegration and Threshold * Adjustment," JBES, vol. 19, no. 2, 166-76. * * Revision Schedule: * 05/2011 Written by Tom Doan, Estima * procedure EndersSiklos u start end type series u type integer start end * option integer lags 1 option series threshold option real tau option real pi .15 option string title * local series du local vec[series] xsplit local integer i ttau local vector testvalues local integer startl endl pistart piend local real rssbest ltau taubest local report esreport local string ltitle * set du = u-u{1} * inquire(reglist) startl<=ltau) set xsplit(2) = u{1}*(threshold=ltau) set xsplit(2) = u{1}*(threshold