* PS10.2, using DATA10-2, for Koyck lag model in Example 10.2
SAMPLE 1 744
READ(data10-2) DAYHOUR LOAD TEMP
*
* Estimate the Static Model.
*
OLS LOAD TEMP
*
* Estimate the Transformed Model from Equation 10.4 of the textbook.  Before
* this model can be estimated, the dependent variable, LOAD, must be lagged
* back one time period and the new variable, LOAD1, will be included as an 
* independent variable in the Transformed Model.  The LAG function on the
* GENR command is used.
*
GENR LOAD1=LAG(LOAD)
*
* Since the variable LOAD was lagged back one time period, the sample range 
* must be adjusted to reflect this change so the OLS estimate will begin with
* observation 2 rather than 1.
*
SAMPLE 2 744
OLS LOAD LOAD1 TEMP / COEF=COEF
*
* Calculate the Long-Run Multiplier.  The COEF:2 function on the GEN1 command
* takes the coefficient stored in Row 2 of the vector COEF from the previous
* OLS regression for the calculation.  Notice the COEF:1 function also takes
* the coefficient stored in Row 1 of the vector COEF as well.  In this case,
* the stored coefficient in Row 2 is TEMP and Row 1 is LOAD1.
*
GEN1 LRMULTI=(COEF:2)/(1-COEF:1)
*
* Print out the calculated Long-Run Multiplier.
*
PRINT LRMULTI
DELETE / ALL
STOP