MIXVAR Procedure

@MIXVAR computes estimates for a single equation using the mixed estimation procedure used by the RATS command ESTIMATE. It can be helpful in dealing with the following situations:

1.You have a variable you wish to forecast which does not fit into a full-blown VAR, since it has no predictive content for the other variables.

2.You have a limited amount of data for one variable, so you need to use fewer coefficients in that equation than in others.

@MIXVAR( options ) depvar start end

# list of other endogenous variables

# list of deterministic variables(if MORE option)

Parameters

depvar	dependent variable
start, end	range to estimate, defaults to maximum range permitted by all variables involved in the regression.

Options

NUMLAGS=number of lags (for all variables) [1]

TIGHT=tightness [.2]

LAGTYPE=[HARMONIC]/GEOMETRIC

DECAY=parameter of lag decay

These are identical to the options on SPECIFY

OTHER="weight" on other variables [1.0]

This is the symmetric prior weight

ZEROLAG/[NOZEROLAG]

If ZEROLAG, the other variables are put in with lags 0 to NUMLAGS-1

DEFINE=equation to define

MORE/[NOMORE]

Use MORE and an extra supplementary card if you want deterministic variables other than CONSTANT.