Forecasting estimation

[Yiyi123]

Guys
I would appreciate if any of you could explain to me how to solve a problem like this.

A housing completions (COMPS) model is estimated as follows:

LS // Dependent Variable is COMPS
Sample(adjusted): 1968:1 1991:4

Variable Coefficient Std Error t-Statistics Prob
C 1500.000 600.000 2.558045 0.0100
Starts (-1) 0.074724 0.035711 2.092461 0.0373
Starts (-2) 0.040047 0.042406 0.944377 0.0258
COMPS(-1) 0.236774 0.059893 3.953313 0.0001
COMP(-2) 0.206172 0.060554 3.404742 0.0008


where STARTS is housing starts. If housing starts are 2300 and completions are 1700 in 1991:3, and housing starts are 2100 and completions are 1800 in 1991:4, what is the housing completions forecast for 1992:1?

Cheers