The more parameters there are to be estimated, the more memory and time are required to estimate the model. Also affecting these resources are the estimation method chosen and the method to calculate variance components. If the model has parameters including the intercept, there are at least numbers being held in the memory.

If the Arellano and Bond GMM approach is used, the amount of memory grows proportionately to the number of instruments in the INSTRUMENT statement. If the ITGMM (iterated GMM) option is selected, the computation time also depends on the convergence criteria selected and the maximum number of iterations allowed.