The Anderson-Moore Algorithm AIM index

The Anderson-Moore Algorithm was developed at the Federal Reserve Board by Gary Anderson and George Moore. Two papers, written in 1983 and 1985, describe this method for solving linear saddle point models. Since then, undocumented changes in approach have improved the efficiency and reliability of the algorithm.

This site is meant to be a reference center for those wishing to learn more about the algorithm, as well as those who are experienced with the algorithm and wish to download the newest versions of code.

This site provides access to:

• Papers describing the present state of development of the algorithm

• Downloadable code implementing the algorithm and model preprocessors that allow users to specifty models as systems of equations. This Matlab code works for both UNIX and Windows platforms

• Demonstrations of how the code can be used to solve models