Solves a simple AK-OLG-model for a closed economy in Matlab
Shows how to solve a simple deterministic overlapping-generations model (OLG) of Auerbauch-Kotlikoff type, solving for the transition path between two steady-states. The code is not optimized for speed, makes excessive use of global variables, and is mainly meant for instructional purposes. However, the algorithm as such is efficient and written with parallelization in mind (at every iteration the household problems of all finitely lived representative cohorts could be computed in parallel). For larger models, it is recommended to use a more efficient language that supports shared memory multithreading and pass-by-reference, e.g. C++ with the Armadillo library or Julia.
A model description can be found here: https://github.com/solveCGE/solveOLG_doc.
Parameters can be set in calib.m. Policy shocks are defined in run.m (or just uncomment some of the predefined exemplary shocks). The model is then solved by just running run.m.
Philip Schuster