Code is for Matlab (.m)

For Matlab, run qmle_CMS.m in the MATLAB folder to conduct QMLE estimation for the BPP model in levels with the option of a bootstrap bias correction. This calls the lik_fcn.m, scr_fcn.m, trans.m, selif.m, shocks.m, dgp.m, and numeric_jacobian.m files. Run qmle_CMS_FD.m to conduct QMLE estimation for the BPP model in first differences with the option of a bootstrap bias correction. This calls the lik_fcn_FD.m, scr_fcn_FD.m, trans.m, selif.m, shocks_FD.m, dgp_FD.m, and numeric_jacobian.m files. The remaining files are data files. See program file for details about the data files.

Note: the calculation of Huber-White SEs will differ from one computer or program to the next due to precision issues with numerical derivatives.

James Morley

Email: james.morley@sydney.edu.au

This version: 11 November 2020