Questions I had while working on the DC-EGM replication are collected below. Please do document, in case you find answers while working on the project!
[ ] According to the paper, when no shocks are present (as is the case in m0), the DC step should be required already in T-1. In the code, it does not seem to be so, seemingly contradicting theory. Results for T-1 and T-2 can be replacing using EGM only. What can be the reason for this? Is it that a discontinuity is actually occurring, but in the credit constrained region, where handling boils down to leaving the EGM solution as it is?
[ ] Why does the Gaussian quadrature enter the retirees problem? There is no income involved, quadrature is there to integrate out the shocks/errors in the income process, there are none here?
[ ] Paper claims that solution for the retiree is the same as the solution for the worker who decides to retire. However, this does not seem to be the case for the default parametrisation of model_retirement.m. Investigate!
Questions I had while working on the DC-EGM replication are collected below. Please do document, in case you find answers while working on the project!