Interactive guide to Fernández-Villaverde, Hurtado, and Nuño (2019): "Financial Frictions and the Wealth Distribution".
- Paper (slides)
- Krusell-Smith economy: regression (code) (description)
- Krusell-Smith economy: neural networks (code) (description)
- A model with financial intermediaries: neural network (code) (description)
- Replication files. Click here to download replication files as a zip.
The first example. A comprehensive tutorial in the context of a simple, well-known HA model. It shows how to compute the model using both a linear regression (standard approach) and a neural network. The computation takes less than 2 hours on average (depending on the computer).
The second example. This is the model that we solve in the paper. We omit an extensive description as it can be found in the Appendix, but we provide a description of the files. The computation takes around 12 hours on average.
Extra features. The complete replication files, including the estimation and all the exercises and robustness analyses. Just launch a0_superlaunch.m and relax for three weeks, which is the average time to run the whole code in a normal computer.