03.12.2024Felix Kübler and Yucheng Yang receive SNSF Grant

In this project, the researchers plan to utilize some of the recent advances in deep learning to develop global solution methods for three important and popular heterogeneous agent macro models: A large-scale stochastic overlapping generations (SOLG) model, a heterogeneous agent New Keynesian (HANK) model, and a search and matching model. The plan is to develop reliable solution methods and make the code suites available to the wider profession.  

The project owners propose a series of applications of these models. The first is tied to existing work by Kubler on climate change macro and the green transition; the second application is tied to existing work by Yang on the distributional consequences of monetary policy. The third application is related to Yang's ongoing work on how aggregate risks are transmitted into markets with search frictions and heterogeneous agents.  

In all the researchers' existing work on these problems, the bottleneck comes from a lack of a computational solution method. The main goal of the project is to overcome this. Finally, the researchers plan to explore synergies between the models, such as introducing monetary policy into the SOLG model or search frictions into the HANK model. 

“Machine learning-based AI methods have provided unique opportunities for scientific fields facing high-dimensional modeling challenges, including economics. These methods have the potential to transform quantitative macroeconomics and finance. UZH has been a hub for a series of influential work in this field, and with the support of the SNSF, we look forward to further pushing the frontier of this agenda.”
-Felix Kübler, Yucheng Yang

More information: https://data.snf.ch/grants/grant/10003091  
Swiss National Science Foundation (SNSF): https://www.snf.ch/en