Macroeconomic Models of Financial Frictions

By Lawrence J. Christiano

  

 

Overview

These lectures review three models of banks and financial frictions in a macroeconomic context. The first model is of a rollover crisis. It is a recent model by Gertler and Kiyotaki (AER2015), which is a descendant of the seminal bank run model of Diamond and Dybvig (1983). A rollover crisis in the US shadow banking system (i.e., the many banks not covered by the protective umbrella of the Fed) is reputed to be a major cause of the financial crisis (for background, see this Journal of Economic Perspectives (JEP) manuscript and this). We discuss informally, the implications of the Gertler-Kiyotaki model for macro prudential policy and for unconventional monetary policy. We then turn to a model of financial frictions in heterogeneous nonfinancial firms that are constrained by how much collateral they have. It is a model of how credit might be misallocated among firms if firm collateral is not perfectly correlated with firm productivity. Finally, we consider another model of banking in which the financial friction is a moral hazard problem inside the bank. We derive explicitly (using the tools of Ramsey analysis) the implications for macroprudential policy. For some general discussion about DSGE models, beyond the JEP article, see this interview).

 

 

Lectures

1)        Rollover Crisis in Financial Institutions

a)                  Two-period exposition of Gertler-Karadi/Gertler-Kiyotaki model in which the financial frictions stem from bankers’ ability to ‘run away’ (section 3 in reading). This model is an important ingredient in the rollover crisis studied in part (b).

b)                  Extending the analysis in (a) to multiple periods and to bank runs (‘rollover crises’), using Gertler-Kiyotaki. Here is a highly informal version of the same notes.

2)        Collateral constraints.  Financial Frictions in Non-financial Firms: Buera and Moll’s model of deleveraging as a cause for a drop in interest rates, employment, GDP and TFP. The MATLAB code used to compute the example in the lecture notes is in this zip file. In the model, deleveraging can depress the economy by reallocating financial resources away from high productivity firms whose collateral is limited, and towards low productivity firms which have collateral. This discussion will illustrate one example of how DSGE models are evolving not just by integrating financial frictions, but also by incorporating heterogeneity of firms and households.

3)        A model of financial frictions in banking, in which the agency problem reflects ‘hidden’ effort inside the banks (lecture notesmanuscript). In a two-period version of the model, Ramsey methods show that a leverage restriction on banks raises welfare because it forces them to internalize a particular negative externality associated with issuing deposits. An infinite-horizon version of the model is used to display the implication of the model for dynamics. Here is zip file containing the lyx file for the slides, which contain some derivations in the form of lyx `notes’.

4)        Role of costly state verification credit markets, in US business cycles: an informal discussion.