Money, Banks and Finance in Economic Thought

Modelling financial crises as DSGE phenomena: a genuine step forward?

Croisier Huguette, Université Saint-Louis Bruxelles

The prevailing approach to macroeconomic thinking, “dynamic stochastic general equilibrium” (DSGE), has been the subject of vivid criticisms following the 2008 global financial crisis. Several of these criticisms were not new, in fact they dated back from the early development of the DSGE framework (1980s). These include in particular the permanent equilibrium assumption, the use of representative agent(s) with rational expectations, and the role of exogenous productivity shocks in driving business cycles. There were nevertheless (at least) two serious criticisms specific to financial crises voiced in the aftermaths of 2008. First, DSGE models did not include an explicit description of the financial sector, and were thus unable to account for a distinct role of financial intermediation in the economy. Second, they were unable to reproduce the depth of the 2008 crisis without assuming the occurrence of « unusually large », unexpected exogenous shocks to the economy. In this work, we examine how DSGE practitioners have addressed these criticisms. Several models now include a banking sector, and are able to reproduce severe recessions in the absence of « large » exogenous shocks, thanks to the presence of nonlinear features. But do these features improve our understanding of financial crises, and provide us with tools (policy advice) to prevent new ones? In particular, in these models, the existence of credit constraints is typically motivated by the existence of moral hazard problems afflicting financial intermediaries: how plausible are these moral hazard issues, especially compared to other potential causes of financial instability which are left out? How do the « microfoundations » of financial frictions in the new models differ from those which were usually assumed in the pre-crisis credit-constraint models (the so-called « financial accelerator » literature), in which financial intermediaries were absent? Also, what does it mean exactly that « large » exogenous shocks are no longer needed to account for financial recessions? Has the role of exogenous shocks in business cycle models qualitatively changed since the 1980s, with the introduction of nonlinear features? What we propose in short is a critical analysis the recent DSGE literature in light of the history of the field and of the criticisms it faced since the 1980s.


Keywords: financial crises, DSGE, exogenous shocks, nonlinearity, financial accelerator, credit constraints, microfoundations