Financial crises occur out of prolonged and credit-fueled boom periods and, at times, they are initiated by relatively small shocks that can have large effects. Consistent with these empirical observations, this paper extends a standard macroeconomic model to include financial intermediation, long-term loans, and occasional financial crises. Within this framework, intermediaries raise their lending and leverage in good times, thereby building up financial fragility. Crises typically occur at the end of a prolonged boom, initiated by a moderate adverse shock that triggers a liquidation of existing investment, a contraction in lending, and ultimately a deep and persistent recession.

Introduction

The 2007-09 financial crisis revealed the need for macroeconomic models to incorporate connections between the financial sector and the macroeconomy that can amplify economic shocks and lead to occasional deep economic downturns. Over the past few years, rapid advances have been made to extend standard macroeconomic models and include financial intermediation to account for episodes of severe financial distress. At the same time, a quickly growing empirical literature has revealed several stylized facts about financial crises. Crises are rare events that are usually preceded by prolonged boom periods and a buildup of macro-financial imbalances. For example, in the run-up to crises, credit usually rises rapidly, and credit growth is a robust early-warning indicator of crises (e.g., Schularick and Taylor, 2012; see also Figure 13 in Appendix A.4.2). Financial crises are associated with severe recessions that are typically deeper than normal recessions, particularly if they are preceded by a buildup of credit (Jordà, Schularick, and Taylor, 2013). However, the ultimate triggers of crises can be relatively small. With respect to the 2007-09 financial crisis, Gorton and Ordoñez (2014) argue that losses from mortgage-backed securities − the relevant shock for the financial sector around that time − were actually quite modest (see also Ospina and Uhlig, 2018). These empirical facts about crises pose challenges to current macroeconomic models. Why does financial fragility build up in good times? What is the propagation mechanism that turns shocks that are not particularly large into severe macroeconomic events? In this paper, I develop a quantitative macroeconomic model that addresses these questions. In my model, crises are as frequent and severe as in the data, financial fragility endogenously builds up during booms when credit expands, and crises are usually initiated by a moderate adverse shock.