Real Credit Cycles

Pedro Bordalo; Nicola Gennaioli; Andrei Shleifer; Stephen Terry

doi:10.1257/aer.20211820

Real Credit Cycles

Pedro Bordalo¹; Nicola Gennaioli²; Andrei Shleifer³; Stephen Terry⁴

¹ University of Oxford (email: ) · ² University of Bocconi (email: ) · ³ Harvard University (email: ) · ⁴ University of Michigan (email: )

American Economic Review 2026

We embed diagnostic expectations in a workhorse neoclassical model with heterogeneous firms and risky debt. A realistic degree of overreaction estimated from US firms’ earnings forecasts generates realistic credit cycles. Good times produce economic and financial fragility, predicting future disappointment of expectations, low bond returns, and investment declines. To generate the size of spread increases observed during 2007–2009, the model requires only moderate negative shocks. Diagnostic expectations offer a realistic, parsimonious way to produce financial reversals in business cycle models. (JEL D84, E13, E22, E32, E44, G12, G32)

DOI: 10.1257/aer.20211820
Volume: 116 (4)
Pages: 1274-1308
Language: en
Export: BibTeX
Sources: crossref openalex