1 Introduction
In a world a` la Modigliani and Miller (1958) with no financial market imperfections,
the capital structure of firms is indeterminate and the aggregate mix of debt versus
equity is irrelevant for the evolution of the real economy. Models of economic fluctu-
ations such as IS-LM, real business cycle models, and the canonical new-Keynesian
model have relied upon this theorem to justify their abstraction from financing deci-
sions.
Nevertheless, since Fisher’s (1933) explanation of the Great Depression1, economists
have recognised that financial factors can play a crucial role in the amplification and
propagation of macroeconomic shocks and in the transmission of monetary policy. As
shown in Figure 1, the spread between Baa and Aaa corporate yields—an indicator
of the deterioration of financial conditions—widened during most of the NBER-dated
recessions since the 1920s. Although this could simply reflect higher expected de-
fault costs, it seems unlikely that these costs alone can account for the sharp rise in
the spread—given actual defaul rates for Aaa and Baa bonds.
Starting with the seminal work of Bernanke and Gertler (1989), financial market
imperfections have been formalized in micro-founded models to explore the role of
financial factors during economic downturns.2 There is, however, a lack of consensus
on the quantitative significance of these frictions for business cycle fluctuations. In
the context of dynamic general equilibrium models, financial market frictions have
been evaluated solely using calibrated values of the agency cost parameters. The
empirical literature, instead, has investigated the extent to which financial balance
sheet variables influence investment, production, and employment decisions of firms.
Though not without controversy, the general finding is that the real decisions of firms
that are most likely to be facing severe informational problems in credit markets
1Fisher (1933) argued that the deterioration of borrowers’ balance sheets stemming from “debt
deflation” contributed importantly to the severity and persistence of the Great Depression. More re-
cent work on that era by Bernanke (2000) and Christiano, Motto, and Rostagno (2004) has identified
a host of other financial factors that reinforced the contraction phase of the early 1930s.
2 Informational asymmetries between borrowers and lenders have been introduced under various
guises into general equilibrium models by Carlstrom and Fuerst (1997), Kiyotaki and Moore (1997),
Bernanke, Gertler, and Gilchrist (1999), and Cooley, Marimon, and Quadrini (2004). Implications
of financial market frictions for the transmission of monetary policy are discussed by Bernanke and
Gertler (1995). Following the 1997 Asian financial crisis, a number of economists started considering
implications of credit market imperfections in open-economy settings; see, for example, Krugman
(1999), Aghion, Bacchetta, and Banerjee (2000), Cespedes, Chang, and Velasco (2000), Caballero
and Krishnamurthy (2000), and Gertler, Gilchrist, and Natalucci (2003).