Following a thorough analysis of banks’ and firms’ behaviour during the 2007-2009 financial crisis, Paolo Colla (Department of Finance) formulates a model of credit which encompasses empirical features so far neglected by the extant literature, in Which Financial Frictions? Parsing the Evidence from the Financial Crisis of 2007-2009 (paper for the NBER Macro Annual Conference, April 20-21, 2012, with Tobias Adrian, Federal Reserve Bank of New York, and Hyun Song Shin, Princeton University).
The initial observation is that while most models posit that borrowers suffer a contraction in the quantity of credit during a downturn, the evidence suggests that bond financing actually increases to make up part of the gap. Although the decline in issuance of new loans is sharp (-75% from peak to trough of the recent financial crisis), it’s partly balanced by a remarkable increase in bond issuance (+100%). In both cases, interest rates shoot up, with a four-fold increase for new loans and a three-fold increase for bonds.
Firms with access to both the loan and the bond markets substitute loans for bonds, leaving the total amount of new financing unaltered, even if at higher costs, while the smaller, bank-dependent firms bear the brunt of the crisis, suffering a marked reduction in the amount of new credit. “Our analysis”, the authors write, “shows that firms that are larger or have more tangible assets, higher credit ratings, better project quality, less growth opportunities, and lower leverage, were better equipped to withstand the contraction of bank credit during the crisis”.
Colla and his colleagues sketch, then, a model of direct and intermediated credit which takes into consideration five relevant stylized facts: the contrast between loan and bond financing points to the importance of accomodating both direct and intermediated credit; during the recent downturn, loan financing contracted but bond financing increased to make up some of the gap; even as the two categories of credit diverged in quantity, the spreads of both types of credit rose; bank lending changes dollar for dollar with a change in debt, with equity being sticky , consistent with credit supply by banks being the consequence of their choice of leverage for a given level of equity; as a consequence, bank leverage is procyclical.
Bank credit supply is modelled as the flip side of a credit risk model, where banks adjust lending so as to satisfy a risk constraint (they maximise profit maintaining the probability of bank failure under a threshold, and this probability is a function of financed projects’ probability of failure); direct credit (bonds) is subject to investors’ risk aversion.
In times of crisis the probability of projects’ failure increases and the market for loans tightens; the risk premium, then, has to spike in order to convince the risk-averse investors to buy bonds.
“The leverage of the banking sector emerges as being a key determinant (and reflection) of financial conditions”, the authors conclude. “As such, understanding how the leverage of financial intermediaries fluctuates over the cycle emerges as perhaps the most pressing question in the study of macroeconomic fluctuations”.