Empirical Support for the ABCT?

Vasso P. Ioannidou, Steven Ongena, Jose Luis Peydro write:

An excellent setting to econometrically identify the impact of short-term interest rates on bank risk-taking is Bolivia. In recent years, the boliviano was pegged to the US dollar and the financial system was highly dollarised. During this period, the proper measure of short-term interest rates in Bolivia was the US federal funds rate, which is exogenous to Bolivian economic conditions. Hence, using the Bolivian credit registry, we analyse on a loan-by-loan basis the impact of the US federal funds on risk-taking and credit risk. The registry contains detailed contract information on all loans issued by any bank operating in the country as well as several measures of bank risk-taking such as ex-post loan performance, internal credit ratings, loan rates, and borrower credit history. The analysis draws from the 1999-2003 period, when the funds rate varied between 0.98% and 6.5%, and the boliviano was pegged to the US dollar.

We find that short-term interest rates affect risk-taking and credit risk. In particular, low interest rates encourage ex-ante risk-taking. Prior to loan origination, low interest rates imply that banks soften their lending standards for new loans – banks give more loans to borrowers with lower credit score and/or with bad credit history. Not only do banks take loans with higher ex-ante risk but also grant new loans that have higher ex-post credit risk, which we measure using a loan’s hazard rate, i.e. the default rate per unit of time. In addition, banks do not seem to price these extra risks they take. This finding suggest that our results are not driven by a higher demand for loans from risky firms (vis-à-vis less risky firms) when interest rates are low. All in all, low short-term interest rates seem to increase the banks’ appetite for risk.

A couple of observations:

• This seems to confirm one of the main assertions of the Austrian business cycle theory in that banks increase their risk taking in response to lower interest rates. The authors note (p. 20):
  

  “Firm fixed effects control for firm specific risk that is constant over the sample period. Consequently, when the federal funds rate is low, banks not just simply start financing risky firms that were excluded otherwise, but also engage in funding riskier projects (i.e., firms that would only have obtained loans for their safer projects when rates were high, are able to obtain financing for their riskier projects when rates are low).”

• The previous point seems to provide weak evidence against recent claims by Evans and Baxendale (which were first advanced by Block and Barnett) that the lower interest rates attract adverse selection. Rather, it seems that the effects are supply-driven.
• This paper also serves to answer some of the criticism leveled by recent critics of the ABCT (namely, Tyler Cowen and Bryan Caplan) who question why individuals make systematic errors in response to falling interest rates. This paper suggests that falling interest rates lead to increased access to credit for riskier firms and riskier projects.
• Despite presenting some evidence in support of the ABCT, however, it fails to explain two things*: (1) why banks increase their risk-taking, and (2) why the duration (or time structure) of production increases. (I currently have a paper under review that, in my opinion, provides an explanation for these two points and which might be of interest.)

For those interested, here is a non-gated link to the paper.

*It should be of note that the authors are not conducting an analysis of the Austrian theory and therefore have no reason to care about these issues.