Discussion of “ Interbank Lending , Credit - Risk

The paper by Heider and Hoerova (2009) is ambitious. It studies the interaction between secured interbank lending, unsecured interbank lending, and banks’ portfolio choices. It is motivated by a puzzling empirical observation, namely the “decoupling of secured and unsecured lending rates” in the Great Financial Crisis of 2007–09. The observation made by Heider and Hoerova is that on August 9, 2007, not only did secured three-month interbank rates start to exhibit historically unprecedented discounts to unsecured rates, but the time-series behavior of both rates began to diverge as never before. Furthermore, the discount of secured interbank rates fluctuated more strongly in the United States than in the euro area. While the evidence consists of only two plots of time series, it is obvious without any further econometric analysis that the phenomenon is there and is significant. The simple observation that unsecured lending became more expensive than secured lending in a period of financial turbulence is not surprising. Any model of lending will produce this result. What is puzzling is the decoupling of the time-series behavior. In order to address this puzzle, Heider and Hoerova propose a model of interbank borrowing that imbeds both types of interbank markets into a model of bank portfolio management under investment risk. This model exhibits a key feature of what seems to have occurred after August 2007: the dramatic change in the risk of bank portfolios has affected secured and unsecured interbank rates differently. While this result is derived in an essentially static model, it is plausible and contributes greatly to our understanding of the crisis.