This paper tests the monitoring concentration hypothesis that loan participants can transfer monitoring incentives back to lead arrangers by purchasing credit default swaps (CDSs) from lead arrangers with the capacity to sell such contracts. Monitoring concentration appears to play a significant role in reducing lead arrangers’ loan allocation, expanding syndicate size resulting in larger loans for borrowers, and increasing the usage of covenants. Other loan CDS dealers participating in the same syndicates dilutes the role of CDS in monitoring concentration because these dealers can also sell CDSs to non-dealer participants. The evidence is further supported by two robustness tests — the analysis of traditional term loans and the London Whale incident. Among the three types of loan facilities, traditional term loans have the highest likelihood that CDSs are purchased by the participant lenders because banks tend to keep these loans on their balance sheets. On the other hand, institutional term loans, which are arranged with the intention to be securitized or sold to institutional investors, are less likely to have CDS protections purchased by original participant lenders. In contrast to term loans, revolvers are lines of credit, i.e., the amount of loans may not be drawn. Therefore, participant lenders also have less need to purchase CDS protections. The above main findings are driven by traditional term loans. However, there are spillover effects to revolvers that are in the same loan packages with traditional term loans. The London Whale incident, which caused JP Morgan more than $6 billion trading losses in the CDS market, presents a unique opportunity to examine the effect of CDS on lending practices. During the event quarter, there is a clear shift of lending activities from CDS selling banks to non-CDS selling banks. In addition, arranger allocation increases and syndicate size reduces among CDS selling banks relative to non-CDS selling banks. The fact that not only the lending practices of JP Morgan but also those of other CDS selling banks are affected suggests a general mistrust of using CDS protections, which in term affects the loan syndicate structures of CDS selling banks. However, the effect is transitory and reversed during the following quarter. Overall, the findings indicate that financial innovations, such as CDSs, can profoundly impact banking practices.
Read full abstract