Fabio Panetta, Ph.D, Economic Outlook and Monetary Policy Department, Banca d'Italia, via Nazionale 91, Roma, 00184, Italy and Alberto Franco Pozzolo, Ph.D., Scienze Economiche Gestionali e Sociali, Università degli Studi del Molise, via de Sanctis, Campobasso, 86100, Italy.
In the last decade, the use of credit risk transfer (CRT) techniques by banks has increased spectacularly, pulled by major improvements in screening and monitoring technologies and in financial engineering techniques. For banks to choose to transfer credit risk, the benefits must therefore overcome the costs. A number of factors are likely to affect this trade-off. For example, according to risk management theories, hedging is more profitable for firms that bear greater costs of financial distress (Smith and Stulz, 1995), while banks with a more concentrated loan portfolio are more likely to benefit from the diversification opportunities offered by CRT. Clearly, credit risk transfer also has ex-post effects on bank policies and characteristics. For example, by reducing the impact of highly concentrated loans, banks may reduce their ex-post risk. Or, in contrast, they could augment their default probability by removing safer loans from their balance sheets.
In this paper we study the determinants and the consequences of banks’ use of credit risk transfer techniques using a large data set of bank holding companies and analyzing separately the ex-ante characteristics of banks that make for the first time use of securitization techniques, and the ex-post effect of the securitization on those same characteristics. Moreover, we do not concentrate on banks from a single country, but we analyze a large number of banks from both developed and developing countries and we are therefore able to verify whether specific features of a country, such as the structure of its financial system or its regulatory setting, make it more or less common for its banks to use CRT.
We proxy banks’ use of CRT by the amount of assets it has securitized in each year. The ex-ante characteristics of bank using CRT are estimated using a survival-time data model. The ex-post effects are estimated using a difference-in-difference specification, verifying the robustness of the control sample using propensity score matching.
Preliminary results show that larger and riskier banks are more likely to make use of CRT techniques, and that ex-post they turn out having a lower probability of default, but have no effects on profitability. Expected results will analyze more in detail the ex-ante charactersitcs of banks using CRT technmiques and of their countries of activity, as well as the ex-post effects on characteristics such as asset liquidity and attitude towards risk.