Securitization and risk: empirical evidence on US banks

Purpose - This paper aims to offer a comprehensive comparison of the characteristics between banks that securitize and banks that do not and to provide evidence of the capital arbitrage theory of securitization. Design/methodology/approach - First, the fundamental financial similarities and differences between banks that securitize assets and banks that do not participate in the securitization market are tested. Second, variables that help predict whether a bank securitizes assets are analyzed. Third, the determinants of securitization extent in banks that securitize assets are investigated – for general securitization extent and for specific type of asset securitized. Using a sample of 112 banks that securitize different assets, a matched sample of banks that do not securitize based on entity type and size is created. A quarterly panel data set of these banks dating back to 2001 is used. Findings - The results indicate that bank size is a significant determinant of whether a bank securitizes. Further, overall securitization extent is negatively related to the bank's capital ratio (in support of capital arbitrage theory), but this result is primarily driven by credit card securitization. Originality/value - Utilizing a unique data set of quarterly data from bank Call Reports; the panel data set is large relative to past studies. A matched sample approach was used to test fundamental financial similarities and differences between securitizing and non‐securitizing banks. In addition to aggregated securitization, an examination was made of how different classes of assets affect the banks' risk‐based capital ratios and test the capital arbitrage theory of securitization.