Regulatory Capital for Securitizations in a Basel Setting

Abstract

In the first part of the thesis, we study the credit risk profiles of a publicly traded CLO (the Ares IIIR/IVRCLO) and a fictitious CLO of ABS. We demonstrate how the Vasicek model and the multi-poolcorrelation model can be used to evaluate the credit risk of these products by determining appropriateminimum attachment points given a set of desired credit ratings. The results of this part, and the twoproducts themselves, are used to concretize our findings regarding regulatory capital for securitizationsunder the Basel frameworks throughout the rest of the thesis.In the second part of the thesis, we study and evaluate the treatment of regulatory capital for both theAres IIIR/IVR CLO and the CLO of ABS under the Basel Accords I, II, and III. We demonstrate howBasel I’s simplistic “one-size-fits-all” approach for assigning risk weights to securitization exposures couldbe used by banks to engage in regulatory capital arbitrage (RCA) by securitizing corporate loans of highcredit quality while providing the SPV with a subordinated loan. Next, we show how the improved creditrisk-sensitivity of the Basel II framework prevented banks from exploiting the previously investigated typesof RCA. However, we argue that the look-up tables and overreliance on external credit assessmentinstitutions associated with the simplistic Standardized Approach (SA) and, to some extent, the RatingsBased Approach (RBA) still offered inadequate credit risk-sensitivity. In contrast, we argue that the moreadvanced Supervisory Formula Approach was adequately credit risk-sensitive as it is based on a widelyaccepted statistical measure of credit risk. Nevertheless, this approach was plagued by significant levels ofcliff-effects. Additionally, we demonstrate how Basel III attempts to rectify the weaknesses associated withBasel II by 1) greatly enhancing the credit risk-sensitivity of the SA by making regulatory capital dependenton the credit ratings of the underlying obligors rather than on the credit ratings of the tranches themselvesand 2) by introducing the Simplified Supervisory Formula for allocating regulatory capital across tranches,which considerably reduced the levels of cliff-effects that were observed from Basel II.In the third part of the thesis, we investigate the issue of the Basel framework amplifying the procyclicalityof the banking sector with respect to securitizations and the potential issue of capital requirements forsecuritizations being portfolio-independent.Our results show that the fluctuation of regulatory capital for securitizations throughout time heavilydepends on the reinvestment behavior of CLO managers. We propose two initiatives aimed at mitigatingthe procyclical tendencies of the Basel framework with respect to securitizations: 1) an autoregressive filterthat smooths the output of the current capital requirements across time and 2) basing capital requirementson through-the-cycle rather than point-in-time risk statistics. Moreover, we show that both of theseinitiatives are able to reduce the volatility of capital requirements for securitizations throughout timewithout significantly altering the long-term average mean.With respect to portfolio-invariance, we propose a two-factor setting that explicitly accounts for thepotentially imperfect correlation between the risk factors specific to the pool underlying a securitizationand the risk factors specific to the bank’s existing portfolio.