Securitization and financial solvency: empirical evidence from Portugal

ABSTRACT

This paper analyses the effect of securitization issues on the solvency of Portuguese financial institutions. For this purpose, we use an unbalanced panel model estimated using GMM methods and find that securitization has a slightly positive impact on the soundness of the issuing entity. We study 35 financial entities and 60 traditional securitizations issued by 9 originators between 2001 and 2013. The analysis reveals that the financial entities’ soundness improved slightly, showing that securitization enhanced the quality of the originators’ portfolios and increased the regulatory capital requirements. We also found that efficiency and profitability improve the risk-adjusted ROAA and that efficiency increases regulatory capital requirements. The robustness analysis confirms the positive effect of securitization on solvency, where both credit quality and liquidity are shown to be significant variables.

KEYWORDS:

• Securitization
• Z-score
• soundness
• Portugal
• financial entities

JEL CLASSIFICATIONS:

• codes: G28
• G21

Disclosure statement

No potential conflict of interest was reported by the authors.

ORCID

Carmen López-Andión http://orcid.org/0000-0002-7994-5335

Ana Iglesias-Casal http://orcid.org/0000-0002-2393-4696

Maria Celia López-Penabad http://orcid.org/0000-0001-8738-4366

Jose Manuel Maside-Sanfiz http://orcid.org/0000-0001-5621-6546

Notes

1. The European Commission has already introduced incentives for properly structured securitization in the Delegated Regulations for Solvency II (2015/35) and the Liquidity Coverage Ratio (LCR – 2015/62), adopted in October 2014. Both the ECB and the BoE have carried out work on this area. A task force led by the BCBS and IOSCO has been involved in developing International standards to facilitate the identification of simple, transparent, comparable securitizations, while the European Banking Association (EBA) has been fulfilling a similar role for European banking standards.

2. To the best of our knowledge, empirical, Amaral Loureiro (2013) and theoretical, Pinto and Marques (2007), Almeida and Crespo (2010, 2011) and Almeida, Crespo, and Santa (2012).

3. Of the 42 entities, only 35 were included in the estimations, since no data were available for any of the variables for the remaining 7 entities.

4. Following Michalak and Uhde (2012) and López-Andión et al. (2015), our model included: the rate of GDP growth, inflation, short-term interest rates and the Herfindall Index, but neither was found to be significant. As a result we did not include these variables in the final model. The results will be made available to those interested.

5. The analysis has been carried out for the referred periods. The results will be made available to those interested.