Matematisk modellering og empirisk analyse af CoCo-obligationer

Frederik Bøhme & Maria Lund Sørensen

Student thesis: Master thesis


The main focus of this master's thesis is on mathematical modelling and empirical analysis of Contingent Convertible bonds (CoCos). At its core, CoCos are a security that recapitalizes troubled nancial rms without the use of taxpayer funds. They do so by converting into equity when the rm reaches nancial distress. CoCos have grown increasingly popular in the wake of the recent nancial crisis as they present a way of strengthening the rms' protection against losses as well as a possible solution to the SIFI (Systemically Important Financial Institutions) problem. We review the main structural features of CoCos as well as the existing CoCo-market, and argue how CoCos t into the Basel guidelines. We use a structural credit risk model from the article "CoCos, Bail-in, and Tail-risk" from 2012 by Chen, Glasserman, and Nouri to model a nancial rm that uses CoCos as part of its funding base. The model incorporates rm- and market speci c jumps, debt rollover, and endogenous default. We use the model to derive closed-form expressions for valuing the rm and its liabilities, and in our empirical analysis we implement the model in Matlab to examine how issuing Co- Cos a ects the capital structure of the rm. Subsequently we apply the model to two major Danish banks (Danske Bank and Jyske bank), using data from their nancial statements. Given that the banks had issued CoCos, we examine whether they would have converted in the period from 2005 to 2012, and nd that the CoCos issued by Danske Bank would have converted by the end of 2008. Our ndings indicate that CoCos can provide a number of positive e ects. They create incentive for shareholders to raise new capital in the form of CoCos, exactly when the rm needs it the most. Moreover, at conversion, they can reduce the liabilities of pressured banks, reducing the need for government support.

EducationsMSc in Business Administration and Management Science, (Graduate Programme) Final Thesis
Publication date2013
Number of pages156