The research estimates a competing risk model of mortgage terminations on samples of UK securitised subprime mortgages. Given the argued role of these types of loan in the recent financial crisis then it is important to better understand their performance and supposed idiosyncratic behaviour. The methodological and empirical advance is the use of a general, flexible modelling of unobserved heterogeneity over several dimensions, controlling for both selection issues involving initial mortgage choices and dynamic selection over time. Moreover, we estimate specific coefficients for this unobserved heterogeneity and determine the correlation between the unobserved components of default and prepayment. The paper demonstrates the need for researchers and practitioners to jointly estimate household choices whiles controlling for selectivity through unobserved heterogeneity.