Abstract/Summary

Residential mortgages represent a crucial segment within the financial and lending industry across major economies. Their significance stems primarily from the market’s size, a result of the combination of substantial loan amounts and widespread presence in the consumer credit sector. Consequently, residential mortgages constitute a significant proportion of the assets managed by financial institutions. In addition, this type of loans is central to numerous policy and governmental initiatives, reflecting the importance of the underlying collateral for a substantial number of citizens in most developed countries. Lastly, given their pivotal role, residential mortgages have been at the heart of various crises over the years and continue to be closely monitored due to their critical importance in contributing to overall financial stability. A substantial corpus of academic literature emerged following the Global Financial Crisis, exploring various dimensions of mortgage financing, including its relationship with the broader economy, and the events and market disruptions that caused the collapse of this sector. Nevertheless, as the crisis gradually receded, certain pivotal questions and research areas have seen diminished interest, overshadowed by other subjects. This Thesis aims to enrich the existing body of literature by offering new insights into two significant aspects of residential mortgages that remain pertinent today, despite receiving scant attention in recent years. The first research domain focuses on correlation; the second investigates mortgage dynamics in response to specific lifecycle events, such as modification and default. Correlation plays a crucial role in determining both regulatory and economic capital, as it quantifies the interconnectedness of loans within the same asset class. Hence, the precision of the correlation parameter is vital for accurate risk assessment and management. Within the regulatory framework, the Basel Committee on Banking Supervision (BCBS) has set the correlation for residential mortgages at a fixed value of 15%. Although this value is deemed to be sufficiently conservative, there is limited evidence supporting its flat nature. This research primarily seeks to ascertain the validity of this assumption by investigating the presence of significant variations in default correlation across different segments of the residential mortgage market. Moreover, the study examines the impact of correlation on lenders’ loan pricing strategies and questions whether current regulation implicitly encourages regulatory capital arbitrage. Throughout the lifetime of a mortgage, various events can alter the standard repayment trajectory, which in turn has significant implications for lenders, borrowers, and stakeholders. The second and third empirical chapters of this Thesis analyse the determinants that characterise mortgage resolutions following loan modification and default, respectively. Specifically, this Thesis enhances existing research by examining how policy changes and mortgage market breaks have impacted consumers behaviour after the occurrence of these events. In particular, the second empirical study analyses the outcomes of post-modification and its determinants following the cessation of the Home Affordable Modification Program (HAMP), implemented by the US government to address the escalating defaults triggered by the Global Financial Crisis. Conversely, the third empirical chapter examines post-default resolutions and their determinants across three periods: before, during, and after the crisis. This study utilises mortgages originated in the United States by the Federal Home Loan Mortgage Corporation (Freddie Mac). The data under examination encompasses over 20 years and offers a nationwide scope, thereby providing a novel viewpoint within the existing literature, which has predominantly focused on either sub-prime portfolios or state-specific mortgages.