Macro-financial stability under a semi-nonparametric approach

Abstract

This thesis proposes a set of tools for measuring and managing financial risks related to the stability of banking systems and for establishing macroprudential policies aimed at preventing the materialization of systemic risks. It is based on accurately modeling probability density functions associated with banking stability indicators. The methodologies used respond to the second pillar of the Basel Committee on Banking Supervision agreement, which states the need to determine and monitor the Economic Capital banks' need to cover losses caused by the materialization of financial risks with a certain level of confidence and for a given time horizon. Semi-nonparametric statistics were used to parameterize stylized facts such as asymmetries and heavy and wavy tails observed in the empirical probability distributions of financial stability indicators. Analytical and simulated solutions for probability measures and economic capital settings are proposed. Applications are made on aggregate solvency indicators and their components, the bank leverage indicator for developed and emerging economies, and interactions between these indicators and monetary policy were analyzed. The results point to the need to model the skewness and kurtosis of the probability distributions of the financial stability indicators for not to underestimate risk and the level of economic capital. The hypothesis of an interaction between prudential and monetary policy and the need to jointly consider the decision-making of both policies is confirmed.