Theoretical Development and Application of Robust Copula Methods for Modelling Multivariate Financial Risk in Nigerian Markets

📖 ABSTRACT/OVERVIEW

Financial risk in Nigerian markets involves complex dependence structures between asset classes, exchange rates, and commodity prices that conventional multivariate normal models inadequately capture, necessitating the theoretical development of robust copula methods tailored to the extreme tail dependence and skewness characterising emerging market return data. This dissertation develops and applies robust copula methods for modelling multivariate financial risk in Nigerian markets, making original theoretical contributions to copula estimation under model misspecification. New robust estimation procedures for copula parameters that are resistant to outlier contamination are derived using M-functional theory. Influence function analysis quantifies the breakdown point of standard versus robust copula estimators. Empirical applications cover multivariate dependence modelling between NGX equity returns, naira exchange rates, and crude oil prices using daily data from 2010 to 2023. Robust Student-t copula estimation achieved 40 percent lower contaminated-sample parameter bias than maximum likelihood across simulation experiments. Tail dependence coefficient estimates from robust procedures were 15 to 22 percent higher than MLE estimates under contaminated data, implying that standard methods underestimate joint extreme loss probability. Value at Risk computed from robust copula models exceeded standard copula-based VaR by 18 percent at the 99.5 percent confidence level, with implications for regulatory capital calculations. The dissertation contributes robust copula theory with specific recommendations for CBN and SEC adoption in Nigerian financial institution risk modelling standards. Keywords: robust copula estimation, multivariate financial risk, tail dependence, Nigerian markets, M-functional theory

Need Complete Chapters of the Above Topic?

Get high-quality, Zero-AI research materials with current citations.

Request via WhatsApp 💬
Departments# Statistics