Work

Modeling tail risks and systemic risks using copulas

Year: 2018
Type: dissertation
Abstract: Understanding the relationship between financial variables is an essential part of managing financial risk. Such relationships in the context of risk management are of sensation interest in the modern world in which modeling risk has become a popular idea. The \n2007/08 financial meltdown has accentuated the importance of this because it showed that \nwith a distant probability of an event happening, say 0.01%, it should not be overlooked. \nThe existence of dependency between the two variables is underscored in their established \njoint distribution. Conventional financial models intensely depend on a relationship to \nelucidation the dependence that exists between tow variables. Nevertheless, as a means of \nmeasuring dependence, correlation entirely defines the dependence structure for normal \ndistributions. This is more so in the case of elliptical distributions as opposed to other \nclasses. Even within elliptical distributions, correlation remains delicate because it is easily \ndistorted by outliers. \nOne of the biggest shortfalls of current financial models is in their assumption and oversimplification \nthat leads to false parameterization. Many financial models assume that \nasset returns and risk variables follow normal distribution and as such pay a_ention the \njudiciously possible (central moments), undermining the value of what is remotely probable \n(extreme events). Yet, the financial crisis showed that remote events are understated. Thus, \nthis paper focuses on copula model- that characterized the dependence structure across \nthe entire distribution. Also, copulas have the advantage of being able to portray more \ninformation concerning the dependence structures between the two variables because \nthey are distributions themselves. \nFrom the viewpoint systemic and tail risk, the copula is motivating. In this regard, it \nenables us to disengage the dependence structure (linked with systemic risk) from the \nmarginal distribution (linked with tail risk) and archetypal both distinctly with a bigger \ndegree of precision. Fundamentally, copulas act as magnifying glasses that permit us to \nexamine and model financial variables with greater precision- in this case the connections \nbetween systemic and tail risk
Language: English
Subfield: Finance
Open Access status: green
OpenAlex ID: W2904949987
Cites: 25
Cited by: 2
Related to: 10