New heavy tailed distributions for risk measurements

Project: Research

Project Details


Characterization and quantification of risks associated with stocks(share prices, crude oil prices, stock prices, etc) movements is vital and of great interest to researchers, regulators and fund managers, due to the dramatic negative developments that the occurrence of extreme events such as a global financial crises causes on financial markets. Financial market risk quantification is fundamentally concerned with describing price/return uncertainty resulting from market movements. Two of the most common techniques for financial risk assessment are the value at risk ( VaRp(X) ) and expected shortfall ( ESp(X) ), where value at risk is defined as an amount lost on a portfolio with a given small probability over a time horizon and expected shortfall is defined as the average of all losses which are greater than or equal to value at risk. However, the performance of these risk assessment techniques depends on the probability distributions assumed for the behavior of the daily stock market returns. For instance, assuming a normal distribution for financial returns can be dangerous as it has a relatively low kurtosis and cannot better account for tail risk. This could lead to a wrong choice of portfolio as well as underestimation of losses. In recent years, financial institutions including banks have witnessed pockets of financial crises around the globe, irrespective of the existing risk measure. Consequently, speculators and researchers have started criticizing and questioning the credibility of existing risk measures that were only proposed over the last few decades. These crises have underlined the shortcomings of the much-celebrated risk measures like Value at risk (VaR), and Expected shortfall (ES), hence the need for a better risk measure. Given that the performance of these existing risk measures is hinged on the choice of probability distributions assumed for the financial indices, it is vital to propose a new heavy tailed probability distribution which will in turn be employed to develop a new novel risk measure that will satisfy as many of the properties of a robust risk measure.
Effective start/end date1/10/2131/12/22


Explore the research topics touched on by this project. These labels are generated based on the underlying awards/grants. Together they form a unique fingerprint.