Abstract
This study examines extreme tail behavior in Asian currency markets for the period of 2005-2018. Value-at-Risk (VaR) is estimated through Extreme Value Theory (EVT) approach to forecast losses incurred in a day in Asian currencies. Initially EVT approach is used to estimate extreme losses on the left tail of the distribution. Then, the VaR estimation of this approach is back tested through traditional and advance back testing methods to ascertain the accuracy of the models used. Results indicate that the estimation of GPD static model is relevant for extreme risk forecasting in EVT approach at both 95% and 99% confidence intervals. The used method is recommended for use by market players.
Highlights
Risk management is an integral part of the decision making process
The results provide that least risky currencies are Argentine Peso (ARG), Indian Rupee (INR), New Peruvian Sol (PEN) and Chinese Yuan (CNY)
At 99% confidence interval in the 2nd column Generalized Extreme Value Distribution (GEV) (Block Maxima), there are 99% chances that loss of Irani Rial will not increase 63.92% followed by Iraqi Dinar with 44.68% and Kazakhstani Tenge with 13.33% loss in a day
Summary
Risk management is an integral part of the decision making process. VaR is a common method for risk measurement. To estimate the risk behavior of the extreme left tail of the Asian currency market, Extreme Value Theory (EVT) is used. With time in worldwide financial markets, evaluating the extreme events probability, has become the main concern in the management of financial risks These market conditions require the quantication of worst losses in all types of financial markets. EVT provides more accurate estimation of VaR than other traditional methods, tail dependence decreases when filtering out heteroscedasticity and serial correlation by multivariate GARCH models (Fernandez 2005). The generalized Pareto distribution (GPD) is another EVT approach which is mostly used to estimate extreme tails behavior This method does take a maximum value and captures all extreme values above the threshold
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
