Font Size: a A A

Risk Measures Of Ruble’s Exchange Rate And Brent Based On The Extreme Value And The Analysis Of Timevarying Correlation

Posted on:2016-04-30Degree:MasterType:Thesis
Country:ChinaCandidate:B SunFull Text:PDF
GTID:2309330482465719Subject:Statistics
Abstract/Summary:PDF Full Text Request
Since 2014, the price of crude oil occurred the huge drop, and since the financial crisis, oil prices never fallen so violent like current situation. As oil exporting countries, Russia is greatly affected by the huge drop of crude oil prices, as Russian currency, Ruble also occurred the unusual huge drop, the devaluation of Ruble and the recently huge drop in oil prices have corresponding relation.Faced the crisis of extreme financial, for example, the huge drop of crude oil prices and the crisis of Ruble, the precise measurement of risk is particularly necessary and urgent. Existing research show the VaR model based on the extreme value theory can well estimate the extreme risk value of financial market, however, the real financial market data can not meet the assumption of independent and identical distribution, so this paper first use GJR model、EGARCH model and GARCH model combined with t distribution、GED distribution and SKST distribution to deal with Brent crude oil and Ruble’s exchange rates’the logarithm rate of return sequence, obtain the sequence of the standard residuals, to the the sequence of the standard residuals which meet the assumption of independent and identical distribution, the VaR value and CVaR value are calculated by using the threshold POT model, finally obtain the risk value of the single asset. Considering the Kupiec test ignores the time variation characteristics of the data which is usually used in the model test, this text uses the Christofferson test that considers the time variability of over threshold sequence based on the Kupiec test. Empirical research shows: in the lower confidence level, to extreme risk of the two asset sequence, every model can not effectively measure, however, in the higher confidence level, every model is significantly effective. On the upper tail of the two asset sequence, the test values of every model are close, and every model is effective in the higher confidence level; however, on the lower tail of the asset sequence, to the measure of the extreme risk, the GJR-SKST-POT model is most effective, and the test values in this model are significantly smaller than other models, shows to the risk measure of the lower tail of the two asset sequence, the GJR-SKST-POT model is really better than other models.In order to study the dependence structure between the oil market and the Ruble’s exchange rates market, the influence degree that the Ruble crisis is affected by the huge drop of the crude oil, and considering the correlation of the financial market is always changing with time, this paper uses three kinds of time-varying copula models and three kinds of constant copula models to study the correlation between the two markets. Because the copula models have an advantage that can not be limited by the marginal distribution, we can study by separated the marginal distribution and copula models, this paper uses the GJR-SKST-POT model which is the integrated optimal model in the previous article as the marginal distribution, combine copula models measure the correlation of the assets. Empirical research shows:the SJC copula model is used to describe the dependent structure among the asset sequences is the most accurate, and the average correlation coefficient of the upper tail is greater than the lower tail which is measured by the time-varying SJC copula model, shows that the two asset markets are more prone to appear the phenomenon of combined extreme value during the bull market. By the correlation coefficient, the correlation between the two asset sequences is not as big as the imagination, but under common influence in other factors, such as some the West take the sanction to Russia and Dollar strength, the correlation between the two asset sequences has been considerable, shows the huge drop of crude oil prices is indeed one of the main reasons for the huge drop of Ruble’s exchange rates.
Keywords/Search Tags:Extreme Value Theory, Time-varying Copula, Risk Measure, Crude Oil Slump, Ruble Crisis
PDF Full Text Request
Related items