Archives-September 2020

4-(naphthalen-2-yl)thiazol-2-amine (I) on reaction with 2-bromo-1-arylethanone(IIa-f)yields6-Aryl-3-(naphthalen-2-yl)imidazo[2,1-b]thiazole (III a-f). The structures of all the compounds series (IIIa-f) were characterized analytically. The compounds were also monitored for anti microbial activity

Value at Risk (VaR) is one of the most popular measures of risk associated with financial instruments. The generalized Pareto distribution (GPD) is useful in modeling values exceeding a high threshold in estimation of VaR and Expected Shortfall (ES). It has been observed that none of the existing methods for estimating parameters of GPD performs uniformly better than others (P. Z. Bermudez and S. Kotz, 2010), and even if one develops a method, it would suffer from heavy computational requirements. Recently, P.Chen et.al (2017) proposed a method of estimating parameters of GPD based on minimum distance approach and M-estimation. In literature survey it has been observed that a study on effect of outliers on estimation of parameters of GPD has not been carried out, though presence of outliers is common in financial data. In this regard this paper focuses on the effect of outliers on estimators of parameters of GPD, on estimators of VaR and estimators of ES under Peaks over Threshold (PoT) framework obtained from different methods. A simulation study is carried out to compare performances of five robust and seven non-robust methods for estimating parameters of GPD, VaR and ES in the presence of a single additive outlier under PoT set up and it is found that robust methods suggested by P.Chen et.al (2017) for estimating GPD parameters perform on par with other estimators when shape parameter k > 0 (for thin tailed distributions), in estimating VaR in presence of outlier, especially when sample size is moderately large

Table of contents for the month of September 2020