Abstract:
This research aims to assess economic capital for underwriting risks by using loss ratio and aggregation portfolio of risks via guassian copula and students t copula with degree of freedom 1, 3, 5, 10, 15, 20 and 25. This paper uses value at risk (VaR) and tail value at risk (TVaR) to measure risks. The data used in this research are drawn from the income statement of a non-life insurance company reported monthly from march 2008 to december 2010. This research classifies business units into 7 groups : fire insurance, marine & transportation insurance, compulsory automobile insurance, voluntary automobile insurance, miscellaneous insurance, industrial all risks insurance and personal accident insurance. The result shows that economic capital of aggregation portfolio of risks using students t copula with 1 degree of freedom gives highest value. This is able to explain correlation of tails distribution. However, guassian copula yields the lowest economic capital since this ignores the tail dependence of each business line. Considering the diversification benefit it seems that students t copula with 1 degree of freedom yields maximum diversification benefit while guassian copula yields minimum diversification benefit. Therefore, students t copula with 1 degree of freedom gives the best performance if firm has an insurance dependent risks between each line of business and especially when firm faces catastrophe.