COPULA FRANK UNTUK PERHITUNGAN VALUE AT RISK PORTOFOLIO BIVARIAT PADA MODEL EXPONENTIAL GENERALIZED AUTOREGRESSIVE CONDITIONAL HETEROSCEDASTICITY

Stocks are one type of investment that promises return for investors but often carries a high risk. Value at Risk (VaR) is measuring tool can calculate the amount worst loss occurs in stock portfolio with certain level confidence and within time period. In general, financial data have volatility value, which causes residuals not normally distributed. ARCH/GARCH modoel used to solve heteroscedasticity problem. If also an asymmetric effect, it modelled Exponential GARCH model. Copula-Frank part Archimedian copula empirical cases. The on this study were BBCA KLBF price observation period 30 December 2011 – 6 2019. Furthermore, test validity VaR model, backtesting will be carried out using Kupiec Test. results showed best model stocks was ARIMA (1,0,1) EGARCH (1,1) (1,2). risk 95% combination models 2.233% today's investment. Based Test, obtained declared valid.