Journal of Probability and Statistics

Research Article

VaR: Exchange Rate Risk and Jump Risk

Table 7

Model accuracy using christoffersen’s independence test for a large portfolio.
StatesDay before
𝐼 𝑡 − 1 = 0 𝐼 𝑡 − 1 = 1 Unconditional
Historical simulations
Current dayPanel A: Period I
𝐼 𝑡 = 0 58088668
𝐼 𝑡 = 1 8725112
Total667113780
Panel B: Period II
𝐼 𝑡 = 0 43666502
𝐼 𝑡 = 1 661985
Total50285587
Analytical VaR
Current DayPanel C: Period I
𝐼 𝑡 = 0 484112596
𝐼 𝑡 = 1 15331184
Total637143780
Panel D: Period II
𝐼 𝑡 = 0 43656492
𝐼 𝑡 = 1 801595
Total51671587
Note that this table displays Christoffersen’s independence test [8] for the weights of domestic assets of 0.5. The data samples are split into two parts. The 1-day VaRs are estimated from January 1, 2004 to July 31, 2007 (Period I). The second part is used to forecast VaRs from August 1, 2007 to November 27, 2009 (Period II). 𝐼 𝑡 = 0 represents an indicator function if VaR is not exceeded at time 𝑡 ; otherwise; 𝐼 𝑡 = 1 if VaR is exceeded at time 𝑡 . The critical value is 3.84 at a significant level of 5 percent.