Using the Fréchet VaR formula derived from the Generalized Extreme Value (GEV) distribution for block maxima: $VaR_p = \mu - \frac{\sigma}{\xi} \left[ 1 - (-n \ln p)^{-\xi} \right]$ Given parameters: $\mu = 2$, $\sigma = 0.6$, $\xi = 0.4$, and $n = 100$.

For **$95%$VaR** ($p = 0.95$): $-n \ln p = -100 \ln(0.95) = -100 \times (-0.051293) = 5.1293$ $(5.1293)^{-0.4} \approx 0.51867$ $VaR_{95\%} = 2 - \frac{0.6}{0.4} [ 1 - 0.51867 ] = 2 - 1.5 \times 0.48133 = 2 - 0.722 = 1.278 \approx 1.28$

For **$99.5%$VaR** ($p = 0.995$): $-n \ln p = -100 \ln(0.995) = -100 \times (-0.0050125) = 0.50125$ $(0.50125)^{-0.4} \approx 1.3184$ $VaR_{99.5\%} = 2 - \frac{0.6}{0.4} [ 1 - 1.3184 ] = 2 - 1.5 \times (-0.3184) = 2 + 0.4776 = 2.4776 \approx 2.477$

The correct estimations are exactly presented in Option C.