An assumption of normality when returns data have fat tails leads to:

An assumption of normality when returns data have fat tails leads to:

I. underestimation of VaR at high confidence levels

II. overestimation of VaR at low confidence levels

III. overestimation of VaR at high confidence levels

IV. underestimation of VaR at low confidence levels
A . I and II
B . I, II, III and IV
C . I, II and III
D . II, III and IV

View Answer

Answer: A

Explanation:

When returns are non-normal and have fat tails, an assumption of normality in returns leads to underestimation of VaR at high confidence levels. At the same time, at lower confidence levels the normal distribution may give higher VaR estimates. Therefore Choice ‘a’ is correct. The other choices are incorrect. Also refer to the tutorial about VaR and heavy tails.