# Last Minute VAR Calc Question

From Schweiser: If expected monthly return and daily volatility are 1.05% and 6.20%. Compute the monthly VAR at 95% confidence Schweiser answer for downside exposure is: “To compute the VAR at the 95% confidence level, we will construct a confidence interval at the 90% level. The 90% interval spans 1.645 standard deviations on either side of the mean” 1.05%-1.645(6.20%)= -9.15% my question: is there a typo in the question, should it read test at the 90% level? Obviosuly the answer is using the t stat for 90% Should we not be testing for 95% confidence using the 95% (1.96 stat)?

I might be mistaken… but since the VAR is a one side… aka 5% chance the loss could be greater… that leaves 5% on each side so you have to use the 90% confidence interval. If you used the 1.95 there would be 2.5% on each side of the distribution.

Hiya yes soccertom is correct. If you look at the z table for a 95pc 1 tailed test you use 1.65. This is because 45pc of values lie within 1.65 standard deviations of the mean in the normal distribution.So if you use 1.65 on one tailed test you leave 5pc on the tail ( this is 95 pc confidence because you have 50pc above the mean and 45pc lying 1.65 sd below).If you use a two tailed test ie mean plus or minus 1.65 stdev you leave 5pc in each tail so it woul be 90pc confidence level

I see. makes sense. thanks guys. I think I’m one of last to sit for this test…friday morning, just in time for the weekend. Hope to see you guys on the Level II forum next spring or fall. cheers

I am writing it tomorrow morning… I have been getting 70-75% on all practice tests so I dunno if I am ready!

Good luck to you both. How you have something nice planned for after