What is the logic of calculation of TVAR as the arithmetic average between VAR at 99.5 and VAR at 99.999? Also how would you calculate TVAR in this case using the formula and obtaining an answer of 578? I seem to be getting a different answer....
Hi V, The formula approach follows the formula for TVAR under a normal distribution shown in sweeting.
The approach we use in the ActEd solution (TVaR = 578) is an “exact” calculation of the TVaR using the formula in the Course Notes and values from the Tables. The exam question asked us to “estimate” and so the examiners did not require such a precise approach. One way to estimate the TVaR is to take some sort of average of the values at risk in excess of the 99.5% VaR calculated in part (i). Using this approach (averaging VaRs at 99.5% and 99.999%) we ended up with an "average VaR" of 684 The examiners accepted a wide range of reasonable answers from this crude approach to the more precise approach given above.
Is it? Or does this simply reflect the difficulty and uncertainty of trying to place a value on events on the tails?!