Jenny are you talking about question four? If so here is what I did ... I could be wrong, who knows!
The question went something along of the lines of some dude is thinking of investing in something with payoff R with pdf
f(t) = 0 for t < 0.5
and
f(t) = c/t^4 for x => 0.5
with c = something and cost = 0.7.
The mean of the net payoff, or whatever it is called, can be worked out as
E[R] = E[t-0.7] = integral from - infinity to infinity of (t-0.7)f(t)dt
then you split this up into two integrals between -inf and 0.5 and 0.5 and +inf and one is zero and evaulate the other.
Then you can work out the semi variance using the mean you have now and some tricks. And again similar for the value at risk.
If this wasn't the question you were refering to then oops and the only other similar type question just involved the lognormal model distribtuion and some rearranging and evaluating.
If this question didn't appear on the paper and I am just making it up then it's an even bigger oops..........
I hope you guys do ok!
Last edited by a moderator: Apr 14, 2007