E(a) = 200*.01=2
E(a2) = 2002*.01=400
Var(a)=E(a2) - (E(a))2 = 400 - 4 = 396
E(b) = 100*.05=5
E(b2) = 1002*.05=500
Var(b)=E(b2) - (E(b))2 = 500 - 25 = 475
The variance of total claims is 500*396+300*475 = 340,500
The standard deviation of total claims is 3405001/2 =~ 583.52
Expected total claims is 500*2 + 300*5 = 2500
Pr(total claims <= total revenue) = .95
Pr(C<=R) = .95 (where C=total claims, R=total revenue)
Pr(C<=R) = .95
Pr(C-m<=R-m) = .95 (where m is the expected total claims)
Pr(C-2500<=R-2500) = .95
Pr((C-2500)/s<=(R-2500)/s) = .95 (where s is the standard deviation)
Pr((C-2500)/583.52<=(R-2500)/583.25) = .95
We can now use the central limit theorem because (C-2500)/583.52 has a normal distribution with mean of 0 and stanrdard deviation of 1.
(R-2500)/583.25 = 1.645
R = 3459.89
So the insurance company needs 3459.89 in revenue. The expected costs of claims is 2500. So k = 3459.89/2500 = 1.38
This is a Society of Actuaries sample problem for course 100.
Michael Shackleford, ASA - March 17, 2000
MathProblems.info home