Clark (Experience Rating)

[PANIC]

Clark discusses Casualty Per Occurrence Excess Treaties on pages 22-26. He states that "the steps in experience rating procedure follow those of property, but with some additional complications".

Step#3 (pg23) states that "the trended losses must then be capped at applicable policy limits".

Will someone please explain why this is not a problem when experience rating property? In other words, doesn't a similar problem exist when property values change over time?

[ReserveRage]

Just guessing, and I know very little about property insurance but perhaps it's because policy limit in property usually equals the maximum potential loss, since you can't suffer a loss to a building/home more than the insured value of that property and the policy limit for most properites is the insured value, but in casualty, there are potential losses far beyond the policy limits bought.

[PANIC]

thanks for the reply. i actually considered that already. if that was the case, you should not cap losses at all when pricing a casualty treaty.

Some property insurance has "inflation guard" where the value automatically adjusts upward. Property values may also decline over time (ie Chicken Houses). Hence, the issue with ITV runs parallel with this issue.

As a result, I think property treaty pricing shares the same problem as casualty in regard to changing limits/coverage. In fact, i would argue that property is more complicated in this step because the "drift" can go up or down.

[ReserveRage]

Regarding Clark stating how to deal with capping trended losses at the policy limit in Casualty experience rating...

His second suggestion "apply trend factor to historical losses without applying a policy limit cap. This assumes policy limit drifts upward to precisely match inflation

Can someone explain this better? Why couldn't you apply inflation trend directly to the policy limit cap? How does not capping losses at all equate the policy limit to drift upward matching inflation?

[frank_exams]

As a result, I think property treaty pricing shares the same problem as casualty in regard to changing limits/coverage. In fact, i would argue that property is more complicated in this step because the "drift" can go up or down.

There is no way the trend for property is more complex than the trend for casualty. If my reputation depended on accurate forecasting, I'd much rather try to predict construction costs than medical bills.

Frank

[frank_exams]

Regarding Clark stating how to deal with capping trended losses at the policy limit in Casualty experience rating...

His second suggestion "apply trend factor to historical losses without applying a policy limit cap. This assumes policy limit drifts upward to precisely match inflation

Can someone explain this better? Why couldn't you apply inflation trend directly to the policy limit cap? How does not capping losses at all equate the policy limit to drift upward matching inflation?

Remember what you're pricing here isn't at the primary level, where you have all the distributions, but at the reinsurance level. I'm sure you know that if you trend capped losses, you end up understating the effect of inflation (assuming inflation is positive). So that's clearly not a solution.

What you suggest is trending the cap. However, if you're going to do this, you also have to know what ILF's would be used by the cedant for these new trended ILF's. This is because, again you want to trend premium so as not to overstate the effect. But now, there are two problems. The first is, does the cedant plan on trending ILF's? If not, you're really back to square one, because you don't know exactly what kind of premium increase will result from policyholders selecting possibly different limits from before. The second problem is, what new limits will the cedant plan on offering? Will they match your trended caps?

Hopefully, the "second solution" seems more appropriate now. You get a good approximation of the inflation effect and it would take a whole lot more work to get a better one.

Frank