WC Severity Trend calculation

[silverfox]

I'm trying to figure out the best way to break out trend groupings.

I'm thinking

Indemnity
Medical on lost time claims
Med Only

I'm not sure where I should put ALAE. Should it be allocated proportionally, combined with indemnity, combined with medical, or something else? Thoughts?

[MountainHawk]

Exactly what we use. We ignore ALAE for trends, or trend it separately if we really need it. I find that most of the high severity WC ALAE is driven by EL anyway.

[UCSDKID]

Originally Posted by MountainHawk

Exactly what we use. We ignore ALAE for trends, or trend it separately if we really need it. I find that most of the high severity WC ALAE is driven by EL anyway.

I agree with MountainHawk.

We separate the 3 types of claims in the way that you did as indemnity has different cost drivers than medical. We also segment the medical into medical on lost time vs medical only claims as the severities are drastically different between the two and so any shift in the proportions of each claim type would really skew the indicated trends if analyzed together.

We treat ALAE completely separate and rarely look at the trends associated with them. We look at everything gross of any subrogation recoveries as well.

[silverfox]

What do you do with subro recoveries? Do you also ignore it or do you do some kind of allocation?

[MountainHawk]

All of our recoveries are associated with a line of coverage (EL, voc rehab, indemnity, medical, or expense), so I can just do it net of recoveries.

[silverfox]

But someone has to decide whether they were indemnity or medical or ALAe recoveries right? Is it the claims person that makes that judgment call?

[MountainHawk]

Depends on the type of recovery. For recoveries not associated with a particular payment (most of subro), I think the claim system follows a set of rules. For subro, I think it goes against Medical first, but I'm not 100% positive of that.

[silverfox]

When you look at losses for pricing purposes, are you guys all using 0% ALAE trend? That doesn't sit well for me.
What's the rationale in excluding subro? If you completely recover all costs, to me, that's not a claim at all.

[UCSDKID]

Originally Posted by silverfox

When you look at losses for pricing purposes, are you guys all using 0% ALAE trend? That doesn't sit well for me.
What's the rationale in excluding subro? If you completely recover all costs, to me, that's not a claim at all.

I appologize for not catching that you were looking at trends for pricing purposes. I was speaking in regards to reserving (year-end reserves on the brain).

With regards to pricing...

Our ALAE to loss ratio has actually been pretty consistent (consistent enough to not warrant further analysis at least) and so when pricing we would use an ALAE to Loss Ratio and thus the ALAE will basically trend with the underlying losses (some composite average of the three claim type trends).

With regards to subrogation, occasionally we have looked at losses net of subrogation and at other times we have looked at losses gross of subrogation and then applied a subrogation factor (some percent of losses) to account for it.

[MountainHawk]

Originally Posted by silverfox

When you look at losses for pricing purposes, are you guys all using 0% ALAE trend? That doesn't sit well for me.
What's the rationale in excluding subro? If you completely recover all costs, to me, that's not a claim at all.

We just use an ALAE/loss ratio and tack that on, so essentially we are assuming the same trends as the loss.

Unless you have a HUGE WC book, the ALAE trend is going to be dominated by a few large EL cases, and therefore not very meaningful.


Edit: What UCSDKID said.