Mind Your Manual (and Eat Your Vegetables)

Speaking of vegetables, when trying to maintain the motivation to lose weight, one theory holds that it's best to avoid stepping on the scale on a daily basis.

Instead, the theory goes, it is better to set short-term achievable goals that can be rewarded with a non-food item or event. 

This makes good sense; we all like to be rewarded for good behavior. Just look at the currently-running advertisements for car insurance. Insurance companies are now formally recognizing and encouraging carrier loyalty and safe driving with a reward of lower premium payments as drivers remain accident-free. The reasoning behind these discounts is that drivers will be safe, loyal, and continue to save more premium dollars as long as they do.

It shouldn't come as a surprise that an employer group may expect to receive a reward (a lower rate) for good claims experience, especially if they have been with the same carrier for more than three years.

In those instances when claims experience is running better than expectations, the long term disability industry as a whole seems too often to be disregarding some of the fundamentals upon which our pricing model is built. We will all too often ignore the manual rating and the out-of-proportion total insured exposure that it's telling us we are no longer just rewarding employer groups for good experience. We are giving exposure coverage for free. Let's look at an example:

We've been asked to quote on a 780 life law firm based in the Midwest with two benefit classes.

We received 4.5 years of premium and claims, current census, and the current rate, but did not receive lives or rate history. The experience exhibit is constructed as best we can to look something like the following:

With a 32.6% overall loss ratio, this looks like a good group to go after. Based on the 3,112 total life years, a typical credibility formula would say that this group's experience is somewhere between 30-40% credible, so we need to run a manual rating. Let's assume the manual rating tells us the following:

If the target or desired loss ratio is 82%, the experience rate would calculate to be $.099 and the blended rate with 40% credibility would be $.85. Of immediate concern is the disparity between experience and manual. Based on how well experience is running my manual looks overinflated and my blended rate is apparently highly out-of-sync with the in-force rate. The average premium per certificate at $472 is also well above the industry average of $238. If we are going to go after this group, then something needs to be done, right? I mean, what do those actuaries and their manual rates know anyways?

Given the significant number of high wage earners and large exposure, though, is the manual actually overinflated? If we received a claim with a $25,000 monthly benefit and the average duration of a LTD claim is 60 months, the claim cost would be $1,500,000 with a needed annual premium of $1,829,268. If the claimant is out longer than five years that claim cost could be significantly higher. Granted we would not expect to see this large of a claim every year, but even with only one, given the current annual premium of $370,000, it would take more than five years to make up the difference.

On the other hand, many underwriters may argue that someone at such a high level of salary would be out for less than 60 months. The high earners who qualify for this large of a benefit are typically highly motivated to return to work. Even with a $25,000 benefit this individual may still be receiving less than 60% income replacement. So it appears that a discount to the manual, maybe 25%, would be a defensible argument. It's still not going to get us to the in-force rate.

Adjusting the credibility is another area underwriters will typically look at to help gain that competitive edge. Based on the size of the group and the 90 day elimination period we would expect to see about one to two claims per year, which is actually where the group's experience is running. Increasing the credibility from the 40% formula could be considered a justifiable action. However, I would need to increase my credibility to 83.5% just to match the in-force rate using the discounted manual. Is discounting my manual 25% and increasing credibility by 110% an appropriate deviation from standard underwriting practice? My hope is that all underwriters would shout "no!" collectively; that this goes against all underwriting principles. Alas, time and time again, we are seeing this kind of pricing in the marketplace.

So let's assume that we won the account from the incumbent by giving the customer a 50% discount to the in-force rate and guaranteed it for three years. It's now Year 8 and we are looking to renew the account. Experience and manual now look like the following:

A few observations can be made:

• After our rate reduction, the average premium per certificate is more in-line with the industry average.
• Claims incidence has stayed at expected, but the size of the claims has increased.
• The experience rate and manual, while still not totally in sync have moved closer together.
• When we originally priced the business, the blended rate using our standard assumptions ($.85) was not too far off the mark of what could have happened given this scenario as presented three years later.
• We just lost more than two million dollars on a group that we thought was a good group to go after.

While the case study presented may be viewed as an over-dramatization of pricing patterns in the market, would you look at this case differently if I told you that the incurred claims presented in the renewal scenario were the actual incurred claims for this group in Years 6 – 8? Well, they are.

And what if I had presented this case in the opposite chronological order, such that the group had two large claims in Years 1 – 3 whereby experience began to improve and run at a lower loss ratio? How deep would you have cut the rates to gain this business? And what if I also told you that those two large claims were non-subjective: 1) a skiing accident in which the claimant was paralyzed and 2) a Multiple Sclerosis diagnosis. Would you have thrown out one or both large claims as being an anomaly or outlier?

Maybe the 50% rate discount and the gamble we took on this group would have been acceptable if the plan had a more reasonable max, thus lowering our exposure and volatility. Maybe. That doesn't change the fact that we should be recognizing the total potential exposure and charging for the risk of that coverage instead of giving it away for free. We are in the business of predicting future results based on past experience, AND also predicting future risk based on the assumptions that have been collected and accumulated within our own organizations and throughout the industry. Maintaining disciplined pricing through our pricing tools is the foundation for continued profits. So mind your manual (and while you're at it, be sure to eat your vegetables!)