Medicare Supplement (Med Supp) carriers often ask, "What's going on with my Med Supp experience?" It's a good question. And the answer isn't always directly in front of you in a report or a chart.
In response to this question, my first question is, "What are you referring to: Loss ratio? Claims costs?" Loss ratio experience is impacted by both sides of the ledger, premium revenue and claims. There may be issues on both sides.
My second question is, "What were you expecting?"
It's a difficult question to answer. Often, Med Supp carriers reply, "Our claims are going through the roof!" That may be so. But what is critically important to understand is the following:
1) Too high relative to what?
2) Where are claims too high?
Without answers to these questions, a carrier is left without a clue about the source, a potential solution, or a basis for future action.
The approach I often use includes measuring historical claims experience relative to benchmark (or expected) values at a refined level and then rolling up the results in total as well as by key risk characteristics. This type of actual-to-expected (A/E) analysis does two things to answer the questions above:
1) It provides an overall claims level measure (referred to here as morbidity level) together with the variation over time as well as specific risk characteristics (which we will expand on later).
2) It can provide the inherent underlying pure claims trend experienced after adjusting the measurements for other influences such as changes in the demographic and/or geographic mix, etc.
The hypothetical case study of InsureU Insurance Company provides a simple example.
Case study: InsureU Insurance Company
InsureU's actuary, Cliff Diver, provides his boss, CEO Wanda Profits, with quarterly experience reports of among other things the new Med Supp line of business that InsureU has been selling for the last three years in the state of Bliss. InsureU has spent significant resources and capital to enter this market in order to expand membership base and hedge its risk in the commercial Patient Protection and Affordable Care Act (ACA) market. Given the lack of expertise in-house, InsureU turned to the support of consultant Sonny Days to price the product line. Premiums were priced based on various assumptions and expected to yield initial loss ratios in the low 70s. Projected financial results look something like Figure 1.

The launch of the product line had been seen as providing mixed results at best with regard to sales targets. Original sales targets were 800 policies in the first year but only 400 were issued. Competition has been fierce in the state of Bliss but InsureU weathered the storm and, on a bright note, the financial results were fantastic! So fantastic that Wanda convinced Cliff that rates were fine where they were and no rate increase in the first year was necessary. This was in spite of the recommendation by Sonny Days to file for a nominal rate increase equal to expected claims trend.
Fast-forward to year three. As sales have grown, Cliff nervously notices the quarterly reports showing financials going in the wrong direction. See Figure 2.


So, dear reader: if there was no rate increase in 2017 why did average premium drop in 2017? (For an immediate answer, scroll to the end of this blog.*)
Wanda is not happy and directs Cliff to call Sonny to find out what's going on and fix this. On the surface, while sales have definitely grown, it looks like InsureU has an issue with growth in claims. What could be the cause? Out of control utilization or Medicare cost levels? Claims processing? Fraud? Before investing in valuable time and resources, Sonny suggests an experience and trend analysis against the underlying experience records using industry benchmarks that reflect the following:
- Plan
- Year
- Age
- Gender
- Area
- Enrollment category¹
- Policy duration
By application of specific benchmarks to the detailed claims and exposure experience records, Sonny calculates "expected" claims. In addition, the benchmarks can be used to adjust or "normalize" the actual claims levels to a common morbidity level (such as a female age 70 under open enrollment at the nationwide cost level) so that the inherent true claims trend can be measured. Summary results of the analysis are presented in Figure 3.

In addition to the unadjusted experience, Figure 3 provides a) "Expected Claims" along with the resulting A/E ratio, as well as b) "Normalized Claims" costs and the inherent claims trend. Viewed in this light, one sees that the overall A/E is close to 100% and actually started out below 100% and has crept up but not dramatically. The inherent claims trends are roughly 4.5% to 5.0%, which, while higher than inherent expected levels, are more of a reflection of the better than expected morbidity reverting to 100%.
Based on these observations, Sonny convinces Cliff and Wanda that it's too early to investigate or audit detailed claims payments and claims procedures. By segmenting the results in finer detail it was determined that, while in 2016 the proportion of underwritten business was higher than expected, the proportion has dropped in 2017 and 2018. (And here, dear reader, is the answer to the question above. This change in enrollment type mix drives the average premium change noted above.) Note that in this case study underwritten business was issued at older ages than guaranteed issue business. Therefore, an increase in guaranteed issue business results in a decrease in average premium. This together with the initially lower 2016 morbidity level contributed to the initial favorable (though not very credible) experience and apparently higher claims trends. However, high increases in claims costs were driven primarily from a shift in enrollment type mix rather than high claims cost trend. That doesn't mean InsureU doesn't have any financial concerns and a need to address rates (beyond making up for the 2017 rate increase not taken) but at least Cliff and Wanda know where to focus their priorities; in this case, on distribution and why the enrollment type mix deviates so much from pricing assumptions.
Conclusion
The above case study is just an example of how an actual-to-expected analysis can be used to uncover the overall claims levels relative to benchmark levels, as well as estimate the true underlying claims trend over time. Whether the goal is to serve as a reasonability check against an internal analysis or undercover outliers and specific drivers to focus on, Medicare Supplement carriers can benefit from this exercise.
Limitations
This case study was presented for demonstration purposes only and does not reflect the actual experience of any Medicare Supplement business.
Guidelines issued by the American Academy of Actuaries require actuaries to include their professional qualifications in actuarial communications. I, Kenneth L. Clark, am a consulting actuary for Milliman, Inc. and am a member of the American Academy of Actuaries. I meet the qualification standards of the American Academy of Actuaries to render the analysis contained herein.
The opinions expressed in this report are those of the author alone and do not necessarily reflect the opinions of Milliman or other employees of Milliman.

* It's the change in enrollment type mix that drives the average premium change noted.
1 Open enrollment at age 65(OE); Subject to Guaranteed Issue provisions (GI);Medically underwritten (UW).