Of all the features de-risked during this latest cycle in LTCI product design, lifetime benefit periods stand out in the hearts and minds of veteran agents, as well as their best clients. Only a decade ago, sales of lifetime benefits alone, as measured in premium volume, equaled the total stand-alone industry production for all of 2014. Yet, as the distribution of this product feature has dwindled, the need for lifetime benefits has only grown.
It looks doubtful that existing carriers are positioned to bring back one of the benefits that inflicted many of the wounds on their in-force policies. In part, this is a casualty of the overall trend to put the product into the smallest box of potential outcomes so that the product will not lose money in any plausible scenario. This is largely a reaction to the pressures on corporate management from analysts, shareholders and rating agencies who have developed a particularly vitriolic view of lifetime benefits. Offering protection forever sounds scary, but is it really?
What if the root cause of the problem was not really the lifetime benefit itself? Lifetime benefits can now be viewed in a different light as the result of the largest, most recent study to date compiled by the Society of Actuaries (SOA)1 The results from this study shocked many of the experts in the LTCI actuarial community when they were first announced at the Intercompany Long Term Care Insurance Conference held in late March 2015.
This article will help explain why bringing lifetime benefits back to the market has the potential to be a game changer. In particular, there has been divided opinion for years on two key actuarial assumptions. In the process of examining these assumptions, the SOA study has begun to point to the true causal factors. By dispelling the myths and replacing impressions with facts, we can break the “carrier barrier” and once again offer lifetime benefits in stand-alone LTCI products.
Carrier Barrier Concern #1:
Do Lifetime Benefits Result in More
Claims Than Limited Benefits?
Carriers have had differing experiences with lifetime benefits among their more troubled products of the past. For some carriers, lifetime benefits had more claims (in actuarial lingo: higher “incidence rates”) than limited benefits sold on those same products. One common explanation for this phenomenon was that policyholders wanted to conserve limited benefits, so that they would be available later in life when really needed. Yet, for other carriers with healthier lives, data showed that those blocks had better incidence rates on lifetime benefits than expected. It was then argued, from this observation, that people who could afford the higher premiums also had other resources at their disposal and would use their LTCI policy as a last resort. How could both interpretations be correct?
Using a statistical technique more common to the property and casualty industry, the recent SOA study revealed insight on the contradictory data:
3.1.5 Sample Model Discussion
“Certain results of the model were discussed in depth among the steering committee and other team members. An example of this is the impact of benefit period on incidence rates.
The final model has a slight difference between limited and lifetime benefit periods. The expectation prior to the study was that lifetime benefit periods would have materially higher incidence rates than limited benefit periods. Though this relationship was observed in some of the participating companies, not every participant’s data exhibited this relationship. When all of the data was aggregated, the relationship of higher incidence for lifetime benefit period was no longer present.
This was an example of the strengths of the model, but illustrates its limitations as well. The models and relationships therein were developed based on the aggregate data without regard for the relationships inherent in specific companies’ data. For this reason the same relationships may not exist for each company in the long term care industry.
There was significant discussion around this topic that provided potential explanations for the differences of this relationship from expectation. A (non-exhaustive) sample of potential drivers of the benefit period relationship in this model include:
• The mix of companies participating in the study; and
• Differences in underwriting protocols by company.
This relationship is an example of a potential difference between an individual company’s result and the aggregate model. Other differences could be length of underwriting differences by policy duration. Some companies may have differences for five or 10 or 25 years, for example, based on their underwriting.”
The data demonstrates that underwriting plays a deeper role in the utilization of LTCI benefits than originally thought. For the first time, the SOA Long Term Care Experience Study2 subdivided the data based on the underwriting style of the issuing company. As can be seen in Table 1 in column four, annual incidence rates for policies that were not fully underwritten were as much as three times higher than fully underwritten policies. As policyholders get older, the underwriting selection effect reduces in percentage terms, but never fully disappears. On the other hand, the difference in the absolute claims rates continues to increase by age. As seen in column five, at age 50, fully underwritten business has one in 1,000 fewer claims per year, but by age 75, the difference is almost 10 in 1,000 fewer claims per year.
Carrier Barrier Concern #2:
Do Lifetime Benefits Have an
Unlimited Cost per Claim?
The need for stand-alone LTCI is evident and has only grown larger, because long term care services are expensive. According to the recent cost of care surveys, the average cost of a nursing home is $90,000–$100,000 per year, and this cost seems to grow at a rate greater than overall inflation each year. Lifetime benefits appeals to consumers who can afford the coverage because it allows them to protect against a catastrophic multiyear event by replacing an unknown cost with a fixed premium. While it is understood that long term care needs are common, the chance of a catastrophic long term care event is not as well understood. We know that people are living longer and that Alzheimer’s—one of the most costly diseases in the United States—is a growing epidemic. On the other hand, most uses of LTCI are shorter in duration as people either recover quickly from acute conditions or use LTCI for end-of-life care. Once again, we must turn to the data to understand the cost per claim. The largest uncertain component of cost per claim is the amount of time spent on claim (actuarially termed “continuance”).
Historically, it has been difficult for actuaries to measure credible data for the amount of time spent on claim. Not only does it often take many years before policyholders use LTCI, but additional time is needed to collect data about the people who have already been on claim for several years. The new SOA study contained significantly more insured experience simply because more years have passed since the prior study. In particular, the first three years on claim has enough data to help quantify the unlimited risk potential of lifetime benefits.
It turns out that after satisfying the elimination period, roughly 50 percent of individual claims last less than one year. One of the reasons is that recovery from claim in the first year is very common, particularly for people who require care at younger ages. About 80 percent of males and 70 percent of females have claims lasting less than three years. Once policyholders survive the first three years in a disabled state, the remaining length of claim is highly dependent on the age that they went on claim. Younger people already on claim for three years could spend an average of six more years on claim, whereas older claimants might average only two more years or less.
As was the case with incidence rates, underwriting plays a critical role in controlling the risk of lifetime benefits. Full underwriting largely limits claims until they occur at ages at which the catastrophic risk is minimized. To measure the impact of the delay of claim on continuance, I calculated an average length of claim in years from the SOA study continuance rates. The average length of claim was calculated for both the full lifetime claim with no cap, as well as the average length of claim for the first three years only; thus designed to approximate a three-year benefit period (BP).
As can be seen in column two and column five of Table 2, male claimants average between 1.5 and 2.6 years on claim. Female claimants have roughly 50 percent longer claims than males at most ages, with a range of 1.7 to 2.8 years on claim. The differential between males and females has been well understood for many years, and the more recent trend toward gender-distinct pricing captures this fact. However, less commonly understood is the difference between the lengths of the unlimited claim compared to the three year capped claim. As can be seen in column four and column seven, claims at younger ages have much more risk with an unlimited BP, but the difference between lifetime and three-year BP claims converges quickly at older ages. This differential is magnified considering that the policyholders who recover at younger ages have a strong likelihood of requiring care again later in life.
Conclusion
Lifetime benefits have traditionally been most popular among consumers, and especially the well-to-do consumer who represents the most frequent LTCI sale. In the past, more liberal underwriting caused lifetime benefits to be far too risky for carriers to offer. However, a deeper analysis demonstrates that lifetime benefit can still be profitably sold for fully underwritten products because strict underwriting controls both the number of claims and the length of those claims.
Full underwriting significantly reduces incidence rates at younger ages, which is also the epicenter of the catastrophic length of claims for lifetime benefits. Going one step further, there are several other positive impacts, from the carrier’s perspective, for delaying claims by using full underwriting, such as:
• More premiums collected to fund reserves
• More investment income earned on money held in reserves
• Less chance that a rate increase will ever be needed
• More time for the impact of medical improvements in healthcare to reduce the future cost of care, and in particular,
• If a breakthrough in Alzheimer’s treatment occurs, as everyone hopes will happen, the resulting morbidity improvement will benefit lifetime benefits the most.
The views expressed here are those of the author, and not of LifeCare Assurance or the Society of Actuaries. Any inaccuracies in interpreting the data are those of the author alone.
Footnotes:
1. https://www.soa.org/Research/Experience-Study/Ltc/2000-2011-ltc-experience-basic-table-dev.aspx-Release – Phase 2: Jan – Apr 2015
2. https://www.soa.org/Research/Experience-Study/Ltc/research-ltc-study-2000-11-aggregated.aspx
