The interest rate challenge is perhaps best seen in relation
to automatic compound inflation adjustment.
Since the early 1990s, an inflation adjustment factor of 5 percent
compounded had become the gold standard. Agents and brokers,
who also wanted to protect their clients, made it de rigueur. But
a daily maximum benefit and a lifetime benefit pool increasing
by 5 percent compounded per year presupposes a healthy interest rate environment. State insurance departments disallow rate
increases based on interest rate assumptions, and soon insurers
found themselves in the rough water, caught between the Scylla of
ballooning liability and the Charybdis of asset-liability mismatch.
This problem was further exacerbated when the Federal Reserve
began implementing massive bond-buying programs to hold down
interest rates as economic stimulus.
Mortality has been an issue, with older persons living even
longer on claim than expected. The 1994 Group Annuity Mortality Table remains the basis for all mortality projections with
impact for selection based on gender and underwriting class.
But the bigger concern is morbidity, comprising incidence,
utilization, recovery, and continuance. This subject has long
remained something of a black box. The data needed to assess
the biological endpoint of natural human life are not yet available, although some longevity experts have argued that there
will eventually be a reduction in morbidity with death arriving
relatively soon after a reasonably healthy and active life.
most LTCI actuaries continue to assume that the elimination
of disease will result in many more persons living to advanced
age in a state of debility, and/or remaining in a state of debility
for a longer period of time. In this assumption they are looking at insured experience rather than general population data.
Disability at older ages has actually been improving among the
non-insured, while disability among insured persons, particularly those with dementia, seems to be lengthening.
Whatever the reality that will one day be borne out by research, actuaries must give more thought to what Rachel
Brewster and Sam Gutterman have termed “a properly determined provision for risk and uncertainty” with respect to both
premiums and reserves.
Citing the classic North American actuarial distinction between process risk, which is subject to random fluctuations that
nevertheless fall within certain probability distributions that
can be measured, and parameter risk, which involves random
occurrences that resist prediction because they follow no known
probability pattern or model, Brewster and Gutterman argue
for approaches that “decompose” and quantify risks. In another
recent paper, Roger Loomis, Christopher Churchill, and others
similarly underscore the importance of achieving an accurate
understanding of the “boundaries of normal variation,” i.e., how
much variance is entailed for each operational or financial metric—and estimating the probability of random events occurring
in certain intervals.
Not surprisingly, the latter requires the assignment of a con-
fidence factor and comes particularly into play with lapse and
claims utilization, where reserves may fluctuate widely within
a normal period. With the benefit of stochastic modeling, which
employs Monte Carlo simulation methods, the authors exam-
ine a host of features and provisions to ascertain outliers from
risks that merely appear to be so. What needs to take place, they
conclude, is an examination of all the features and provisions that
pose parameter risks and magnify uncertainty. Such a reckoning
would help to ensure that each risk assumed aligns with an insur-
er’s marketing strategy and capacity for or tolerance of risk—and
conversely, to identify and avoid risks that may only make sense
in light of social or other non-business objectives. Failure to dif-
ferentiate among risks that are reasonably predictable from those
that are not has resulted in confrontations with state insurance
departments as insurers seek approval for premium increases
that may not correspond with the actual risks assumed.
Finally, there is the matter of reinsurance. Reinsurance is an
important tool in other risk lines. Yet it is almost absent in LTCI.
There are many reasons for this, including changing assumptions about the amount and length of claims, rate increases that
do not get approved, unwillingness of reinsurers to assume risk
without information that insurers find hard to obtain, and action by the Federal Reserve to depress interest rates. But it is
unsurprising that reinsurance has not played a major role in the
development of private LTCI. Insurers initially thought the risk
was manageable, and early reinsurance contracts tended to be
written as coinsurance, which lowered total risk assumed but
also reduced premiums. Today, however, the use of reinsurance
is likely to increase as carriers look to manage volatility in their
books and reduce capital strain. If the experience of Genworth
and other insurers is to carry weight, reinsurance would seem a
necessary tool that should be employed on a wide basis.
III. Operational Challenges
As noted, LTCI poses administrative challenges, as might be
expected of a permanent product in flux. Take, for example,
inflation adjustment, especially with respect to future or guaranteed purchase option (FPO or GPO) offers. These adjustments
must be orchestrated regularly (generally every one to three
years, depending on the plan design) and can pose problems
■ ■ Assumptions: morbidity, mortality, lapse,
recovery, continuance, interest rates, inflation
■ ■ Unlimited/lifetime policies
■ ■ Inability to get rate increases approved by state
■ ■ Capital strain
■ ■ Profitability