Multiple Triggers and Partial Benefit Eligibility

Why Multiple Triggers Need a Payment Rule

A long-term care contract may recognize more than one route to eligibility. A policyholder might qualify through functional impairment, cognitive impairment, or both. Those tests describe conditions that can overlap; they do not, by themselves, say whether the contract pays one benefit, the larger of two benefits, or a sum subject to a cap. That payment rule must come from the contract.

The actuarial task is to represent the possible combinations of impairment, assign each combination the benefit specified by the contract, and then value payments without counting the same policyholder twice. The earlier material on activities-of-daily-living eligibility, cognitive impairment triggers, and state-dependent benefit amounts provides the ingredients. Here the focus is how to coordinate those triggers when more than one can apply.

Definition: Let $F$ denote satisfying a functional-impairment trigger and $C$ denote satisfying a cognitive-impairment trigger. The mutually exclusive impairment categories are $N$ (neither trigger), $F$ only, $C$ only, and $B$ (both triggers). A contract's benefit rule assigns one payment $b(s)$ to each category $s$. Death, if included, is a separate state with no care benefit unless the contract explicitly provides otherwise.

The categories are mutually exclusive even though the original trigger events are not. For example, $F$ and $C$ overlap in category $B$. Therefore, the probability that at least one trigger is met is $P(F\cup C)=P(F)+P(C)-P(F\cap C)$. The subtraction removes the overlap. In a state-based valuation, the same principle is applied by assigning the combined category its own probability and payment.

Specify the Contract Before Calculating

Common coordination rules include paying the larger applicable benefit, paying a specified combined amount, or adding eligible amounts subject to a limit. These rules are not interchangeable. In particular, “two triggers are satisfied” does not imply “two full benefits are payable.” The contract's wording determines whether the combined-state benefit is a maximum, a stated amount, or a capped sum.

Contract rule	Payment in the combined state	Valuation implication
Alternative triggers; one benefit	One specified benefit, potentially the larger applicable benefit	Do not add payments merely because both triggers apply
Partial benefits that coordinate	A stated combined amount	Use that amount, not an inferred sum
Additive benefits with a cap	The lesser of the sum and the contractual cap	Apply the cap after adding the specified components

A useful setup is to list every mutually exclusive state and write its payment beside it before calculating an expected value. If a benefit is paid at time $t$, the expected present value is the sum over states of the state probability times the state payment, discounted to the valuation date. This calculation requires probabilities for the joint states, not just separate probabilities for each trigger.

\operatorname{EPV}=\sum_t v^t\sum_s P(S_t=s)\,b_t(s), \qquad v=\frac{1}{1+i}.

Here $S_t$ is the policyholder's state at payment time $t$, $b_t(s)$ is the contract payment in state $s$, and $i$ is the effective annual interest rate when $t$ is measured in years. The formula assumes the stated payment timing and a benefit amount determined by the state at that time. If benefits are paid monthly or continuously, the timing and discounting must instead match those payment terms.

Why the State-Based Formula Works

At a fixed payment time, define the benefit random variable $Y=b(S_t)$. The mutually exclusive states partition all possible outcomes, so exactly one state occurs. The expected value of $Y$ is therefore the sum of each possible payment multiplied by the probability of its state. Discounting a payment at time $t$ multiplies it by $v^t$; summing across payment times gives the stated expected present value.

E[Y]=\sum_s b(s)P(S_t=s), \qquad E\!\left[\sum_t v^tY_t\right] =\sum_t v^t E[Y_t] =\sum_t v^t\sum_s P(S_t=s)b_t(s).

The second equality follows from linearity of expectation; it does not require payments at different times to be independent. The formula does require that the state probabilities and benefit amounts correspond to the same time and contractual definition. For example, probabilities conditional on survival must not be used as though they were unconditional probabilities if death can occur before payment.

Worked Example: Alternative Triggers With a Larger-Benefit Rule

A hypothetical policy pays $1,000 per month for functional impairment and $800 per month for cognitive impairment. When both triggers are met, it pays the larger amount, not the sum. At the annual payment date, the probabilities of functional-only, cognitive-only, both, neither, and death are 0.20, 0.10, 0.15, 0.50, and 0.05, respectively. The annual benefit is paid at year-end. Use an effective annual interest rate of 4%; the stated probabilities include mortality, so no additional mortality adjustment is made. Find the expected present value of this year's benefit.

Method. First map each mutually exclusive state to the contract payment. The annual payments are $12,000 for functional-only, $9,600 for cognitive-only, $12,000 for both, and $0 for neither or death. The combined-state payment is $12,000 because the contract pays the larger amount.

The probabilities sum to $0.20+0.10+0.15+0.50+0.05=1.00$. The expected year-end payment is $0.20(12{,}000)+0.10(9{,}600)+0.15(12{,}000)+0.50(0)+0.05(0)$. The terms are $2,400, $960, $1,800, $0, and $0, giving an expected payment of $5,160.

Since the payment is at year-end, $v=1/1.04$, and its expected present value is $5{,}160/1.04=$$4,961.538... . Rounded to the nearest cent, the answer is $4,961.54.

Check and common mistake. The five probabilities total 1, and the combined-state payment is no larger than the functional benefit. Adding both monthly benefits in state $B$ would incorrectly use $21,600 for that year's payment. That wrong calculation raises the expected year-end payment by $0.15(9{,}600)=\$1{,}440$, because it counts an extra cognitive benefit in every combined case.

Valuing Benefits Across Multiple Dates

When impairment can change over time, the state distribution at a later date depends on transitions between states. A transition matrix gives the probability of moving from each current state to each next state. To value a benefit at a future date, first propagate the state probabilities through the matrix, then apply the contract's payment rule to the resulting distribution. Death may be included as an absorbing state: once entered, it remains the state and produces no later care benefit.

Define exclusive states.
Separate neither, functional-only, cognitive-only, both, and (if relevant) death.

Translate the contract.
Record one payment for each state, applying any larger-benefit rule or cap exactly as written.

Find state probabilities at each payment time.
Use the initial distribution and transition assumptions, including death where applicable.

Weight and discount.
Multiply each state's probability by its payment, sum within each date, and discount each date's expected payment.

Worked Example: Two Annual Payments With Recovery and Combined Eligibility

A hypothetical contract pays $12,000 at each year-end in states $F$ and $B$, $9,600 in state $C$, and nothing in $N$ or $D$. In $B$, the larger benefit applies. At time 0 the policyholder is in $N$. The one-year transition probabilities, with rows and columns ordered $N,F,C,B,D$, are:

\begin{pmatrix} 0.70&0.12&0.08&0.05&0.05\\ 0.10&0.55&0.05&0.20&0.10\\ 0.10&0.04&0.56&0.20&0.10\\ 0.04&0.06&0.05&0.70&0.15\\ 0&0&0&0&1 \end{pmatrix}.

The entries in each row sum to 1. Death is absorbing. Use an effective annual interest rate of 4%, and value the two year-end payments. The transition probabilities specify the mortality assumption as part of the model; do not apply a separate mortality factor.

Year 1. Starting from $N$, the state probabilities are the first row: $N=0.70$, $F=0.12$, $C=0.08$, $B=0.05$, and $D=0.05$. Thus the expected year-end benefit is $0.12(12{,}000)+0.08(9{,}600)+0.05(12{,}000)=1{,}440+768+600=\$2{,}808$.

Year 2. Multiply the year-1 distribution by the transition matrix. For example, $p_2(F)=0.70(0.12)+0.12(0.55)+0.08(0.04)+0.05(0.06)=0.1562$. Calculating the other components gives $p_2(N)=0.5120$, $p_2(C)=0.1093$, $p_2(B)=0.1100$, and $p_2(D)=0.1125$. These probabilities sum to 1.0000.

The expected year-2 benefit is $0.1562(12{,}000)+0.1093(9{,}600)+0.1100(12{,}000)$ $=1{,}874.40+1{,}049.28+1{,}320.00=\$4{,}243.68$.

Present value. The year-1 expected payment has present value $2{,}808/1.04=\$2{,}700.00$. The year-2 expected payment has present value $4{,}243.68/(1.04)^2=4{,}243.68/1.0816=\$3{,}923.52$ to the nearest cent. The total expected present value is $6,623.52.

Check and common mistake. The second-year probabilities include entrants from every non-death state, not just people who were already receiving a benefit in year 1. Also, the $B$ payment is $12,000, not $21,600: the contract specifies the larger benefit. Reusing the year-1 state probabilities for year 2 would ignore transitions, recovery, and death.

Additive Partial Benefits and Caps

Not every contract uses a larger-benefit rule. A contract may specify partial amounts for each trigger and add them when both are met, while limiting the total payment. In that case, the combined-state amount is found by applying the contractual formula, not by choosing the maximum automatically. For component amounts $b_F$ and $b_C$ and a cap $K$, an explicitly additive rule gives $b(B)=\min(b_F+b_C,K)$. The states must still be mutually exclusive so that this combined payment is applied once.

Worked Example: Partial Benefits With a Combined Monthly Cap

A hypothetical plan pays $1,100 per month for functional impairment and $900 per month for cognitive impairment. The contract adds the amounts when both triggers are met, subject to a combined cap of $1,800 per month. For a one-year benefit period, the year-end probabilities of functional-only, cognitive-only, both, neither, and death are 0.18, 0.12, 0.10, 0.55, and 0.05. These probabilities include mortality. The annual benefit is paid at year-end, and the effective annual interest rate is 5%. Find its expected present value.

Method. The annual functional-only payment is $12(1{,}100)=\$13{,}200$; the cognitive-only payment is $12(900)=\$10{,}800$. In state $B$, the uncapped monthly sum is $1{,}100+900=\$2{,}000$, but the cap makes it $1,800 per month, or $12(1{,}800)=\$21{,}600$ annually. The payments in $N$ and $D$ are zero.

The expected year-end payment is $0.18(13{,}200)+0.12(10{,}800)+0.10(21{,}600)$ $=2{,}376+1{,}296+2{,}160=\$5{,}832$. The probabilities sum to 1.00. Discounting gives $5{,}832/1.05=\$5{,}554.2857...$, or $5,554.29 to the nearest cent.

Check and common mistake. The combined payment uses the cap after adding the two contractual components. Using $2,000 per month ignores the cap and overstates the expected year-end payment by $0.10(12)(200)=\$240$, or the expected present value by $\$240/1.05=\$228.57$ to the nearest cent.

Common Exam Traps

Adding probabilities for overlapping triggers without removing the intersection. Use exclusive state probabilities, or use $P(F\cup C)=P(F)+P(C)-P(F\cap C)$.
Assuming a payment rule that the contract does not state. The benefit in the both-trigger state may be the larger amount, a specified combined amount, or a capped sum.
Counting a combined case twice. Once the policyholder is assigned to state $B$, apply the single payment $b(B)$; do not also include that person in the $F$-only or $C$-only category.
Mixing conditional and unconditional probabilities. If trigger probabilities are conditional on survival, multiply by the probability of survival before treating them as unconditional payment probabilities.
Discounting at the wrong time. A year-end payment is discounted by $v$, while a payment at the end of year 2 is discounted by $v^2$.

Key takeaway: Define mutually exclusive impairment states, apply the contract's exact coordination rule once to each state, and value each payment using probabilities and discount factors that match its timing. This framework is aligned with the SOA Exam ALTAM (Advanced Long-Term Actuarial Mathematics) syllabus.

Check Your Understanding

For each problem, identify the state-level payment rule before calculating. Answers are provided for checking.

A hypothetical policy pays $900 per month for functional-only impairment, $600 for cognitive-only impairment, and the larger amount when both triggers apply. At year-end, the probabilities of functional-only, cognitive-only, both, neither, and death are 0.16, 0.09, 0.10, 0.60, and 0.05. Benefits are paid at year-end, mortality is included in these probabilities, and $i=5\%$. Find the expected present value. (Answer: $3,291.43)
At a specified assessment date, $P(F)=0.24$, $P(C)=0.18$, and $P(F\cap C)=0.07$. Find the probability that at least one trigger is satisfied. (Answer: 0.3500)
A hypothetical contract pays $800 per month for functional-only impairment and $500 for cognitive-only impairment; for both, it adds the two amounts subject to a cap of $1,100 per month. The year-end probabilities of functional-only, cognitive-only, both, neither, and death are 0.20, 0.10, 0.15, 0.50, and 0.05. The annual benefit is paid at year-end, mortality is included, and $i=4\%$. Find the expected present value. (Answer: $4,326.92)
A hypothetical contract pays $12,000 at year-end in states $F$ and $B$, and $9,000 in state $C$. A policyholder starts in $F$. Over the year the transition probabilities from $F$ to $N,F,C,B,D$ are 0.10, 0.50, 0.10, 0.20, and 0.10. Death pays no benefit, and $i=3\%$. Find the expected present value of the year's benefit. (Answer: $9,029.13)
Among policyholders who survive to the assessment date, suppose $P(F)=0.25$ and $P(C)=0.15$, and the triggers are independent. A hypothetical contract pays the larger of $1,200 per month for $F$ and $800 per month for $C$, if either or both triggers are met. Survival probability to the date is 0.96, death produces no payment, the annual benefit is paid at that date one year from now, and $i=4\%$. Find the expected present value. (Answer: $4,320.00)

What You'll Learn