A Guide to Finite Term Insurance and Its Actuarial Value

Ever bought something with an expiration date? A gallon of milk, a subscription, or even a coupon? You pay for a benefit that is only valid for a specific timeframe. Life insurance has a parallel concept in the world of finance, and it’s called Finite Term Insurance. It’s one of the simplest and most essential products in the actuarial toolkit, yet its underlying mathematics holds surprising depth. If you’ve ever wondered how insurance companies in India, or anywhere else, can offer substantial cover for comparatively small premiums, the answer lies in understanding this concept. Let’s peel back the layers and explore this powerful form of pure protection and its complex actuarial valuation.

Finite term insurance: The pure protection play 🔗

Imagine you’re an ambitious young professional in Mumbai, just bought a home, and have a five-year education loan for your child. Your biggest financial concern is what happens if the unexpected occurs before your mortgage is paid off or your child finishes school. You need a financial safety net, but only for that specific, critical period.

This is where finite term insurance, or $n$-year term insurance, steps in. It is a straightforward contract: the insurer promises to pay a fixed death benefit only if the insured person dies within a specified term, $n$ years, from the policy’s start. If the insured survives past the term, the policy simply expires, and no benefit is paid. This makes it a pure protection policy, contrasting sharply with traditional bundled products that include a savings or investment component, like an endowment plan.

In the Indian context, term insurance has gained significant traction, allowing consumers to choose a high sum assured for a limited period at an affordable premium, separating their insurance and investment needs. Because it lacks a survival benefit, the insurer’s liability is strictly contingent on a single event (death) occurring within a defined window (the term), simplifying the risk profile compared to whole life products.

The net single premium: Actuarial present value ($A^1_{x:n}$) 🔗

To the layperson, the premium is just the monthly or annual payment. To an actuary, the foundation of this payment is the Net Single Premium (NSP). The NSP is the single, lump-sum amount of money an insurer must collect at the start of the policy from the insured to exactly cover the expected future benefit payment, assuming no expenses or profit loading, and a specific investment return. It represents the Actuarial Present Value (APV) of the contingent death benefit.

Calculating the expected cost: The integral of discounted probability 🔗

For an $n$-year term insurance policy issued to a person currently aged $x$, the APV is denoted by $A^1_{x:\overline{n}|}$. Since the death benefit is a single, contingent cash flow, its value is derived by calculating the Expected Present Value ($E[Z]$) of the payment. The payment amount is fixed (say, 1 unit), but the timing of the payment, $T$, is a random variable (the insured’s future lifetime), and the payment is only made if $T \le n$.

Assuming the benefit is payable immediately on death (the continuous case, often used as a theoretical benchmark), the present value random variable $Z$ is:

$$Z = \begin{cases} v^T & \text{if } T \le n \\ 0 & \text{if } T > n \end{cases}$$

where $v^t$ is the discount factor for time $t$.

The Net Single Premium, $A^1_{x:\overline{n}|}$, is the expected value of $Z$, which is computed by integrating the discounted benefit over the term of the policy, from time 0 to $n$:

$$A^1_{x:\overline{n}|} = E[Z] = \int_{0}^{n} v^t \cdot t p_x \cdot \mu_{x+t} \, dt$$

In simpler terms:

• $v^t$ is the present value of 1 unit of money payable at time $t$.
• $t p_x$ is the probability that a life aged $x$ survives to time $t$.
• $\mu_{x+t}$ is the force of mortality at age $x+t$ (the instantaneous probability of death).
• The term $t p_x \cdot \mu_{x+t} \, dt$ represents the infinitesimal probability that the person aged $x$ dies at time $t$.

The integral sums up the present value of the death benefit for every possible instant of death within the $n$-year term. This final figure, $A^1_{x:\overline{n}|}$, is the pure risk cost that must be covered by the initial single premium payment.

Volatility of the policy cost: The variance of the present value ($Var(Z)$) 🔗

For an insurer, simply knowing the average expected cost (the NSP) isn’t enough. They need to understand the volatility or risk associated with that cost. The policy cost is a random variable, $Z$, and its value could be anything from $v^T$ (if death occurs early) to 0 (if the insured survives the term). The degree to which the actual claim payout deviates from the expected cost is measured by the variance of the present value random variable, denoted by $Var(Z)$.

The need for the second moment ($^2A^1_{x:n}$) 🔗

The general formula for variance is $Var(Z) = E[Z^2] – (E[Z])^2$.

In our case:

1. $E[Z]$ is the Net Single Premium, $A^1_{x:\overline{n}|}$.
2. $E[Z^2]$ is the second moment of the present value random variable, denoted by $^2A^1_{x:\overline{n}|}$.

The second moment is calculated similarly to the NSP, but instead of discounting the benefit once by $v^t$, we discount the square of the benefit, which means discounting by $v^{2t}$. Recall that the benefit is 1 unit, so the squared benefit is $1^2 = 1$. The discount factor $v^{2t}$ can be viewed as the standard discount factor $v^t$ but calculated using a different, higher interest rate, $i^*$, such that $v^* = v^2$.

The formula for the second moment is:

$$^2A^1_{x:\overline{n}|} = E[Z^2] = \int_{0}^{n} (v^t)^2 \cdot t p_x \cdot \mu_{x+t} \, dt$$

The Variance of the Present Value for a unit benefit is then:

$$Var(Z) = ^2A^1_{x:\overline{n}|} – (A^1_{x:\overline{n}|})^2$$

Why variance matters in a story 🔗

Think of an insurance company like a chef running a large kitchen. The NSP ($A^1_{x:\overline{n}|}$) is the average ingredient cost per dish-what they *expect* to pay. However, some days a shipment of costly ingredients is spoiled (an early claim), and some days they pay nothing for that specific dish because the customer cancels (the insured survives the term). The variance tells the chef (the insurer) how much their actual costs might swing from the average. High variance means high volatility and high risk.

Understanding this volatility is crucial for:

• Pricing (Gross Premium): Insurers must add a risk loading to the NSP to cover unexpected spikes in claims. Higher $Var(Z)$ means a higher loading, leading to a higher gross premium for the customer.
• Solvency Capital: Regulatory bodies, like the Insurance Regulatory and Development Authority of India (IRDAI), mandate that insurers hold a certain amount of capital to ensure they can pay even a statistically unlikely wave of claims. This capital requirement is directly influenced by the calculated risk, which the variance helps quantify (Source).

Relating term insurance to its whole life counterpart 🔗

Finite term insurance can be seen as a building block for more complex products. For instance, an $n$-year Endowment Assurance, which pays a benefit upon death within $n$ years OR upon survival to the end of $n$ years, can be mathematically decomposed:

$$A_{x:\overline{n}|} = A^1_{x:\overline{n}|} + E_{x:\overline{n}|}$$

Where $A_{x:\overline{n}|}$ is the endowment APV, $A^1_{x:\overline{n}|}$ is the term insurance APV, and $E_{x:\overline{n}|}$ is the APV of the Pure Endowment (the survival benefit). This relationship shows that finite term insurance provides the ‘death’ component for all time-limited life insurance products.

Moreover, the term insurance risk often declines over time for the insurer, especially in long-duration contracts. As the term approaches its end, the probability of a claim within the remaining time decreases, and so does the remaining APV.

What do you think? Given that the Net Single Premium for a whole life policy ($A_x$) covers the risk of death at any future time, why do you think its actuarial variance might sometimes be considered lower, on a relative basis, than the variance for a short-term policy like $A^1_{x:\overline{5}|}$?

References

1. https://www.actuariesindia.org/sites/default/files/2022-05/Valuing_the_Term%20Insurance%20Products%20in%20the%20Indian%20Market.pdf
2. https://www.hdfclife.com/insurance-knowledge-center/about-life-insurance/what-is-single-premium-insurance-policy
3. https://en.wikipedia.org/wiki/Actuarial_present_value
4. https://www.investopedia.com/terms/t/timevalueofmoney.asp
5. https://www.sbilife.co.in/sites/SBILife/Annual-Report/FY24/pdf/Additional.pdf