Future loss random variables for life insurance and annuities

Q: What is the definition of the future loss random variable at issue (\b(L\b) or \b(_0L\b)) for an insurance or annuity contract?

\b(L = PV(\text{Future Benefits}) - PV(\text{Future Premiums})\b).

Q: What is the Equivalence Principle and how is it used to determine net premiums?

It sets the net premium such that the expected future loss at issue is zero: \b(E[L] = 0\b), meaning the expected PV of future benefits equals the expected PV of future premiums.

Q: How can the future loss \b(L\b) for a fully discrete whole life policy be expressed as a linear function of the present value of benefits \b(v^{K+1}\b)?

\b(L = \left(1 + \frac{P}{d}\right)v^{K+1} - \frac{P}{d}\b)

Q: What is the formula for the variance of the future loss random variable \b(Var(L)\b) for a fully discrete whole life policy?

\b(Var(L) = \left(1 + \frac{P}{d}\right)^2 \left[^2A_x - (A_x)^2\right]\b)

Q: How is the future loss random variable \b(L\b) defined piecewise for a discrete \b(n\b)-year term insurance policy?

\b(L = v^{K+1} - P \ddot{a}_{\overline{K+1|}}\b) if \b(K

Q: What are the linear expression and variance formula for the future loss \b(L\b) of an \b(n\b)-year endowment insurance?

\b(L = \left(1 + \frac{P}{d}\right) v^{\min(K+1, n)} - \frac{P}{d}\b) and \b(Var(L) = \left(1 + \frac{P}{d}\right)^2 \left[^2A_{x:\overline{n|}} - (A_{x:\overline{n|}})^2\right]\b).

Q: What is the future loss random variable \b(L\b) and its variance for a single premium whole life annuity-due?

\b(L = \ddot{a}_{\overline{K+1|}} - \Pi\b), and under the Equivalence Principle (\b(\Pi = \ddot{a}_x\b)), \b(Var(L) = \frac{^2A_x - (A_x)^2}{d^2}\b).

Q: What is the variance of the future loss random variable for an \b(n\b)-year temporary life annuity-due funded by a single net premium?

\b(Var(L) = \frac{^2A_{x:\overline{n|}} - (A_{x:\overline{n|}})^2}{d^2}\b)

Q: How is the second moment actuarial notation \b(^2A\b) evaluated with respect to the interest and mortality assumptions?

\b(^2A\b) is calculated by doubling the force of interest \b(\delta\b) (replacing \b(v\b) with \b(v^2\b) or using \b(i* = 2i + i^2\b)) while keeping mortality probabilities unchanged.

Exam FAM – Fundamentals of Actuarial Mathematics · SOA (Society of Actuaries) · 5 min read · Updated: 7 Nov 2026

Welcome to the Core of Actuarial Pricing!

Hello future actuaries! Today we are diving into one of the most important concepts in Exam FAM: Future Loss Random Variables. While the word "loss" sounds scary, in the actuarial world, it's just our way of keeping score. Think of it as a balance sheet for a single insurance or annuity contract.

By the end of these notes, you will understand how insurance companies define loss variables and calculate premiums so that, on average, they break even or meet target profit goals. This is the foundation of Premium and Policy Value Calculation for Long-Term Insurance Coverages.

1. What is a Future Loss Random Variable?

Imagine you are the insurance company. When you sell a policy, two cash flow streams take place:
1. You promise to pay Benefits at future dates contingent on survival or death (money going out).
2. You receive Premiums from the policyholder (money coming in).

The Future Loss Random Variable, denoted by $L$ (or $_0L$ at issue), is the difference between the present value of future benefits (and expenses, if any) and the present value of future premiums:

The Basic Formula:
$L = PV(\text{Future Benefits}) - PV(\text{Future Premiums})$

Analogy: Think of it like a see-saw. On one side, you have the present value of money the company owes. On the other side, you have the present value of money the customer pays. Our goal is to assess the balance of those cash flows today.

Quick Review:
- If $L > 0$: The contract produced a loss for the insurer (benefits exceeded premiums in present value).
- If $L < 0$: The contract produced a profit for the insurer (premiums exceeded benefits in present value).

2. The Equivalence Principle

How does an actuary determine the "fair" net premium? Under the Equivalence Principle, the premium is set such that at issue, the Expected Value of the future loss random variable is exactly zero.

The Rule: $E[L] = 0$
This means: Expected PV of Benefits = Expected PV of Premiums.

Example: Whole Life Insurance (Continuous)
For a whole life insurance policy paying \$1 at the moment of death with a continuous premium rate $\bar{P}$:
$L = v^T - \bar{P} \cdot \bar{a}_{\overline{T|}}$
Setting $E[L] = 0$:
$E[v^T] - \bar{P} \cdot E[\bar{a}_{\overline{T|}}] = 0$
$\bar{A}_x - \bar{P} \cdot \bar{a}_x = 0 \implies \bar{P} = \frac{\bar{A}_x}{\bar{a}_x}$

3. The "Linear Trick" for Future Loss Variance

Calculating the Variance of the future loss is a classic Exam FAM exam question. For standard whole life insurance policies, we can rewrite $L$ as a simple linear function of the present value of benefits $Z = v^{K+1}$.

For a Discrete Fully Discrete Whole Life Policy:

For a unit benefit whole life policy with level annual premium $P$ paid at the beginning of each year:
$L = v^{K+1} - P \ddot{a}_{\overline{K+1|}}$
Since $\ddot{a}_{\overline{K+1|}} = \frac{1 - v^{K+1}}{d}$, substituting gives:
$L = v^{K+1} - P \left( \frac{1 - v^{K+1}}{d} \right) = \left(1 + \frac{P}{d}\right)v^{K+1} - \frac{P}{d}$

Variance Formula:

Because $Var(aX + b) = a^2 Var(X)$, the variance of $L$ simplifies to:
$Var(L) = \left(1 + \frac{P}{d}\right)^2 \cdot [^2A_x - (A_x)^2]$

When $P = P_x = \frac{A_x}{\ddot{a}_x}$, remember that $1 + \frac{P_x}{d} = \frac{1}{d \ddot{a}_x} = \frac{1}{1 - A_x}$.

4. Loss Variables for Life Insurance Policies

The form of $L$ depends on the benefit trigger and coverage term:

Term Insurance:

The benefit is paid only if death occurs within $n$ years. Premiums are paid up to $n$ years while alive.
For a discrete $n$-year term insurance with annual premium $P$:
- If $K < n$: $L = v^{K+1} - P \ddot{a}_{\overline{K+1|}}$
- If $K \ge n$: $L = 0 - P \ddot{a}_{\overline{n|}} = -P \ddot{a}_{\overline{n|}}$

Endowment Insurance:

An $n$-year endowment insurance pays 1 at the end of the year of death or at maturity $n$, whichever comes first:
$L = v^{\min(K+1, n)} - P \ddot{a}_{\overline{\min(K+1, n)|}}$
Using the identity $\ddot{a}_{\overline{\min(K+1, n)|}} = \frac{1 - v^{\min(K+1, n)}}{d}$, this also reduces to the linear shortcut:
$L = \left(1 + \frac{P}{d}\right) v^{\min(K+1, n)} - \frac{P}{d}$
$Var(L) = \left(1 + \frac{P}{d}\right)^2 [^2A_{x:\overline{n|}} - (A_{x:\overline{n|}})^2]$

5. Loss Variables for Annuity Contracts

Future loss random variables also apply to life annuities. Here the cash flow signs are inverted: the insurer pays out annuity benefits while the annuitant survives, funded by a single upfront premium or temporary premiums.

Single Premium Whole Life Annuity-Due:

Suppose an annuitant aged $x$ purchases a whole life annuity paying \$1 at the beginning of each year while alive, with a single net premium $\Pi$.
The present value of benefits paid is $Y = \ddot{a}_{\overline{K+1|}} = \frac{1 - v^{K+1}}{d}$.
The future loss at issue is:
$L = \ddot{a}_{\overline{K+1|}} - \Pi = \frac{1 - v^{K+1}}{d} - \Pi$

Applying the Equivalence Principle:
$E[L] = E[\ddot{a}_{\overline{K+1|}}] - \Pi = \ddot{a}_x - \Pi = 0 \implies \Pi = \ddot{a}_x$

To find the variance of the future loss for this annuity:
$Var(L) = Var\left(\frac{1 - v^{K+1}}{d} - \Pi\right) = Var\left(-\frac{v^{K+1}}{d}\right) = \frac{1}{d^2} Var(v^{K+1}) = \frac{^2A_x - (A_x)^2}{d^2}$

Term (Temporary) Life Annuity:

For an $n$-year temporary life annuity-due paying \$1 at the start of each year lived, up to $n$ payments, funded by a single premium $\Pi = \ddot{a}_{x:\overline{n|}}$:
$L = \ddot{a}_{\overline{\min(K+1, n)|}} - \Pi = \frac{1 - v^{\min(K+1, n)}}{d} - \Pi$
$Var(L) = \frac{^2A_{x:\overline{n|}} - (A_{x:\overline{n|}})^2}{d^2}$

6. Common Mistakes to Avoid

1. Mixing up Continuous and Discrete: Always verify whether premiums and benefits are paid discretely (using $d$, $\ddot{a}_x$, and $K$) or continuously (using $\delta$, $\bar{a}_x$, and $T$).
2. Forgetting the Annuity Linear Relation: Remember that $\ddot{a}_{\overline{K+1|}} = \frac{1 - v^{K+1}}{d}$. This transforms annuity variance directly into insurance second moments $[^2A_x - (A_x)^2]$.
3. Premium Timing: In discrete models, ensure the number of premiums matches the timing of death. If an insured dies in year $K+1$, exactly $K+1$ annual premiums were paid at times $0, 1, \dots, K$.
4. The Double Force of Interest: $^2A_x$ is evaluated by doubling the force of interest $\delta$ (or replacing discount factor $v$ with $v^2$), while keeping mortality rates unchanged.

7. Summary and Key Takeaways

Key Points to Remember:
- Future Loss ($L$) = PV of Benefits - PV of Premiums.
- Equivalence Principle: Set $E[L] = 0$ to calculate the net premium.
- Insurance Variance: For whole life and endowment, $Var(L) = \left(1 + \frac{P}{d}\right)^2 [^2A - (A)^2]$.
- Annuity Variance: For a single premium life annuity-due, $Var(L) = \frac{^2A - (A)^2}{d^2}$.
- $L$ is a random variable driven by the policyholder's future lifetime ($T_x$ or $K_x$).

References

Fundamentals of Actuarial Mathematics Exam—October 2026 — Society of Actuaries
Exam FAM: Fundamentals of Actuarial Mathematics — Society of Actuaries

Quick check

Can you answer these now?

Open each question to check the key ideas from this chapter.

What is the definition of the future loss random variable at issue (\b(L\b) or \b(_0L\b)) for an insurance or annuity contract?

\b(L = PV(\text{Future Benefits}) - PV(\text{Future Premiums})\b).

What is the Equivalence Principle and how is it used to determine net premiums?

It sets the net premium such that the expected future loss at issue is zero: \b(E[L] = 0\b), meaning the expected PV of future benefits equals the expected PV of future premiums.

How can the future loss \b(L\b) for a fully discrete whole life policy be expressed as a linear function of the present value of benefits \b(v^{K+1}\b)?

\b(L = \left(1 + \frac{P}{d}\right)v^{K+1} - \frac{P}{d}\b)

What is the formula for the variance of the future loss random variable \b(Var(L)\b) for a fully discrete whole life policy?

\b(Var(L) = \left(1 + \frac{P}{d}\right)^2 \left[^2A_x - (A_x)^2\right]\b)

How is the future loss random variable \b(L\b) defined piecewise for a discrete \b(n\b)-year term insurance policy?

\b(L = v^{K+1} - P \ddot{a}_{\overline{K+1|}}\b) if \b(K < n\b), and \b(L = -P \ddot{a}_{\overline{n|}}\b) if \b(K \ge n\b).

What are the linear expression and variance formula for the future loss \b(L\b) of an \b(n\b)-year endowment insurance?

\b(L = \left(1 + \frac{P}{d}\right) v^{\min(K+1, n)} - \frac{P}{d}\b) and \b(Var(L) = \left(1 + \frac{P}{d}\right)^2 \left[^2A_{x:\overline{n|}} - (A_{x:\overline{n|}})^2\right]\b).

What is the future loss random variable \b(L\b) and its variance for a single premium whole life annuity-due?

\b(L = \ddot{a}_{\overline{K+1|}} - \Pi\b), and under the Equivalence Principle (\b(\Pi = \ddot{a}_x\b)), \b(Var(L) = \frac{^2A_x - (A_x)^2}{d^2}\b).

What is the variance of the future loss random variable for an \b(n\b)-year temporary life annuity-due funded by a single net premium?

\b(Var(L) = \frac{^2A_{x:\overline{n|}} - (A_{x:\overline{n|}})^2}{d^2}\b)

How is the second moment actuarial notation \b(^2A\b) evaluated with respect to the interest and mortality assumptions?

\b(^2A\b) is calculated by doubling the force of interest \b(\delta\b) (replacing \b(v\b) with \b(v^2\b) or using \b(i* = 2i + i^2\b)) while keeping mortality probabilities unchanged.

Ready to test yourself?

Turn these notes into exam-style practice. Get unlimited AI questions on this topic with instant marking and explanations.

Practise This Topic

More Exam FAM – Fundamentals of Actuarial Mathematics chapters

* The content provided by thinka is generated by AI and may not always be accurate or up-to-date. Please use it as a supplementary resource and verify with official materials.

Put These Notes into Practice

Reading the notes is a great start. Now practise with unlimited AI-generated questions and get instant feedback. 100,000+ students are already improving their grades.

Start Practising Now View Pricing

Done reading? Test yourself with AI practice questions

Practise This Topic Now