Actuarial Mathematics for Modelling · Gross random future loss

Future Loss Random Variable Basics: How to Define L_t

Updated 11 October 2026 · Fact-checked

The future loss random variable at time t, written L_t, is the present value at time t of future benefits and expenses minus the present value at time t of future premiums. You define it by writing each cash flow, its timing and its payment condition, then discounting each one to time t.

Understand Future Loss Random Variable Basics

An insurer collects premiums and pays benefits. Both depend on how long the policyholder lives. So the profit or loss on a policy is uncertain. The future loss random variable captures this uncertainty at a chosen time.

The definition is simple: future loss = PV of future outgo − PV of future income. Outgo means benefits and expenses. Income means premiums. A positive value is a loss for the insurer. A negative value is a profit.

The loss is a random variable because the timing of payments depends on future lifetime. For a person aged x, the future lifetime is T_x (continuous) or K_x (curtate, whole years). Every present value is a function of T_x or K_x.

L_0 is the loss at policy start. L_t is the loss at time t, given the policy is still in force. Always say "given alive at time t" when t > 0. The remaining lifetime from t is then T_(x+t).

The word gross means you include expenses and use the gross premium. A net loss ignores expenses and uses the net premium. The setup is the same. Only the cash flows change.

Once L is written down, you can find its mean, variance and percentiles. Those are covered in later topics. This topic is about writing L correctly.

Key rules to remember

Future loss at time t: L_t = PV of future benefits and expenses − PV of future premiums; All present values are taken at time t, on the stated basis, for a policy still in force at t.
Whole life assurance, sum assured S, death benefit paid at end of year of death, annual premium P in advance, initial expense e0, renewal expense e per year: L_0 = S v^(K_x + 1) + e0 + e ä_(K_x + 1) − P ä_(K_x + 1); Here K_x is curtate future lifetime. Premiums are paid at times 0, 1, ..., K_x, which is K_x + 1 payments. Renewal expenses are assumed to be paid with each premium, including the first, unless e0 is stated to replace it.
Whole life assurance, benefit paid at moment of death: L_0 = S v^(T_x) + expenses − P ä_(⌈T_x⌉); Premiums are annual in advance, so the number of payments is the number of policy years started, ⌈T_x⌉ (T_x rounded up). If premiums are payable continuously, use ā_(T_x) instead.
n-year term assurance: L_0 = S v^(K_x + 1) − P ä_(K_x + 1) if K_x < n; L_0 = − P ä_n if K_x ≥ n; Benefit is paid only on death within the term. Premiums stop at death or at time n.
n-year endowment assurance: L_0 = S v^(K_x + 1) − P ä_(K_x + 1) if K_x < n; L_0 = S v^n − P ä_n if K_x ≥ n; The sum assured is paid either on death in the term or on survival to n.
Immediate life annuity-due of 1 pa, premium paid as a single sum π: L_0 = ä_(K_x + 1) − π; A single premium has no premium annuity. It is just π at time 0.
Loss at later time t: L_t uses K_(x+t) or T_(x+t) in place of K_x or T_x, and discounts to time t; Condition: the policy is in force at t, so the life has survived to age x + t.

How to solve Future Loss Random Variable Basics questions

Use this method for any question that asks you to define or set up the future loss random variable.

1Identify the contract: type of benefit, term, premium pattern (single, annual, limited pay, continuous) and when the death benefit is paid.
2Define the lifetime variable. Use K_x for payments at year ends and annual premiums. Use T_x for benefits paid at the moment of death. State the age at the valuation time.
3List the outgo cash flows: benefit amount, expenses (initial and renewal), and when each is paid. Note any condition such as death within the term or survival to the end.
4List the income cash flows: premium amount, frequency, and the number of premiums paid under each outcome of the lifetime.
5Discount each cash flow to the valuation time t using v = 1/(1 + i), or e^(−δ) for force of interest. Write the benefit as S v^(K_x + 1) or S v^(T_x).
6Combine as L_t = PV outgo − PV income. Split into cases if the formula changes at the end of the term.
7State the valuation time, the survival condition and the basis (interest rate, mortality, expenses) clearly in your answer.
8Check the sign and the premium count: n premiums paid means ä_n, and death in year k+1 means k+1 premiums if premiums are in advance.

Quickest way: Timeline check for writing L

When to use it: Use this when you have a few minutes and a standard contract in a written or multiple-choice question.

Draw a line from time 0. Mark the term n.
Write the death benefit at its payment time (K_x + 1 or T_x) and the survival benefit at n, if any.
Mark each premium date. For annual in advance, ask: how many dates fall at or before the payment time of the benefit?
Write L as one expression for death in the term and one for survival, if the contract has both.
Add expenses on the same timeline, then stop. Do not simplify until the question asks you to.

Common mistakes in Future Loss Random Variable Basics

Writing the loss as premiums minus benefits.
Students think of profit instead of loss.
Fix: Remember loss = outgo − income. Write "PV benefits + PV expenses − PV premiums" every time.
Using the wrong number of premiums, such as ä_(K_x) instead of ä_(K_x + 1).
Premiums in advance are paid at the start of each year, including the year of death.
Fix: If death occurs in year K_x + 1, premiums are paid at times 0 to K_x. That is K_x + 1 payments.
Discounting the benefit with v^(K_x) instead of v^(K_x + 1).
Students forget that the end-of-year benefit is paid at the end of the year of death.
Fix: Use v^(K_x + 1) for end-of-year benefits and v^(T_x) for benefits paid at death.
Ignoring the survival case in term or endowment contracts.
Students write only the death formula.
Fix: Give L for K_x < n and for K_x ≥ n. Term assurance has no benefit on survival but premiums continue to time n.
Taking L_t at time t but still discounting to time 0.
Students copy the L_0 formula and change only the lifetime.
Fix: At time t, use the remaining lifetime from age x + t, count only future premiums, and discount to time t.
Leaving out expenses in a gross loss, or treating initial expense as paid at the wrong time.
Students recall the net version.
Fix: Add initial expenses at time 0 and renewal expenses at the stated dates. Use the gross premium, not the net premium.

Worked examples

Example 1

A life aged 40 buys a 20-year term assurance with sum assured ₹10,00,000, paid at the end of the year of death. Level annual premium P is payable in advance for the term. Initial expenses are ₹2,000, and renewal expenses are ₹300 at the start of each year from year 2 to year 20. Write the gross future loss random variable L_0 at issue in terms of K_40 and annuity functions.

Show the solution

Let K = K_40, the curtate future lifetime. Death in the term means K ≤ 19, since K < 20.
Benefit outgo if K < 20: 10,00,000 v^(K + 1).
Premiums are paid at times 0, 1, ..., K if the life dies in year K + 1, so the premium annuity is P ä_(K + 1).
Initial expense is 2,000 at time 0. Renewal expenses of 300 are paid at times 1, 2, ..., K, which is K payments. Their present value is 300 (ä_(K + 1) − 1).
So for K < 20: L_0 = 10,00,000 v^(K + 1) + 2,000 + 300 (ä_(K + 1) − 1) − P ä_(K + 1).
If the life survives the term, K ≥ 20, there is no benefit. Premiums are paid at times 0 to 19, and renewal expenses at times 1 to 19. So L_0 = 2,000 + 300 (ä_20 − 1) − P ä_20.

Answer: L_0 = 10,00,000 v^(K + 1) + 2,000 + 300 (ä_(K + 1) − 1) − P ä_(K + 1) for K = 0, 1, ..., 19. L_0 = 2,000 + 300 (ä_20 − 1) − P ä_20 for K ≥ 20.

Example 2

A whole life assurance is issued to a life aged 50. The sum assured of ₹5,00,000 is payable at the end of the year of death. A single premium of ₹1,20,000 is paid at issue. Ignore expenses. The effective annual interest rate is 5%. Find the value of L_0 if the life dies in the 3rd policy year, to the nearest rupee.

Show the solution

Death in the 3rd policy year means K = 2, so the benefit is paid at time K + 1 = 3.
L_0 = 5,00,000 v^3 − 1,20,000, because the single premium is paid at time 0 and has present value ₹1,20,000.
v^3 = 1 ÷ 1.05^3. Now 1.05^3 = 1.157625.
So v^3 = 0.863838 (to 6 decimals).
5,00,000 × 0.863838 = 4,31,919.
L_0 = 4,31,919 − 1,20,000 = 3,11,919.

Answer: L_0 ≈ ₹3,11,919, a loss to the insurer.

Exam tips

Define K_x or T_x in your first line. Marks are often given for correct definitions.
Always write the loss in cases when the contract has a term. Examiners expect the death case and the survival case.
State the time of valuation and 'given alive at time t' when t > 0.
In MCQs, check the number of premiums first. Wrong counts are the commonest source of wrong options.
In computer-based questions, show the formula in notation before the code or spreadsheet formula, then the numerical result.

Practice questions from Gross random future loss

Future Loss Random Variable Basics: frequently asked questions

What is the future loss random variable in CM1?

It is the present value at a given time of future benefits and expenses minus the present value of future premiums, for a policy still in force. It is random because payment times depend on future lifetime.

What is the difference between gross and net future loss?

Gross future loss includes expenses and uses the gross premium. Net future loss ignores expenses and uses the net premium. The method of setting up the variable is the same.

When do I use K_x and when T_x?

Use K_x when benefits are paid at the end of the year of death and premiums are annual. Use T_x when the benefit is paid at the moment of death. Premiums paid annually in advance with a T_x benefit need the number of years started, written ⌈T_x⌉.

How is L_t different from L_0?

L_t is the loss at time t, given the policy is still in force. It uses the remaining lifetime from age x + t, only future cash flows, and discounts to time t instead of time 0.