Negative Reserves from Acquisition Expense Strain

Why an Early Reserve Can Be Negative

A policy can have a negative reserve early in its term even when its benefits and premiums are defined consistently. One reason is acquisition expense strain: a large expense at issue can exceed the first premium, leaving too few resources at the start of the year to support the expected year-end obligations. Later premiums and investment income may reduce or remove that deficit.

A negative reserve is a signed actuarial valuation, not a claim that an insurer holds a literal negative amount of cash for one policy. It indicates that, under the stated assumptions and timing, the present value of future premiums and other income exceeds the present value of future benefits and expenses. The size and duration of a negative reserve depend on the premium pattern, expenses, mortality, benefits, and interest.

This tutorial uses a gross premium recursion, so expenses are included explicitly rather than absorbed into a net premium. All product details and data below are hypothetical. The treatment is aligned with the SOA Exam FAM (Fundamentals of Actuarial Mathematics) syllabus.

Gross Premium Recursion and Sign Condition

Definition: Let $V_k$ be the reserve at the start of policy year $k+1$, just before that year's premium and expense are paid. Let $G_k$ be the gross premium, $e_k$ the expense, $B_{k+1}$ the death benefit payable at the end of the year of death, and $i$ the effective annual interest rate. Conditional on being alive at age $x+k$, write ${}_1p_{x+k}$ for the probability of surviving one year and ${}_1q_{x+k}$ for the probability of dying within that year. If the premium and expense are paid at the start of the year, the death benefit is paid at the end of the year, and survivors carry reserve $V_{k+1}$, then

(V_k+G_k-e_k)(1+i) = {}_1q_{x+k}B_{k+1} + {}_1p_{x+k}V_{k+1}.

The formula assumes the stated timing and that ${}_1p_{x+k}>0$. It does not directly apply if, for example, the premium or expense is paid at year-end, or if the death benefit is paid at a different time; the cash flows must then be accumulated or discounted to a common time point. Solving for the next reserve gives

V_{k+1} = \frac{(V_k+G_k-e_k)(1+i)-{}_1q_{x+k}B_{k+1}} {{}_1p_{x+k}}.

Since the survival probability in the denominator is positive, the next reserve is negative precisely when the accumulated premium and reserve, net of expense, are less than the probability-weighted death benefit:

V_{k+1}<0 \quad\Longleftrightarrow\quad (V_k+G_k-e_k)(1+i)<{}_1q_{x+k}B_{k+1}.

This condition separates two sources of strain. A large expense reduces the amount available to accumulate, while a large probability-weighted benefit increases the expected end-of-year outgo. A low premium or a negative starting reserve can also contribute.

Why the Recursion Holds

The recursion follows by balancing amounts at the end of the policy year. The reserve $V_k$, premium $G_k$, and expense $e_k$ are all at the beginning of the year. Their net amount, $V_k+G_k-e_k$, accumulates for one year at rate $i$. At year-end, death occurs with probability ${}_1q_{x+k}$, in which case the insurer pays $B_{k+1}$; otherwise the policy continues with reserve $V_{k+1}$. The probability-weighted year-end amount is therefore the right-hand side of the recursion. Equating the accumulated beginning amount to that expected year-end amount gives the formula.

This is an actuarial balance on the survival-conditioned policy, not a claim that the same cash flow occurs on every path. On a death path, the benefit is paid; on a survival path, the next reserve remains. The probabilities combine these outcomes in the reserve equation.

Worked Example: Measuring First-Year Acquisition Strain

A hypothetical one-year term policy has an initial reserve $V_0=0$, a start-of-year gross premium of $100, and a start-of-year acquisition expense of $500. The effective annual interest rate is $i=10\%$. The insured has a one-year death probability of $0.10$, and the end-of-year death benefit is $1,000. Calculate $V_1$, then find the largest issue expense that would make $V_1$ nonnegative with all other assumptions unchanged.

Method. Apply the gross premium recursion, then set the numerator equal to zero to find the expense boundary. This avoids treating the expense as though it were paid at year-end.

The survival probability is ${}_1p_x=1-0.10=0.90$. The amount remaining at the start of the year after premium and expense is

$$ V_0+G_0-e_0=0+100-500=-400. $$

After accumulating at 10%, this becomes $(-400)(1.10)=-440$. The probability-weighted death benefit is $(0.10)(1{,}000)=100$. Thus

V_1 = \frac{-440-100}{0.90} = \frac{-540}{0.90} = -600.

Check the balance: the expected year-end amount is $(0.10)(1{,}000)+(0.90)(-600)=100-540=-440$, which equals the accumulated amount. To find the largest expense $e_0$ for which $V_1\geq 0$, use the zero-reserve boundary:

(100-e_0)(1.10)=100, \qquad e_0=100-\frac{100}{1.10}=9.0909\ldots.

The boundary expense is $9.09 to the nearest cent. Any larger expense makes the first-year reserve negative under these assumptions. The stated $500 expense is well above that boundary, producing a reserve of $-$600$.

How Later Premiums and Interest Change the Reserve

A negative reserve does not by itself determine the next reserve. The next year's premium and expense enter before interest is credited, and the next year's mortality and benefit determine the expected year-end claim amount. A sufficiently large later premium can offset an existing deficit; investment growth then applies to the net amount after the premium and expense.

Worked Example: Finding a Later Premium That Restores a Zero Reserve

At the start of a later policy year, suppose a hypothetical policy has reserve $V_k=-$600$. The effective annual interest rate is $i=10\%$, the one-year death probability is $0.05$, and the end-of-year death benefit is $1,000. A $50 expense is paid at the start of the year. Find the gross premium $G_k$ that makes the next reserve $V_{k+1}=0$.

Method. Substitute the target reserve into the gross recursion and solve for the premium. This accounts for interest on the net amount after both the existing reserve and the expense are included.

The survival probability is ${}_1p_{x+k}=0.95$, and the probability-weighted death benefit is $(0.05)(1{,}000)=$50$. The recursion becomes

$$ (-600+G_k-50)(1.10)=50+(0.95)(0). $$

Divide both sides by $1.10$:

G_k-650=\frac{50}{1.10}=45.4545\ldots, \qquad G_k=695.4545\ldots.

The required premium is $695.45 to the nearest cent. Before interest, reserve plus premium less expense is $-600+695.4545\ldots-50=$45.4545\ldots$; after 10% interest it is $50. That equals the expected death benefit, while the surviving policies carry zero reserve. The balance checks. Rounding the premium before calculating would cause a tiny discrepancy, so the exact unrounded value is used in the check.

Level Premiums Can Coexist with Negative Early Reserves

A level gross premium may be chosen to satisfy a terminal reserve condition while still producing a negative reserve at an earlier duration. The first premium need not cover a large issue expense by itself: the level premium is evaluated across the full contract, including later premiums, benefits, expenses, and investment income.

Worked Example: A Two-Year Policy with a Negative First-Year Reserve

A hypothetical two-year term policy charges the same gross premium $G$ at the start of each year. It pays a $1,100 death benefit at the end of either year of death. The one-year death probability is $0.10$ in each year, so the survival probability is $0.90$. The effective annual interest rate is $10\%$. A $200 acquisition expense is paid at issue, with no expense at the start of year 2. The issue reserve is zero and the terminal reserve is $V_2=0$. Find $G$ and the first-year reserve $V_1$.

Method. First use the terminal-year recursion to express $V_1$ in terms of $G$; then substitute into the first-year recursion. This enforces the terminal condition rather than assuming that the first-year reserve must be nonnegative.

For year 2, the probability-weighted death benefit is $(0.10)(1{,}100)=$110$. Since $V_2=0$ and there is no year-2 expense,

(V_1+G)(1.10)=110+0.90(0)=110, \qquad V_1=100-G.

For year 1, the $200 issue expense is paid at the start of the year. The recursion is

$$ (G-200)(1.10)=110+0.90V_1. $$

Substitute $V_1=100-G$:

$$ 1.10G-220 = 110+0.90(100-G) = 200-0.90G. $$

Therefore $2G=420$, so $G=$210$. The first-year reserve is

$$ V_1=100-210=-$110. $$

Check year 1: accumulated net funds are $(210-200)(1.10)=$11$, and expected year-end outgo and reserve are $110+0.90(-110)=11$. Check year 2: $(-110+210)(1.10)=$110$, equal to the probability-weighted death benefit of $110. Both recursions balance, and the terminal reserve is zero. The level premium supports the two-year contract even though the acquisition expense creates a negative first-year reserve.

Common Timing and Sign Traps

Applying interest before the start-of-year expense. With the timing used here, the accumulated amount is $(V_k+G_k-e_k)(1+i)$, not $(V_k+G_k)(1+i)-e_k$. In the first worked example, the incorrect expression gives $(100)(1.10)-500=-390$ before the death-benefit adjustment. The correct accumulated net amount is $(100-500)(1.10)=-440$; the $50 difference is the interest incorrectly credited to the $500 expense.
Dropping the survival probability. The numerator in the solved recurrence is divided by ${}_1p_{x+k}$. In the first example, dividing $-540$ by $1$ would give $-540$, but the correct reserve is $-540/0.90=-600$.
Treating a negative reserve as an unsigned amount. The sign is part of the valuation. Replacing $-600$ with $+600$ in a later recursion reverses the effect of the prior reserve.
Assuming terminal balance implies nonnegative earlier reserves. A zero reserve at expiry is a boundary condition; it does not require intermediate reserves to be positive, as the two-year example demonstrates.

A practical check is to calculate both sides of the recursion after solving. If the accumulated beginning amount does not equal the probability-weighted year-end benefit and continuation reserve, recheck the timing, signs, and survival probability.

Key takeaway: With start-of-year premiums and expenses, a reserve becomes negative when the net beginning resources accumulated with interest are less than the probability-weighted year-end benefit. Later premiums and investment growth can reduce that deficit, but the result must be evaluated using the contract's stated cash-flow timing and mortality assumptions.

Check Your Understanding

For each problem, use the stated start-of-year premium and expense timing and the gross premium recursion.

A one-year policy has $V_0=0$, $G_0=$120$, $e_0=$300$, $i=5\%$, death probability $0.02$, and an end-of-year death benefit of $5,000. Calculate $V_1$. (Answer: $-$294.90$)
A one-year policy has $V_0=0$, $G_0=$150$, $i=8\%$, death probability $0.04$, and a $2,000 death benefit. What is the largest start-of-year expense that leaves $V_1\geq0$? (Answer: $75.93)
A policy has reserve $V_k=-$250$, $i=4\%$, death probability $0.02$, a $1,000 death benefit, and a $20 start-of-year expense. Find the premium that makes $V_{k+1}=0$. (Answer: $289.23)
A two-year term policy has $V_0=0$, $V_2=0$, a $1,000 death benefit at the end of either year, death probability $0.10$ in each year, $i=5\%$, a level start-of-year gross premium $G$, a $100 issue expense, and no second-year expense. Find $G$ and $V_1$. (Answer: $G=$149.08$, $V_1=-$53.85$, to the nearest cent)

What You'll Learn