Break-Even NPV and Zero-Profit Design

Break-Even Design from a Profit Signature

A design parameter can change several annual profits at once. For example, changing a premium, expense allowance, or benefit amount may affect both early and later profits. A break-even design chooses the parameter so that the discounted value of the entire profit signature is zero. This is a statement about the signature’s net present value (NPV), not a requirement that profit be zero in every year.

The calculation depends on the profit timing convention. In this tutorial, each signature entry is an expected profit amount emerging at the end of its stated year, and all entries use the same time-zero valuation basis. Any survival or in-force weighting has already been incorporated in the stated signature. The examples use hypothetical products and specified signatures; no mortality model is needed to solve the resulting equations.

Definition: For a contract with end-of-year profit signature $q_1(\theta),\ldots,q_n(\theta)$, where $\theta$ is a design parameter, and effective annual interest rate $i$, define $v=1/(1+i)$. Its time-zero net present value is $$ \operatorname{NPV}(\theta)=\sum_{t=1}^{n}v^t q_t(\theta). $$ A break-even value is a permitted value of $\theta$ satisfying $\operatorname{NPV}(\theta)=0$. It is feasible only if it also satisfies the stated design constraints.

The exponent is determined by when the profit emerges. An end-of-year 1 profit is multiplied by $v$, an end-of-year 2 profit by $v^2$, and an end-of-year $t$ profit by $v^t$. If a problem specifies a different timing convention, the corresponding discount times must be used instead. A break-even value depends on the complete stated signature, interest basis, and timing; it cannot be found by simply adding the annual profits.

Why an Affine Signature Gives a Direct Test

Suppose every entry is an affine function of the parameter: $q_t(\theta)=a_t\theta+b_t$, where $a_t$ and $b_t$ are constants. Substituting these entries into the NPV definition gives a single affine equation.

\begin{aligned} \operatorname{NPV}(\theta) &=\sum_{t=1}^{n}v^t(a_t\theta+b_t)\\ &=\left(\sum_{t=1}^{n}v^t a_t\right)\theta +\sum_{t=1}^{n}v^t b_t\\ &=A\theta+B. \end{aligned}

This follows from distributing each discount factor over its duration profit and collecting the terms that multiply $\theta$ separately from the constant terms. Thus, when $A\ne0$, the only algebraic candidate is $\theta=-B/A$. That candidate is a valid design only if it lies in the permitted range.

Break-even test for an affine NPV: If $\operatorname{NPV}(\theta)=A\theta+B$, then $A\ne0$ gives exactly one algebraic root, $-B/A$. There is exactly one feasible break-even value if and only if this root is in the permitted range. If $A=0$ and $B\ne0$, there is no root; if $A=0$ and $B=0$, every permitted value is a root.

For a nonlinear signature, NPV need not be affine. It can have several roots, or none in the permitted range. A sign change across an interval guarantees at least one root when NPV is continuous, but does not by itself establish uniqueness. Uniqueness requires an additional argument, such as strict monotonicity over the permitted range, or direct analysis of all roots.

Worked Example: Solving for a Premium Parameter

A hypothetical three-year contract has the following end-of-year profits, in dollars, as functions of an annual premium parameter $p$: $q_1(p)=1.5p-900$, $q_2(p)=0.5p-300$, and $q_3(p)=-100$. The permitted range is $0\le p\le1000$. The effective annual interest rate is $25\%$. Find the break-even premium and determine whether it is unique and feasible.

Method. Form the NPV using the end-of-year timing, collect the coefficient and constant, then solve the resulting affine equation. Here $v=1/1.25=0.8$, so the year 1, 2, and 3 discount factors are $0.8$, $0.64$, and $0.512$, respectively.

\begin{aligned} \operatorname{NPV}(p) &=0.8(1.5p-900)+0.64(0.5p-300)+0.512(-100)\\ &=(1.2p-720)+(0.32p-192)-51.2\\ &=1.52p-963.2. \end{aligned}

Set NPV equal to zero. Then $1.52p=963.2$, so $p=963.2/1.52=12040/19$ dollars. The exact value is approximately $633.6842$, giving a break-even premium of $633.68 to the nearest cent. The coefficient $1.52$ is nonzero, so this is the unique algebraic root; it lies between 0 and 1000, so it is also the unique feasible value.

Check. At the unrounded root, the signature entries are $q_1=960/19$, $q_2=320/19$, and $q_3=-100$ dollars. Their discounted sum is $0.8(960/19)+0.64(320/19)+0.512(-100)$, which is $40.4211+10.7789-51.2=0$ dollars, up to the displayed rounding. The check uses the unrounded root; the cent-rounded premium need not give an exactly zero NPV.

Worked Example: Finding a Permitted Expense Allowance

A hypothetical policy has a design parameter $e$, an expense allowance in dollars, with $0\le e\le600$. Its end-of-year profit signature is $q_1(e)=e-500$, $q_2(e)=0.5e-300$, and $q_3(e)=0.25e+100$. The effective annual interest rate is $25\%$. Find the zero-NPV allowance, assess uniqueness and feasibility, and check the result.

Method. Discount each duration’s profit to time zero before solving. With $v=0.8$, use factors $0.8$, $0.64$, and $0.512$ for the three end-of-year entries.

\begin{aligned} \operatorname{NPV}(e) &=0.8(e-500)+0.64(0.5e-300)+0.512(0.25e+100)\\ &=(0.8e-400)+(0.32e-192)+(0.128e+51.2)\\ &=1.248e-540.8. \end{aligned}

The break-even equation is $1.248e=540.8$. Therefore $e=540.8/1.248=16900/39$ dollars, approximately $433.3333$. The allowance is $433.33 to the nearest cent. Since the coefficient $1.248$ is positive, the affine NPV has one algebraic root. The root is in the permitted interval $[0,600]$, so it is the unique feasible break-even allowance.

Check. At the unrounded value $e=16900/39$, the profits are $-200/3$, $-250/3$, and $625/3$ dollars. Their discounted sum is $0.8(-200/3)+0.64(-250/3)+0.512(625/3)$ $=-160/3-160/3+320/3=0$. As an additional range check, at $e=400$ the signature is $(-100,-100,200)$, and its NPV is $0.8(-100)+0.64(-100)+0.512(200)=-80-64+102.4=-41.6$ dollars, not zero.

Worked Example: A Nonlinear Design with One Feasible Root

A hypothetical design parameter $x$ is restricted to $0\le x\le3$. Its three-year signature, in thousands of dollars, is $q_1(x)=x^2-3$, $q_2(x)=-1.25x+1.25$, and $q_3(x)=0$. Profits emerge at the ends of the corresponding years, and the effective annual interest rate is $25\%$. Find all break-even values in the permitted range and decide whether the feasible value is unique.

Method. Because one signature entry is quadratic, first form the NPV polynomial. Then find all its roots and apply the parameter constraint.

\begin{aligned} \operatorname{NPV}(x) &=0.8(x^2-3)+0.64(-1.25x+1.25)+0.512(0)\\ &=0.8x^2-2.4-0.8x+0.8\\ &=0.8(x^2-x-2)\\ &=0.8(x-2)(x+1). \end{aligned}

The algebraic roots are $x=2$ and $x=-1$. Only $x=2$ satisfies $0\le x\le3$, so there is exactly one feasible break-even design. At $x=2$, the signature is $q_1=1$, $q_2=-1.25$, and $q_3=0$, in thousands of dollars. Discounting gives $0.8(1)+0.64(-1.25)+0.512(0)=0.8-0.8+0=0$ thousand dollars. The negative root is mathematically valid but not an allowed design.

This example illustrates why an NPV equation should not be assumed to have a unique root merely because a break-even value is requested. The polynomial must be solved and its roots checked against the design constraints.

Timing Errors and Feasibility Checks

A common exam trap is to discount the first profit by 1, then use $v$ for year 2 and $v^2$ for year 3, even though the profits emerge at the ends of years 1, 2, and 3. That calculation shifts every duration’s discount factor one year earlier. For the first worked example, the incorrect expression would be $\operatorname{NPV}(p)=(1.5p-900)+0.8(0.5p-300)+0.64(-100)=1.9p-1204$. It is not the NPV under the stated end-of-year convention. The correct expression is $0.8(1.5p-900)+0.64(0.5p-300)+0.512(-100)=1.52p-963.2$.

The incorrect expression can even produce a plausible-looking root, but that root solves a different discounted equation. In general, moving all the factors by the same power of $v$ may leave a zero unchanged; that does not make the displayed NPV correct. Write the timing factor alongside each duration before simplifying.

A reliable break-even calculation follows these steps:

State the basis.
Record the interest rate, whether profits are at year-end or another time, the units of profit, and the permitted parameter range.

Build NPV from the signature.
Multiply each profit by the discount factor for its actual emergence time. If the entries are conditional profits rather than already weighted signature entries, first determine whether the required in-force weighting is absent.

Solve and assess roots.
Set NPV to zero, solve the equation, and distinguish algebraic roots from roots that satisfy the design constraints. Check whether the equation permits one root, several roots, or none.

Substitute to verify.
Use the unrounded parameter value to recompute the discounted sum. Round the reported parameter only after the check.

The parameter’s effect on NPV is determined by its discounted effect across all durations, not by one annual profit viewed in isolation. A parameter may improve some years’ profits and worsen others. The break-even equation combines all those changes, and the feasibility check ensures the mathematical answer corresponds to a design that can actually be selected.

Key takeaway: A break-even design solves the correctly timed discounted equation $\sum_{t=1}^{n}v^tq_t(\theta)=0$. A complete answer identifies the root or roots, tests the permitted range, establishes uniqueness when possible, and verifies the result by substitution.

Check Your Understanding

For each problem, use the stated end-of-year timing and check both the equation and the design constraints.

A hypothetical two-year contract has $q_1(c)=2c-300$ and $q_2(c)=100-c$, in dollars. The effective annual interest rate is $10\%$, and $0\le c\le300$. Find the feasible break-even value and state whether it is unique. (Answer: $c=800/3$, or $266.67 to the nearest cent; unique and feasible.)
A three-year signature in dollars is $q_1(a)=a-200$, $q_2(a)=2a-500$, and $q_3(a)=100$, with $0\le a\le250$ and effective annual interest rate $25\%$. Find the break-even value and determine uniqueness. (Answer: $a=2690/13$, or $206.15 to the nearest cent; unique and feasible.)
At effective annual interest $25\%$, a hypothetical design has $q_1(k)=2k-400$, $q_2(k)=-k+150$, and $q_3(k)=50$, all in dollars. The parameter must satisfy $0\le k\le300$. Find the feasible zero-NPV value. (Answer: $k=248$; unique and feasible.)
At effective annual interest $25\%$, a design parameter satisfies $0\le x\le4$. Its signature, in thousands of dollars, is $q_1(x)=x^2-4$, $q_2(x)=-5x+8.75$, and $q_3(x)=0$. Find all feasible break-even values and state whether the solution is unique. (Answer: $x=1$ and $x=3$; both feasible, so the break-even value is not unique.)
A two-year signature is $q_1(z)=2z-100$ and $q_2(z)=-2.5z+50$, in dollars, at effective annual interest $25\%$. The permitted range is $0\le z\le100$. Determine whether a break-even value exists. (Answer: No value; the parameter coefficient cancels, leaving $\operatorname{NPV}(z)=-48$ dollars throughout the permitted range.)

What You'll Learn