Sales Volatility and Operating Risk

Operating Risk Begins With Uncertain Sales

Operating risk is the uncertainty in a firm’s operating income arising from its business operations. In this tutorial, the focus is the way uncertain sales interact with the firm’s contribution margin and fixed operating costs. The central point is that sales volatility alone does not determine operating-income risk: the cost structure and the firm’s position relative to break-even matter too.

This analysis is about operating income before financing charges. It therefore isolates operating risk rather than measuring how borrowing changes shareholders’ earnings. The distinction matters because two firms facing the same sales uncertainty can have different operating-income variability if their margins or fixed costs differ.

Definition: Let $S$ be sales revenue over a stated period. Assume variable operating costs equal a constant proportion of sales, so the contribution margin ratio $m$ is the proportion of each sales dollar remaining after variable operating costs. Let $F$ be fixed operating costs for that period. Operating income before financing charges is $\mathrm{EBIT}=mS-F$. The model assumes $m>0$, $F$ is fixed over the relevant range, and the same period and currency units are used for $S$ and $F$.

For example, if $m=0.35$, then each dollar of sales contributes 35 cents toward fixed operating costs and operating income. The other 65 cents represents variable operating costs under this model. Fixed costs do not vary with sales within the relevant range, although actual costs can become step-fixed or change outside that range.

How Sales Uncertainty Translates Into Income Uncertainty

Suppose sales have expected value $E[S]=\mu_S$ and variance $\operatorname{Var}(S)=\sigma_S^2$. If $m$ and $F$ are constants, then expected operating income and its variance are

E[\mathrm{EBIT}]=m\mu_S-F,\qquad \operatorname{Var}(\mathrm{EBIT})=m^2\sigma_S^2,\qquad \operatorname{SD}(\mathrm{EBIT})=m\sigma_S.

The standard deviation measures the typical scale of variation around expected operating income. It is not the probability of a loss, and it does not describe the entire shape of the income distribution. A loss probability additionally depends on the sales distribution and on how far the break-even point lies into that distribution.

Derivation: Since $F$ is constant, subtracting it shifts every possible value of operating income but does not change its variance. Starting from the definition of variance,

\begin{aligned} \operatorname{Var}(\mathrm{EBIT}) &=E\left[(mS-F-E[mS-F])^2\right]\\ &=E\left[(mS-F-(mE[S]-F))^2\right]\\ &=E\left[m^2(S-E[S])^2\right]\\ &=m^2\operatorname{Var}(S). \end{aligned}

Because $m>0$, taking the square root gives $\operatorname{SD}(\mathrm{EBIT})=m\sigma_S$. This also explains an important cost-structure effect: for the same sales standard deviation, a larger contribution margin ratio produces a larger absolute operating-income standard deviation. Fixed costs affect the income level and loss threshold, but, when deterministic, they do not affect this standard deviation.

Break-even sales are the sales level at which operating income equals zero. Solving $mS-F=0$ gives $S_{\mathrm{BE}}=F/m$. If sales are below this level, operating income is negative; if sales are above it, operating income is positive. At exactly break-even, income is zero, not a loss.

Result: If sales are normally distributed as $S\sim N(\mu_S,\sigma_S^2)$, with $\sigma_S>0$, then operating income is also normally distributed. The probability of an operating loss is $P(\mathrm{EBIT}<0)=P(S<S_{\mathrm{BE}})=\Phi\left((S_{\mathrm{BE}}-\mu_S)/\sigma_S\right)$, where $\Phi$ is the standard normal cumulative distribution function. This probability formula requires the normal-sales assumption; the mean and variance alone do not determine a loss probability for an arbitrary distribution.

The standardized distance from expected sales to break-even is $(\mu_S-S_{\mathrm{BE}})/\sigma_S$. A larger positive distance means break-even is farther into the lower sales tail under a normal model, so the chance of an operating loss is smaller. This standardized measure complements the dollar distance: a gap of $100,000 represents a different amount of protection when sales standard deviation is $20,000 than when it is $200,000.

Worked Examples

Worked Example: Normal Sales and the Chance of a Loss

A hypothetical equipment supplier models quarterly sales $S$ in thousands of dollars as normal with mean 1,200 and variance 40,000 in thousands of dollars squared. Its contribution margin ratio is $m=0.35$, and quarterly fixed operating costs are 315 thousand dollars. Find break-even sales, expected operating income, the standard deviation of operating income, and the probability of an operating loss.

Method. Apply the linear operating-income model, then use the normal distribution to calculate the loss probability. The sales standard deviation is the square root of the stated variance: $\sigma_S=\sqrt{40{,}000}=200$ thousand dollars.

Break-even sales. Set operating income equal to zero and solve for sales: $S_{\mathrm{BE}}=F/m=315/0.35=900$ thousand dollars. At sales of 900, contribution is $0.35(900)=315$, exactly equal to fixed costs.

Expected operating income. Since $E[S]=1{,}200$, $E[\mathrm{EBIT}]=0.35(1{,}200)-315=420-315=105$ thousand dollars.

Operating-income standard deviation. Scale the sales standard deviation by the positive margin ratio: $\operatorname{SD}(\mathrm{EBIT})=0.35(200)=70$ thousand dollars.

Probability of a loss. A loss occurs when sales are below 900. Standardizing the break-even value without rounding the z-score early gives $z=(900-1{,}200)/200=-1.5$. Thus $P(\mathrm{EBIT}<0)=\Phi(-1.5)=0.0668$, or 6.68% to four decimal places as a probability.

Check and common mistake avoided. At mean sales, the model gives operating income of 105 thousand dollars, consistent with the expected-income calculation. The loss probability is not found by dividing expected income by its standard deviation; it is the lower-tail probability at break-even. The inputs and outputs are all in thousands of dollars, except the dimensionless probability.

Worked Example: Comparing Two Cost Structures

Two hypothetical service businesses face the same annual sales distribution, $S\sim N(1{,}000,10{,}000)$, with sales and variance measured in thousands of dollars and thousands of dollars squared, respectively. Firm A has $m=0.40$ and fixed costs of 300 thousand dollars. Firm B has $m=0.25$ and fixed costs of 150 thousand dollars. Compare their expected operating incomes, operating-income standard deviations, break-even sales, and probabilities of loss.

Shared sales quantities. Both firms have expected sales of 1,000 thousand dollars and sales standard deviation $\sqrt{10{,}000}=100$ thousand dollars.

Firm A. Expected operating income is $0.40(1{,}000)-300=100$ thousand dollars. Its operating-income standard deviation is $0.40(100)=40$ thousand dollars. Break-even sales are $300/0.40=750$ thousand dollars. Therefore the standardized break-even value is $(750-1{,}000)/100=-2.5$, giving a loss probability of $\Phi(-2.5)=0.0062$ to four decimal places.

Firm B. Expected operating income is $0.25(1{,}000)-150=100$ thousand dollars. Its operating-income standard deviation is $0.25(100)=25$ thousand dollars. Break-even sales are $150/0.25=600$ thousand dollars. The standardized break-even value is $(600-1{,}000)/100=-4$, giving a loss probability of $\Phi(-4)\approx0.0000317$, or 0.0000 to four decimal places.

Comparison and check. Both firms have the same expected operating income, but Firm A has greater operating-income variability and a higher probability of a loss under the stated model. Firm B’s loss probability is small but not exactly zero; rounding it to four decimal places produces 0.0000. This comparison shows why equal expected income does not imply equal operating risk. The calculations depend on the normal model and on the stated cost structures remaining applicable across the sales range.

Worked Example: A Discrete Sales Model and an Exact Loss Probability

A hypothetical parts distributor has quarterly sales, in thousands of dollars, of 700 with probability 0.25, 1,000 with probability 0.50, and 1,300 with probability 0.25. Its contribution margin ratio is 0.40 and fixed quarterly operating costs are 320 thousand dollars. Calculate expected sales, sales variance, expected operating income, operating-income standard deviation, and the probability of an operating loss.

Expected sales. Weight each sales value by its probability: $E[S]=0.25(700)+0.50(1{,}000)+0.25(1{,}300)=175+500+325=1{,}000$ thousand dollars.

Sales variance. Deviations from the mean are $-300$, $0$, and $300$. Therefore, $\operatorname{Var}(S)=0.25(300^2)+0.50(0^2)+0.25(300^2)=22{,}500+0+22{,}500=45{,}000$ in thousands of dollars squared. The sales standard deviation is $\sqrt{45{,}000}\approx212.13$ thousand dollars.

Operating income in each outcome. Using $\mathrm{EBIT}=0.40S-320$, the three outcomes are $0.40(700)-320=-40$, $0.40(1{,}000)-320=80$, and $0.40(1{,}300)-320=200$, all in thousands of dollars. Expected operating income is $0.25(-40)+0.50(80)+0.25(200)=-10+40+50=80$ thousand dollars.

Operating-income standard deviation. By the variance result, it is $0.40\sqrt{45{,}000}\approx84.85$ thousand dollars. As a check, the variance calculated directly from the three operating-income outcomes is $0.25((-40-80)^2)+0.50((80-80)^2)+0.25((200-80)^2)=3{,}600+0+3{,}600=7{,}200$; its square root is approximately 84.85.

Probability of a loss. Break-even sales are $320/0.40=800$ thousand dollars. Only the 700-thousand-dollar outcome is below break-even, so the exact loss probability is 0.25. The outcome at 1,000 produces positive income; no outcome is exactly at break-even.

Method and check. For a discrete distribution, sum the probabilities of the outcomes that cause a loss rather than applying a normal-tail formula. The weighted outcome probabilities total $0.25+0.50+0.25=1$, and the direct variance check agrees with the scaled-standard-deviation result. Using a normal approximation here would give a different answer and is not needed.

Interpreting Operating Risk Carefully

The model separates three related quantities. Expected operating income describes the average level under the sales distribution. Operating-income standard deviation describes the absolute spread when the contribution margin ratio and fixed costs are deterministic. The probability of an operating loss depends on the break-even threshold and the distribution of sales. These measures answer different questions, so one should not be substituted for another.

For normally distributed sales, moving break-even farther below expected sales relative to the sales standard deviation reduces the loss probability. Increasing fixed costs while holding the margin ratio constant raises break-even sales and lowers expected operating income, but it does not change the operating-income standard deviation in this model. Increasing the margin ratio raises the standard deviation when sales are uncertain, while it also changes break-even sales to $F/m$. The net effect on loss probability must be assessed using both the new threshold and the sales distribution.

The assumptions limit the conclusions. If variable costs are not proportional to sales, if fixed costs change in steps, or if the contribution margin ratio itself is uncertain, then $\mathrm{EBIT}=mS-F$ with constant $m$ and $F$ may not represent the business. A normal model also assigns some probability to negative sales; it is a useful approximation only when that feature is reasonable for the application. For a non-normal or discrete sales distribution, calculate loss probability from the actual distribution rather than inferring it from mean and variance alone.

Common exam trap: If sales standard deviation is $\sigma_S$, do not report $m^2\sigma_S$ as the operating-income standard deviation. The variance is multiplied by $m^2$, but taking its square root gives $m\sigma_S$. Also distinguish a loss, defined by $\mathrm{EBIT}<0$, from break-even, where $\mathrm{EBIT}=0$. If the sales distribution has probability mass at break-even, that mass is not part of the loss probability.

A useful diagnostic sequence is to state the operating-income model and units, calculate break-even and expected income, translate sales variability into income variability, and then calculate the loss probability using the specified distribution. This produces a risk assessment grounded in both the cost structure and the uncertainty in sales, rather than in a single leverage measure.

Check Your Understanding

For each problem, keep sales and fixed costs in the same units. Probabilities are rounded to four decimal places where needed.

A hypothetical retailer has $S\sim N(900,14{,}400)$, with sales and variance measured in thousands of dollars and thousands of dollars squared. Its margin ratio is 0.30 and fixed costs are 210 thousand dollars. Find break-even sales, expected operating income, operating-income standard deviation, and probability of a loss. (Answer: Break-even sales = 700 thousand dollars; expected operating income = 60 thousand dollars; standard deviation = 36 thousand dollars; loss probability = 0.0478.)
A hypothetical clinic-services firm has normal annual sales with mean 800 thousand dollars and standard deviation 80 thousand dollars. If $m=0.50$ and fixed operating costs are 360 thousand dollars, find expected operating income, its standard deviation, and the probability of an operating loss. (Answer: Expected operating income = 40 thousand dollars; standard deviation = 40 thousand dollars; loss probability = 0.1587.)
A hypothetical distributor’s sales are 500, 800, or 1,100 thousand dollars with respective probabilities 0.20, 0.50, and 0.30. For $m=0.40$ and fixed costs of 300 thousand dollars, find expected sales, sales standard deviation, expected operating income, operating-income standard deviation, and loss probability. (Answer: Expected sales = 830 thousand dollars; sales standard deviation = 210 thousand dollars; expected operating income = 32 thousand dollars; operating-income standard deviation = 84 thousand dollars; loss probability = 0.2000.)
Two hypothetical businesses have the same annual sales, $S\sim N(2{,}000,62{,}500)$, in thousands of dollars and thousands of dollars squared. Firm A has $m=0.30$, $F=450$; Firm B has $m=0.50$, $F=850$, with fixed costs in thousands of dollars. For each firm find expected operating income, operating-income standard deviation, break-even sales, and loss probability; identify which has the greater loss probability. (Answer: Both expected incomes = 150 thousand dollars; A standard deviation = 75 thousand dollars, break-even sales = 1,500 thousand dollars, loss probability = 0.0228; B standard deviation = 125 thousand dollars, break-even sales = 1,700 thousand dollars, loss probability = 0.1151; Firm B has the greater loss probability.)
A hypothetical operator has $S\sim N(1{,}200,22{,}500)$, with sales and variance measured in thousands of dollars and thousands of dollars squared, and $m=0.25$. Compare fixed costs of 240 thousand dollars and 300 thousand dollars. For each case calculate break-even sales, expected operating income, operating-income standard deviation, and loss probability. (Answer: At $F=240$, break-even sales = 960 thousand dollars, expected income = 60 thousand dollars, standard deviation = 37.5 thousand dollars, and loss probability = 0.0548; at $F=300$, break-even sales = 1,200 thousand dollars, expected income = 0, standard deviation = 37.5 thousand dollars, and loss probability = 0.5000.)

What You'll Learn