Gross vs Net Revenue: Why Margins and EV/Sales Can Mislead

A company reporting a 100% gross margin is not necessarily keeping more money than one reporting 20%. In the example below, each earns 20 million before other operating expenses and 12 million afterwards. Yet their enterprise-value-to-revenue multiples are 12 times and 2.4 times. Ranking them by either headline would confuse the revenue denominator with the economics.

The useful distinction is between gross revenue earned as a principal and fee revenue earned as an agent. First establish what the company has promised to deliver. Then reconcile the transaction amount, the income retained and the costs needed to earn it. Only after that can revenue growth or a valuation multiple support a comparison.

Who controls the promised product before the customer receives it?

Under the IFRS 15 framework, the assessment concerns a specified good or service. A principal controls it before transfer to the customer; an agent arranges for another party to provide it. The principal recognises the relevant gross consideration as revenue, while the agent recognises its fee or commission. A company cannot choose whichever presentation makes its growth or margins look better.

Responsibility for fulfilling the promise, inventory risk and pricing discretion can help establish control. They are evidence, not a three-box vote. Sending the invoice, collecting the customer's money or describing the business as a platform does not settle the question by itself. Different promises within the same company can require different conclusions.

For an investor, the starting documents are the revenue policy, the description of performance obligations and significant judgements. If they do not establish the relevant facts, a ratio cannot fill the gap. A comparison may need to remain inconclusive rather than force every business onto a presumed net basis.

Follow 100 of customer spending through two different businesses

Consider two hypothetical businesses whose different contract facts support their respective principal and agent classifications. All amounts are millions of the same currency and cover the same reporting period. Each handles 100 of end-customer transaction value; 80 is a purchase cost or an amount passed to the supplier. Each retains 20 before other operating expenses of 8.

To isolate the denominator, the agent has no additional cost of sales in this model. Taxes, refunds, financing, timing differences and other costs are excluded. These are not two optional accounting treatments for one unchanged contract, and the assumptions do not establish equal business risk.

The principal's gross margin is 20/100, or 20%. The agent's is 20/20, or 100%. Both have gross profit of 20 and operating profit of 12. Their operating margins are also different—12% and 60%—because revenue remains the denominator.

The agent's 100% figure is a consequence of the deliberately simplified cost structure, not a rule for agency businesses. An actual agent may incur direct fulfilment or other costs. Equally, the 20 retained here is an analytical subtotal: it is neither net income nor free cash flow and should not be relabelled as a universal accounting measure called “net revenue”.

A lower revenue multiple need not mean a cheaper business

Give each business an enterprise value of 240 at the same valuation date. This is an input for the comparison, not a finding that the businesses deserve the same value. Dividing by reported revenue produces 240/100 = 2.4 times for the principal and 240/20 = 12 times for the agent. The fivefold gap does not come from a difference in the assumed operating profit.

Dividing by operating profit produces 240/12 = 20 times in each case. That removes this particular revenue-base mismatch, but does not finish the valuation. Capital expenditure, working capital, reinvestment needs, growth and contractual exposure can justify different values even when current operating profit matches. Nor is operating profit interchangeable with EBITDA, net income or distributable cash.

The numerator needs its own consistent scope. Reconciling enterprise value with market capitalisation and net debt avoids mixing an operating valuation with an equity-only measure. Here the separate question is whether the revenue beneath that numerator describes the same economic activity. Use the same reporting window and consolidation perimeter, and identify historical versus forecast denominators explicitly.

Separate a presentation difference from a shrinking fee

Now keep the agency basis unchanged. Let G be the consistently defined transaction value, t the fee revenue divided by G, and O the other operating expenses. With the same simplified cost assumptions, revenue is G × t and operating profit is G × t − O.

Suppose transaction value rises from 100 to 120 while the fee ratio falls from 20% to 15%. Revenue falls from 20 to 18. With expenses still 8, operating profit falls from 12 to 10: a decline of one-sixth, approximately 16.666667%, despite 20% growth in transaction value. There is no gross-to-net accounting switch in this case. The business is retaining less income from each unit of activity.

To preserve the original operating profit of 12, the required fee ratio is t = (12 + O) / G. At transaction value of 120 and expenses of 8, that is exactly one-sixth, approximately 16.666667%. If expenses instead rise to 10, the threshold becomes 22/120, or 11/60, approximately 18.333333%. The table uses the exact fractions for its money amounts; the displayed percentages are rounded.

A ratio above the relevant threshold improves profit beyond 12; equality merely preserves it. Avoid confusing that hurdle with avoiding a loss: at G = 120 and O = 8, zero operating profit requires only 8/120, or approximately 6.666667%. Neither percentage is an industry benchmark. Both depend on the stated transaction value and cost structure.

Business mix can halve revenue without halving profit

A third, separate example combines genuinely different principal and agency arrangements. Hold total transaction value at 100, the amount retained before other expenses at 20% in both activities, and total other expenses at 8. Let a be the principal activity's share of transaction value.

Reported revenue is 100a + 20(1 − a), or 20 + 80a. Reported cost of sales is 80a. Gross profit is therefore 20 and operating profit 12 throughout this simplified mix experiment.

If the principal share falls from 75% to 25%, reported revenue falls from 80 to 40. Gross margin rises from 25% to 50%, and operating margin from 15% to 30%. Revenue has halved and the margins have doubled, but neither gross profit nor operating profit has improved. This is a change in the mix of different activities, not permission to change the accounting for an unchanged contract.

The real-world test is whether transaction value, retention and cost assumptions actually hold. A shift in contracts may change returns, inventory, credit exposure and funding needs. Without a sufficient segment or transaction bridge, do not manufacture a “like-for-like” growth number by subtracting an estimated supplier share from all reported revenue.

Revenue retained is not cash available to shareholders

Receiving 100 into a bank account while owing 80 to a supplier does not make all 100 available for distribution. Conversely, reporting 20 of revenue does not establish that 20 has been collected or is unrestricted cash. Collection timing, settlement obligations and refunds need their own reconciliation. The principal–agent revenue assessment is not, by itself, a rule for net presentation in the cash flow statement.

Comparability also fails if one company includes taxes, shipping or cancelled orders in its transaction metric while another excludes them. A disclosed “take rate” may divide a different income measure by a different activity measure. Reproduce the definitions before interpreting a change. Discounts, subsidies and customer incentives also need consistent treatment rather than an assumed universal deduction.

Even matched gross profit can conceal cost-classification differences. Staff, technology and fulfilment charges may sit in different lines. An agency model may use less inventory but still face customer, supplier, operational or contractual risk. Higher margins are evidence to investigate, not proof of a stronger business or wrongdoing.

Build a comparison that survives the next results release

1. Identify the specific goods or services and the disclosed control judgement. Separate activities with different roles instead of assigning the whole company one permanent label.
2. Bridge end-customer activity to reported revenue using the company's definitions. Record taxes, refunds, exclusions, currency, period and any missing supplier or fee information.
3. Reconcile revenue to gross profit and operating profit on a common cost and consolidation basis. Keep amounts alongside percentages so denominator changes remain visible.
4. Test the explanation with fixed-basis scenarios. Calculate how much fee compression, cost growth or business-mix change would preserve or reduce the profit being compared.
5. Match valuation dates and numerator scope, then document the evidence that would change the conclusion. If the disclosures cannot support a bridge, mark the comparison unresolved and update it when new information arrives.

A useful conclusion states what changed: a reporting denominator, the mix of distinct activities, or the income earned after costs. “Twelve times revenue is expensive” and “100% gross margin is superior” skip that distinction. The more defensible comparison follows the contract, the amounts and the profit they can actually support.