How Lost Profits Are Calculated: A Practical Guide

A lost profits claim asks a court to accept a number describing something that never happened. There are no invoices for the sales that were never made and no payroll records for the crew that was never hired. Everything rests on a counterfactual, and the credibility of that counterfactual is what the whole exercise is about.

The mechanics below are common to contract, patent, antitrust and insurance matters. Where the legal standard differs by context — and it does — that is flagged as it comes up.

What “lost profits” means precisely

Lost profits are the net, incremental profits the claimant would have earned but for the conduct. Three words in that sentence do real work.

Net. Revenue that never arrived did not carry the costs of earning it. Those costs have to come out.
Incremental. The relevant costs are the ones that would have changed with the lost activity, not an allocated share of everything the business spends.
But for. The comparison is against a world identical in every respect except the defendant’s conduct — not against the claimant’s budget, and not against its best year.

Gross revenue is not a damages figure and never has been. A surprising number of claims are pleaded as though it were.

Building the but-for revenue line

Three approaches are generally accepted. Most credible analyses use one as the primary method and at least one other as a cross-check.

Before-and-after

Compare performance after the conduct with performance before it. Strongest where the business has a stable operating history and the conduct is the only material thing that changed. It fails when something else changed at the same time — a lost anchor customer, a lease, a regional downturn — and the analysis quietly attributes the whole gap to the defendant.

Yardstick, or comparable benchmark

Use similar businesses, other locations of the same business, or industry data over the same window. This is what carries a claim where the claimant’s own history is short or already disrupted. It is only as good as the comparability of the benchmark, and comparability is where the cross-examination goes.

Projection

Build the counterfactual from contracts, backlog, pipeline, budgets or a business plan. The persuasive version rests on documents created before the dispute for some purpose other than the litigation. A forecast first prepared after the complaint was filed is worth very little, however carefully it is built.

Whichever is chosen, the same judgments have to be made explicitly: the underlying growth trend, seasonality, cyclicality, and what the wider market was doing over the loss period. A but-for line that runs flat through a quarter the whole industry lost is not neutral. It is an assumption, and it should be defended as one.

Costs: where two experts diverge most

If two competent analysts produce very different numbers from the same records, the cost treatment is usually why.

The question is not how an accountant classified a cost but how that cost behaves as output changes. Chart-of-accounts labels are unreliable guides. “Salaries” often contains a fixed management layer and variable production labor in one line. Utilities are usually part fixed and part volume-driven. Some costs move in steps rather than smoothly — a second shift, another delivery vehicle, a further lease — and a model that treats them as linear will misstate the loss in both directions depending on where the volume lands.

The work is to establish the cost-volume relationship empirically from this business’s own records, then apply it to the revenue that did not occur. Where the records will not support that, the honest output is a range with its drivers identified, not a point estimate that implies precision nobody has.

One recurring error: deducting all overhead. If the overhead would have been incurred anyway, it is not avoided, and subtracting it understates the loss. The test is avoidability, not category.

The loss period

Both endpoints are contestable and both are frequently asserted rather than analyzed.

The period begins when the conduct began to affect the business, which is not always the date of the breach or the filing. It ends when the business returned to the position it would have occupied anyway — not when it reopened, not when the claimant says it felt recovered, and not at the trial date because that is convenient.

In some contexts the endpoint is set for you. A first-party insurance claim runs for the policy’s period of restoration. A patent claim runs to the expiry of the patent or the end of infringement. A contract claim may be bounded by the term or by a notice period. Getting this wrong is a bigger source of error than most of the arithmetic.

Mitigation and offsetting benefits

Sales deferred rather than lost, customers who returned, capacity redeployed to other work, costs genuinely saved: each reduces the claim, and an analysis that ignores them invites the trier of fact to discount the whole opinion rather than just that line.

The mirror-image error is over-deducting — treating unrelated new business as mitigation when the claimant could have earned it anyway. The question is whether the benefit arose because of the conduct.

Present value, and interest

Losses already incurred and losses still to come are treated differently. Future amounts are discounted to present value; past amounts are generally not discounted, and may carry prejudgment interest depending on the jurisdiction and the cause of action.

The discount rate is genuinely contested. One view holds that a lost profits stream should be discounted at a rate reflecting the claimant’s own cost of borrowing, since the loss is of cash the business would have held. The competing view is that the rate should reflect the risk of those profits, which for a volatile stream is materially higher. What is not defensible is importing a weighted average cost of capital from a valuation prepared for another purpose without asking whether the risks match. The rate frequently moves the number more than the revenue projection does.

Reasonable certainty: the filter the number has to pass

Courts do not require mathematical precision. They require a reasonable basis, and they require it as to both the fact of the loss and its extent — an expert can be entirely right that profits were lost and still be excluded for having no defensible way to size them.

California instructs this through CACI 3903N, and the gatekeeping standard comes from Sargon Enterprises, Inc. v. University of Southern California (2012) 55 Cal.4th 747. Courts also draw a long-standing distinction between established and unestablished businesses: past volume makes an established business’s losses ascertainable, while a new venture has to carry the claim through comparables and economic evidence. That is a higher bar, not an absolute one.

What changes by context

Breach of contract. Damages are bounded by what was foreseeable at contracting. Consequential losses may be limited or excluded by the agreement itself — read the limitation-of-liability clause before modeling anything.
Patent infringement. Lost profits are typically established through the Panduit framework: demand for the patented product, absence of acceptable non-infringing substitutes, capacity to have met the demand, and the profit that would have been made. Price erosion is a separate component, and it must account for the volume that would have been lost at the higher price — see price erosion.
Antitrust. The but-for world has to be constructed at market level, not just at the claimant’s own accounts, and damages are subject to trebling — which raises the scrutiny the underlying model attracts.
First-party insurance. The policy governs, not the common law measure. Coverage responds to actual loss sustained over a defined period of restoration, subject to the policy’s own definitions and exclusions.

The documents that decide it

Almost always available early, and almost always dispositive: monthly financial statements for several years either side of the loss; the general ledger at transaction level; point-of-sale or order data; customer-level revenue history; contracts and purchase orders; any budget or forecast prepared before the dispute; and tax returns as a reconciliation check against the internal figures.

Where internal statements and tax returns disagree materially, that gap will be found. Better to explain it in the report than to be asked about it for the first time in a deposition.

Six ways these calculations fail

Gross revenue treated as profit.
Cost behavior assumed rather than established from the records.
A loss period with an unexamined endpoint.
Mitigation ignored — or over-applied.
A projection built after the complaint and presented as though it predates the dispute.
No back-test. If the method cannot reproduce the claimant’s actual results for a period before the conduct, it is not a measurement. This is the most useful question in the whole exercise and the least often asked.

Where to start

Earlier than most people do. The theory of damages shapes what should be requested in discovery, and a head of loss that cannot be proven is much cheaper to identify before it is pleaded than after. If you are evaluating whether a lost profits claim can be established — or testing the other side’s — the underlying records are usually enough to give a preliminary view.

How Lost Profits Are Calculated