Why percentage-of-completion revenue recognition is one of the highest-judgment areas in FDD, how it flatters EBITDA, and what to test in construction deals.
Give a construction business to two analysts and one will hand back a tidy EBITDA bridge that took the reported numbers at face value; the other will hand back a warning that a chunk of that EBITDA was manufactured by an estimate nobody has verified. The difference between them is whether they understood percentage-of-completion accounting - the method that governs revenue recognition on long-term construction, engineering and project contracts, and one of the highest genuine judgment risks anywhere in financial due diligence. Most revenue recognition problems you meet elsewhere are about timing on short, clean contracts. Long-term contracts are a different animal, because profit is booked today against a guess about costs that will not be known for another eighteen months.
A construction project spanning eighteen months cannot sensibly recognise all its revenue when the project finally completes. Doing so would badly distort every intervening year for a business that is genuinely working, spending and earning throughout the whole period - reporting nothing, nothing, then a giant spike. So instead, revenue is recognised progressively as the contract is fulfilled.
The mechanism is percentage-of-completion accounting: revenue and profit are recognised in proportion to how much of the contract has been completed. The most common way of measuring that proportion is the cost-to-cost method - costs incurred to date divided by total expected costs - though physical milestones or surveys of work performed are also used.
Put concretely, if a contract is expected to cost €8.0m in total and €4.0m has been incurred, the contract is judged 50% complete, and 50% of the contract's expected revenue and profit is recognised. Simple enough on the surface. The problem is the denominator.
The reason percentage-of-completion deserves scepticism becomes obvious the moment you change one input. Take a fixed-price contract worth €10.0m in revenue, originally expected to cost €8.0m - a healthy €2.0m margin. Suppose €4.0m of cost has genuinely been incurred to date. Watch what happens to recognised profit when the estimate of total cost moves.
| Scenario | Total expected cost | Cost incurred | % complete | Revenue recognised | Profit recognised |
|---|---|---|---|---|---|
| Original estimate | €8.0m | €4.0m | 50% | €5.0m | €1.0m |
| Optimistic estimate | €7.0m | €4.0m | 57% | €5.7m | €1.6m |
| Realistic (overrun) | €10.0m | €4.0m | 40% | €4.0m | €0.0m |
Nothing has changed in the real world between these three rows. The same €4.0m of cost has been spent on the same project. The only thing that moved is management's estimate of what the contract will ultimately cost - and recognised profit swings from €0.0m to €1.6m on that estimate alone. Shave the cost-to-complete estimate and you pull revenue and margin forward; face reality on an overrun and the recognised profit collapses.
The core risk in one line: in percentage-of-completion, the estimate of total cost is the lever that sets today's profit - and it is management's estimate, not an audited fact.
That is why this area sits at the heart of any quality of earnings exercise on a project business. You are not reviewing a number; you are reviewing the reliability of a process for producing numbers.
The single mechanic above already shows the exposure, but in practice the judgment risk clusters in three specific places.
Total expected contract cost is an estimate, not a fact. It is the denominator that drives how much profit is recognised at any point. A contractor that under-estimates total cost over-recognises profit early in the project's life, only to book a catch-up loss later when reality asserts itself. A business with a history of late-project cost overruns is telling you its estimation discipline is weak - and weak estimation discipline is precisely what flatters early-stage margins.
Contract modifications and variations. Long-term projects routinely acquire scope changes mid-flight - extra work, altered specifications, client-requested variations. How these are priced, and whether the associated revenue is recognised before it is contractually agreed, is a genuine area of discretion. Recognising margin on a claimed variation that the customer has not yet approved is one of the classic ways project revenue gets pulled forward optimistically.
Loss-making contract recognition. Accounting standards require a business to recognise the full expected loss on a contract as soon as that loss becomes probable, even if the contract is nowhere near finished. This is the mirror image of the optimism problem. A project business with several contracts turning loss-making has a real, immediate EBITDA hit coming - an onerous contract provision - whether or not it has yet been booked in the numbers in front of you. Finding an unbooked probable loss is one of the highest-value catches available on a construction deal.
Because percentage-of-completion leans so heavily on management's own cost estimates, it is genuinely susceptible to earnings management. Deliberately optimistic cost-to-complete assumptions can flatter reported EBITDA for several consecutive periods before reality forces a correction - and a business being prepared for sale has every incentive to present its strongest possible margin profile.
This is a meaningfully different risk from the non-recurring items review that dominates most FDD engagements. There you are stripping one-off noise out of a settled number. Here you are stress-testing whether an ongoing estimation process deserves trust at all. The two require different instincts. A revenue-quality lens built for recurring, subscription-style businesses does not transfer cleanly; project revenue is lumpy, estimate-driven and back-loaded with risk, and it demands its own scepticism as part of any revenue quality assessment.
There is a cash dimension too. Percentage-of-completion detaches recognised revenue from cash received, which shows up in the balance sheet as contract assets (amounts earned but not yet billed - sometimes called under-billings) and contract liabilities (amounts billed ahead of work done - over-billings). A business that consistently over-recognises revenue tends to accumulate contract assets faster than it collects cash, and that divergence between reported profit and cash generation is often the clearest tell. It also complicates the net debt and working capital picture, because contract balances sit awkwardly between the two and are easy to misclassify.
The work is specific and testable. Start with the largest contracts by value and compare original budgeted cost-to-complete against actual cost incurred to date. A consistent pattern of overruns across multiple contracts is a red flag about estimation discipline in general, not just about one troubled project.
A practical checklist:
Each of these connects back to the financial due diligence process as a whole: the point is not to re-estimate every contract yourself, but to judge whether the estimation machine producing the reported EBITDA is trustworthy.
Construction and project accounting is a favourite interview topic precisely because it separates candidates who memorised definitions from those who understand where profit really comes from. A strong answer walks through the lever.
"Under percentage-of-completion, you recognise revenue and profit in proportion to how complete a contract is, usually measured by costs incurred to date over total expected costs. The risk in FDD is that total expected cost is an estimate set by management, and it's the denominator - so if they shave the cost-to-complete, the contract looks more complete than it is and profit gets pulled forward. I'd test that by comparing original budgets to actual costs on completed contracts to see how accurate their estimates historically were, looking for catch-up adjustments where earlier margins were later reversed, and checking whether any in-progress contracts are probably loss-making without a provision booked, because standards require the full expected loss to be recognised as soon as it's probable. I'd also watch the contract asset balance - if earned-but-unbilled revenue is growing faster than cash collection, that's a sign revenue's being recognised aggressively. So I'd treat a project business's reported EBITDA with more scepticism than a simple transactional one, because the judgment risk genuinely sits in the estimates."
That answer lands because it names the lever (total expected cost), explains the direction of the distortion, gives concrete tests, and connects the accounting to cash. Interviewers are listening for whether you understand why the method creates risk, not just that it exists.
Percentage-of-completion is a perfectly legitimate and necessary method - you genuinely cannot account for an eighteen-month project any other way without badly distorting the picture. But it leaves real room for both honest estimation error and deliberate optimism, and those two are hard to tell apart from the outside. Treating a construction or project business's reported EBITDA with more scepticism than you would a simpler, transaction-based business is not unfair. It is appropriately calibrated to where the judgment risk genuinely lives - in an estimate about the future, made by the people selling the business, that sets how much profit gets booked today.
The Transaction Services Interview Programme (€119.99, one-time) includes a full walkthrough of percentage-of-completion accounting, worked cost-to-cost examples, and the exact tests to run on a construction QoE. Enrol today.
Hundreds of candidates prepared their interviews with this programme. Those who landed the role have one thing in common: they worked the cases before walking into the room.
