Discontinued operations and held-for-sale assets force you to restate the perimeter, not just add back noise. Here is how FDD analysts handle disposals cleanly.
A target that sold a division eighteen months ago, mothballed a factory last quarter, or plans to spin off a segment the week after your deal closes hands you a problem that looks like a normalisation but isn't. A standard add-back cleans noise out of a stable business. A disposal changes the shape of the business itself. Get the two confused and you will hand a buyer a number that describes a company that no longer exists.
Every piece of earnings analysis rests on a silent assumption: that the entity earning the profit last year is the same entity the buyer will own next year. Discontinued operations break that assumption. The moment a division has been sold, closed, or earmarked for disposal, the historical profit and loss account is a blend of two things - the business the buyer is acquiring and a business the buyer will never touch.
Consider the arithmetic. If a target disposed of a loss-making division a year ago, group EBITDA in the earlier periods is depressed by losses the buyer will never inherit. The remaining business is worth more than the reported trend suggests. Flip it around: if a profitable division was carved out and sold before the process began, historical EBITDA flatters the go-forward business, and a buyer pricing off that trend will overpay.
The fix is not an add-back. It is a restatement. You are not removing noise from within a fixed perimeter - you are redrawing the perimeter and rebuilding the trend inside the new boundary.
This is why discontinued operations deserve their own workstream rather than a line buried in the EBITDA adjustments schedule. An add-back says "this cost was abnormal, ignore it." A restatement says "this whole activity was never part of the business you are buying, so let us show you what the last three years looked like without it."
The two terms travel together but mean different things, and the distinction drives your analysis.
A discontinued operation (IFRS 5, or ASC 205-20 under US GAAP) is a component of an entity that has either already been disposed of or is classified as held-for-sale, and which represents a separate major line of business or geographical area. When something qualifies, the income statement is re-presented: the discontinued activity is stripped out of every line above and shown as a single net figure - "profit/(loss) from discontinued operations" - usually right at the bottom, after tax.
Held-for-sale classification applies earlier, while the asset is still owned. It kicks in when management has committed to a disposal plan, the asset is available for immediate sale in its present condition, and the sale is highly probable - normally expected to complete within twelve months. Held-for-sale assets and liabilities are presented separately on the balance sheet, and the disposal group is measured at the lower of carrying value and fair value less costs to sell, with depreciation halted.
| Feature | Held-for-sale | Discontinued operation |
|---|---|---|
| Timing | Still owned, sale pending | Sold, or held-for-sale AND a major line |
| Where it shows | Separate balance sheet lines | Single net line in P&L |
| Measurement | Lower of carrying value / FVLCS | Results re-presented, prior years restated |
| FDD signal | Why now? Will it close before we do? | Rebuild the trend without it |
For diligence, held-for-sale is the louder alarm. It tells you a disposal is in flight right now. The questions write themselves: why is management selling this asset at this precise moment, does the timing connect to your transaction process, and - most practically - will the sale complete before your deal signs, quietly changing the perimeter mid-process?
Numbers make this concrete. Take a group that sold a loss-making logistics division in FY24. The buyer is acquiring only the continuing manufacturing business. Here is the reported group EBITDA against the restated continuing-only view.
| (£m) | FY22 | FY23 | FY24 | Notes |
|---|---|---|---|---|
| Reported group EBITDA | 18.0 | 19.5 | 22.0 | Includes logistics |
| Less: logistics EBITDA | (2.5) | (3.0) | (1.5) | Loss-making, disposed FY24 |
| Continuing EBITDA | 20.5 | 22.5 | 23.5 | Perimeter the buyer gets |
| Reported growth | - | +8.3% | +12.8% | Flattered by exit |
| Continuing growth | - | +9.8% | +4.4% | The real underlying trend |
Two things jump out. First, the continuing business earns more than the reported group in every year, because the discontinued logistics arm was dragging the total down. A buyer anchoring on the reported £22.0m in FY24 would understate the earnings base by £1.5m - and if the deal is priced at, say, 9x, that is over £13m of enterprise value left on the table. Second, the shape of the trend changes. Reported EBITDA accelerates into FY24 (+12.8%); but a chunk of that acceleration is simply the loss-making division disappearing, not the core business improving. The continuing trend actually decelerates to +4.4%, which is a very different story to underwrite.
This is the whole point. Restatement doesn't just move the level of EBITDA - it changes the growth narrative, and the growth narrative is what a buyer pays a multiple for.
The temptation is to net off a single lump sum - "logistics lost £1.5m in FY24, add it back." Resist it. That gives you one corrected year and a broken trend. Proper restatement means rebuilding each historical period as if the current perimeter had always existed.
Do this properly and the discontinued-operations analysis becomes one of the cleaner sections of the report. Do it lazily and you get a single-year adjustment that anyone can pick apart.
Here is the subtlety that separates a competent analyst from a sharp one. When a division is sold, its direct costs leave with it - but the central overhead that used to be allocated to it does not vanish. Group finance, IT, HR, the CFO's time, the shared ERP licence: those costs were partly borne by the discontinued operation, and now they land entirely on the remaining business.
These are stranded costs, and they are the opposite of a synergy. If you restate continuing EBITDA by simply removing the discontinued division's revenue and its directly attributable costs, you will overstate go-forward profitability, because you have quietly left behind the central overhead the division used to help carry.
A disposal that removes £3.0m of direct cost but leaves £0.8m of previously-allocated central overhead stranded on the continuing business means your restated EBITDA is £0.8m too generous unless you flag it.
Stranded costs are not a historical adjustment - you cannot rewrite the past to pretend the overhead was always fully borne by the continuing business, because it wasn't. They belong in the forward-looking commentary: a note that go-forward standalone costs will be higher than the restated historical continuing view implies. This is precisely the same mechanic you meet in a carve-out financial due diligence, where separating a business from its parent leaves the parent - or the carve-out - carrying costs that used to be shared.
Sometimes the discontinued operation is the transaction. You are running buy-side FDD on a business being carved out of a larger group, and from the parent's perspective your target is its held-for-sale disposal group. Now the two analyses collide.
The historical financials you receive were never prepared standalone. Shared services, intercompany trading, group-allocated costs and centrally-negotiated supplier contracts all knit the target into the parent. Separating genuinely dedicated activities from shared ones - and pricing the standalone cost base the target will actually run - is the core carve-out challenge, and it feeds straight into the net debt and working capital deliverables as well as EBITDA. The discipline is the same as the disposal case in reverse: establish exactly what belongs inside the perimeter, and rebuild the numbers on that basis.
Interviewers use discontinued operations to test whether you understand the difference between cleaning a number and redefining a business. Expect a prompt like: "A target sold a loss-making division last year. Reported EBITDA is growing nicely. What do you do?"
A strong answer sounds like this:
"My instinct is that reported EBITDA is misleading here, but not in the usual direction. Because the disposed division was loss-making, the reported trend is actually understating what the continuing business earns - and part of the reported growth is just that loss disappearing rather than the core business improving. So I wouldn't treat this as a simple add-back. I'd restate the full historical trend on a continuing-only basis, going back to divisional management accounts to strip the discontinued piece out of every line, and reconcile each year back to the statutory numbers. Then I'd look separately at stranded costs - the central overhead that used to be allocated to the disposed division doesn't leave with it, so my restated continuing EBITDA needs a forward-looking note that standalone costs will be a bit higher than the pure restatement suggests. The output I'd want to hand the deal team is a clean continuing-perimeter EBITDA trend plus a clear view of the go-forward cost base."
That answer wins because it does three things at once: it spots the counter-intuitive direction of the distortion, it insists on restatement over adjustment, and it catches stranded costs. Most candidates get one of the three.
Most FDD errors misstate a figure. Getting the perimeter wrong misstates what the buyer is valuing. You can normalise every one-off perfectly, build a flawless net debt to equity bridge, and structure a beautiful report - and it is all built on sand if the earnings trend describes a business the buyer will never own. Discontinued operations and held-for-sale assets are the moment the ground shifts under the analysis. Treat them as a distinct exercise, restate rather than adjust, chase the stranded costs, and always ask what perimeter is actually crossing the line at completion. Anchor to the reported number without asking that question, and you will have priced the wrong company beautifully.
The Transaction Services Interview Programme (€119.99, one-time) includes a full module on perimeter restatement - rebuilding continuing-only EBITDA trends, spotting stranded costs, and handling held-for-sale assets that move mid-deal, with worked exercises and model interview answers. 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.
