The multiple is the last step, not the first: the defensible value comes from the normalisation and from adjustments that can each be challenged on their own.
Multiply normalised EBITDA by the median multiple of comparable listed companies, adjusted for the documented differences between them and the subject, then deduct net debt. On an illustrative manufacturer, reported EBITDA of 42.0 normalises to 38.8 and the peer median of 8.0x adjusts to 7.3x, giving an enterprise value of 283.24 and equity of 193.24. Applying the raw median to reported EBITDA would have produced 246.00 of equity, 27.3 per cent too much.
Worked in full in The Private Markets Valuation Specialist by Julian R. Sterling, with every figure reproduced in a free workbook.See the book on Amazon →
The market approach looks like one multiplication. In a fair value file it is four judgments in a row: what EBITDA a buyer would pay for, which companies are comparable, how the subject differs from them, and what sits between enterprise value and equity. Each moves the answer, and each has to survive the auditor's question "why?".
| Listed peer | EV / EBITDA |
|---|---|
| Northmoor Industrial | 7.1x |
| Calderon Components | 7.8x |
| Halstead Precision | 8.0x |
| Verrier Manufacturing | 8.6x |
| Ostheim Werke (larger, faster-growing) | 9.5x |
| Median / mean | 8.0x / 8.2x |
The subject is smaller than every peer, at 53.2 per cent of the median peer's EBITDA, earns a 21.2 per cent margin against a peer median of 18.2 per cent, grows 1.4 per cent a year against 2.4, and one customer is 28 per cent of revenue. Net debt is 90.0.
Enterprise value = normalised EBITDA × (peer median + documented adjustments)
Equity value = enterprise value − net debt and other claims
In Excel: =MEDIAN(PeerMultiples)+SUM(Adjustments) for the multiple, then =Multiple*NormEBITDA-NetDebt.
Step 1, normalise EBITDA. Start from the 42.0 in the management accounts. Add back 2.4 of restructuring cost that is genuinely non-recurring, and deduct a 5.6 gain on a property sale, which is not the earnings power a buyer acquires. Normalised EBITDA is 38.8, 7.6 per cent below reported. A restructuring that recurs every year is a run-rate cost, not an add-back.
Step 2, take the median. The median of the five peers is 8.0x. The mean, 8.2x, is pulled up by Ostheim, which is kept in the set on probation because its scale and growth put it in a different bracket.
Step 3, adjust for named differences.
| Component | Reason | Turns |
|---|---|---|
| Peer median | Starting point, never the answer | 8.0 |
| Size | About half the median peer | −0.45 |
| Customer concentration | One customer at 28% of revenue | −0.35 |
| Growth | 1.4% against 2.4% | −0.20 |
| Margin | 300 bp above the peer median | +0.30 |
| Applied multiple | 7.3 |
The four judgments net to minus 0.7 of a turn. A single "private company discount" of 0.7 would give the same number, but it could only be accepted or rejected whole. Four signed components can each be evidenced, and the margin premium shows that the build is allowed to cut both ways.
Step 4, the value. 7.3x × 38.8 = 283.24 of enterprise value. Less 90.0 of net debt, equity is 193.24. Expressed on reported EBITDA, the company is valued at 6.74x, a number worth showing because it is the one a reader of the accounts will compute.
The fund paid 264.65 two years ago for 33.5 of normalised EBITDA, 7.9x, when the same peers traded at a median of 8.6x. The price paid implied a discount of 0.7 of a turn to the set. The bottom-up build and the transaction agree, which is the strongest evidence a market approach can have. When they do not, the gap must be explained, as set out in how to calibrate a valuation to the entry price.
| Variant | Multiple | EBITDA | Enterprise value | Equity | Change in equity |
|---|---|---|---|---|---|
| Base | 7.3x | 38.8 | 283.24 | 193.24 | |
| Mean instead of median | 7.5x | 38.8 | 291.00 | 201.00 | 4.0% |
| Ostheim excluded | 7.2x | 38.8 | 279.36 | 189.36 | −2.0% |
| No adjustments | 8.0x | 38.8 | 310.40 | 220.40 | 14.1% |
| Reported EBITDA | 7.3x | 42.0 | 306.60 | 216.60 | 12.1% |
| Both shortcuts | 8.0x | 42.0 | 336.00 | 246.00 | 27.3% |
| Half a turn lower | 6.8x | 38.8 | 263.84 | 173.84 | −10.0% |
Leverage magnifies every error: enterprise value is 1.47 times equity, so one turn of multiple, 38.8 of enterprise value, is 20.1 per cent of the equity mark. Removing the probation peer moves the median to 7.9x and changes equity by only 2.0 per cent, which is the reassurance a stable set should give.
The common mistake is to treat the multiple as the valuation and the EBITDA as a given. Here the normalisation alone is worth 23.36 of enterprise value, almost as much as the 27.16 the whole adjustment build takes off. The second mistake is to pick the multiple first and reason backwards to it; a file that cannot show where 7.3x came from, component by component, will not survive a valuation committee, let alone an audit.
Normalise, take the median, adjust by named and signed judgments, then bridge to equity: 38.8 × 7.3x = 283.24, less 90.0, gives 193.24. Every input is in a cell you can overwrite in the free workbook for this case, and the items that belong between enterprise value and equity are worked in what goes in the EV to equity bridge.
The median, because one outlier pulls the mean. In the illustrative set of five peers the median is 8.0x and the mean 8.2x, pulled up by a larger, faster-growing peer at 9.5x. On 38.8 of EBITDA the mean adds 7.76 to enterprise value, 4.0 per cent of equity, for no reason that relates to the subject company.
Not a single round number. Build it from named differences: here a size discount of 0.45 of a turn, 0.35 for one customer at 28 per cent of revenue, 0.20 for slower growth and a premium of 0.30 for a margin 300 basis points above the peers. They net to 0.7 of a turn, and each can be evidenced or challenged separately.
More than enterprise value, because net debt does not move. On the illustrative company one turn is 38.8 of enterprise value and 20.1 per cent of the 193.24 of equity. Half a turn either way moves equity by 10.0 per cent, which is why the multiple build has to be documented line by line.
The market approach and its calibration are the subject of chapter 5 of The Private Markets Valuation Specialist, and the companion workbook carries this exact chain. The book takes the same case from first principles to the decision, chapter by chapter, and every figure it prints is a live formula in the free companion workbooks.
Get the book on Amazon →Free companion files
Also on Amazon UK · Amazon Germany · Amazon France · Amazon Canada
Reading guide: private equity and private markets → · All 681 articles →
If this book helped, or didn’t, a few lines on Amazon are worth more than they look: they are what the next reader goes on. Write a review. The workbook stays free either way.