Comparable Companies Analysis Template (Trading Comps)

Comparable Companies Analysis: [Subject company]

A trading comparables study (also called "comps" or "peer multiples"): purpose and level of value, peer selection, data conventions, calendarisation and normalisation, enterprise value build, multiples, statistics and outliers, positioning of the subject, implied value range, an optional precedent transactions extension and a football field summary. It documents one valuation method. Structure follows common market practice and is consistent with the market approach named in the International Valuation Standards (IVS 103 Valuation Approaches, effective 2025-01-31), cited by number only. The worked example uses "Halvard Tools", a fictional listed toolmaker, and five fictional peers; every figure is invented. Replace every [bracketed] value and every row marked Example. This is a document format, not financial, investment, tax, legal or accounting advice. Obtain the standard or law for the normative text.

Document control

Approvals

Revision history

Contents

1. Purpose, scope and level of value
2. Subject profile and metrics
3. Peer selection
4. Data sources and conventions
5. Calendarisation and normalisation
6. Enterprise value build
7. Multiples table
8. Statistics and outliers
9. Positioning the subject
10. Implied valuation range
11. Precedent transactions extension
12. Football field summary
13. Limitations and review checklist
14. Glossary and references
15. Standard and licence note

1. Purpose, scope and level of value

The market approach estimates value by reference to prices paid for comparable assets. A comps study answers one question: at what multiples does the market price businesses with similar risk, growth and returns, and what does that imply for the subject? It does not replace a fundamental view of the subject's cash flows, and the market can misprice an entire sector.

2. Subject profile and metrics

At the reference price of EUR 42.00 the subject has a market capitalisation of EUR 5,040.0m and an enterprise value of EUR 5,640.0m, which is 12.5x LTM EBITDA and 10.9x NTM EBITDA (Example).

3. Peer selection

3.1 Criteria

3.2 Screening funnel

Show how many companies passed each criterion, in order, so the selection can be repeated at the next refresh (Example):

3.3 Selected peers

Record companies considered and rejected with the reason, so a reviewer can see that the selection was not tuned to the result.

Peer selection is the step with the most judgement and the least arithmetic, so write down the reasoning. Start from what drives value for the subject: end markets and their cycle, the business model (product, service or recurring revenue), growth, margins, capital intensity, leverage and size. A company with the same industry code but a different model, for example a distributor rather than a manufacturer, is usually a poor peer however close its name. Aim for five to ten peers; fewer makes every statistic fragile, more usually means the later additions are weak. When no listed company is a close match, use a broader set and lean harder on the positioning step in section 9 rather than adding weak peers to reach a number. Keep the same set across refreshes unless a peer is acquired, restructures or changes its model, and record any change in the revision history so that a movement in the median can be explained.

4. Data sources and conventions

Treasury stock method:

where $K$ is the weighted average exercise price and $P$ the share price.

Common data errors to check before computing any multiple: a share count taken from the cover of the annual report rather than the latest filing, treasury shares not deducted, convertibles ignored when they are in the money, net debt taken from a balance sheet date before a large acquisition or disposal, consensus figures that mix adjusted and reported earnings across brokers, and currency translation at different dates for price and forecasts. Each of these can move a single multiple by a full turn, which is larger than most of the differences the study is trying to explain.

5. Calendarisation and normalisation

NTM calendarisation for a peer whose financial year ends in month $m$ (pricing at 31 December):

where $FY_1$ is the forecast year ending within the next twelve months and $m$ the months of it still to run. Worked example for Corlett Hardware (year end 31 March, so 3 months of FY Mar-2027 remain) (Example):

LTM for the same peer (Example): FY Mar-2026 EBITDA 114.0 plus nine months to December 2026 of 94.0 less nine months to December 2025 of 88.0 equals 120.0.

Apply one normalisation policy to every company. If you add back restructuring costs for the subject, add them back for each peer that reports them; if you accept a peer's own adjusted EBITDA, read its reconciliation and remove items you would not accept for the subject, such as recurring share-based payment or acquisition costs in a serial acquirer. Normalise LTM figures only: consensus forecasts are usually already on an adjusted basis, so check the provider's definition rather than adjusting them again.

6. Enterprise value build

Keep a per-peer bridge in the workbook showing each adjustment and its source line in the filing.

The EV build must match the metric it divides. A consolidated EBITDA includes the full result of subsidiaries that are not wholly owned, so add non-controlling interests to EV; equity-accounted associates are excluded from EBITDA, so deduct their value. Treat pension deficits, leases and similar debt-like items the same way for every company, and decide once whether EBITDA is before or after lease costs. Use market values for listed debt where they differ materially from book values, and say so. When a peer has announced but not completed a transaction, either adjust both EV and the metric on a pro forma basis or neither.

7. Multiples table

Peer operating metrics (EUR m, calendarised) (Example):

Multiples (Example):

P / NTM EPS equals market capitalisation divided by NTM net income attributable to owners, which gives the same result as price divided by diluted EPS when the share counts match.

Check that each peer's metrics hang together before using its multiples. In the example, Ostrand's NTM net income of EUR 175m follows from NTM EBIT of EUR 250m less about EUR 17m of interest on EUR 392m of net debt, taxed at 25%: (250 - 17) x 0.75 = 174.75 (Example). A net income that is close to EBIT, or a margin far from the company's history, is usually a data error or a one-off item in consensus and should be traced before it reaches the statistics.

8. Statistics and outliers

Percentiles use linear interpolation between ranked observations (the default in common spreadsheet functions). With five peers, state that the sample is small and that each percentile rests on one or two observations.

9. Positioning the subject

A simple cross-sectional view plots EV / NTM EBITDA against NTM growth or margin. With five peers, use it as a sanity check, not as a regression model with statistical meaning.

Base the positioning on evidence that investors pay for. Growth and margins usually explain most of the spread in EV/EBITDA within a sector; returns on capital, cash conversion, cyclicality and leverage explain much of the rest. Where the subject is better on some factors and worse on others, say which ones matter more for this sector and why, and avoid premiums or discounts expressed to a decimal place that the evidence cannot support. If you place the subject away from the median, show the multiple you apply and its source (for example the 75th percentile) rather than an adjusted median.

Positioning conclusion (Example): Halvard sits close to the median; its higher forecast margin offsets slightly lower growth, a larger retail share and higher concentration. Apply the median, with the interquartile range as the valuation range.

10. Implied valuation range

Worked check at the EV / NTM EBITDA median (Example): 11.5 x 519.75 = EUR 5,977.1m EV; less EUR 600.0m net debt = EUR 5,377.1m equity; divided by 120.0m shares = EUR 44.81 per share. Implied values use unrounded multiples from the workbook.

Selected range (Example): EUR 42.21 to EUR 47.84 per share, point EUR 44.81 (EV / NTM EBITDA). EV / NTM EBIT corroborates it. P / NTM EPS gives lower values because the peer P/E median is held down by the lower-growth peers, while Ostrand, which carries a 26x P/E on recurring service revenue, sits at the top; record such differences rather than averaging them away.

11. Precedent transactions extension

Median 14.4x LTM EBITDA, applied to FY2026 EBITDA of EUR 450.0m, implies EUR 48.99 per share at control level (Example). Report it separately from the trading range.

Precedent multiples are LTM multiples at announcement, so apply them to LTM EBITDA, not to a forecast. Older deals reflect a different credit cycle and valuation level; show the year next to each multiple and test whether the conclusion changes when deals older than three years are dropped. The one-day premium median of 26.5% is a cross-check: the undisturbed price of EUR 42.00 plus that premium is EUR 53.13, about 8% above the precedent multiple value of EUR 48.99, so the two control references broadly agree. A large gap between them would suggest that the precedent targets traded on different multiples before their deals, and should be explained.

12. Football field summary

The bars show point values and the line shows the current price; show the low to high ranges in the table or in a floating bar chart in the presentation version.

13. Limitations and review checklist

Limitations (Example):

1. Five peers is a small sample; each statistic is sensitive to one company.
2. No peer matches the subject's mix of hand tools, cordless and accessories.
3. Multiples reflect market sentiment at the pricing date, which may be cyclical.
4. Consensus forecasts may be stale or biased; dispersion is not shown here.
5. Accounting differences (IFRS against US GAAP, capitalised development costs) are adjusted only where material and disclosed.

Review points for the reviewer: recompute at least one peer from raw inputs (price, shares, net debt, metric) to the multiple; confirm that the subject's metrics are on the same definition and calendar window as the peers; check that the statistics row uses the method stated (linear interpolation, median rather than mean); and read the commentary against the numbers, because text written before a data refresh often survives it. A study that will feed a published report or a board document should also record who extracted the data, from which provider and when, so the next refresh can reproduce it.

14. Glossary and references

14.1 Glossary

14.2 References

Standard and licence note

Structure aligned with common market practice for trading comparables and with the market approach named in the International Valuation Standards (IVSC, effective 2025-01-31), cited by standard number and name only. No text from any standard or vendor publication is reproduced; headings, guidance, formulas in our own notation and the worked example are original. Halvard Tools, its peers and the transactions are fictional and all figures are invented; nothing here is a view on any real security. Check the current official versions of any standard you rely on and the licence terms of your market data provider before use. This is a document format, not financial, investment, tax, legal or accounting advice. Template licence: MIT.