Trading Comps: How to Value a Company Using Comparable Public Multiples

A DCF tells you what a business is worth in theory — based on your assumptions about future cash flows, growth, and risk. A trading comps model tells you something different and equally important: what the market is currently paying for similar businesses right now. Investment bankers almost always present both side by side. The gap between them is often where the most interesting conversations happen.

In a first-round investment banking, private equity, or equity research interview, there is a near-certain chance you will be asked to “walk me through a trading comps model.” The question is not testing whether you can recite a formula. It is testing whether you understand the logic, the limitations, and the mechanics well enough to have a genuine conversation about valuation. By the end of this post you will be able to do exactly that — and you will have built a complete comps table for our case study company, WidgetCo Ltd.

---

The Logic of Relative Valuation

The core assumption behind trading comps is straightforward: similar companies in the same industry, with similar growth rates and risk profiles, should trade at similar multiples of earnings or revenue. If every comparable business in a sector trades at 8–10× EBITDA and your target trades at 6×, either the target is cheap or there is a good reason for the discount — and your job is to work out which.

Relative valuation is powerful because it grounds you in market reality. A DCF can produce any number you want if you choose optimistic enough assumptions. A comps analysis is anchored to observable prices in liquid markets.

But it has real limitations. The method assumes the peer group is genuinely comparable. For a highly unusual business — one with a unique operating model, an unmatched growth rate, or a dominant market position — there may simply be no valid peer. Relative valuation also inherits any mispricing in the market: if an entire sector is overvalued, the comps will tell you the target is fairly priced when it is, in fact, expensive. And it is backward-looking by default — public market prices reflect sentiment today, not necessarily the fundamental value five years out.

The distinction between trading comps and precedent transactions is also worth establishing here. Trading comps use current public market prices — the price a minority investor pays for a small stake. Precedent transactions use prices paid in actual acquisitions — the price a buyer pays to take control of an entire business. Acquisition prices are almost always higher, because control is worth more than a minority stake. We will come back to this when we build the football field.

---

Selecting the Peer Group

Peer selection is the most subjective — and most debated — part of any comps analysis. There is no algorithm for it. It requires judgement, and experienced bankers will argue over it.

The screening criteria, in rough order of importance:

1. Same industry and sub-industry. Use SIC codes, Bloomberg industry classifications, or GICS as a starting point, then refine by reading the business descriptions. A SaaS company and a legacy enterprise software business both fall under “software” — they are not comparable. Be specific about the sub-sector.

2. Similar business model. For WidgetCo — a UK industrial machinery manufacturer with a mix of product sales and aftermarket service revenue — the right peers are other industrial manufacturers with a similar product/service split, not pure-play component suppliers or capital equipment rental businesses.

3. Similar size. Revenue or market cap within roughly 0.3× to 3× of the target. A £10m revenue business and a £500m revenue business face different cost structures, financing options, and competitive dynamics. They are not directly comparable even if they operate in the same sector.

4. Similar growth profile. A company growing revenues at 30% per year will command a meaningfully higher multiple than one growing at 3%. Mixing high-growth and low-growth peers without acknowledging this creates a misleading range. If the peer set genuinely spans a wide growth spectrum, the multiple should be adjusted — or the peers should be separated.

5. Similar geography. UK-listed peers for a UK target where possible. International peers are acceptable in global sectors, but note that UK multiples have historically traded at a discount to US equivalents for similar businesses.

Aim for 5–10 peers. Fewer than five gives you too little data to construct a meaningful range. More than ten often means you have included poor comparables simply to pad the list.

Handling outliers: identify any peer that is clearly anomalous — a distressed company trading at 2× EBITDA when the rest trade at 8–10×, or a recent acquisition target trading at an inflated price ahead of deal close — and exclude it from the mean and median, noting the exclusion clearly. Transparency here is important: a comps analysis that quietly drops inconvenient data points is a comps analysis that no one should trust.

---

Building the Comps Table

The standard investment banking comps table has a consistent column structure. Here is the layout for WidgetCo’s peer group of six UK industrial machinery companies:

Company	Share Price (p)	Diluted Shares (m)	Market Cap £m	Net Debt £m	EV £m	LTM Revenue £m	LTM EBITDA £m	EV/Revenue	EV/EBITDA	P/E
Peer A	420	85.0	35.7	4.2	39.9	28.4	5.1	1.4×	7.8×	14.2×
Peer B	615	62.0	38.1	6.8	44.9	31.0	5.8	1.4×	7.7×	15.1×
Peer C	1,240	28.5	35.3	(1.2)	34.1	22.6	4.0	1.5×	8.5×	16.8×
Peer D	875	44.0	38.5	8.1	46.6	34.5	5.6	1.4×	8.3×	14.9×
Peer E	320	110.0	35.2	12.4	47.6	38.2	5.0	1.2×	9.5×	17.3×
Peer F	590	55.0	32.5	3.5	36.0	25.0	3.9	1.4×	9.2×	16.1×
Mean								1.4×	8.5×	15.7×
Median								1.4×	8.4×	15.6×
25th pct								1.4×	7.8×	14.6×
75th pct								1.5×	9.4×	16.6×

The key calculations:

Market Cap = Share Price × Diluted Shares Outstanding
           (use diluted shares — include options and convertibles)

Enterprise Value = Market Cap + Net Debt + Minority Interest − Associates

Net Debt = Total Financial Debt + Lease Liabilities − Cash & Cash Equivalents
           (post-IFRS 16: lease liabilities must be included)

EV/Revenue    = Enterprise Value / LTM Revenue
EV/EBITDA     = Enterprise Value / LTM EBITDA
P/E           = Share Price / LTM Earnings Per Share

A note on LTM vs NTM: LTM (Last Twelve Months) uses the most recently reported 12 months of financials and is backward-looking. NTM (Next Twelve Months) uses analyst consensus forecasts and is forward-looking. In practice, most investment banking pitches use NTM multiples because they reflect where the business is going, not where it has been. Build both if you have consensus estimates available; lead with NTM in live M&A work, and use LTM when published estimates are unavailable or unreliable.

---

Calculating and Applying the Multiples

Once you have the peer set statistics, the application is mechanical. For WidgetCo, we have:

• LTM EBITDA: £950k (£0.95m)
• LTM Revenue: £6.5m

Applying the peer median EV/EBITDA of 8.4×:

Implied EV  = 8.4 × £950k = £7,980k

Less: Net Debt           = (£800k)
Equity Value             = £7,180k

Implied share price = £7,180k / 985,714 shares = £7.28 per share

Applying the peer 25th–75th percentile range (7.8× to 9.4×):

Low EV    = 7.8 × £950k = £7,410k  →  Equity Value £6,610k  →  Share price £6.71
High EV   = 9.4 × £950k = £8,930k  →  Equity Value £8,130k  →  Share price £8.25

Trading comps implied share price range: £6.71 – £8.25, with a central estimate of £7.28.

It is worth noting that our DCF analysis (from Post 2 in this series) implied a share price of approximately £8.47 — above the midpoint of the comps range. This gap is worth discussing. Possible explanations: the DCF assumptions may be optimistic; WidgetCo may be genuinely undervalued by the market; or the peer group may be pricing in sector-level concerns that do not apply to WidgetCo specifically. The football field does not resolve this tension — it surfaces it.

---

The Football Field

The football field is a horizontal bar chart that places the valuation range implied by each methodology side by side. It is the standard output in an investment banking pitch book whenever multiple valuation methods are used.

For WidgetCo, the full football field looks like this:

Methodology	Low (£/share)	High (£/share)
Trading Comps (25th–75th pct EV/EBITDA)	6.71	8.25
Precedent Transactions	8.00	10.50
DCF (base case ±10% WACC sensitivity)	7.50	11.00

The precedent transactions range sits above trading comps — as it almost always does — because acquisition prices include a control premium. The DCF range is the widest because it is the most sensitive to assumptions. Where the three methodologies overlap is where the market is most likely to converge on a price.

How investment banks use this: in an IPO, the football field anchors the price range. In a sell-side M&A mandate, it sets the floor (trading comps) and ceiling (DCF / precedent transactions) for negotiations. A buyer paying below the trading comps range needs a compelling explanation.

---

Common Mistakes in Trading Comps

Using reported EBITDA without normalising. Many companies include restructuring charges, legal settlements, or stock-based compensation in their reported numbers. These items are real costs, but they distort comparability. Always check the notes to the accounts and adjust for one-offs before building your multiples.

Using basic shares instead of diluted shares. Diluted shares include the impact of options, warrants, and convertible instruments. Using basic shares understates the market cap and therefore the enterprise value — making the multiple look lower than it really is.

Forgetting lease liabilities post-IFRS 16. Since 2019, IFRS-reporting companies capitalise operating leases on the balance sheet. These lease liabilities are financial in nature and belong in net debt. Omitting them understates net debt and understates enterprise value, giving artificially low multiples.

Mixing LTM and NTM without flagging it. If you apply an NTM multiple to LTM EBITDA, you are using a forward-looking price with a backward-looking earnings figure. The implied value will be wrong. Always be explicit about whether each multiple and each financial metric uses the same time reference.

---

Key Takeaways

• Trading comps answer the question: “What is the market currently paying for similar businesses?” — they ground valuation in observable market prices rather than modelled assumptions
• EV/EBITDA is the most widely used multiple in M&A and investment banking; P/E is most common in public equity research
• Peer selection is the most subjective part of a comps analysis — be prepared to defend every inclusion and exclusion
• Always present a range (25th–75th percentile) rather than a single multiple; a point estimate of value implies false precision
• Normalise EBITDA for one-offs, use diluted shares, and include lease liabilities in net debt — these are the three most common mechanical errors
• Use trading comps alongside DCF and precedent transactions in a football field; the three methodologies together tell a much more complete story than any one alone

Practice

Build a trading comps table for the six peer companies in the table above. Calculate EV/Revenue, EV/EBITDA, and P/E for each peer. Compute the mean, median, 25th percentile, and 75th percentile for each multiple. Apply the peer median EV/EBITDA to WidgetCo (EBITDA £950k, net debt £800k, shares outstanding 985,714) to derive an implied share price. Then overlay this with the DCF-implied share price of £8.47 from Post 2 and sketch a simple two-bar football field. Where do the two methodologies agree — and where do they diverge?