The target
Harlow Tools, a private maker of industrial tools.
- LTM EBITDA
- $80 million
- Debt and cash
- $220 million debt, $60 million cash
- Diluted shares
- 40 million
Comparable company analysis values a company at the multiples similar public companies trade at today. Pick the peers, build each one's enterprise value and EBITDA on the same basis, take the median and middle range of EV/EBITDA, apply it to the target's EBITDA, then bridge to equity value and a price per share. In the first example below, peers at 9.5x to 11.0x put an invented target at $15.00 to $18.00 a share. Soreno, an AI voice interview practice app for consulting, investment banking and other finance candidates, runs a free enterprise value calculator with no sign-up, and mock technical interviews where an AI interviewer asks you to walk through comps out loud.
Updated
Every company and number here is invented by Soreno for practice. Amounts are in millions of dollars. Conventions we use, and say so: share prices are today's, balance sheet items are from the latest filing, and the middle range is the 25th to 75th percentile as Excel's QUARTILE.INC computes it.
Harlow Tools, a private maker of industrial tools.
Five listed toolmakers of similar size, growth and margin.
1. Multiples
Divide each peer's enterprise value by its LTM EBITDA. Brantwood is $1,320 million ÷ $120 million = 11.0x, and so on down the list.
2. Summarize
Sorted, the median is the middle value, 10.0x. The 25th and 75th percentiles are 9.5x and 11.0x. The mean is 10.30x, pulled up by Delmar's 12.0x. Lead with the median, since one outlier moves it less.
3. Enterprise value
Apply the range to the target's $80 million of LTM EBITDA: 9.5x × 80 = $760 million, 10.0x × 80 = $800 million, 11.0x × 80 = $880 million.
4. Equity value
Net debt is $220 million of debt less $60 million of cash, $160 million. Equity value runs $600 million to $720 million, $640 million at the median.
5. Per share
Divide by 40 million diluted shares: $15.00 to $18.00, $16.00 at the median. Say it as a range and name the multiple behind each end.
Brantwood Machine's fiscal year ends June 30, so its annual report is six months old. Its share price is $40.00. It has 27 million basic shares, 2 million in-the-money options with a $25.00 strike, 0.25 million unvested restricted stock units, $340 million of debt, $160 million of cash and $20 million of noncontrolling interest. EBITDA was $112 million for the year to June 30, $63 million for the six months to December 31 this year and $55 million for the same six months last year. No preferred stock, no one-offs.
1. Options
Treasury stock method: exercising the options brings in 2 × $25.00 = $50 million, which buys back $50 million ÷ $40.00 = 1.25 million shares. Net new shares: 2 − 1.25 = 0.75 million.
The treasury stock method, step by step2. Diluted shares
27 basic + 0.75 from options + 0.25 RSUs = 28 million. Counting RSUs in full is a common convention in comps. Basic shares alone would understate equity value by $40 million.
3. Enterprise value
Equity value 28 million × $40.00 = $1,120 million. Add $340 million of debt and $20 million of noncontrolling interest, take off $160 million of cash: $1,320 million. Noncontrolling interest goes in because Brantwood's consolidated EBITDA includes the whole of the part it doesn't fully own.
4. LTM EBITDA
Last twelve months = the last full fiscal year, plus the year-to-date period, less the same period a year earlier. That lines Brantwood up with peers whose years end in December.
5. Multiple
$1,320 million ÷ $120 million = 11.0x, the figure in example 1's first row. Using the six-month-old annual EBITDA instead would give 11.8x.
A sixth candidate, Gresham Precision, reports under IFRS. A data screen shows 5.85x: enterprise value $1,200 million (market value of equity $1,000 million + borrowings $250 million − cash $50 million) over the company's own "adjusted EBITDA" of $205 million. The notes say adjusted EBITDA adds back a $15 million restructuring charge, and Gresham has booked one every year for five years. EBITDA also includes a $10 million gain on selling a plant. Gresham has $300 million of lease liabilities, and its yearly lease cost (depreciation of the leased assets plus interest on the lease liabilities) is $40 million. The five US peers report under US GAAP and their leases are operating leases, so a single lease cost sits inside their EBITDA. By our convention, common in practice, their lease liabilities stay out of enterprise value. Using Gresham's $40 million as the cost a US filer would book is our simplification.
1. Restructuring
A charge that comes every year is a cost of running the business. $205 million − $15 million = $190 million. The SEC staff's own guidance treats adjusted measures that drop normal, recurring cash operating expenses as possibly misleading.
2. One-off gain
Selling a plant won't happen again, so it doesn't belong in a recurring profit figure: $190 million − $10 million = $180 million.
3. Leases
FASB's summary of the US lease standard says IFRS 16 accounts for all leases the way US GAAP treats finance leases, where interest is recognized separately from amortization of the leased asset. Both sit below EBITDA. A US operating lease instead books a single lease cost. To match the US peers, take the $40 million off: $180 million − $40 million = $140 million, and keep lease liabilities out of enterprise value. The multiple is $1,200 million ÷ $140 million = 8.57x.
4. Check
Leave the lease cost out of EBITDA and add the lease liabilities to enterprise value instead: ($1,200 million + $300 million) ÷ $180 million = 8.33x. Either consistent basis puts Gresham between 8.3x and 8.6x. Only the mixed one makes it look cheap. The two consistent answers differ a little because Gresham's lease liabilities are 7.5x its lease cost, not 8.6x.
5. What it changes
Added at the screen's 5.85x, Gresham drags the six-peer mean to 9.56x. Cleaned, the mean is 10.01x. The median is 9.75x either way, which is one reason to lead with it. Fix the number anyway: in a smaller set, one bad peer moves the range.
From SEC staff: non-GAAP financial measures interpretations, IFRS Foundation: IFRS 16 Leases and FASB: ASU 2016-02, Leases (Topic 842).
In our experience the question starts as a walk-through and turns into follow-ups on judgment: which peers, which multiple, what you cleaned and why.
Pick peers, gather the numbers, compute multiples, apply the median and range to the target, then bridge to equity value and a price per share.
Enterprise value pairs with revenue, EBITDA or EBIT, which belong to all capital providers. Equity value pairs with net income. A mismatch is the most common slip we hear.
Same industry and business model first, then size, growth, margins and geography. Leave out a company with a takeover bid on it, since its price includes a premium.
Comps give what the market pays today for similar companies, without a control premium. Say how it sits against precedent transactions and a DCF.
The short answers are ours. Under each is what the rest of a strong answer covers and the follow-ups we would expect.
Walk me through a comparable company analysis.
Short answer. Pick a set of public companies like the target. Calculate each one's enterprise value and equity value, and its revenue, EBITDA and earnings for the same periods. Compute the multiples, take the median and the middle range, apply them to the target's numbers, and bridge from enterprise value to equity value and a value per share.
Names the multiple you would lead with for this industry, and ends on a range rather than one number. Practice it, see a model answer
How do you pick comparable companies?
Short answer. Start with the same industry and business model, then narrow by size, growth, margins and where they sell. Five to ten good peers beat twenty loose ones. Drop companies with a pending takeover or a distorted year.
Explains a trade-off: a tight set may be too small, a loose one compares unlike businesses. Practice it, see a model answer
When would you use EV/EBITDA instead of P/E?
Short answer. When peers carry different amounts of debt or different tax and depreciation profiles. EV/EBITDA looks at the whole business before financing. P/E reflects leverage and tax, so it compares well only among companies with similar capital structures.
Adds where P/E is still the norm, such as banks and insurers. Practice it, see a model answer
Why use LTM and forward multiples, and which do you show?
Short answer. LTM uses reported numbers, so it is factual but backward looking. Forward multiples use analyst estimates of next year, which matter more for a growing company. Most comps tables show both.
Says how to calendarize a peer whose fiscal year ends in June, as in example 2.
What is the difference between enterprise value and equity value?
Short answer. Equity value is what the shares are worth. Enterprise value is the value of the operating business to all capital providers: equity value plus debt, preferred stock and noncontrolling interest, less cash.
Links each to the multiples that use it. Practice it, see a model answer
How do you calculate diluted shares with the treasury stock method?
Short answer. Assume in-the-money options are exercised and the company uses the cash to buy back shares at today's price. Add the net new shares to the basic count, along with restricted stock units.
Works one number through, like Brantwood's options in example 2. Practice it, see a model answer
Why do you add noncontrolling interest to enterprise value?
Short answer. Because the parent consolidates 100% of a subsidiary's EBITDA even when it owns less than all of it. Adding the outside owners' stake keeps the numerator on the same basis as the denominator.
Mentions the matching problem in reverse for equity method investments. Practice it, see a model answer
A peer reports under IFRS and the others under US GAAP. What do you change?
Short answer. Leases. Both standards put leases on the balance sheet, but under IFRS 16 the cost shows up as depreciation and interest, below EBITDA, while a US operating lease books a single lease cost inside EBITDA. Put every peer on one basis before comparing multiples.
Shows both consistent fixes and why the mixed version flatters the IFRS peer.
Two companies have the same earnings but trade at different P/E multiples. Why?
Short answer. Faster expected growth, lower risk, higher returns on capital or a different amount of debt. The market pays more per dollar of earnings it trusts to grow.
Turns it into what that means for picking peers. Practice it, see a model answer
Which valuation method usually gives the highest value?
Short answer. Usually precedent transactions, because deal prices include a control premium. Comps reflect minority trading prices. A DCF can land anywhere, depending on its assumptions.
Uses words like usually, and gives a case where comps come out higher, such as a hot sector. Practice it, see a model answer
What multiples would you use for a bank, a REIT and a software company?
Short answer. Price to book or P/E for a bank, where debt is the raw material. Price to funds from operations for a REIT. EV/revenue or EV/EBITDA for software, depending on whether it is profitable yet.
Explains why EV/EBITDA breaks for banks. Practice it, see a model answer
How would you show comps next to the other methods?
Short answer. On a football field chart: one bar per method, each the range of implied values, with the current price or offer as a line across.
Says which bars you would trust most for this company and why. Practice it, see a model answer
EBITDA belongs to lenders and shareholders, so it needs enterprise value on top. Equity value over EBITDA mixes two groups of owners.
Options and RSUs are claims on the equity. In example 2 basic shares miss $40 million of Brantwood's equity value.
A June year end next to December year ends, or one peer on LTM and another on a forecast. Calendarize first.
Read what was added back. A charge that recurs every year stays in.
An IFRS peer's EBITDA excludes lease cost. Compare it with US peers only after putting both on one basis.
Comps give a range. Lead with the median and give the 25th to 75th percentile, then say where in it the target belongs and why.
A loss-making peer has no meaningful EV/EBITDA. Show it as not meaningful, and use revenue multiples if the set is full of them.
Soreno's AI interviewer asks the walk-through, then the follow-ups: why that peer, why the median, what you would do with an IFRS reporter. You get a scored report on what you said.
A relative valuation method. It values a company at the multiples, such as EV/EBITDA or P/E, that similar public companies trade at today.
Yes. Bankers say trading comps, or just comps, to tell it apart from transaction comps, which are precedent transactions.
In our experience five to ten is common. Fewer than four makes the median fragile. More than fifteen usually means some peers aren't very similar.
The median, as a rule, because one outlier moves it less. In example 3 a bad peer moves the mean from 10.01x to 9.56x and leaves the median at 9.75x.
No. The SEC staff covers EBITDA in its interpretations on non-GAAP measures. Earnings in EBITDA means GAAP net income, a company that shows EBITDA as a performance measure should reconcile it to net income, and a measure calculated another way should carry a different name, such as Adjusted EBITDA.
No. Share prices reflect trading in small stakes. To value a sale of the whole company, bankers look at precedent transactions, where the prices paid include a premium.
The same way, using its EBITDA and public peers, then bridge with its own net debt. Some practitioners apply a discount for its size or for shares that can't be sold easily, which is a judgment, not a formula.