Comparable Company Analysis: How to Value a Stock Against Its Peers

Ask five people what a stock is "worth" and you'll get five answers — and most of them are really just a guess dressed up with a number. Comparable company analysis is the discipline that turns that guess into something you can defend: instead of valuing a business in isolation, you price it the way the market prices everything else in its neighborhood.

The logic is simple. If similar companies — same industry, same rough size, same growth and risk profile — all trade at roughly the same multiple of their earnings or cash flow, then a stock trading far above or below that pack is telling you something. Either the market sees something you don't, or there's a gap worth understanding. Comparable company analysis (often just called "comps" or relative valuation) is how you find that gap.

It's also the most misused tool in valuation, because it's so easy to do badly. Pick the wrong peers, grab the wrong multiple, ignore the differences that actually matter, and you get a precise-looking number built on sand. This guide walks through doing it properly — and, just as important, where even a perfect comp set can quietly lead you astray.

What comparable company analysis actually measures#

Comps answer one question: what is the market willing to pay for a dollar of this company's earnings, cash flow, or sales — and how does that compare to what it pays for similar businesses?

You're not calculating an intrinsic value from scratch the way a discounted-cash-flow model does. You're reading a relative price. A comp doesn't tell you a stock is cheap in some absolute, cosmic sense — it tells you the stock is cheap or expensive relative to a specific group of peers, right now, at today's prices. That distinction is the whole ballgame, and it's the source of both the method's power and its biggest blind spot.

Done well, relative valuation is fast, transparent, and grounded in real market prices rather than a spreadsheet's assumptions. Done badly, it launders a bad peer group and a mismatched multiple into a confident conclusion. The four steps below are the difference.

Step 1 — Build a true peer set#

Everything downstream depends on this. A comparable company analysis is only as honest as its peer group, and the most common mistake is a group that's comparable in name only.

A "true" peer isn't just another company in the same broad sector. You're looking for businesses that share the characteristics that actually drive valuation:

• Business model — how the company makes money. A software company that sells subscriptions and one that sells one-time licenses are not comps, even under the same industry label.
• Size — a $200B mega-cap and a $2B small-cap in the same industry face different growth ceilings, capital access, and risk. Market cap and revenue in the same ballpark matter.
• Growth profile — a business growing 30% a year and one growing 3% will never trade at the same multiple, and shouldn't.
• Margins and profitability — a high-margin business and a thin-margin one are different animals financially, even selling similar products.
• Geography and end markets — where revenue comes from changes the risk and the growth story.

Look for: four to eight genuinely similar businesses. Fewer than that and one outlier distorts everything; far more and you're diluting the set with loose matches to pad the list. Be wary of: the "everyone in the index" peer group, or cherry-picking peers that happen to make your stock look cheap. If you find yourself excluding the obvious competitor because it "doesn't fit," ask whether it doesn't fit or whether it just doesn't help your case.

A quick sanity check: could you defend each peer to a skeptic in one sentence? If not, it doesn't belong. When you're orienting on a specific name, a company research page is a fast way to see who a business genuinely competes with before you commit to a list. And if you're really weighing two names head-to-head, our guide on how to compare two stocks drills into the like-for-like discipline this step depends on.

Step 2 — Pick the right multiples for the sector#

Once you have a peer set, you need the yardstick to measure them by. Not all multiples fit all businesses, and using the wrong one is how comps go sideways. Here are the workhorses and where each earns its keep.

Price-to-earnings (P/E) — share price divided by earnings per share. The most quoted multiple, and useful for stable, profitable companies. Its weakness: it's based on net income, which sits at the bottom of the income statement after interest, taxes, and every accounting choice — so it's easily distorted, and it's useless for a company with no earnings.

EV/EBITDA — enterprise value divided by earnings before interest, taxes, depreciation, and amortization. This is the analyst's default for a reason. Because it uses enterprise value (market cap plus debt, minus cash) rather than just share price, it compares companies on an apples-to-apples basis regardless of how they're financed. That matters when peers carry very different debt loads. If the distinction between EV and market cap is fuzzy, our explainer on enterprise value versus market cap is worth a detour — it's foundational to almost every serious comp.

EV/Sales — enterprise value divided by revenue. When a company isn't yet profitable (young growth companies, turnarounds), earnings-based multiples break down and you fall back to revenue. It's a blunter tool — a dollar of high-margin recurring revenue is worth far more than a dollar of low-margin one-time revenue — so it only means something within a group of similar-margin businesses.

Price-to-book (P/B) — share price divided by book value per share. Built for businesses where the balance sheet is the business: banks, insurers, and other financials whose assets are marked close to market value. For an asset-light software company, book value is nearly meaningless, and P/B tells you almost nothing.

A rough map of which multiple tends to fit where:

Treat this as a starting point, not a law. The right move is usually to look at two or three multiples together — if a stock looks cheap on one and expensive on another, that tension is itself information. For a sector-specific walk-through, how to analyze SaaS stocks shows why revenue multiples carry the load where earnings haven't arrived yet.

Step 3 — Adjust for growth, margins, and risk#

Here's where amateur comps and real ones part ways. Two companies trading at "12x EBITDA" are not equally priced if one is growing twice as fast, earns higher margins, and carries less debt. The raw multiple is the starting point; the reason for any gap is the actual analysis.

A higher multiple is often completely justified. Faster growth, fatter margins, more durable competitive advantages, and lower risk all earn a premium — and should. The mistake is seeing a stock trade above its peers and concluding it's "expensive" without asking whether the premium is earned. The reverse trap is just as common: a stock trades at a discount and looks like a bargain, when the discount is the market correctly pricing in slower growth, thinner margins, or a real risk you haven't spotted. That's the classic setup for a value trap.

The single most important adjustment is for growth. This is why the PEG ratio exists — the P/E ratio divided by the earnings growth rate — as a rough way to put a fast grower and a slow grower on comparable footing. It's imperfect, but it captures the core idea: a 30x P/E on a company growing 30% is a very different proposition from 30x on a company growing 5%. Our PEG ratio explainer covers what it does and doesn't capture.

A quick, clearly illustrative example to make the mechanics concrete (round numbers, not a real company):

Suppose your peer group trades at a median EV/EBITDA of 12x. Your target company generates $500 million of EBITDA. Applying the peer median implies an enterprise value of 12 × $500M = $6.0 billion. But the target is growing EBITDA at 20% a year versus 8% for the peers, and earns higher margins. That faster growth and quality is a reason it might reasonably trade above 12x — so a straight median-multiple read would understate it. If instead the target were the slowest grower with the weakest margins, the same 12x would probably be too generous.

The number falls out of arithmetic in seconds. The judgment — how much premium or discount the differences justify — is the part that takes real work, and the part no formula hands you.

Look for: a clear, defensible reason for every gap between your stock and its peers. Be wary of: applying a peer median mechanically and calling the difference "upside" or "downside," as if growth, margins, and risk were all identical.

Step 4 — Read where the stock trades versus peers#

Now you put it together. Line up your stock's multiples against the peer group — typically the median and the range — and see where it sits. Then interpret, using everything from Step 3.

• Trading at a discount to peers? Ask why. Is the market missing something (a genuine opportunity), or pricing in something real you should respect (slower growth, a legal overhang, a structurally weaker position)?
• Trading at a premium? Same question in reverse. Is the premium earned by superior growth, margins, and durability — or is it enthusiasm that outran the fundamentals?
• Trading roughly in line? Then relative valuation isn't your edge here, and any thesis has to come from believing the whole group is mispriced, or from something comps don't capture.

Comps are strongest as a screening and framing tool — they tell you where a stock sits and what question to ask next. They're weakest when treated as a verdict. The output of good relative valuation isn't "buy because it's cheap"; it's "this trades at a discount to peers, and here are the three things that would tell me whether that discount is an opportunity or a warning." For the broader toolkit around that judgment, how to tell if a stock is overvalued or undervalued pulls comps together with the other valuation lenses.

The limitation nobody mentions: the whole group can be wrong#

Here's the flaw baked into the method, and you have to hold it in mind the entire time: comparable company analysis only tells you if a stock is cheap or expensive relative to its peers — not whether the peers themselves are sanely priced.

If an entire sector is caught up in a bubble, every company in it can look "reasonably valued" relative to the others while the whole group is detached from reality. Relative valuation would have called plenty of dot-com stocks "fairly priced against peers" in 1999 — right up until the peers collapsed together. The method is structurally blind to a mispricing that afflicts the entire comp set, because the comp set is its reference point.

This is why comps should never be your only valuation lens. Pair relative valuation with an absolute method — a discounted-cash-flow view, a look at the stock versus its own historical multiples, a hard look at whether the sector's growth assumptions are even plausible. When comps and an absolute approach disagree, that disagreement is a gift: it's telling you the peer group might be the thing that's mispriced. Two anchors are much harder to fool than one.

Be wary of: any comp conclusion that would have looked identical at the top of a sector bubble and the bottom of a crash. If your method can't tell those two moments apart, it's not measuring what you think it's measuring.

Where a research desk fits in#

Running proper comparable company analysis by hand is genuinely time-consuming — assembling a defensible peer set, pulling clean and current financials for each, normalizing for one-time items, choosing the right multiples per business, and then doing the qualitative work of explaining every gap. That last part is where most DIY comps quietly break down, because it's judgment, not arithmetic.

This is a slice of what Valarn is built to handle as an educational research tool. Rather than one AI confidently declaring a stock "undervalued," it convenes up to about 25 specialist analysts across areas like fundamentals, valuation, financial quality, cash flow, ownership, sentiment, and sector context — with a dedicated valuation analyst building the relative-valuation picture and a risk committee stress-testing it. The analysts run a structured bull-versus-bear debate and only then synthesize a single neutral research view — Bullish, Cautious Bullish, Neutral, Cautious, or Bearish — never a buy or sell instruction.

Crucially, the multiples and peer comparisons are traceable to filings and licensed data with an as-of date, so you can check the inputs rather than trust the tone. Instead of a single price target, a report frames a Scenario Range (bear, base, bull) with a Reference Price and a Risk Level — a range you can reason about. And two 0–100 scores travel with it: a confidence score reflecting the quality of the underlying data (not a price prediction), and an agreement score showing how much the analysts actually converged. You can see the whole shape of that in a complete sample research report.

Just remember what a comp is and isn't. Multiples like EV/EBITDA and P/E are descriptive screens — they measure how a stock is priced relative to peers today. They don't predict where the price goes next, and no honest process pretends otherwise. Whether the tool is a spreadsheet or a research platform, comps frame better questions; they don't hand you a verdict.

The bottom line#

Comparable company analysis is one of the most useful tools in valuation and one of the easiest to fool yourself with. Get the peer set honest, pick multiples that fit the business, adjust rigorously for growth, margins, and risk, and read where a stock sits versus its pack — and you have a fast, transparent, market-grounded read on relative price. Skip the judgment, or forget that the whole group can be mispriced together, and you've built a confident-looking number on a shaky foundation.

The multiple is where the analysis starts, never where it ends. The real work is explaining the gap — and staying honest about the one thing comps can never see. Want to watch relative valuation run as part of a full research process? Explore a sample report or generate a free one and grade the comps yourself.

Valarn is an educational research tool, not investment advice. It does not tell you to buy, sell, or hold anything, and nothing here is a recommendation or a promise of results. Always do your own research and consider consulting a licensed financial professional.