Is a Stock Overvalued or Undervalued? 8 Valuation Checks

A stock's price tells you what people are paying. It doesn't tell you what the company is worth. The whole art of valuation is closing that gap — figuring out whether the price in front of you is reasonable, rich, or a bargain hiding in plain sight.

If you've ever wanted to know how to tell if a stock is overvalued (or undervalued), the honest starting point is this: there's no single number that settles it. A stock trading at 40 times earnings can be perfectly fair, and one at 8 times can be a trap. Valuation is a set of checks read together, against the right context — not a verdict you read off one ratio.

This guide walks through eight of those checks, in plain language. By the end you'll know what each metric measures, what it quietly ignores, and — the part most people skip — why a "cheap" stock is not the same thing as a good one.

Before you start: this is an educational framework, not investment advice. Nothing here tells you a stock is a buy or a sell. The goal is to help you judge price against value and know how much to trust your own conclusion.

Why one ratio is never enough#

Every valuation metric is a shortcut. It compresses a whole business — its growth, margins, debt, risk, and durability — into a single number so you can compare it to something else. That compression is useful, but it always throws away information.

The P/E ratio ignores debt. Price-to-sales ignores whether the company makes any profit at all. EV/EBITDA flatters businesses with heavy real capital costs. Each metric is blind in a different direction, which is exactly why you use several. When four different lenses all point the same way, that's a signal worth noticing. When they disagree, that disagreement is telling you something about the business — and it's usually the most interesting part of the analysis.

So treat what follows as a panel, not a menu. You're not looking for the one metric that gives you the answer you like. You're triangulating. If a term is new, keep the Valarn glossary open in another tab, and if you want to see how these checks look applied to a real company, skim a complete sample research report first.

Check 1 — The P/E ratio: price versus earnings#

The price-to-earnings ratio is the most quoted valuation number in existence, and the most misunderstood. It's simple arithmetic: the share price divided by earnings per share (EPS) — the company's annual profit split across every share outstanding.

A stock at $100 with $5 of EPS trades at a P/E of 20. Read plainly, that means you're paying $20 for every $1 of current annual profit. Flip it around and it's an earnings yield of 5% (1 ÷ 20), which makes it easier to compare against other places you could park money.

The number itself means nothing in isolation. A P/E of 20 is unremarkable for a steady, profitable business, expensive for one that's shrinking, and cheap for one growing earnings 30% a year. That's the whole trap: people hear "high P/E" and think "expensive," but a high multiple is often the market correctly pricing in fast future growth.

• Watch for: whether the "E" is trailing (last 12 months, factual) or forward (an analyst estimate, a guess). They can differ wildly.
• Watch for: one-time items — a legal settlement or asset sale — that distort a single year's earnings and make the ratio look artificially high or low.
• Watch for: companies with no earnings at all, where P/E simply doesn't compute. That's where the next check comes in.

Check 2 — Price-to-sales, for when there's no profit yet#

Plenty of real companies — early-stage growth businesses, or established ones passing through a rough patch — have little or no net income. P/E breaks down entirely there. Price-to-sales (P/S) steps in: the company's market value divided by its annual revenue.

Because revenue sits at the top of the income statement, before any costs, P/S is much harder to distort than earnings. That's its strength and its weakness. It works when profits are temporarily absent or noisy. But it's also the crudest yardstick, because it says nothing about whether those sales ever turn into cash. A dollar of revenue at a software company that keeps most of it as profit is worth far more than a dollar at a grocery chain running on thin margins.

The rule of thumb: never compare P/S across different business models. A 10x price-to-sales multiple might be ordinary for high-margin software and absurd for a low-margin retailer. Use it to compare a company against genuinely similar peers, or against its own past — and always ask what margin those sales carry before you conclude anything.

Check 3 — Free-cash-flow yield: what the business actually generates#

Earnings are an accounting opinion; cash is a fact. That's why many careful investors anchor on free cash flow — the cash a business produces after paying to maintain and grow itself (operating cash flow minus capital expenditures).

Free-cash-flow yield turns it into a valuation gauge: free cash flow divided by the company's market value, expressed as a percent. A company generating $2 billion of free cash flow with a $50 billion market value has a 4% FCF yield ($2B ÷ $50B). The higher the yield, the more cash you're getting relative to the price — and because it's grounded in real cash rather than adjustable earnings, it's one of the harder numbers to fake.

• Watch for: yields that look great only because the company stopped investing in itself. Slashing capital spending inflates free cash flow today and hollows out the business tomorrow.
• Watch for: lumpy cash flows. Some businesses swing year to year, so look at a multi-year average rather than one snapshot.

Free cash flow is one of the most abused terms in investing — sources define it differently — so check which version a number refers to before you trust it. The free cash flow glossary entry breaks down the variations.

Check 4 — EV/EBITDA: valuation that accounts for debt#

Here's a blind spot in P/E: it ignores debt entirely. Two companies can have identical earnings and identical P/Es, but if one is drowning in borrowings and the other has cash to spare, they are not equally priced. Enterprise value fixes this by adding debt and subtracting cash from the market value — it's roughly what it would cost to buy the whole business, debts and all.

EV/EBITDA compares that to EBITDA (earnings before interest, taxes, depreciation, and amortization — a rough proxy for operating cash generation). Because it puts debt-heavy and debt-light companies on the same footing, it's the multiple professionals reach for when comparing businesses with very different balance sheets, and it's popular for capital-intensive industries.

The catch: EBITDA deliberately ignores real costs. Depreciation represents actual machinery and equipment wearing out and needing replacement. For a business that eats capital — factories, fleets, networks — EBITDA can paint a rosier picture than the cash reality. Warren Buffett has been famously blunt that EBITDA can mislead precisely because it pretends those costs don't exist. Use it alongside free-cash-flow yield, and the two together tell you more than either alone.

Check 5 — Growth relative to the multiple#

This is the check that resolves most "is it expensive?" arguments, because it connects price to the thing price is supposed to reflect: future growth.

A multiple only makes sense next to a growth rate. Paying 30 times earnings for a company growing profits 5% a year is very different from paying 30 times for one growing 40% a year — the second may be reasonable, the first probably isn't. One quick sanity tool is comparing the P/E to the earnings growth rate (sometimes formalized as the PEG ratio, P/E divided by growth). Loosely, a P/E of 20 against 20% growth is more defensible than a P/E of 20 against 5% growth.

Don't treat any such rule mechanically — growth can decelerate, and estimates are frequently wrong — but the underlying logic is sound: a premium multiple has to be earned by durable growth and strong margins. When you see a rich valuation, the question isn't "is this too high?" It's "what growth is baked into this price, and is that plausible given everything else I know about the business?"

• Watch for: revenue growth and earnings growth. A company can grow sales fast while profits stall, which quietly undermines the multiple.
• Watch for: growth that leans on one product, one customer, or one region — more fragile than the headline rate suggests.

Check 6 — The stock's own historical valuation range#

Every stock trades within its own personality. Some businesses have commanded a P/E in the mid-20s for a decade; others rarely leave the low teens. Pulling up where a company's multiples sit today versus its own five- or ten-year history is one of the fastest reality checks available.

If a stock normally trades around 18 times earnings and now sits at 30, something has to justify the re-rating — accelerating growth, a new product cycle, expanding margins. If nothing has, you may be looking at enthusiasm rather than value. Equally, a stock trading well below its historical range might be genuinely mispriced… or the market may have decided its best days are behind it. History gives you the reference point; it doesn't tell you which of those two stories is true.

That's the discipline here: the historical range frames the question, it doesn't answer it. A multiple can look "cheap versus its own past" for the entirely rational reason that the business has gotten worse. Which leads directly to the most important idea in this whole guide.

Check 7 — Valuation versus peers (and the value-trap warning)#

No multiple means anything until you compare it. The natural comparison set is a company's close competitors — businesses with similar economics, growth, and risk. If a company trades at 12 times earnings while comparable peers sit at 20, that gap is worth understanding.

But here's the nuance that separates careful investors from bag-holders: a low valuation is not automatically "cheap." Sometimes a stock trades at a discount because the market has correctly figured out that its fundamentals are deteriorating — falling market share, shrinking margins, a dying product, mounting debt. The low multiple isn't an opportunity the market missed; it's the market pricing in decline. That's a value trap: a stock that looks cheap on the numbers and keeps getting cheaper because the business really is getting worse.

The only way to tell a bargain from a trap is to go back to the fundamentals. A genuinely undervalued stock is cheap despite a stable or improving business. A value trap is cheap because the business is decaying, and the discount is a warning, not a discount. Ask why the gap exists before you assume you've found value the whole market overlooked.

• Look for: a discount to peers that the fundamentals don't justify — steady revenue, healthy margins, manageable debt.
• Be wary of: a "cheap" multiple paired with falling sales, eroding margins, or a shrinking market. That's the market being right, not wrong.

Peer comparison is a skill in its own right; our guide on how to compare two stocks walks through building a fair comparison set, and the deeper you go, the more this becomes a bull-versus-bear stress test rather than a single number.

Check 8 — How interest rates change the whole board#

Valuations don't happen in a vacuum. They move with interest rates — and understanding why explains a lot of otherwise baffling market swings.

A stock is worth the future cash it will generate, valued in today's money. When interest rates are low, safe alternatives like government bonds pay little, so investors accept lower yields on stocks and are willing to pay higher multiples. When rates rise, those safe alternatives suddenly pay more, the "discount rate" applied to future cash goes up, and the same future earnings are worth less today. Higher rates tend to compress valuation multiples across the board — no company's fundamentals changed, but what investors will pay for them did.

This hits growth stocks hardest, because most of their value sits in profits expected years out, and distant cash is discounted most heavily when rates climb. It's why a whole market can re-rate lower even as the underlying businesses keep performing.

The practical takeaway isn't to predict rates — nobody reliably does. It's to recognize that the multiple you're paying today is partly a product of the rate environment, and that environment can change. A valuation that looks reasonable in a low-rate world can look stretched in a higher-rate one, without a single thing changing at the company itself.

Putting the eight together#

Notice what none of these checks did: hand you a verdict. Each one answers a narrow question, and the judgment lives in reading them together against the business you've already studied. A stock can look expensive on P/E, fair on EV/EBITDA, cheap on free-cash-flow yield, and reasonable versus history — and your job is to understand why those disagree, not to average them into a single score.

This is exactly the kind of multi-angle work that's tedious to do by hand — pulling the multiples, building the peer set, checking the history, factoring in growth and rates. It's also the work that a full research process is built to run for you. As an educational research tool, Valarn convenes up to about 25 specialist AI analysts — including dedicated valuation, financial-quality, and peer-comparison analysts — across categories like core research, market structure, financial quality, and macro. They stage a structured bull-versus-bear debate, then synthesize a single research view (Bullish, Cautious Bullish, Neutral, Cautious, or Bearish — never a buy or sell instruction).

Instead of inventing a single price target, it reports a Scenario Range (bear, base, and bull reference levels), a Reference Price, and a Risk Level — and it shows third-party Wall Street consensus separately, labeled as an external fact rather than its own view. Every claim is traceable to a filing or licensed source with an as-of date, a quality gate runs before you see it, and two 0–100 scores travel with the report: a confidence score (how complete and reliable the data is — not a price prediction) and an agreement score (how much the analysts converged). You can start from a company research page for any ticker, and this valuation work slots into the broader 12-step research checklist as just one piece of a fuller picture.

The bottom line#

There is no magic number that tells you a stock is overvalued or undervalued. There's a panel of checks — P/E, price-to-sales, free-cash-flow yield, EV/EBITDA, growth relative to the multiple, the stock's own history, its peers, and the interest-rate backdrop — that only make sense read together, against the fundamentals of the actual business.

And the single most valuable habit is the skeptical one: a low valuation is not automatically cheap. Sometimes the market is discounting a stock for a very good reason. Learning to tell a bargain from a value trap — cheap despite the business versus cheap because of it — is most of what separates real valuation work from staring at a low number and hoping.

Want to see all eight checks applied to a real company? Explore a full sample report or run your own free research report and read the valuation section for 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.