Sector-Relative Valuation: Why You Should Compare P/E to the Sector, Not the Market

Ask most investors what counts as a "cheap" stock and you'll get a number: P/E under 15, maybe under 20 if they're feeling generous. It's an intuitive rule, and it's also the single most common valuation mistake retail investors make. A price-to-earnings ratio only means something in context — and the context that matters most is the sector the company competes in, not the market as a whole.

This article explains why a single, market-wide P/E cutoff misleads, what sector-relative valuation fixes, and where the method itself breaks down.

Why one P/E cutoff can't work for the whole market

The P/E ratio is a distilled version of a discounted cash flow: price divided by earnings compresses a company's growth rate, capital intensity, risk profile, and cost of capital into one number. Two companies can have wildly different earnings quality, growth trajectories, and balance-sheet risk and still land on the same P/E by coincidence — or, more commonly, two companies in different industries can have very different "correct" P/Es for reasons that have nothing to do with mispricing.

Consider two real structural categories:

• Software companies tend to carry high gross margins (often 70%+), recurring subscription revenue, low incremental capital needs, and — when the business model works — high returns on invested capital. Markets are willing to pay a premium multiple for a dollar of earnings that's more durable and more scalable than a typical dollar of industrial earnings.
• Utilities operate under rate regulation that caps how much they're allowed to earn on their asset base. Growth is slow and predictable, cash flows resemble a bond coupon more than a growth stock, and the multiple the market assigns reflects that: steady, capped, and bond-like.

A 25x P/E software stock and a 25x P/E utility are not "equally valued" in any useful sense. The software stock, at 25x, might be trading at a discount to how software companies typically get priced. The utility, at the same 25x, might be trading at a significant premium to how utilities typically get priced. The absolute number is identical; the signal it carries is close to opposite.

This is why a market-wide screen — "show me everything under 15x P/E" — tends to return the same handful of structurally low-multiple sectors over and over: banks, integrated energy, auto manufacturers, some insurers. It isn't finding mispriced stocks. It's finding sectors that always trade at low multiples because of their underlying economics — leverage, cyclicality, capital intensity, regulatory caps. Running the filter in reverse ("P/E over 40") just as reliably surfaces software, semiconductors, and biotech, sectors that structurally command a premium. Calling one group "cheap" and the other "expensive" based on the raw number alone measures the sector, not the opportunity.

What sector-relative valuation fixes

Sector-relative valuation reframes the question from "is this stock cheap?" to "is this stock cheap relative to the businesses it actually competes with for capital?" That's a narrower, more answerable question, and it controls for the biggest driver of multiple differences before looking for a real signal.

The mechanics are simple:

1. Group. Assign each stock to its sector (most commonly using a standard classification like GICS, at either the sector or industry-group level).

2. Find the sector's typical multiple. Compute the median P/E across the sector's constituents — a median rather than a mean, since P/E distributions are skewed by extreme values (loss-making companies produce meaningless or negative P/Es, and a handful of high-flyers can pull an average far from where most peers actually trade).

3. Compute the discount or premium. Compare the individual stock's P/E to that sector median:

   relative discount = (stock P/E − sector median P/E) / sector median P/E

   A negative number means the stock trades below its sector's typical multiple; a positive number means it trades above.

That single relative number does something an absolute P/E cutoff structurally cannot: it separates "this stock is cheap because its sector is cheap" from "this stock is cheap relative to the companies it's actually being compared against by the market."

A worked example

Assume — these are illustrative, typical levels, not a live quote — that the technology sector's median forward P/E currently sits around 34x, while the utilities sector's median sits around 17x. Now look at a single stock trading at a 25x P/E in each context.

The absolute multiple didn't move. The read on it flipped entirely, because the reference point changed. A 25x P/E tech company would sit comfortably below where its peer group typically trades — worth a closer look for why the market is discounting it. A 25x P/E utility would sit well above where its peers typically trade — worth asking what's driving the premium, and whether it's justified.

This is the core argument for sector-relative valuation: the absolute number is nearly meaningless in isolation. The relative number, computed against the right peer group, is where an actual signal starts to show up.

Limitations — where sector-relative valuation breaks down

Sector-relative valuation is a real improvement over an absolute cutoff, but it isn't a free lunch. A few places it can mislead:

• Sector definitions are imperfect. Standard classifications draw boundaries that don't always match economic reality. A company can be formally classified in one sector while its actual business model, margin structure, and competitive dynamics look like another sector entirely. Comparing it to its formal peer group in that case understates or overstates the "true" discount.
• A handful of mega-caps can distort the median. In sectors where a few giant companies dominate total market cap, their multiples — and the sentiment attached to them specifically — can pull the reference point away from what a typical mid-sized peer looks like. Trimming extreme values before computing the median helps, but doesn't eliminate the effect entirely.
• Cyclical earnings distort the ratio, not just the sector. A cyclical company (commodities, autos, homebuilders) can show a deceptively low P/E right before earnings roll over, because the "E" is a peak-cycle number about to fall — and a deceptively high P/E right after a trough, because the "E" is temporarily depressed. Sector-relative comparison doesn't fix this on its own; it needs to be paired with some sense of where a company sits in its own earnings cycle, not just where it sits versus peers today.
• An entire sector can be mispriced. Sector-relative valuation assumes the peer group, on average, is priced sensibly — and just measures deviation from that average. If the whole sector is caught in a bubble or a structural rerating, "cheap relative to sector" can still mean "expensive in any absolute or historical sense." A famous case: in 1999, plenty of internet stocks looked "cheap" relative to their internet-stock peers. The peer group itself was the problem.

None of these limitations argue for going back to an absolute cutoff — they argue for treating sector-relative valuation as one well-constructed input, not a standalone verdict.

How to apply it

A workable checklist for using sector-relative valuation in practice:

• Use a standard, consistently-applied sector classification rather than an ad hoc grouping.
• Use a median, not a mean, and consider trimming extreme values so a few outliers don't distort the reference point.
• Check the discount against a reasonable liquidity or size floor — micro-cap multiples are noisy and can produce misleadingly large "discounts" that reflect thin trading, not opportunity.
• Sanity-check the stock's own multiple against its own historical range, not just its sector peers — a discount that has persisted for years is a different situation than a discount that just opened up.
• Ask whether the entire sector looks stretched (or depressed) relative to its own long-run history before trusting the relative read at face value.
• Treat the discount as a starting point for research, not a buy signal on its own — cheapness relative to peers says nothing about whether the underlying business is actually sound.

For a live view of how sector-median P/E levels compare across the market today, see P/E ratio by sector. For more on how a systematic quality and valuation process fits together, see the methodology overview.

FAQ

Is sector-relative valuation the same as relative valuation in general? Sector-relative valuation is a specific, common form of relative valuation. "Relative valuation" more broadly can mean comparing a stock to any peer group — competitors of a similar size, companies with similar growth rates, or even historical multiples of the same company. Comparing against the sector median is the most standard and widely available version of the idea, since sector classifications are consistent and public.

What's a good sector-relative discount to look for? There's no universal number, because it depends on how noisy the sector's multiples are and how the stock has historically traded relative to peers. A double-digit percentage discount to the sector median is a common starting threshold used in screening, but the discount alone doesn't tell you whether it reflects a real opportunity or a real problem — that requires looking at the underlying business.

Does sector-relative valuation work for unprofitable companies? Not directly. P/E is undefined or meaningless for companies with negative earnings, which is common in early-stage biotech, pre-profit tech, and cyclical companies at the bottom of a downturn. For those cases, other relative multiples — EV/Sales, EV/EBITDA, or price-to-book — computed the same sector-relative way, are more useful than P/E.