Archetype-adjusted P/E across sectors

Why archetype before sector

The convention in sell-side equity research is to compare a company against its sector median. That works when the sector is internally homogeneous, and falls apart when it isn’t. Healthcare contains both pre-revenue biotech and ninety-year-old dividend payers. Communication Services contains both ad-network growth machines and turnaround telcos with falling ARPU. Consumer Cyclical contains both Hermès and a regional auto-parts wholesaler. Quoting one P/E for any of these sectors is averaging numbers that don’t belong in the same bag.

Our methodology classifies every covered company into one of ten archetypes — hyper-growth, growth infrastructure, mature compounder, mature dividend, cyclical, turnaround, financial, pre-profit, REIT, and Chinese-listed ADR. The archetype determines which valuation models we apply, which terminal-growth assumptions we allow, and what a “normal” multiple even looks like. A 25× forward P/E on a mature compounder is a different statement from a 25× forward P/E on a turnaround. Once the archetype is fixed, the sector becomes the second-order axis: it changes the cyclicality, the capital intensity, and the regulatory drag, but it no longer mixes growth profiles.

The archetype is the prior; the sector is the conditioning variable. Multiples mean different things in different priors.

Methodology

For every (archetype × sector) pair with at least four covered tickers, we report three quantities: the 25th, 50th, and 75th percentiles of the bucket’s forward P/E, the median consensus forward EPS growth rate, and the ratio of the two — what we call the archetype-adjusted P/E. The ratio follows the PEG convention used throughout our valuation stack: forward P/E divided by the forward growth rate as an integer percent. A bucket trading at a forward P/E of 25 with median growth of 12% has an adjusted P/E of 2.08; a bucket at 50× and 30% growth scores 1.67 — and is therefore the cheaper bucket on a growth-adjusted basis, even though its absolute multiple is twice as high.

We deliberately exclude buckets with fewer than four covered tickers — the dispersion is too high for a single percentile to mean anything — and we mark pre-profit buckets as not meaningful, because a P/E with a denominator near zero is not a number, it is an opinion. Pre-profit names should be valued on EV/Sales or implied unit economics, not on this metric. The dataset notes which buckets fall into each exclusion class so downstream consumers can audit our reporting choices.

The forward P/E and forward growth inputs are point-in-time consensus means from our published archive on 2026-05-08, sourced from the underlying coverage feed (Yahoo Finance forward EPS, divided by the closing price of the same day). The growth integer is the consensus 1-year forward EPS growth from the same source. We use forward, not trailing, because the point of an archetype is to characterize a forward profile, not a historical one — and because for the cyclical and turnaround archetypes in particular, trailing earnings are a hostile proxy.

Five findings worth fighting about

1. Hyper-growth Technology is not the most expensive bucket. Its archetype-adjusted P/E sits below 2 — cheaper, on a growth-adjusted basis, than mature compounders in Consumer Defensive. The reason is arithmetic: 30% growth divides through a lot of multiple. Whether the growth is real for any individual name in the bucket is a separate question, and one that the archetype does not answer for you. (See: every Reverse-DCF we publish.)
2. Mature dividend buckets cluster near 3. Across Consumer Defensive, Utilities, Healthcare, and Industrials, the mature-dividend archetype-adjusted P/E sits in a tight 2.7–3.4 band. The market is paying roughly the same growth-adjusted multiple for any reliable payer, regardless of sector — which is what you would expect if the marginal buyer of these names is allocating to a single mental bucket called “bond proxy.”
3. Turnaround names look pricey on this metric, and that is the point. Turnaround buckets in Healthcare and Communication Services score above 4.5 because depressed earnings inflate the denominator. This is why we don’t valuate turnarounds on P/E in the first place — sum-of-parts and normalized earnings are the right lenses, and the headline multiple is misleading. The dataset preserves the misleading number to make the point visible.
4. Cyclicals are cheap; cyclicals are always cheap. Cyclical Energy at 2.30 and Cyclical Financial Services at 1.73 represent buckets valued near the trough of an earnings cycle. The trick — and what our cyclical archetype lens explicitly demands — is normalizing through the cycle before assuming the cheap multiple is a real opportunity rather than a value trap.
5. REITs at adjusted-P/E 7 is a methodology artifact, not a warning. P/E for REITs is FFO-distorted. The right lens is P/FFO and dividend coverage. The bucket is in the dataset because excluding it would hide a real population of covered names, but the cell should not be read as “REITs are 4× as expensive as banks.” It should be read as “P/E is the wrong tool here.”

The dataset

The full table is published below as a public CSV under a CC BY-NC 4.0 license. Every row is one (archetype × sector) bucket, with bucket-size, forward P/E percentiles, median forward growth, and the derived archetype-adjusted P/E. The schema is stable across quarterly snapshots — column order and slug spellings will not change without a versioning bump in the file name.

• archetype-adjusted-pe-2026-q2.csv — 33 buckets with reported P/E, 7 excluded (small sample or pre-profit).
• Schema: archetype_slug, archetype_label, sector_slug, sector_label, ticker_count, forward_pe_p25, forward_pe_median, forward_pe_p75, forward_growth_pct, archetype_adjusted_pe, data_as_of, note.
• Cadence: one snapshot per quarter. The next release ships at /research/archetype-adjusted-pe-2026-q3.

Caveats

Three. First, consensus forward growth is a moving target — analysts revise estimates constantly, and the snapshot freezes them on a single day. The quarterly cadence is meant to track that drift, not paper over it. Second, the bucket counts here are point-in-time and reflect our coverage universe, not the global investable universe; the U.S. mega-caps are well-represented, the long-tail emerging markets less so. Third, archetype assignment is a judgment our model makes, calibrated against historical revenue growth and ROIC. Borderline cases — a hyper-growth name decelerating into mature-compounder territory — could land in either bucket on a different methodology cut. The full classification rules are in the methodology.

Citing this work

Plain-text citation: StockMarketAgent.AI. (2026). Archetype-adjusted P/E across sectors — 2026 Q2. Available at https://stockmarketagent.ai/zh/research/archetype-adjusted-pe-2026-q2. For attribution in academic or commercial contexts please link the page; the CSV is licensed under CC BY-NC 4.0.

Related: Archetype hubs · Sector hubs · Methodology · Blog