← Taiwan Stock Snapshot Database

Home / Indicator library / Valuation percentiles

Valuation percentiles

Is a P/E of 18 high or low?

On its own the number means nothing. Eighteen times earnings is punchy for a mature industrial and cheap for a company compounding at thirty percent, so the only comparison that settles anything is against the same company's own past. This page puts each ticker's current ratios back into its own history and reports where they sit, as a percentile rather than an adjective.

See it

Five tickers, priced from the official panel, each with its own one-year P/E band and where today sits inside it. Rendered on the server with no key, no signup and no quota.

Five tickers · current ratios against their own historyGET /v1/valuation/latest · /v1/factors
TickerP/EP/E percentileP/E range (1y)P/BDividend yield
2330 TSMC28.525122.47 – 34.879.920.89%
2454 MediaTek77.799917.21 – 78.6117.721.14%
2317 Hon Hai (Foxconn)16.513513.83 – 21.951.842.86%
2308 Delta Electronics55.223446.28 – 93.5615.340.67%
2412 Chunghwa Telecom28.599626.16 – 29.133.083.57%
Snapshot dated 2026-09-18 · one-year window 2025-09-18 to 2026-09-18 · percentile is the rank of today's P/E inside that ticker's own one-year history, where 0 is the cheapest it has been and 100 the most expensive · source: Taiwan official exchange public data (GODL v1) · historical statistics, not a signal.
What we publish is data and tooling — not investment advice, not stock picks, and no promise of returns. Every figure is either an official public source reproduced as published, or a derivation we document well enough for you to recompute it. What you trade on it is your call. A percentile is a statement about the past distribution, not a forecast. Cheap relative to history can stay cheap, or get cheaper.

How it is built

A percentile is only worth reading if it can be taken apart. Every number below can be recomputed from the same public inputs.

01

Ratios come from the official panel

Price-to-earnings, price-to-book and dividend yield are read from the official public source for the whole market and reproduced as published, with no vendor in between doing its own adjustments.

02

Each ticker is ranked against itself

The current ratio is placed inside that company's own trailing history and its rank reported as a percentile — zero being the cheapest it has traded in the window, one hundred the most expensive. Cross-company comparison of raw multiples is the mistake this replaces.

03

The whole market is ranked the same way

The same percentile logic runs across every covered ticker at once, which is what produces the cheap and expensive lists rather than a hand-picked watchlist.

04

Every historical point is as-of, not restated

Look at a past date and the answer is the version that was on the record that morning. Applying today's restated earnings to a price from two years ago produces a P/E history nobody could have computed at the time, and a backtest built on it is reading the future.

Terms

Four words carry this page:

Price-to-earningsP/E
Share price divided by earnings per share. In short: the number of years of current profit it would take to earn back what you paid. A high multiple means the market is paying up, either for quality or for growth it expects to arrive.
Price-to-bookP/B
Share price divided by book value per share. In short: what you pay relative to what the balance sheet says the company owns. Below one is cheap against book, though there is usually a reason.
Dividend yield
Cash dividends over the last year divided by the price. In short: the cash return on the money you put in, before any change in the share price. A high yield can mean generosity or a falling price.
Percentile
The rank of a value inside its own distribution. In short: the twentieth percentile means it has been cheaper than this only twenty percent of the time — far more informative than the raw number.

Where the line is

A percentile answers one question well: is this unusual for this company. It says nothing about whether the current level is deserved. A business whose earnings power has structurally improved should trade above its own history and will read as expensive here; one in decline will look cheaper every quarter as the denominator falls. The statistic is a prompt to look, not a verdict, and the range printed above is deliberately shown next to the percentile so the underlying spread stays visible rather than being compressed into a single rank.

FAQ

Is a P/E of 18 high or low?

Neither, until it is compared with something. Eighteen times earnings is expensive for one business and cheap for another, so this page ranks the current multiple inside that same company's own trailing history and reports the percentile: zero is the cheapest it has been in the window, one hundred the most expensive. The raw range is printed next to the rank so the spread stays visible.

How is the percentile computed?

The ratios are read from the official public panel and reproduced as published. The current value is then ranked inside that ticker's own trailing series, so the answer is a position in a distribution rather than a score from a model. Every percentile can be taken apart into the historical observations that sit below it.

Does the history read the future?

No. Asking about a past date returns the version of the ratio that was on the record that morning, not today's restated earnings applied backwards. That distinction is the whole point: a conventional valuation history is rebuilt from current figures and is therefore unavailable to anyone trading at the time.

Which tickers are covered?

The valuation panel covers the liquid subset of the Taiwan listed market rather than every registered ticker, and a ticker outside it returns a coverage answer instead of an estimate. Nothing on this site fills a gap with a modelled number and presents it as data.

What does the free tier include?

Everything on this page renders server-side with no key and no signup, and the current ratios are free through the API as well. A paid tier adds reach rather than different numbers: percentile history against arbitrary past dates, and full-universe coverage called programmatically.

Run it yourself

The free tier needs no signup — call this feature's API straight away with the demo key. Unlock as-of history and full ticker coverage on a paid tier.