Taiwan Health Scorecard — the rubric is ours, so we publish all of it
Any composite score raises one question: who chose the weights? Our answer is: we did, and the entire rubric is published. Four categories of 0–25 points, every sub-score annotated with the reasoning, the raw values it cites and its weight. You are free to disagree with our weights and rebuild the score from the same raw numbers.
What the four categories measure
Earnings momentum comes from nine binary year-over-year signals; financial safety uses a four-variable distress measure chosen because it needs no market cap and carries the least industry bias; peer position takes percentile placement within the official industry classification; cash quality reuses cash-flow and accrual signals already computed upstream, reading no new data.
Every input is an existing, separately validated module on this platform, so the scorecard can never be more accurate than its parts — and each part can be opened and inspected on its own.
Incomplete beats fake-complete
If any category cannot be computed, its score is null and it is removed from the denominator. The total is always reported as achieved-over-actual-maximum together with the list of scored and missing categories and a completeness flag.
Scoring a missing category as zero turns “unknown” into “bad”; filling it with a peer average invents data. Both produce a number that pretends to be a complete 100-point score, which is precisely what we refuse to emit.