HomeBlogHow Zenith scores a stock (August 2026)
MethodMethodcomposite signalAlpha Enginefundamentals

How Zenith scores a stock (August 2026)

Zenith Market AI TeamAugust 21, 20268 min readRSS

The short version

A Zenith score is a published ruleset, not a daily rescreen of the whole market and not a human analyst. Public pages blend three legs. Alpha Engine (Premium) adds a fourth.

LegWhere it livesWhat it actually is
TechnicalPublic analyze, screener, heatmapA deterministic technical-style map (same function everywhere)
FundamentalPublic analyzeGrade A–F from valuation, growth, profitability, leverage, efficiency. Live vendor metrics when we have them; otherwise published estimates
AI disruptionPublic analyzeA stock or sector overlay: Beneficiary, Low, Medium, High
Headline toneAlpha Engine onlyKeyword tone on ~30 days of company news, plus a small overlay from today's % move
Microsoft is then capped at Hold. That rule fires last.

If a blog post from April 2026 told you something fancier — live 13F flow, options, a secret fourth model on the free tier — that post is stale. This one is the desk copy.


Leg 1 — Technical

Every ticker gets a technical-style reading: Strong Buy through Strong Sell, plus a confidence number.

It is deterministic. The same symbol produces the same technical map on the homepage, the screener, /analyze/NVDA, and Alpha. A handful of large names have written reasoning (Apple, Microsoft, NVIDIA, and the rest of the featured set). Everything else gets a generic sentence.

It is not a live RSI/MACD recalculation from the last candle. Do not trade it as if it were.


Leg 2 — Fundamentals

The fundamental grade is a 0–100 blend:

  • Valuation 20% (P/E bands)
  • Growth 25% (revenue growth)
  • Profitability 25% (gross / operating / net margin)
  • Leverage 15% (debt/equity)
  • Efficiency 15% (ROE, FCF yield)
A ≥ 80, B ≥ 65, C ≥ 50, D ≥ 35, else F.

When Finnhub (or the vendor behind our quote stack) returns live metrics, those win. If they do not, we fall back to a stock-specific estimate table, then to a sector average. The UI should say when live metrics were used. If it does not on a given screen, assume estimates.

A high P/E is not automatically "bad" — it just scores lower on the valuation sleeve. NVIDIA can still grade well because growth and profitability dominate. Palantir-type multiples get punished on valuation even when growth is fine. That is the model, not a value-investing lecture.


Try the public tools first

Screener, heatmap, and ticker scores stay free. Premium is the desk on your book — $14/mo early adopter (listed $29), 7-day trial.

Leg 3 — AI disruption

This is an overlay, not a classifier reading today's news.

  • AI Beneficiary (NVIDIA, TSMC, Broadcom, Amazon, Google, Palantir, Cloudflare, …): public composite can notch up.
  • High Risk (Microsoft is the headline case; generic Software/IT defaults here): public composite can notch down.
  • Low / Medium: everything from Apple hardware to Visa to energy.
Stock-specific tags beat sector defaults. If we have not tagged a name, you get the sector.

How the public composite is assembled

Start with the technical reading. Then:

  • Fundamental score ≥ 75 → shift one notch toward Buy
  • Fundamental score < 40 → shift one notch toward Sell
  • High Risk → shift one notch toward Sell
  • AI Beneficiary → shift one notch toward Buy
Then enforce the Microsoft cap. Buy and Strong Buy become Hold; confidence is pulled down. Sell is still allowed.

That composite is what you see on public analyze, the screener, and the heatmap.


What Alpha Engine adds (Premium)

Alpha does not replace the three legs. It re-weights them and adds headline tone.

Default mix: technical 25 / fundamental 30 / headline tone 20 / disruption 25.

You can drag the sliders. The ring is a weighted average, then the same Microsoft cap. Headline tone is a keyword count on 30-day company headlines (beat, upgrade, miss, layoff…), not Twitter, not 13Fs, not options flow.

There is no backtest on that page because we do not publish one. If a number is not on the page, we do not have a number.


What we will not claim

  • "50,000 real-time AI signals"
  • A win-rate
  • That the free screener is the same as a book-aware Co-Pilot
  • That Microsoft is a Strong Buy "this week"
For the product split, see Free vs Premium.

Educational only. Not financial advice. We are not a registered investment adviser.

Public scores, then a desk if you need one

Analyze any ticker, run the screener, read the heatmap — no account. Premium adds Co-Pilot, thesis, Alpha weights, Rotation, Guardian, and the ranked Fund+TA shortlist. $14/mo early adopter (listed $29). 7-day trial.