Missing figures pull the score toward 50
If a company is missing a lot of the usual numbers, the score does not treat that hole as good news or bad news. It sits closer to the middle.
On quote pages
TECHi Forward Model Score
MODELQuote pages show this model number from 0 to 100. A higher number means the inputs TECHi checks look stronger together. The public words are Constructive, Neutral, and Cautious. They are a research note, not a buy, hold, or sell instruction. A horizon percent is a scenario estimate. A directional-match percent is how often a stored price-direction call lined up with that stock’s later close. It is not model accuracy.
This is not the AI Equity Editorial Score on Best AI Stocks. That list is an editorial ranking of AI revenue durability, moat, and valuation risk. A 96 there and a 96 here are not the same calculation. Confidence on a quote page is a separate figure, and evidence quality is a data-quality figure.
When a company is missing a lot of the usual figures, the score moves toward the middle instead of pretending the gap is good or bad news.
Methodology reviewed May 28, 2026
Validation dataset updated Sep 23, 2026
Model v1.0
Those are three different dates. Weights below are the ones in use. The May review did not rewrite the weights. The September 23, 2026 date is the replay of stored closes, and it is not an out-of-sample test.
Read the band first
The number sits in one of five bands. Neighboring scores, such as 93 and 96, are close. Read the breakdown on the quote page before you treat them as different stories.
74-100
Constructive setup
61-73
Positive but selective
47-60
Balanced watch
34-46
Cautious setup
0-33
High-risk setup
Which number you are reading
Every proprietary TECHi number carries one of these labels. The name beside the number says which calculation it is.
- MODEL
- EDITORIAL
- DATA QUALITY
- CONFIDENCE
- MARKET SENTIMENT
- TRACK RECORD
What goes into the number
Each piece below is turned into its own 0 to 100 reading, then combined with the weight shown. Missing pieces pull the final number toward 50.
Technical tape
12%
Trend, realized volatility, range behavior, and price structure.
Momentum
12%
Recent return profile and persistence across short and medium windows.
Historical setup factor
12%
A 0–100 setup score from long-run growth, positive windows, and earnings surprises. It is not the price-tested validation percent.
Quality
14%
Profitability, margins, cash-flow quality, balance-sheet durability, and operating consistency.
Future value
11%
Growth runway, revenue durability, estimate direction, and long-horizon setup.
Valuation
10%
Multiples, yield-style measures, and valuation pressure versus fundamentals.
Analyst and revision
10%
Target range, recommendation mix, revisions, and consensus dispersion.
Sentiment
6%
News and market attention where provider coverage is available.
Leadership
4%
Leadership and governance context when supported by issuer or curated data.
Social and developer traction
2%
Available community, product, and developer attention signals.
Risk brake
7%
Drawdown, volatility, debt, data gaps, and other penalties that reduce conviction.
Why a 74 differs from a 71
On this model, a three-point gap is small. A 74 and a 71 usually differ in one or two of the pieces above, often valuation or the risk weight, not in the whole story. Read the breakdown on the quote page, not just the single number. A 96 on Best AI Stocks is the AI Equity Editorial Score, not this model.
The same bands, in a row
74-100
Constructive setup
61-73
Positive but selective
47-60
Balanced watch
34-46
Cautious setup
0-33
High-risk setup
A high score can still be wrong, and a low score can still be followed by a rally. Read the pieces on the quote page before you rely on the number.
What the score will not do
Backtests are labeled by maturity
Rolling validation only appears after enough stored closes and factor snapshots mature. When accuracy tracking is pending, the UI says so.
Narrative does not create the score
The paragraph next to a score explains it. The paragraph does not set the number. The weights on this page do.
No trade instruction
TECHi Intelligence is a research screen. It is not a broker recommendation, price target, or personalized investment instruction.
This score is not personal advice, a price target, or a promise about what the shares will do next. Check the company's filings and a live quote before you act.