Loading data...
Loading data...
Every psychology score on this site is computed the same way every other number here is: deterministic arithmetic over real, already-fetched data โ price, volume, technical indicators, news sentiment, growth, and fair value. There is no separate psychology data feed, and the AI layer never invents a score โ it only writes a plain-language explanation of a number this engine already calculated.
The pipeline
Each state below is a weighted average of a handful of real, named inputs โ every input is scaled onto a 0-100 range between a documented low and high value, then combined by the stated weights. A state's evidence panel on a stock page always shows the exact raw input values, not just the final score.
Fires when price, volume, and sentiment are all accelerating together โ evidence investors are chasing the move, not just following it.
| Input | Data used | How it's scaled / weighted |
|---|---|---|
| Price momentum | 12-period Rate of Change (ROC) | Scaled 0โ20% ROC to a 0-100 range (30% weight) |
| Volume surge | Current volume รท 20-day average volume | Scaled 1.0xโ2.5x to 0-100 (25% weight) |
| Overbought momentum | RSI(14) | Scaled 50โ80 to 0-100 (20% weight) |
| Price extension | Bollinger %B | Scaled 0.5โ1.0 to 0-100 (15% weight) |
| News sentiment | StockIQ news score | Scaled 50โ90 to 0-100 (10% weight) |
The mirror image of FOMO โ a sharp decline, abnormal volume, and negative sentiment hitting at once.
| Input | Data used | How it's scaled / weighted |
|---|---|---|
| Price momentum | 12-period ROC | Scaled 0โ-20% to 0-100 (30% weight) |
| Volume surge | Current volume รท 20-day average | Scaled 1.0xโ2.5x to 0-100 (20% weight) |
| Oversold momentum | RSI(14) | Scaled 50โ20 to 0-100 (20% weight) |
| Volatility spike | ATR(14) vs. its own 30-day baseline | Scaled 1.0xโ1.8x to 0-100 (15% weight) |
| News sentiment | StockIQ news score | Scaled 50โ15 to 0-100 (15% weight) |
Extreme momentum, a price far above its long-term average, and a large premium over fair value โ all at once.
| Input | Data used | How it's scaled / weighted |
|---|---|---|
| Price momentum | 12-period ROC | Scaled 5%โ30% to 0-100 (25% weight) |
| Distance above trend | Price vs. 200-day SMA | Scaled +5%โ+35% to 0-100 (25% weight) |
| Overbought momentum | RSI(14) | Scaled 60โ85 to 0-100 (15% weight) |
| News sentiment | StockIQ news score | Scaled 55โ90 to 0-100 (15% weight) |
| Valuation premium | Price vs. DCF fair value | Scaled 0%โ60% premium to 0-100 (20% weight) |
Fires when price appreciation has run far ahead of the company's actual fundamental growth.
| Input | Data used | How it's scaled / weighted |
|---|---|---|
| Price-vs-fundamentals gap | 12-period ROC minus EPS (or revenue) growth | Scaled 0โ40pp gap to 0-100 (55% weight) |
| Valuation premium | Price vs. DCF fair value | Scaled 0%โ50% premium to 0-100 (45% weight) |
How closely this stock's move today tracks the average move of its listed peers, weighted by how large the shared move is.
| Input | Data used | How it's scaled / weighted |
|---|---|---|
| Peer alignment | This stock's % change today vs. its peers' average % change | Only scores when both move the same direction; alignment = 100 minus the spread between them, scaled |
| Move magnitude | The smaller of the two % changes | Scaled 0.5%โ4% to 0-100; final score averages alignment and magnitude |
A stricter check than the Narrative Gap alone: price, news sentiment, and valuation all confirming each other upward, while revenue or margins are actively deteriorating.
| Input | Data used | How it's scaled / weighted |
|---|---|---|
| Deterioration trigger | YoY revenue growth < 0%, or operating margin down >0.5pp YoY | Score is 0 unless this trigger is true |
| Narrative Gap | See the Narrative Gap section below | Score scales with the gap size once the trigger fires (capped 0-60, boosted 30% if revenue is also negative) |
How close the current price sits to a real, documented support or resistance level โ a genuine 50-day price extreme, not a same-day statistic.
| Input | Data used | How it's scaled / weighted |
|---|---|---|
| Distance from nearest level | Current price vs. the closer of support/resistance (50-day extremes) | Within 1% of a level scores near 100; beyond ~5% scores near 0 |
A meaningful decline over the medium term, but on below-average current volume โ a slow bleed with no capitulation spike, consistent with holders declining to realize the loss.
| Input | Data used | How it's scaled / weighted |
|---|---|---|
| Trigger | 3-month price change | Score is 0 unless the stock is down 8% or more over 3 months |
| Decline size | 3-month % change | Scaled -8%โ-35% to 0-100 (50% weight) |
| Volume drought | Current volume รท 20-day average | Scaled 1.3xโ0.7x (inverted โ lower volume scores higher) (50% weight) |
A separate metric, not one of the 8 states: how much more positive the story the market is telling (Narrative Score) is than what the underlying fundamentals actually support (Reality Score). Narrative Score blends news sentiment (40%), price momentum (35%), and price-vs-fair-value (25%). Reality Score blends revenue growth (40%), the recent trend in operating margin (30%), and the site's Quality score (30%). The gap is simply Narrative minus Reality โ positive means the story is currently running ahead of the numbers, negative means the opposite.
Every psychology reading also carries a Confidence score โ the share of the underlying inputs (technical indicators, DCF fair value, growth data, at least 3 news articles, peer data) that were actually available for that stock. Confidence measures how much real data went into the classification itself. It is never a measure of, or a proxy for, where the price is headed next.
โ ๏ธ What this can and cannot tell you
This engine infers a psychological state consistent with observable market behavior โ it never claims to know what any individual investor is thinking, and no score here is a buy/sell signal or a price prediction. Four states from the wider behavioral-finance literature (Regret Aversion, Availability Bias, Disposition Effect, Familiarity Bias) are deliberately not built yet โ each would need persisted historical data (e.g. a remembered past price or news event) this site doesn't store today. The model is grounded in established research (Kahneman & Tversky's Prospect Theory, and frameworks the CFA Institute and SEC already document), not psychology invented by the AI layer.
See it in action: the Investor Psychology hub and the Market Psychology Radar.
For how the 0-100 Investor Score itself is calculated, see the main methodology page. The information on this page is presented for transparency only and does not constitute investment advice โ full details on the disclaimer page.