How the ChipSentiment score is calculated
6 August 2026 · 6 min read
Every stock on the board carries a number between 2 and 98. This page explains exactly how it is produced, because a market sentiment score you cannot audit is a market sentiment score you should not trust.
What the score is, and what it isn't
It is a description of how a stock is currently behaving relative to its own history and to its peer group. It is not a valuation, not a forecast, and not a recommendation. A score of 90 does not mean a stock is a good buy - it frequently means the opposite, since it takes a large recent move to get there.
The six components
Each is scored on its own, compressed onto a scale of −1 to +1, then multiplied by its weight. The weights are the whole formula:
| Component | Weight | What it measures |
|---|---|---|
| Momentum | 25% | Is price trending up or down today? |
| Relative strength | 20% | Is it outperforming the rest of the complex? |
| Trend | 20% | Is price above or below its moving averages? |
| Volume | 15% | Is the move attracting unusual participation? |
| Volatility | 10% | Is uncertainty increasing? |
| Drawdown | 10% | How far is it from its recent high? |
The weighted contributions are added to a base of 50, giving a score between 2 and 98. Momentum can therefore move a score by at most 12 points, volatility by at most 4.8.
Momentum, 25%
Today's percentage change against the previous close. The largest single input, because a sentiment read that ignores what just happened is not describing sentiment.
Relative strength, 20%
The stock's five-day move minus the market-cap-weighted five-day move of all 34 names on the board.
This is the component I would defend hardest. In a week where the entire AI silicon complex is up 8%, being up 8% tells you nothing at all. Relative strength is the only input that distinguishes a stock that is leading from one simply floating on the tide, and it is the reason a sector board is more useful than 34 separate quote pages.
Trend, 20%
Where the price sits against its own 10-day and 20-day moving averages, averaged. Two lookbacks rather than one because a single moving average flips on and off around a flat price, making the component chatter from day to day - the opposite of what a trend measure is for.
Volume, 15%
Today's volume against the stock's own 20-day average, signed by the direction of the move. A 3% rise on double normal volume is a different event from a 3% rise on a quiet day.
The sign matters. Measured without it, a stock collapsing on five times normal volume would score this component positively, which is plainly wrong. Heavy volume confirms whatever is happening, so it reads bullish into a rise and bearish into a fall.
Volatility, 10%
Ten-day realised volatility against the same stock's own 30-day baseline. Rising volatility scores negative.
Note that this penalises violent rallies as well as selloffs. That is the intended reading of "is uncertainty increasing" - and a name ripping higher on expanding volatility still scores strongly on momentum, trend and volume, so the net effect is a modest brake rather than a reversal.
Drawdown, 10%
How far below its 52-week high the price sits. Centred on an ordinary drawdown rather than on the high itself: anchored at the high, this component could only ever subtract, and it would drag the entire board below neutral in any normal market.
Why nothing is clipped
Every component is compressed onto its range with a curve that approaches its bound without reaching it, rather than being capped.
This site has walked into the clipping trap three times now. An early version clipped the total, and five different stocks all scored exactly 98 - the ceiling destroyed all resolution at precisely the end of the scale readers look at first. A later version clipped momentum and trend individually, with the same result one level down: on any strong day the top five names had identical component breakdowns. Compression means a stock up 12% still outscores one up 8%, all the way out.
What happens when data is missing
Any component whose input is unavailable contributes zero rather than falling back to an older value, which pulls the score toward neutral instead of inventing a reading. If a provider stops supplying daily history, the trend and volatility components drop out and the affected figures show a dash.
Displaying last week's number in today's column, styled identically to live data, is worse than displaying nothing - the reader has no way to know which they are looking at.
Every score here can be rebuilt from scratch
Because all six inputs come from price and volume, anyone with a daily bar series can reproduce any score this site has ever shown, including historical ones. That is not true of a score that depends on how many articles happened to be published on a given afternoon - article counts cannot be reconstructed after the fact, so a history containing them can only ever be taken on trust.
Known limitations
- It is momentum-weighted. Momentum and volume share the sign of the same day's move, so 40% of the score turns on a single session. On a sharp reversal the score will move hard.
- It has no fundamental input. No earnings, margins or valuation. A company can be scored 90 while getting materially worse.
- Volatility is direction-blind. It marks down uncertainty wherever it appears, including in a rally that turns out to be the start of something.
- It is not backtested yet. We have not published evidence that the score predicts forward returns, and until we do you should assume it does not.
Changes to the formula
When the formula changes, this page changes with it, and the change is noted. If you ever see a score you cannot reconcile with this page, that is a bug worth reporting.
9 August 2026. The score became price-only. A news coverage component was removed and the remaining components were reweighted, with trend switched from a raw five-day change to a moving-average comparison, and volatility and drawdown added.
The coverage component was dropped because it was measured and found wanting: on a typical day it was contributing exactly zero for 23 of the 34 names, since most of these companies are not written about daily. It spoke only for the handful of names already receiving the most press, which is largely the same set already scoring high on momentum. A component that is silent two thirds of the time and correlated with momentum the rest of the time is worse than no component at all. Headlines are still collected and still shown; they are simply no longer scored.