← All insights

Divergence Rising vs Falling: How to Read Disagreement Over Time

A rising divergence score means the four data layers on a stock are moving apart: one of them is heading somewhere the others are not. A falling score means they are coming back together, either because the odd one out rejoined the rest or because the rest moved to meet it. Neither direction says where the price goes next. What direction does tell you is whether the evidence on a stock is getting more settled or less, which no single day's reading can.

Four ways the line moves

Most of what we see falls into a handful of patterns. These are real sequences from our data, bias and divergence only.

Rising slowly, with the score. Skyworks Solutions (SWKS): -17 with divergence 6 on 1 September 2026, +56 with 49 on 8 September, +91 with 84 on 18 September. Price structure climbed session after session while the fundamentals layer stayed soft. Bias and divergence rose together, which is what it looks like when one layer carries the move on its own.

Jumping in a day. Matador Resources (MTDR): +92 with divergence 21 on 16 September, +84 with 58 on the 17th, +78 with 60 on the 18th. News sentiment turned negative in one session and the other three layers held. A jump that size in a day almost always comes from one of the two fast layers, news or price, because analyst coverage and fundamentals rarely move that quickly.

Spiking and settling. AMC Entertainment (AMC): divergence 10 on 18 August, 65 on 20 August, 12 on 25 August. News turned negative and then back inside a week, and fundamentals never moved. The full sequence is in bad news but the numbers have not changed.

Flat and high. Deere (DE): divergence between 72 and 87 on every trading day from 1 to 18 September. Nothing formed and nothing settled. The disagreement between price structure and fundamentals was simply the state of the stock that month, covered in why a stock goes up when its fundamentals look bad. The mirror image is flat and low: Suncor Energy (SU) stayed between 8 and 24 over the same stretch.

Reading a rise

A rising divergence on a stock you follow is a change worth noticing: something moved in one layer and not in the others. The useful question is which one, and whether it is a layer that moves fast or slow.

That second part matters for a mechanical reason. News tone can change in a session, and change back just as quickly. Fundamentals change once a quarter. So a gap opened by news has a way to close within days, while a gap between price and the reported figures usually has to wait for the next report or for price to come back towards them. AMC and Deere above are one of each. That is a statement about how often each layer updates, not a forecast for any stock.

Reading a fall

Falling divergence means the layers are converging, and how they converge matters more than the fact. When the odd layer rejoins the others, as AMC's news tone did, the bias score moves back towards where it was. When the other layers move to meet the odd one, the score moves with them.

Merck (MRK) is the second kind. From 1 to 14 September its divergence eased from 73 to 60 as price structure and news tone cooled, moving part of the way towards where the fundamentals already sat, and its bias fell from +86 to +63 along the way. Two falling divergence lines, AMC's and Merck's, with opposite effects on the score.

What direction cannot tell you

Rising divergence is not a warning, and falling divergence is not an all clear. Skyworks' bias rose as its divergence rose. Matador's fell as its divergence rose. Direction in the disagreement does not map onto direction in the price, and none of our pages present it as if it did.

A single day also proves little. Capital One (COF) read divergence 54, then 5, then 46 on three consecutive sessions in mid September. A move that size in one day is sometimes the story and sometimes one layer updating a day ahead of another. Two or three sessions in the same direction is a better basis for saying anything has changed, and the low end of that is covered in what low divergence means.

Where to see the line

Every covered stock has a public twelve-month record of its bias and divergence, for example Deere's or Skyworks'. The divergence board is the day's snapshot across the market. The history pages are where a snapshot becomes a line.

For the stocks you actually follow, the free watchlist report shows the four layers behind each score, which is what turns "divergence went up" into "news turned while everything else held". The guide to the measure as a whole is what stock divergence means, and the high end of the scale is in what a high divergence score means.

Know which layer moved

All four data layers behind the score, for up to five stocks you choose. Free, no account.

Score my five stocks →

EquityBias is a market data research tool. Nothing here is financial advice. Changes in divergence describe changes in current data and do not indicate future price movement. Readings quoted are historical and dated.

Not investment advice. A Bias Score describes market data that already exists, it does not forecast price, and past readings say nothing about future results. Trading equities carries risk and you can lose part or all of what you put in.

Produced by EquityBias, Jones Inc., CVR DK34325170, Denmark. Responsible: Nicky Jørgensen. Not supervised by a financial authority. Methodology · Interests and conflicts