Regular bearish divergence appears when price records a higher swing high while a chosen momentum measure records a lower one. The pattern is easy to spot after the fact, which makes it tempting to treat the second peak as a timer. Yet an oscillator measures a feature of movement; it does not compel price to reverse on cue.
Deceleration is information, not a verdict
A vehicle can slow and still travel a considerable distance. In the same way, an advancing market can lose momentum while persistent demand carries price through several further highs. The divergence identifies reduced acceleration relative to a previous swing. It does not identify sufficient opposing pressure.
This is why repeated divergence can form during a strong trend. Each signal may be mathematically valid while a premature counter-trend conclusion remains practically poor.
Check the swing comparison
Indicators are sensitive to the points being compared. A minor price high paired with a major oscillator high creates a misleading relationship. Compare swings of similar structural importance, and note whether the indicator settings suit the timeframe under study.
Then inspect the path between peaks. Was the pullback shallow and quiet, or did it show heavy opposing volume and broad price ranges? Did price approach a weekly supply area, or is it breaking from a long base? Divergence gains or loses relevance through these neighbouring facts.
Ask price to participate
Price confirmation need not be a magical candle. It can be a failed attempt to hold above the new high, a break of the sequence of higher lows, or an inability to reclaim a breached reference. Volume may add evidence when activity expands on rejection or contracts during a weak retest.
The key is sequence. First note deceleration. Then identify the price behaviour that would show control changing. Until that behaviour occurs, describe divergence as a warning condition rather than a completed reversal.
A useful annotation
Write: “Price has made a marginal new high while momentum is below its previous peak. The trend structure remains intact. Failure to hold above the breakout area, followed by loss of the last reaction low, would make the divergence more consequential.”
That sentence distinguishes what exists from what might follow. It also makes hindsight less flattering, because your required confirmation was recorded before the outcome appeared.