Bullish versus bearish divergence

Updated 2026-10-09 Español·Deutsch·Русский·Italiano·Français·Türkçe·العربية

Regular bullish divergence compares lows; regular bearish divergence compares highs. The names describe possible weakening, not guaranteed ascent or decline.

Regular divergence · Price

Regular divergence · PriceCompare extremes on the same candles. [110, 100, 108, 94, 103]94110L₁L₂Time →
  • Compare extremes on the same candles.

Indicator

IndicatorThe observation alone defines no entry, exit or position size. [45, 20, 45, 30, 48]2048L₁L₂Time →
  • The observation alone defines no entry, exit or position size.
Teaching diagrams with hypothetical numbers. They explain concepts, not the next price move.

Conditions

Use two comparable same-type price extrema and consistently paired indicator values.

Bearish: price HH, indicator LH. For example, price highs rise from 110 to 115 while RSI falls from 75 to 65.

Bullish: price LL, indicator HL. For example, price lows fall from 100 to 95 while RSI rises from 25 to 32.

These are geometric examples, not trades or performance estimates. State any minimum change beforehand. RSI points cannot be directly compared with absolute MACD values.

Weakening is not reversal

A lower price low with higher indicator value suggests the indicator does not confirm a stronger decline under that measure. It does not establish price reversal: a trend can continue and produce more divergences.

Regular divergence · Price

Regular divergence · PriceCompare extremes on the same candles. [100, 112, 105, 120, 110]100120H₁H₂Time →
  • Compare extremes on the same candles.

Indicator

IndicatorThe observation alone defines no entry, exit or position size. [45, 75, 50, 65, 48]4575H₁H₂Time →
  • The observation alone defines no entry, exit or position size.
Teaching diagrams with hypothetical numbers. They explain concepts, not the next price move.

Same bars or nearby extrema?

Taking the indicator on price-pivot bars differs from finding nearby indicator pivots in a defined window. Do not mix conventions in one test.

An overly broad window can combine unrelated movements. Define matching and confirmation before viewing results.

Distance between points

Close points may belong to one fluctuation; distant ones to different regimes. There is no universal bar count, since scale and indicator period matter. Respect the second pivot’s confirmation delay.

Divergence determines neither movement size, timing nor probability. A trading system would require separately tested confirmation, invalidation and risk rules with costs.

See StockCharts’ RSI documentation and hidden divergence.

Related questions

See it on a live chart

The same markup on a prepared example: the chart, the objects it found and what followed.

Open the example