Direct answer
Divergence in Rmi means that the Rmi line and price are moving in different directions at the same time. In plain terms, price may be making one kind of move (for example, pushing higher), while the Rmi reading is not confirming that move (for example, rising less or turning down).
This is not a promise of a specific future outcome. It is a description of a mismatch between momentum (as measured by Rmi) and price action.
Mechanism or definition
Rmi is typically treated as a momentum indicator derived from recent price changes over a chosen lookback period (the exact formula can vary by provider, but the idea is consistent). Divergence is then a relative observation:
- You compare the direction or slope of Rmi with the direction or slope of price.
- “Bullish-type” divergence is often described when price makes a lower low while Rmi fails to make a lower low of the same quality.
- “Bearish-type” divergence is often described when price makes a higher high while Rmi fails to make a higher high of the same quality.
A key construction detail is that divergence depends on how you define “same quality.” Many people look at visual turns (peaks and troughs), but the timing and the chosen lookback period can shift those turning points.
A simple example model (with clear assumptions)
Assume:
- Rmi is computed from recent price changes using one fixed lookback length.
- You identify a trough when Rmi reaches a local minimum.
If price forms a second trough lower than the first, but Rmi’s second trough is not as low as the first (or occurs while Rmi is flat or rising), you would describe that as divergence.
The “meaning” then is that momentum, as captured by the indicator’s construction, is not matching price strength.
Evidence or example
Divergence can be easier to recognize in hindsight because you can see the full sequence of peaks and troughs. For checking it independently, you can do a forward-looking scan on charts:
- Mark the time where price forms a new swing high/low.
- Then observe whether the Rmi swing high/low around the same time is making a stronger move than before.
- Keep the lookback period and chart settings constant for the scan so you are not changing the indicator mid-observation.
If Rmi and price repeatedly diverge and later converge (or follow-through happens), that may suggest the pattern is sometimes informative in a given context. However, historical repetition alone does not establish a reliable rule for new conditions.
Limitations and risks
1) Confirmation limits
Divergence is often a late indicator of changing momentum. By the time the mismatch is clear, market participants may have already acted. So divergence is usually descriptive, not self-sufficient.
2) Variable inputs and calculation differences
Rmi is not guaranteed to be identical across platforms because providers may use different parameter defaults or slightly different constructions. Divergence results can change if:
- the lookback period differs,
- smoothing or normalization differs,
- the chart timeframe differs.
3) Hindsight bias
Hindsight bias can make divergence look “obvious” after the fact. When future movement is known, it becomes easier to select swing points that make the divergence story fit. This can lead to overconfidence that a mismatch “caused” an outcome.
4) Failure modes
Common failure modes include:
- Divergence that appears during normal volatility and then disappears without meaningful follow-through.
- Divergence that is caused by timing differences (price makes a new extreme before Rmi fully reflects it, or vice versa).
- Over-interpreting tiny indicator wiggles as meaningful turning points.
Verification or next question
To verify the concept accurately for yourself, you can:
- Confirm you are using the same Rmi definition (especially the lookback period) across charts and providers.
- Write down your rule for what counts as divergence (for example, “price makes a lower low while Rmi makes a higher low” using clearly identified swing points).
- Check multiple historical instances without adjusting your rule after seeing the later price outcome.
A good next question is whether the divergence you observe is consistent across timeframes and indicator parameter choices, because that is where many “visible” divergences become unstable.