Direct answer: when Rmi can behave differently
Rmi may appear to “behave differently” when the market regime changes—especially when price shifts between trending and ranging behaviour, when volatility rises or falls, and when the indicator’s input conditions (timeframe, data sampling, and effective execution costs) change. Without assuming real-time data, the key point is conditional behaviour: the same underlying indicator can produce different readings when the statistical properties of price change.
Mechanism or definition: separating what is stable from what varies
First define the concept in general terms. Rmi is a momentum-type measure built from price movement over a lookback window. In practice, its behaviour depends on how current price relates to earlier price within that window.
A stable mechanic is that a momentum measure reacts to directional movement: stronger and more persistent movement over the window tends to produce a more pronounced momentum reading, while weaker or alternating movement tends to reduce it.
Variable conditions include:
- Trend persistence vs. mean reversion: A market that moves in one direction for longer will affect momentum measures differently than one that oscillates around a level.
- Volatility regime: If price swings more sharply, the same lookback window can contain larger changes, which can alter the smoothness and magnitude of momentum readings.
- Noise level: Choppy, short-term reversals can make momentum measures respond more erratically.
- Timeframe and sampling: Changing the candle timeframe (or how frequently the input is sampled) changes what “the last N periods” means in real market time.
- Data quality and execution friction: Even if an indicator is computed from historical prices, real-world interpretation can differ when spreads, slippage, and other frictions are non-trivial.
Evidence or example: conditional comparisons you can verify
Because no live market data is assumed, the most useful “evidence” is a controlled comparison in principle—using historical data you can compute and inspect.
Consider two contrasting market segments where the only difference is the regime:
- Ranging segment: Prices oscillate around a central area. Over many windows, the net directional movement can be small or alternating, so a momentum measure can spend more time near its baseline or fluctuate around it.
- Trending segment: Prices move more persistently in one direction. Over the same type of windows, the net movement is more consistently directional, so the momentum reading can stay more aligned with that direction for longer.
Now add volatility differences:
- In a high-volatility environment, the lookback window can capture larger swings. Even if the market is not strongly trending, momentum measures may show larger swings due to the magnitude of moves.
- In lower volatility, even if there is some direction, movements may be smaller, which can make the indicator look calmer.
To keep the comparison fair, hold the indicator settings constant (lookback length and calculation method) and vary only the regime characteristics you are testing.
Limitations and risks: where conditional behaviour breaks down
At least one material failure mode is regime mismatch. Momentum-style measures can underperform when the market’s statistical structure changes faster than the indicator’s window captures (for example, a shift from trend to range).
Other limitations:
- Historical relationships do not establish future behaviour: Even if you observe that Rmi “works” in past conditions, it does not guarantee similar behaviour in the next regime.
- Parameter sensitivity: Different lookback lengths and timeframes can change how responsive or stable the readings appear.
- Interpretation risk: A changing indicator reading is not automatically a tradable edge by itself; it can reflect normal variability rather than meaningful information.
- Costs and execution differences: Any attempt to connect indicator readings to outcomes can be dominated by trading frictions, which vary by provider, venue, and jurisdiction.
Verification and next questions
To independently verify when Rmi behaves differently, use your own historical dataset and compare regimes by inspection and statistics (for example, trend persistence and volatility). Useful next questions include:
- Does the indicator’s reading change more during volatility shifts or during trend persistence shifts?
- How sensitive are the observed differences to timeframe changes?
- Do the indicator patterns remain qualitatively similar after adjusting for measurement frictions?
If you want, share the exact definition you are using for Rmi (inputs, lookback length, and how it is computed).