Direct answer
Forex indicators are computed from price over a chosen timeframe (for example, 1-minute bars vs 1-hour bars). Changing the timeframe changes what each data point represents, so the indicator’s shape, timing, and apparent strength can change. That is why an indicator reading on one chart does not automatically confirm what you would see on another chart.
Mechanism: what “timeframe” changes in an indicator
A timeframe controls the observation window of your data. If your chart uses a shorter timeframe, each bar summarizes fewer seconds or minutes, so the indicator reacts to smaller, more recent changes. With a longer timeframe, each bar summarizes a larger period, so the indicator reflects a broader swing and tends to be smoother.
Most indicators use one or both of these ideas:
- Aggregation: Inputs like averages, ranges, or returns are calculated from bars. Change the bar duration, and the aggregated values change.
- Smoothing length in “bars”: Many indicators use a fixed number of bars (such as a 14-period moving average). If you move from 15-minute bars to 1-hour bars, 14 bars no longer cover the same amount of time.
Assumption for any example: we are comparing the same underlying market, but we are only changing how we group observations (bar size) and possibly how “periods” map to real time.
Evidence through a realistic scenario (with the material limitation)
Imagine you use a momentum-style indicator that reacts to recent price changes. On a short timeframe, a sudden move may create a quick peak, and then it may fade as price mean-reverts within the same larger move. On a longer timeframe, that same event might be only part of a longer trend, so the indicator could look flatter or shift later.
A second scenario is a volatility indicator that measures typical movement size. Short timeframes often show higher apparent variability because individual bars contain microstructure noise and brief spikes. On longer timeframes, spikes are averaged into fewer bars, so volatility looks more stable.
Material limitation / failure mode: An indicator can appear to “work” on one timeframe because it is tailored—implicitly—to the timing of that timeframe’s noise and the frequency of reversals. If market behavior changes (for example, from ranging to trending), the relationship between the indicator and future movement on that timeframe may weaken.
Limitations and risks to consider
- Lack of transferability: Historical relationships observed on one timeframe do not guarantee similar behavior on another timeframe.
- Different time alignment: “Periods” measured in bars may correspond to different real-world durations when you switch timeframes.
- Noise vs lag trade-off: Shorter timeframes often reduce lag but increase noise, which raises whipsaw risk (frequent false turns). Longer timeframes reduce noise but can lag behind turning points.
- Non-indicator factors still matter: Real outcomes can be influenced by trading costs, execution behavior, and how fast prices move relative to your timeframe. This is true even if the indicator line looks consistent.
Verification: a control point you can use independently
To verify timeframe sensitivity without assuming any indicator will predict the future, do the following conceptually:
- Pick an indicator and note how it maps “periods” to bars.
- Compare at least two timeframes while keeping the same underlying price source.
- Check whether the indicator’s key features (timing of peaks/troughs, relative magnitude, and sign changes) shift meaningfully.
- Treat any apparent consistency as descriptive, not predictive, because correlations can change with market regime.
Next question to explore: When you change timeframe, do you also keep the indicator’s “period” length consistent in real time, or does it unintentionally change? This often explains why two charts produce different conclusions.