Direct answer
Williams %R changes with timeframe because its values depend on a rolling lookback range. A shorter timeframe recalculates the indicator using a smaller recent high–low window, so it tends to move more quickly and hit “near extreme” levels more often. A longer timeframe uses a larger historical window, which typically smooths movement but can delay when the indicator reaches extremes.
Mechanism or definition
Williams %R (often written as %R) is a momentum oscillator that measures where the latest closing price sits relative to the highest high and lowest low in a chosen lookback period. In simplified terms, it compares the current price to a range: the higher the price is within that recent range, the closer %R moves toward its upper side; the lower the price is within the range, the closer it moves toward the lower side.
Timeframe affects the calculation in two closely related ways:
- The observation window (lookback length). If you compute %R using a smaller lookback, the “recent highest high” and “recent lowest low” are likely to update more often. That makes the indicator more responsive to short-lived swings.
- The observation frequency (bar size). Using candles of different sizes changes what “one step” means. A 1-hour bar and a 1-day bar represent different time spans, so the same number of bars covers different real time.
A helpful assumption for independent checking: you must keep the %R lookback length and the price data source definition consistent when comparing readings across timeframes.
Evidence or example (scenario-impact)
Imagine the same underlying market regime, but you watch it with two different bar settings.
- Scenario: Over several days, price makes a sharp intraday spike, then retraces.
- Shorter timeframe result (faster update): If your %R calculation uses a lookback window that is frequently refreshed by intraday highs/lows, the highest high and lowest low in the window may change quickly. %R can therefore move rapidly and may show extreme-like values during the spike.
- Longer timeframe result (smoother and later): With a larger lookback window, that intraday spike may not dominate the entire highest-high/lowest-low range. As a result, %R may move more smoothly and may take longer to approach extremes.
Material consequence: two charts can both be “Williams %R,” yet their readings are not directly comparable because the indicator is anchored to different rolling ranges.
Limitations and risks
A few common failure modes follow from the way timeframe changes the calculation:
- Misinterpreting extremes across timeframes: “Extreme” readings can occur for different reasons depending on whether the lookback window captures only recent volatility or a broader range.
- Noise vs lag trade-off: Shorter timeframes often increase responsiveness but also increase the chance of reacting to brief swings that do not persist. Longer timeframes can reduce noise but may lag behind turning points.
- Non-stationary relationships: Historical behavior of %R under one market regime does not guarantee similar behavior under another. Costs (such as trading costs) and execution realities are outside what the indicator itself tells you.
Uncertainty note: without real-time data and without specifying the exact %R formula parameters and bar definitions, any numerical comparison would be incomplete. You can still verify the directional claim—timeframe changes %R sensitivity—by recomputing %R with consistent rules on the same dataset while varying only the lookback or bar size.
Verification or next question
To verify the timeframe effect independently, check these control points:
- Use the same price definition (for example, closing price) across computations.
- Keep the %R lookback length explicit and change only one variable at a time (lookback length or bar size).
- Compare how quickly %R moves after known swing highs/lows, not just whether it crosses a particular level.
Next question to explore: how %R behaves when volatility is rising versus falling, and whether the same timeframe rule produces similar smoothness and lag in different regimes.