Technical indicator guide
Understanding Daily and Weekly RSI
A practical guide to the Relative Strength Index, its rolling calculation, the difference between daily and weekly inputs, and responsible interpretation.
What is RSI?
The Relative Strength Index, or RSI, is a bounded momentum indicator built from recent positive and negative price changes. Its value summarizes the balance and magnitude of those changes under a declared lookback and smoothing method.
RSI describes its input series; it does not measure an asset against another asset despite the word “relative” in its name. A reading is incomplete unless the timeframe, price field, lookback, smoothing convention and timestamp are known.
How is RSI obtained?
- Define the series: select the instrument, timeframe, price field and valid observations.
- Measure consecutive changes: separate positive changes from negative changes.
- Apply the lookback: summarize gains and losses across the declared rolling window.
- Use the stated smoothing rule: update the average gain and average loss consistently.
- Transform the ratio: convert the relative balance into the bounded RSI scale.
RS = smoothed average gain ÷ smoothed average loss. RSI = 100 − (100 ÷ (1 + RS)). The calculation must define its lookback, seed and zero-loss treatment.
What is daily RSI?
Daily RSI applies the indicator to daily observations. When based on closes, each input change compares one daily close with the preceding daily close. The rolling calculation updates when the method accepts a new daily observation.
How is daily RSI obtained?
Build a consistent daily price series, calculate the positive and negative changes, and apply the declared lookback and smoothing method. The daily boundary, timezone and handling of the current unfinished day affect which values are eligible.
What can daily RSI be used for?
It can describe recent daily momentum, compare the current reading with earlier readings generated by the same method and flag a documented threshold event for further investigation. It does not determine that price must reverse or continue.
What is weekly RSI?
Weekly RSI uses weekly observations. Each input represents a weekly period, so the indicator changes more slowly than a daily version calculated from the same lookback count. The two readings summarize different sequences and should not be treated as interchangeable.
How is weekly RSI obtained?
Construct the weekly series under a documented weekly boundary and timezone, calculate consecutive weekly changes, and apply the same declared RSI procedure. A partial week should be excluded, updated or marked provisional according to a consistent rule.
What can weekly RSI be used for?
It can add broader momentum context, show whether a daily reading is accompanied by a similar weekly condition and support comparisons across equivalent weekly samples. Difference between the timeframes is descriptive evidence, not a prediction.
How should RSI reference levels be read?
Charts often place horizontal reference lines on the RSI scale. Their meaning depends on the documented methodology. A reading above or below a line is an observed classification under that rule; it is not proof of valuation and does not guarantee a reversal.
Color can highlight which band contains an observation, but the numerical value and declared thresholds should remain primary. A model should not rely on color alone, especially when charts are reproduced or viewed with accessibility constraints.
Why compare daily and weekly RSI?
A multi-timeframe comparison separates shorter-horizon momentum from a broader weekly reading. The readings may align or diverge because their input observations aggregate price differently. Neither outcome is inherently bullish or bearish without a declared rule.
Historical comparisons should use compatible instruments, sources, timezones, formulas and sample definitions, and always state the sample size. Historical similarity does not guarantee the same future outcome.
Limitations and common mistakes
- Method dependence: lookback, seed and smoothing choices can change the output.
- Timeframe mismatch: daily and weekly readings summarize different observations.
- Incomplete periods: a live day or week can move before its final value is accepted.
- Threshold overreach: crossing a reference line does not guarantee a price response.
- Path compression: one value cannot describe the entire sequence of gains and losses.
- Causal overreach: RSI reflects price inputs and does not explain why they changed.
State the instrument, source, timeframe, price field, lookback, smoothing and seed rules, timezone, cutoff timestamp, threshold convention and treatment of incomplete observations.