Direct answer
Changing the settings of RSI and MACD mainly changes responsiveness (how fast the indicator reacts), smoothing (how much recent movement is averaged), and scale (how the indicator values and histogram behave). Those changes do not make the indicators “better” in all conditions; they trade off detail versus noise.
Mechanism and definitions
RSI settings
RSI (Relative Strength Index) is typically computed from two quantities over a lookback period: average gains and average losses, which are used to form a relative strength value and then map it to a bounded oscillator (commonly 0 to 100).
When you change the RSI lookback period (for example, a shorter vs. longer number of bars), you change which price history the averages reflect. A shorter period uses fewer observations, so it can react more quickly to recent shifts. A longer period uses more data, which usually makes the RSI move more slowly.
MACD settings
MACD (Moving Average Convergence Divergence) is built from moving averages. In its common form, it compares a faster moving average with a slower moving average, then derives a signal line (often another moving average of the MACD line). The histogram is often the difference between MACD and the signal line.
When you change MACD settings (typically the fast, slow, and signal moving-average lengths), you change:
- Fast vs. slow separation: how large or small the MACD line tends to be.
- Smoothing and timing: how quickly the MACD line responds.
- Histogram behavior: because the signal line moves at its own pace.
Evidence or example (with clear assumptions)
Assume you keep the same price series and only change one parameter.
- If you compute RSI with a shorter lookback, small recent runs of gains or losses influence the averages more strongly. As a result, RSI may swing more rapidly.
- If you compute RSI with a longer lookback, those swings are diluted because each new data point affects a larger rolling window.
For MACD:
- With shorter fast/slow lengths, the MACD line generally tracks the price changes more tightly, and the signal line may lag less or more depending on the chosen signal length.
- With longer lengths, MACD tends to smooth out short-term movements. That can reduce abrupt changes, but it can also delay turning points relative to the underlying price movement.
A practical implication of these mechanics is that settings change the indicator’s “time sensitivity.” More sensitivity often means more signals that may reflect short-term randomness rather than a sustained shift.
Limitations and risks
- Noise vs. responsiveness trade-off. Shorter RSI periods and faster MACD settings often increase reaction speed but can also increase false positives from ordinary fluctuations.
- Market regime changes. The statistical relationship between price behavior and oscillator movement can differ across trending, ranging, high-volatility, and low-volatility regimes.
- Non-comparable backtests. Even if historical patterns look consistent, results can differ because of transaction costs, bid/ask effects, execution timing, and the exact data used.
- Indicator value comparability. Changing settings changes the indicator’s meaning and scale, so thresholds used under one configuration may not transfer cleanly to another.
- No standalone certainty. RSI and MACD are descriptive transforms of price, not guarantees about future direction.
Verification and next question
To independently verify how settings change behavior, compare RSI and MACD on the same dataset using at least two configurations (one more sensitive, one more smooth). Focus on qualitative effects first: how quickly the oscillator turns, how often it oscillates, and whether peaks/troughs align with broader swings.
A useful next question is: how does indicator behavior change across different timeframes? Timeframe changes can alter the effective lookback context and further modify sensitivity.