Direct answer
Settings change an SMA (Simple Moving Average) by changing how much recent price action influences the average. In practice, this alters the average’s sensitivity (responsiveness), its lag (delay), and how often it may appear to “turn” during noisy price movement.
Mechanism and definition
An SMA is a line computed from past observations of a price. The most common setting is the window length (often called the period): it specifies how many recent time points are averaged.
A simple way to see the mechanics is to use a fixed step size (for example, each step is one candle). If the SMA window length is N, then at time t the SMA is the mean of the last N observed prices:
- SMA(t) = (price(t) + price(t−1) + … + price(t−N+1)) / N
Two practical implications follow:
- Responsiveness: With a smaller N, the SMA reacts more quickly when prices change because fewer data points dominate the average.
- Smoothness and lag: With a larger N, the SMA changes more slowly because older prices still weigh into the average, which tends to reduce short-term fluctuation but increases delay relative to the latest price.
Some platforms also allow additional choices that affect results, such as the input series (close vs. another price field) and the time frame (how the price is grouped into intervals). Those choices do not change the averaging logic, but they do change the data being averaged, so the output line can look different.
Evidence and example (with explicit assumptions)
Assume a simplified sequence of daily closes over 10 days: 100, 100, 100, 100, 110, 112, 113, 112, 111, 110.
Now compute two SMAs at the time when the jump starts (day 5), using the window length as the only change:
- Short window (N=3): the SMA on day 5 averages days 3–5: (100 + 100 + 110) / 3 ≈ 103.3
- Long window (N=5): the SMA on day 5 averages days 1–5: (100 + 100 + 100 + 100 + 110) / 5 = 102
Both averages rise, but the shorter window rises more because it includes fewer pre-jump values. If you continue computing through days 6–10, the shorter window will usually show more “wiggle,” while the longer window will usually look smoother and lag the turning points.
This illustrates the general trade-off: settings that increase sensitivity can also increase the number of apparent turns during noise, while settings that smooth more can delay response.
Limitations and risks (material failure modes)
Even with correct calculation, SMA settings can produce misleading interpretations:
- Sideways or range-bound movement: In choppy ranges, an SMA may repeatedly flatten and slope up/down, creating many apparent “changes” that do not reflect sustained movement.
- Regime shifts and breakouts: When the underlying behavior changes (for example, volatility increases), the lag from a longer window can cause the SMA to reflect the new regime after the move is already underway.
- Noise and overlapping observations: Because the SMA is an average, it can smooth away the details you might need to separate signal from noise, depending on your time frame and input series.
- Costs and execution differences: Any real-world comparison of “what would have happened” depends on transaction costs, spreads, slippage, and execution timing. Historical relationships do not establish future results.
Because of these limitations, it’s better to treat SMA output as a descriptive tool rather than a standalone predictor, and to verify using the specific data choices you plan to use (time frame, input field, window length) under your own assumptions.
Verification and next question
To independently verify how settings change an SMA, test multiple window lengths on the same historical dataset and compare:
- how quickly the SMA line turns after a price change,
- how much it fluctuates during periods with no sustained trend,
- and how sensitive the interpretation is to time frame and input choice.
If you want, you can next clarify which setting your platform uses (for example, window length, time frame, and which price field is averaged), then we can map that choice directly to how it affects lag and smoothness.