ALMA: Arnaud Legoux Moving Average

הרעיון במילים פשוטות

ALMA wants both smoothness and speed. It weights the candles with a Gaussian curve: the offset moves the peak of the curve towards recent candles (faster), the sigma sets how wide it is (smoother).

איך זה מחושב

For N candles, give candle k the weight e^(−(k − m)² ÷ (2s²)) with m = offset × (N − 1) and s = N ÷ sigma, then take the weighted average of the closes.

m = offset × (N − 1),  s = N ÷ sigma
w(k) = exp(−(k − m)² ÷ (2 s²)),  k = 0 … N−1 (oldest to newest)
ALMA = Σ w(k) × close(k) ÷ Σ w(k)

איך משתמשים בזה

• As a trend line that reacts quickly without zig-zagging.
• Defaults 9 / 0.85 / 6 are the author's suggestion.

הסקת מסקנות

• Offset closer to 1 = faster, closer to 0 = smoother. Higher sigma = sharper weighting.

ממה להיזהר

• Like every moving average it confirms trends, it does not forecast them.

קצת היסטוריה

Developed by Arnaud Legoux and Dimitrios Kouzis-Loukas in 2009.