What it measures

The Arnaud Legoux moving average is a moving average whose weights follow a Gaussian bell curve across the window instead of a line or an exponential decay. The bell's peak can be pushed toward the most recent bars, which trades smoothness for responsiveness in one dial.

The calculation

For a window of n bars, the weight of the bar i positions back is

w(i) = exp(−(i − m)² / (2 s²)), with m = offset × (n − 1) and s = n / sigma

and the ALMA is the weighted sum divided by the sum of the weights. Offset 0.85 puts the peak near the present; sigma 6 sets how wide the bell is.

Inputs and defaults

  • Window: 9 in the Pine built-in.
  • Offset: 0.85; 1 would make it a near-current-price line, 0.5 a symmetric smoother that lags like an SMA.
  • Sigma: 6; smaller values widen the bell and smooth more.

How it is read

As a smooth trend line, with the usual slope and cross rules. Its appeal is visual: it is noticeably cleaner than an EMA of similar lag.

Where it fails

Three parameters is two more than an EMA has, and each can be tuned to the sample. A clean line also hides how much of the turn it missed; the ALMA is still an average of the past, however elegant the weights. It appears in a small number of the screened strategies.

How the desk reads it

A rule with three smoothing parameters is read with extra suspicion about how many combinations were tried before the published one. The desk asks how far each parameter can move before the result changes; a narrow answer means the result belongs to the parameters, not the market.

In the desk's screen

Of the 614 screened strategies in the library, 11 (1.8%) use this indicator somewhere in their logic.

6 of those were gross-profitable on enough cells to reach the after-fee replay.

The count is read from the compiled code of each strategy in the library; a strategy that computes the indicator by hand under another name is not counted. The desk's own rules are not part of the count.