Constructing simple, exponential, and adaptive averages

Start from one ordered close series

A lookback is the number of sampling intervals used to set the smoother. From an ordered close series, a 14-period simple moving average can be computed. A simple moving average is a fixed-lookback average of ordered closes in which each observation in the window has equal weight.

A 14-period exponential moving average can be computed from the same ordered close series. Exponential smoothing is a recursive average that applies a constant weight to the latest close and the remainder to the previous average.

A 14-period adaptive moving average can be computed from that same ordered close series. An adaptive moving average is a recursive average whose speed is allowed to change with the data so the line can stay nearer to price than a fixed-lookback average.

Plot the three lines on one overlay

The simple, exponential, and adaptive lines can be plotted together so their distance from price can be compared.

On that overlay, the adaptive line stays nearer to price and shows fewer dropoffs and less separation than the simple and exponential lines. A simple moving average can move away from the prices it is averaging.

Adaptive lookback and smoothing-factor

The adaptive construction is specified to use only one prior value of the smoother. A smoothing-factor is a speed constant implied by the chosen lookback.

A 14-period setting yields a 0.15 factor that sits above an earlier adaptive setting, while a 6-period setting sits below that earlier setting. The adaptive line is described as rising when the input series is falling.