What it measures

Wilder's moving average is an exponential average with a smaller smoothing constant than the standard EMA. J. Welles Wilder used it inside every indicator he published in 1978, so it is the smoothing inside the RSI, the ATR and the ADX, even when the chart never shows it. In Pine Script it is ta.rma.

The calculation

RMA today = α × price today + (1 − α) × RMA yesterday, with α = 1 / length

Compare the EMA's α = 2 / (length + 1). The same length gives roughly half the weight to the new bar, so a 14-period RMA responds about as fast as a 27-period EMA. When someone says "14-period RSI", the 14 is this smoothing.

Inputs and defaults

  • Length: inherited from the indicator that uses it, 14 in all of Wilder's own.
  • Source: gains and losses for RSI, true range for ATR, directional movement for ADX.

How it is read

Rarely on its own. Its role is to make the indicators built on it steadier, and understanding it explains why the RSI reacts more slowly than its length suggests and why the ATR takes a long time to come down after a volatility spike.

Where it fails

The slowness is the point and the problem: after a volatility shock the ATR stays elevated for many bars, which inflates every stop sized from it, and the RSI needs a sustained move to reach its thresholds. Strategies that switch the RSI to an EMA or SMA basis get a faster, noisier oscillator and usually a different backtest.

How the desk reads it

The desk does not tune the smoothing inside RSI, ATR or ADX; it tests the rule that uses them. A strategy that swaps Wilder's smoothing for an EMA or SMA has changed the indicator, and its backtest is of the new one.

In the desk's screen

Of the 614 screened strategies in the library, 49 (8.0%) use this indicator somewhere in their logic.

16 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.