Hello everyone,

I’ve been looking for filter to avoid false signals and want to submit this idea for improvement and critical thinking… Please share your thoughts.

Nothing new, just compiling some physics principles. May be something similar exist already.

The two basic principales are the Velocity (https://en.wikipedia.org/wiki/Velocity) and the Momentum (https://en.wikipedia.org/wiki/Momentum), applied to price and volume.

For the price to make a signficant change in a sustainable manner, we need both acceleration in price change with signficant volume.

The problem I found comparing Intraday Volume to previous bars (Volume > Volume[1]) or to Intraday Average Volume or to volume same time on previous day doesn’t work consistently and may be impacted by Geopolitics, Macro, etc. Further, because of the large volume on the X first bars of the day, most of the volume bars are below the the Intraday Average Volume, so the filter is defeated.

Velocity = (Indicator – Indicator[1]) / (T – T[1])

Acceleration = (Velocity – Velocity[1]) / (T – T[1])

One can apply this idea to Close, MedianPrice, Averages, etc. I chose Hull Average in the exemple below because its one of the most sensitive.

Then, to move the price in a sustainable manner => Momentum = Mass * Velocity.

I chose to define Mass as (MedianPrice * Volume) “Mass of Money”.

For Long, if Acceleration (change is Velocity) is higher than its EMA and Momentum is positive and higher than its EMA, then if your Indicator(s), say Moving Average crossing or Stochastics, etc., give a Signal, there is a high probability that it’s not a false signal and you can enter the trade with more assurance. (the opposite for Shorts)

When you’re already in position, this might be used as Exit Signal in the opposite direction.

Please share your thoughts!

DEFPARAM CALCULATEONLASTBARS=500

Period = 1

PRICE = MEDIANPRICE
Mass1  = (PRICE * VOLUME)

av1 = HullAverage[13](close)
av2 = HullAverage[34](close)

Velocity1Av1 = (av1 - av1[Period])/Period
Velocity0Av1 = (av1[1] - av1[Period+1])/Period
Acceleration1Avg1 = (Velocity1Av1 - Velocity0Av1) / Period
EMAAcceleration1 = ExponentialAverage[8](Acceleration1Avg1)

Momentum1 = Mass1 * Velocity1Av1
EMAMomentum1 = ExponentialAverage[8](Momentum1)

Velocity1Av2 = (av2 - av2[Period])/Period
Velocity0Av2 = (av2[1] - av2[Period+1])/Period
Acceleration1Avg2 = (Velocity1Av2 - Velocity0Av2) / Period
EMAAcceleration2 = ExponentialAverage[8](Acceleration1Avg2)


IF Acceleration1Avg1 > EMAAcceleration1 and Acceleration1Avg1 > Acceleration1Avg2 and Momentum1 > 0 and Momentum1 > EMAMomentum1 THEN
DRAWARROWUP(barindex, -1.5) coloured(0,255,0)
ENDIF

IF Acceleration1Avg1 < EMAAcceleration1 and Acceleration1Avg1 < Acceleration1Avg2 and Momentum1 < 0 and Momentum1 < EMAMomentum1 THEN
DRAWARROWDOWN(barindex, 1.5) coloured(255,0,0)
ENDIF

RETURN +2, -2