BardParticipant
Master
Hi
I was reading Shay Campbells’ E-Mini Swing Trading System: http://timingcharts.com/articles/2017/10/20/e-mini-sp-500-swing-trading-system/
where he uses a technique to enter the market based on pull backs of the price based upon a Moving Average Distribution indicator: Pls see image.
“Using a very short moving average of median price does a good job of defining a usable equilibrium point. A distribution can then be calculated around that equilibrium by subtracting the equilibrium point from the market price.
To normalize these readings for volatility, this result is then divided by the recent daily range of price.
Buy Signal= Trend Direction is Up (6 to 9 month Mov Ave) and the MA Distribution registers a significant Pullback, i.e, negative value.
Sell Signal= Trend Direction is Down and the MA Distribution registers a significant Pullback, i.e, positive value.”
I was attempting to re-create that indicator but wondered how as I’m not a coder. My attempt based on other indicators is not working but I hope I’m thinking along the right lines.. unless I have over complicated it?
Cheers
DEFPARAM CalculateOnLastBars = 100
Period = 20
MAPeriod = 20
ATRPeriod = 20
r = 0
g =0
Ave = Average[MAPeriod](MedianPrice)
ATR = AverageTrueRange[ATRPeriod](Close)
S = 0
FOR i = 1 TO Period
TP = TotalPrice[i]
Dist = (TP - Ave)/ATR > 0 // 1 = TP > Ave
IF Dist = 0 THEN
Dist = (TP - Ave)/ATR < 0 * -1 // -1 = TP < Ave
ENDIF
S = S + Dist //Sum up all Dist’s
NEXT
Dist = Dist*(s) //Histogram
Ave = Average[MAPeriod](MedianPrice) //Signal Line
Return Dist coloured(r,g,0) style(histogram) as "Moving Average Distribution", Ave AS "Signal Line", 0 AS "Zero"
this result is then divided by the recent daily range of price.
daily range is not the ATR, it is called the ADR (an average of the last X days range). “Recent” would mean how many period?
BardParticipant
Master
Oh, right, thanks. In the article Shay doesn’t specify exactly other than to say recent. It would make sense to keep the ADR in relation to the moving average for the distribution, no? He writes:
The MA Distribution tells us what’s happening in the very short term using just a couple weeks of data.
So I guess 20 days?
Ok, I took this value in the below code, you can change it of course. The indicator makes sense since it finds easily the pullbacks when you are in a well established trend.
//PRC_MovingAverage Distribution | indicator
//05.03.2019
//Nicolas @ www.prorealcode.com
//Sharing ProRealTime knowledge
// --- settings
ShortMAPeriod = 20
ADRperiod = 20
// --- end of settings
//compute ADR
sum=0
for i = 0 to ADRperiod-1 do
sum=sum+(dhigh(i)-dlow(i))
next
adr=sum/ADRperiod
shortMA=average[ShortMAPeriod](medianprice)
diff=medianprice-shortMA
MAD=(diff*100)/adr
return MAD style(histogram)
If you find it correct according to the original description, I will put it into the library for everyone to benefit from this concept.
BardParticipant
Master
😆
The indicator is now available in the library: Moving Average Distribution
