Hello everyone,

Here is a strategy that mixes the « % Bollinger » indicator, with the trend.

One of my readers made me discover an ebook of Charles Dereeper, who spoke of it. The concept does not come from him, it is well known.

But I found the idea of ​​associating this indicator with the trend very relevant, so I tried to code this strategy.

As a reminder, the « % Bollinger » indicator is defined by :

% Bollinger = (closing price – lower Bollinger) / (upper Bollinger – lower Bollinger)

The ProRealTime indicator is coded by myself like this :

// Indicateur % Bollinger : pB
// paramères : période et déviation : libre à vous de changer ces valeurs
p = 10
dev = 1.5

BollInf = Average[p](close)-dev*std[p](close)
BollSup = Average[p](close)+dev*std[p](close)
pB = (close - BollInf) / (BollSup - BollInf)

return pB

We will associate this indicator with the trend.

In bull market, we go to purchase if :

close > MA100

MA20 > MA200

% Bollinger < 0.2 and % Bollinger < 0.2 on previous candle (consolidation)

We close the trade if :

% Bollinger > 1

As you can see on the following test (S&P500 INDEX, daily chart, spread 1 point), the strategy is very successful in the bull market : profit factor> 7 !

However, there is very few position (very selective criterias), which means that there are not many gross profits, much less for example than the Swing S&P500 strategy that I personally use.

Quite sad, as I’m always looking for concepts that can improve my own algorithms.

Actually, it would be quite possible to attempt to adapt this strategy to short positions, but it seems less profitable.

You are free to try to improve this strategy.

On the S&P500, it was profitable.

It is quite possible that the parameters for the CAC40, the DAX, the stocks, etc. can be adapted.

It’s up to you.

Best regards and happy trading !

// % Bollinger en tendance

Defparam cumulateorders = false

// Taille des positions
n = 1

// Indicateur % Bollinger : pB
// Paramètres 5/1
BollInf = Average[5](close)-1*std[5](close)
BollSup = Average[5](close)+1*std[5](close)
pB = (close - BollInf) / (BollSup - BollInf)

// ACHAT
ca1 = close > average[100](close)
ca2 = average[20](close) > average[200](close)
ca3 = pB[1] < 0.2 and pB < 0.2

IF ca1 and ca2 and ca3 THEN
 buy n shares at market
ENDIF

// SORTIE ACHAT
IF pB crosses over 1 THEN
 sell at market
ENDIF