Risk Management Strategy with d’Alembert pyramid

• 11/30/2020 at 12:14 AM #152156
  Oliviertrader2020

  Hello,

  I want to test the risk management strategy called “Contra d’Alembert” on some of my algos.
  The strategy consists in adding one unit after each victory, and removing one after each defeat.
  Example of a series of 10 trades starting with 5 mini contracts

  1 winning trade (5minis contracts): I add one contract to my position size
  1 winning trade (6minis contracts): I add one contract to my position size
  1 losing trade (7minis contracts): I withdraw a contract at my position size
  1 losing trade (6minis contracts): I withdraw a contract at my position size
  1 winning trade (5minis contracts): I add one contract to my position size
  1 winning trade (6minis contracts): I add one contract to my position size
  1 winning trade (7minis contracts): I add one contract to my position size
  1 losing trade (8minis contracts): I withdraw a contract at my position size
  1 winning trade (7minis contracts): I add one contract to my position size
  1 winning trade (8minis contracts): I add one contract to my position size

  It would be a good idea to have a slider to limit the position size to x mini contracts to avoid having too many contracts in case of a good winning streak.

  Thank you for your help. 🙂

  

  Oliviertrader2020

  Participant

  Topics: 22

  Replies: 49

  Been thanked: 4 times
  11/30/2020 at 12:32 AM #152158
  robertogozzi

  There you go:

  ONCE iLots = 1 //let's start with this number of lots ONCE MinLots = 1 //minimum Lots to be traded (never fall below it) ONCE MaxLots = 50 //maximum Lots to be traded (never exceed it) IF StrategyProfit > StrategyProfit[1] THEN iLots = min(MaxLots,iLots + 1) //winning trade ELSIF StrategyProfit < StrategyProfit[1] THEN iLots = max(MinLots,iLots - 1) //losing trade ENDIF

  1 2 3 4 5 6 7 8

  ONCE iLots        = 1           //let's start with this number of lots ONCE MinLots      = 1           //minimum Lots to be traded (never fall below it) ONCE MaxLots      = 50          //maximum Lots to be traded (never exceed it) IF StrategyProfit > StrategyProfit[1] THEN    iLots = min(MaxLots,iLots + 1) //winning trade ELSIF StrategyProfit < StrategyProfit[1] THEN    iLots = max(MinLots,iLots - 1) //losing trade ENDIF

  iLOTS will retain the correct number of lots to be traded.

  1 user thanked author for this post.

  Oliviertrader2020

  

  robertogozzi

  Moderator

  Topics: 69

  Replies: 17088

  Been thanked: 4006 times
  11/30/2020 at 12:37 AM #152160
  robertogozzi

  This looks more gambling than trading, though!

   

  

  robertogozzi

  Moderator

  Topics: 69

  Replies: 17088

  Been thanked: 4006 times
  11/30/2020 at 9:33 AM #152182
  Nicolas

  The “d’Alembert pyramid” contract size calculation is also available in the ProOrder PDF manual, page 71: https://www.prorealtime.com/en/pdf/probacktest.pdf

  1 user thanked author for this post.

  Oliviertrader2020

  

  Nicolas

  Keymaster

  Topics: 81

  Replies: 20863

  Been thanked: 4773 times
  11/30/2020 at 9:34 AM #152183
  Nicolas

  I modified the Roberto’s code as LOTS is a reserved keyword (changed with iLOTS).

  2 users thanked author for this post.

  Oliviertrader2020, robertogozzi

  

  Nicolas

  Keymaster

  Topics: 81

  Replies: 20863

  Been thanked: 4773 times
  11/30/2020 at 11:06 AM #152191
  nonetheless

  With respect, this seems a rather coarse way of doing money management. A win might be 40€ and a loss might be 400€, but your reaction is the same in each case?

  Try a quick search for money management code – there’s lots of more sophisticated ways of doing it.

  1 user thanked author for this post.

  Oliviertrader2020

  

  nonetheless

  Participant

  Topics: 83

  Replies: 1237

  Been thanked: 978 times
  11/30/2020 at 9:05 PM #152264
  Oliviertrader2020

  Thank you for your help @robertogozzi.:-) You are right, this type of risk management is more like gambling than investing. This is also why I noticed that the objective was only testing and only testing with different strategies. And eventually, I say eventually apply it on a small part of the capital.

  

  Oliviertrader2020

  Participant

  Topics: 22

  Replies: 49

  Been thanked: 4 times
  11/30/2020 at 9:26 PM #152266
  Oliviertrader2020

  Thank you for your recommendation @nonetheless.
  Having said that, I don’t see where the problem is with testing, as indicated in my message.

  Lose 400€ and win 40€?If, as I suggest, you use a limit to the number of contracts, and you have the intelligence to test with setups that have close stop losses, I don’t see how you can make such a difference.

  By trading with the same number of lots in each trade, you can also have losses of €400 if you take a trade with SLs 10 times larger from one trade to the next. Having a setup that provides trades of 20/25pts, and with the added security of a max lot size, a security stop loss at x point/€ and a system stop before the market closes, I don’t understand how you can have such a big difference. I think you’re confusing it with a martingale, a type of system that I disapprove of as much as you do.

  I understand that the title “Risk Management Strategy with d’Alembert pyramid” may mislead people who have not made the effort to read the message. This is why I clearly indicated in my message that it was a Risk Management Strategy based on the “CONTRA d’Alembert”. The “d’Alembert Contra” or “Paroli” risk Management strategy are systems where you increase your investments after gains and reduce them after losses. The goal is to minimize the bad series and optimize the good ones.

  Larry Williams won the 1987 world championship with a similar system that allowed him to transform a $10,000 account into a $1 million capital.

  Of course this type of management is not necessarily applied to all his capital. But I don’t see what’s grotesque about using this type of management on 1 or 2% of the capital by boosting the return on already profitable strategies.

  Sorry if there are errors in this message, English is not my mother tongue. I wish you all a good evening and good trades.

  

  Oliviertrader2020

  Participant

  Topics: 22

  Replies: 49

  Been thanked: 4 times
  12/01/2020 at 12:37 AM #152273
  nonetheless

  No, certainly not grotesque and by all means, run some tests. I am all in favour of altering position size up and down, I just think it should be proportional to the size of the win or loss. In that respect, the d’Alembert seems unrefined, that’s all.

  Do let us know how you get on.

  

  nonetheless

  Participant

  Topics: 83

  Replies: 1237

  Been thanked: 978 times
• Author
  Posts