Position Size Management – Performance based increases

• 09/09/2019 at 1:50 PM #106961
  deleted271219

  How can these indicators be converted into a code for execute automatically to proorder???

  The consecutive winning streak based risk increment, and the fixed fraction money management.

  Thanks!

  

  deleted271219

  Member

  Topics: 3

  Replies: 21

  Been thanked: 0 times
  09/09/2019 at 2:33 PM #106973
  Nicolas

  Add the code snippet at the top of your trading strategy’s code and use the variable POSITIONSIZE for your orders:

  if buycondition then buy POSITIONSIZE contracts at market endif

  1 2 3

  if buycondition then buy POSITIONSIZE contracts at market endif

   

  

  Nicolas

  Keymaster

  Topics: 81

  Replies: 20863

  Been thanked: 4774 times
  09/09/2019 at 7:08 PM #106996
  Fran55

  And which is the buyconditión in the two cases?

  

  Fran55

  Participant

  Topics: 34

  Replies: 210

  Been thanked: 7 times
  09/10/2019 at 7:43 AM #107025
  Nicolas

  It was just an example, “buycondition” is the condition to initiate a buy order in your own strategy.

  

  Nicolas

  Keymaster

  Topics: 81

  Replies: 20863

  Been thanked: 4774 times
  02/15/2020 at 5:03 PM #119696
  nonetheless

  I’ve been playing with the money management code posted above by @Despair based on fixed ratio theory by Ryan Jones http://www.straightforex.com/advanced-forex-course/money-management/fixed-ratio/

  My very minor modifications:

  MM = 1 //select MM = 0 for optimization if MM = 1 then once multiplier=1 once fraction=1400 //set to 10x margin required once newlevel=1400 once oldlevel=1400 once startpositionsize=1 once positionsize=startpositionsize if strategyprofit>newlevel then multiplier=multiplier+1 oldlevel=newlevel newlevel=strategyprofit+multiplier*fraction positionsize=multiplier*startpositionsize elsif strategyprofit<oldlevel and multiplier>=2 then newlevel=strategyprofit oldlevel=strategyprofit-multiplier*fraction multiplier=multiplier-1 positionsize=multiplier*startpositionsize endif endif if MM=0 then positionsize=1 endif

  1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23

  MM = 1 //select MM = 0 for optimization if MM = 1 then once multiplier=1 once fraction=1400 //set to 10x margin required once newlevel=1400 once oldlevel=1400 once startpositionsize=1 once positionsize=startpositionsize if strategyprofit>newlevel then multiplier=multiplier+1 oldlevel=newlevel newlevel=strategyprofit+multiplier*fraction positionsize=multiplier*startpositionsize elsif strategyprofit<oldlevel and multiplier>=2 then newlevel=strategyprofit oldlevel=strategyprofit-multiplier*fraction multiplier=multiplier-1 positionsize=multiplier*startpositionsize endif endif if MM=0 then positionsize=1 endif

  The idea is to keep the required margin for any position size to within 10% of running profit. This code works well at increasing position size for every £1400 of profit (margin in this case = £136), but fails to reduce accordingly. For example, after a £2800 loss in backtest I would expect to see position size reduce by 2; in fact it is only ever reducing by 1 no matter what size the loss and decrease in profits. Position size then becomes unsustainably large and the whole thing runs into trouble, eating into working capital in order to keep going.

  I’ve looked at all the other MM snippets I could find but this is the one that comes closest to what I want; v grateful for any help in improving it!

  

  nonetheless

  Participant

  Topics: 83

  Replies: 1237

  Been thanked: 978 times
  02/15/2020 at 5:42 PM #119700
  nonetheless

  This is another possibility that comes close, based on a percentage system posted by @robertogozzi

  ONCE MyEquity = 10000 ONCE positionsize = 1 IF Not OnMarket THEN IF Strategyprofit >= (MyEquity + 1400) THEN MyEquity = Strategyprofit positionsize = positionsize + 1 if positionsize + 1 > 55 then positionsize =55 endif ELSIF Strategyprofit <= (MyEquity -1400) THEN MyEquity = Strategyprofit positionsize = positionsize - 1 positionsize = max(1, positionsize) //do not allow nLots to fall below 1 ENDIF ENDIF

  1 2 3 4 5 6 7 8 9 10 11 12 13 14 15

  ONCE MyEquity = 10000 ONCE positionsize    = 1 IF Not OnMarket THEN IF Strategyprofit >= (MyEquity + 1400) THEN MyEquity = Strategyprofit positionsize = positionsize + 1 if positionsize + 1 > 55 then positionsize =55 endif ELSIF Strategyprofit <= (MyEquity -1400) THEN MyEquity = Strategyprofit positionsize = positionsize - 1 positionsize = max(1, positionsize)    //do not allow nLots to fall below 1 ENDIF ENDIF

  But in back testing has the same problem, only reduces by 1 no matter what the drop in profit.

  

  nonetheless

  Participant

  Topics: 83

  Replies: 1237

  Been thanked: 978 times
  02/16/2020 at 12:09 AM #119720
  nonetheless

  This seems to work, if some very clever person wouldn’t mind checking the code for me, thanks.

  Turns out that margin at 10% of strategy profit is unworkable; 3 – 5% is more realistic.

  //Money Management MM = 0 //MM = 0 for strategy optimization if MM = 1 then ONCE startpositionsize = 1 ONCE factor = f*margin // insert margin value and optimize for value of f after strategy optimization, typically between 15 and 30 ONCE maxpositionsize = 55 //IG first tier margin limit IF Not OnMarket THEN positionsize = startpositionsize + Strategyprofit/factor ENDIF IF Not OnMarket THEN if startpositionsize + Strategyprofit/factor < startpositionsize THEN positionsize = startpositionsize ENDIF IF startpositionsize + Strategyprofit/factor > maxpositionsize then positionsize = maxpositionsize ENDIF ENDIF ENDIF if MM=0 then positionsize=1 ENDIF

  1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21

  //Money Management MM = 0 //MM = 0 for strategy optimization if MM = 1 then ONCE startpositionsize = 1 ONCE factor = f*margin // insert margin value and optimize for value of f after strategy optimization, typically between 15 and 30 ONCE maxpositionsize = 55 //IG first tier margin limit IF Not OnMarket THEN positionsize = startpositionsize + Strategyprofit/factor ENDIF IF Not OnMarket THEN if startpositionsize + Strategyprofit/factor < startpositionsize THEN positionsize = startpositionsize ENDIF IF startpositionsize + Strategyprofit/factor > maxpositionsize then positionsize = maxpositionsize ENDIF ENDIF ENDIF if MM=0 then positionsize=1 ENDIF

   

  

  nonetheless

  Participant

  Topics: 83

  Replies: 1237

  Been thanked: 978 times
  02/20/2020 at 11:26 AM #120023
  nonetheless

  Final version

  //Money Management MM = 1 // = 0 for optimization if MM = 0 then positionsize=1 ENDIF if MM = 1 then ONCE startpositionsize = 1 ONCE factor = 15 // factor of 15 means margin will increase/decrease @ 6.6% of strategy profit; factor 10 = 10% 20 = 5% etc — optimize for best result, profit vs drawdown ONCE margin = 137 // enter margin value of 1 contract ONCE maxpositionsize = 27 // DOW $2 IG first tier margin limit ONCE minpositionsize = .5 // enter minimum position allowed IF Not OnMarket THEN positionsize = startpositionsize + Strategyprofit/(factor*margin) ENDIF IF Not OnMarket THEN if startpositionsize + Strategyprofit/(factor*margin) < minpositionsize THEN positionsize = minpositionsize //keeps positionsize from going below allowed minimum ENDIF IF startpositionsize + Strategyprofit/(factor*margin) > maxpositionsize then positionsize = maxpositionsize// keeps positionsize from going above IG first tier margin limit ENDIF ENDIF ENDIF

  1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23

  //Money Management MM = 1 // = 0 for optimization if MM = 0 then positionsize=1 ENDIF if MM = 1 then ONCE startpositionsize = 1 ONCE factor = 15 // factor of 15 means margin will increase/decrease @ 6.6% of strategy profit; factor 10 = 10% 20 = 5% etc — optimize for best result, profit vs drawdown ONCE margin = 137 // enter margin value of 1 contract ONCE maxpositionsize = 27 // DOW $2 IG first tier margin limit ONCE minpositionsize = .5 // enter minimum position allowed IF Not OnMarket THEN positionsize = startpositionsize + Strategyprofit/(factor*margin) ENDIF IF Not OnMarket THEN if startpositionsize + Strategyprofit/(factor*margin) < minpositionsize THEN positionsize = minpositionsize //keeps positionsize from going below allowed minimum ENDIF IF startpositionsize + Strategyprofit/(factor*margin) > maxpositionsize then positionsize = maxpositionsize// keeps positionsize from going above IG first tier margin limit ENDIF ENDIF ENDIF

   

  1 user thanked author for this post.

  Nicolas

  

  nonetheless

  Participant

  Topics: 83

  Replies: 1237

  Been thanked: 978 times
  01/08/2021 at 2:45 PM #156917
  Nobody

  “”  Position Size Management – Performance based increases “”

  Personally i just use a dynamic position size based on equity . Equity rises = size rises at least as a dollar value and naturally Equity drops position size as a dollar value decreases  .  As far as consecutive losses/ wins i fail to see a benefit in changing relative risk    If you have an excessive run of losses relative to the expectancy i dont think you need to be thinking of reducing size , more like turn it of .  You need a probability matrix based from the tested winrate to define whats is an abnormally large run of losses  to determine when something is drastically wrong ( broken) where rectification of parameters etc needs attention .  My dynamic sizing is as simple as  initial capital / close + strategyprofit/close  (adjusted by whatever coefficient to provide whatever leverage you require)   . it works for me and at least  reduce risk of ruin  after equity runup  .  Naturally if the consec loses happens from the getgo it wont help risk of ruin but then you got other issues , a sh*t system .  Anyway i think this is something to consider  . cheers

   

  PS sometimes a large run of losses is just a shit happens where nothing is really wrong with system  , but you need to know at what level of consec losses you need to worry  . Reducing size relative to consec losses might keep you in the game longer but it also means you need a lot more consec winners to get back to status quo

  1 user thanked author for this post.

  Bel

  

  Nobody

  Participant

  Topics: 9

  Replies: 192

  Been thanked: 59 times
  01/08/2021 at 3:35 PM #156928
  Monobrow

  Brisvegas wrote:

  what level of consec losses you need to worry

  Interesting thoughts…

   

  in very general terms, or rather in your experience / with your systems, in how many consecutive losses would you be turning off your systems?

   

  

  Monobrow

  Participant

  Topics: 15

  Replies: 157

  Been thanked: 23 times
  01/08/2021 at 4:06 PM #156933
  Nobody

  Monobrow wrote:

  Brisvegas wrote:

  what level of consec losses you need to worry

  Interesting thoughts…

  in very general terms, or rather in your experience / with your systems, in how many consecutive losses would you be turning off your systems?

  Its not a simple ‘ x’ losses number as a number of factors are needed to ascertain the amount . Expectancy is a factor and that is derived from 2 metrics being win rate (WR) & risk reward ratio   .  Your risk tolerance which forms the risk of ruin (RoR) side of it is another factor   .  The higher your expectancy  the more you can risk per trade as a % of equity without increasing RoR .  The better the win rate the less the probability of  higher consec losses over ‘x’ trades is  . This is whay i always prefer a higher WR system over lower WR given same expectancy  .  Statistically you will have smaller consec lose runs which gives you to risk slighty more per trade without increasing RoR  . Now this is where a probabilty matrix comes in  , you can feed in WR as % and it will give you the probabilty of ‘x’ consec losing trades over ‘y’ amount of trades . using this info you can dial in an optimal position size that allows max equity curve growth while keeping RoR within risk parameters you are comfortable with  .  IE how much you are willing to risk per trade  .  You really need to find a consec loss probability matrix excel spreadsheet to do this  . You might be able to find one if you google it . I dont think i can share an excel file in PRC .        You need to explore the expectancy formula as well and thats easily found in google .   I dont have time to write everything you need to know here and now .  I suggest you get some literature on building quantitative trading systems .

  https://vantagepointtrading.com/what-is-trading-expectancy-and-how-it-works/ this the first thing i found in google search , get busy and see if you can work it out , will try and find a probability matrix online if i can . I am actually trading atm so time poor for such a complex thing right now

  2 users thanked author for this post.

  Bel, Monobrow

  

  Nobody

  Participant

  Topics: 9

  Replies: 192

  Been thanked: 59 times
  01/08/2021 at 4:10 PM #156935
  Vonasi

  Brisvegas wrote:

  I dont think i can share an excel file in PRC

  Excel files can be attached to your posts.

  Please consider this when answering posts Brisvegas:

  • Be careful when quoting others in your posts. Only use the quote option when you need to highlight a particular bit of text that you are referring to or to highlight that you are replying to a particular member if there are several involved in a conversation. Do not include large amounts of code in your quotes. Just highlight the text you want to quote and then click on ‘Quote’.

  

  Vonasi

  Moderator

  Topics: 192

  Replies: 5088

  Been thanked: 1811 times
  01/08/2021 at 4:30 PM #156937
  Nobody

   

  This may help , currently  havent got time to check whether this is exactly whats required , will check over weekend and get back to you once i have time

  https://www.daytradinglife.com/trading-spreadsheets/losing-streak-probability-calculator/

  2 users thanked author for this post.

  Bel, Monobrow

  

  Nobody

  Participant

  Topics: 9

  Replies: 192

  Been thanked: 59 times
  01/08/2021 at 4:32 PM #156938
  Monobrow

  Thanks again, your response gave me the answer i needed.

  I found a website with exactly what you described (https://www.newtraderu.com/2020/07/22/risk-of-ruin-calculator/) and found – through more luck than judgement that the 6 consecutive loss limit I set on my 70% WR aystem was more or less what I needed, in fact. i could probably go up to 7 if needed.

   

  cheers

  

  Monobrow

  Participant

  Topics: 15

  Replies: 157

  Been thanked: 23 times
  01/08/2021 at 4:38 PM #156942
  Monobrow

  Brisvegas wrote:

  will check over weekend and get back to you once i have time

  Thanks Bris. Yeah, I found one before I saw your post. My system is a 24/5 reversal system so I wanted to have something in there to protect against significant and continuous movements in one direction. Im not sure of there is a better way to do it, but a quit on consecutive losses seemed like a place to start.

  

  Monobrow

  Participant

  Topics: 15

  Replies: 157

  Been thanked: 23 times
• Author
  Posts