Flag martingale systems vote results

Do you like this suggestion?

Yes No

Flag martingale systems Discussion

Jun 04, 2010 at 21:23
929 Views
15 Replies
Member Since Aug 20, 2009   216 posts
Jun 04, 2010 at 21:23
Some of the systems shows impressive gains, but took excessive risks using a martingale strategy. Using a python script and some Sql one would be able to analyze the trades to uncover such martingale trades or excessive averaging down. Would myfxbook be able to implement some sort of graph to show risk/increase in account ratio to quantify the risk exposure?
Member Since Jul 31, 2009   1444 posts
Jun 07, 2010 at 08:28
Stephanus, it's definitely possible and will be created.

Thanks for the suggestion!
Member Since Oct 28, 2009   1424 posts
Jun 07, 2010 at 09:11
Risk exposure would be good.
11:15, restate my assumptions: 1. Mathematics is the language of nature. 2. Everything around us can be represented and understood through numbers. 3. If you graph these numbers, patterns emerge. Therefore: There are patterns everywhere in nature.
Member Since Aug 20, 2009   216 posts
Jun 08, 2010 at 20:17 (edited Jun 08, 2010 at 20:21)
Another idea is to factor in the time it took until a security rallied. It becomes very difficult to distinguish between trading insight and random market movements as time moves on. If the majority of the trades rallies within a 10min period it would indicate trading insight.

Thus I propose a trading evaluation parameter: Time till rally.... How long did it take for a trade to rally. It could also be combined with some form of individual draw-down per trade.

60% of the time we have a counter trend, 30% a trend and 10% a break-out. The euro was in a massive consistent down-trend for months now, if all you had done was to go short every single trade would eventually have been a winner. Problem comes with a counter trend. Thus we should also flag accounts that shows huge equity gains if all they did was trade in the direction of the trend. Trends don't last, being able to trade the 60% counter trend is where the real trading skill is demonstrated.
Krysztau
forex_trader_8466
Member Since Mar 10, 2010   68 posts
Jun 09, 2010 at 12:19
Well, what would you call martingale here...
I mean doubling positions for times in a row everytime looser appeare is quite clear.
What if someone happend to put bigger trade as his estimation of chances rised in second trade?

What with softer form of martingale, where one put 1.2 size of loosing trade?

what if you open one position with big SL in a morning and then scalp all morning with bigger position size?


I'm absolutely for flagging, just wanted to post some concerns.

This may be good if flagged user may mark that he does not agree with this judge, so he can explain in his own description...
Member Since Aug 20, 2009   216 posts
Jun 10, 2010 at 07:51 (edited Jun 10, 2010 at 07:54)
Krysztau posted:
    Well, what would you call martingale here...
I mean doubling positions for times in a row everytime looser appeare is quite clear.
What if someone happend to put bigger trade as his estimation of chances rised in second trade?

What with softer form of martingale, where one put 1.2 size of loosing trade?

what if you open one position with big SL in a morning and then scalp all morning with bigger position size?


I'm absolutely for flagging, just wanted to post some concerns.

This may be good if flagged user may mark that he does not agree with this judge, so he can explain in his own description...

You have a point, we should decide via voting with staff on what algorithm to use. If you martingale but keep your trades open for only 5min the risk could be less than leaving it open for 4hours. Excessive averaging down should also be flagged. Most important the draw-down per trade in pips and % terms must be displayed like they do on collective2.

Another issue is opening a cluster of 10 trades within 5min, it actually skews the win/loss ratio, there are lots of nuances to this whole business.
Krysztau
forex_trader_8466
Member Since Mar 10, 2010   68 posts
Jun 10, 2010 at 11:49
There is also other thing like not puting SL, some traders show no drowndown, but their positions between open and close [for 10 pips let's say] plunge 200 pips below zero, what makes them in my opinion useless...
Member Since Sep 04, 2009   879 posts
Jun 10, 2010 at 12:57 (edited Jun 10, 2010 at 12:57)
Ohhh come on, discrimination of a certain trading approach? OK than lets flag out scalpers , grid and swing trade systems as well, bc they can blow out accounts too....
Member Since Mar 07, 2010   257 posts
Jun 10, 2010 at 18:50
it's not discrimination ... people can look at the equity curve and figure out a martingale since martingale only looses once ... anyways it's a good parameter to see if someone is using averaging down or martingale all the time or sometimes ...
Member Since Mar 07, 2010   257 posts
Jun 10, 2010 at 18:52
averaging down + martingale is easy to check ... simply a bunch of trades closes at the same time (same minute or in between minute change ... eg: 5 trades closed at 15:01 and 5 trades closed t 15:02)

also martingale trades will have the multiplier of 2x and each higher positions will be initiated at the lower price ... similar with averaging down since averaging can be treated as martingale with a smaller multiplier or simply same position sizing
Member Since Mar 07, 2010   257 posts
Jun 10, 2010 at 18:56
another idea is to show for each trade the max negative pips and max positive pips since the position was open ... of course since everyone here is using a different broker that would be an estimate since you'd have to choose probably one price feed for simplicity ... (zulutrade shows the max negative and positive while the trade was open)

also maybe some summary index showing a avg max neg vs avg max pos for all positions ... that would show systems that use high SL with tiny TP
Member Since Feb 16, 2010   1332 posts
Jun 10, 2010 at 21:07
One of my strategies shows negative pips but it has nothing to do with martingale, it's sometimes a matter of aggressive MM.
"In trading, winning is frequently a question of luck, but losing is always a matter of skill."
Member Since Mar 07, 2010   257 posts
Jun 11, 2010 at 02:01
of course if you place a trade with 0.01 and loose 1000 pips and if you place a trade with 10 lots and win 10 pips then you'll be pip negative but cash positive ... counting pips makes sense if you count the same sizes of the positions ...
Krysztau
forex_trader_8466
Member Since Mar 10, 2010   68 posts
Jun 11, 2010 at 14:50
Another issue with martingale is that when you open a trae and then close part in history it appears as bunch of trades.
Let's say you opened 12 lot's @1.0, then closed 4 lots @ 0.9980, and then 8 @1.0020 - it looks like martingale but is a proper money management...
Member Since Aug 20, 2009   216 posts
Jun 11, 2010 at 17:14 (edited Jun 11, 2010 at 17:21)
Krysztau posted:
    Another issue with martingale is that when you open a trae and then close part in history it appears as bunch of trades.
Let's say you opened 12 lot's @1.0, then closed 4 lots @ 0.9980, and then 8 @1.0020 - it looks like martingale but is a proper money management...

Trades should only be flagged as martingale or excessive averaging down, if a cluster of trades is opened that exceeds 5 , 6 or some arbitrary % of equity. Opening 30 micro lots isn't the same as 30 mini lots within a 5min bar. Thus staff should put the arbitrary % to a vote or allow user input on a webpage to set some arbitrary limit. You as an evaluator of a system should be allowed to set your own arbitrary martingale or averaging down limit, depends on the server load really.
Member Since Aug 20, 2009   216 posts
Jun 13, 2010 at 20:16 (edited Jun 13, 2010 at 20:17)
Another issue would be to flag trades that remain in excessive negative equity for to long a period lets say two days. One needs to differentiate between random market flux and trading insight. If the individual trade draw-downs are to large and period in negative equity is long an arbitrary score should be assigned to a trading system. Collective2.com has a grading index. We should agree on some method of scoring a system. It should not though excessively penalize a system, lets say the system only had a 50% exposure three months ago for one day then the scoring system should 'soften' such negative input etc, because traders gains new insight as time moves on.
Sign In / Sign Up to comment
You must be connected to Myfxbook in order to leave a comment
*Commercial use and spam will not be tolerated, and may result in account termination.
Tip: Posting an image/youtube url will automatically embed it in your post!
Tip: Type the @ sign to auto complete a username participating in this discussion.