Change a losing system into a Winning system | Page 3 | Forex Factory

Options

Go to
Permalinks
Print Thread

Similar Threads

Transition from Losing to Winning 139 replies

Is opposite of losing strategy a winning strategy? 21 replies

Close Losing Position, Set Trailing Stop to Winning Position 2 replies

Script/EA to balance losing trades agianst winning trades 23 replies

Scaling: In/Out. Winning/Losing 45 replies

Change a losing system into a Winning system

  • #41
  • Sep 3, 2009 6:16am Sep 3, 2009 6:16am
helto there Tjpld :


hope you can share some of the properly functioning but non-performing EA's that you have encountered (via e-mail , PM or posting here). Let me have a chance at optimizing them and see where that leads.

Did not get what you mean when you said that's going to take a while.

Cheers any way you responded, so there is still a heartbeat in the thread
  • #42
  • Sep 3, 2009 7:28am Sep 3, 2009 7:28am
Quoting pathumracled
helto there Tjpld :


hope you can share some of the properly functioning but non-performing EA's that you have encountered (via e-mail , PM or posting here). Let me have a chance at optimizing them and see where that leads.

Did not get what you mean when you said that's going to take a while.

Cheers any way you responded, so there is still a heartbeat in the thread
http://www.forexfactory.com/attachme...2&d=1251979712
  • #43
  • Sep 4, 2009 1:26am Sep 4, 2009 1:26am
Quoting Subdude
This really reminds me of my experience as of late. Every time the market moves against me, I wait until I just can't take any more unrealized loss, then reverse the position hoping that the momentum will help me recover some of the loss, get me to BE, or even profit. Soon after I do this, the market reverses its direction and instead of putting me back in the green, exacerbates the situation. Worse yet, I come to realize that had I kept the original position, I would not lose or even gain from it. Effectively, the initial bad decision soon snowballs...
Hey Subdude,

I think you truly understand whipsawing now...right?
Good luck,
Joey
  • #44
  • Sep 4, 2009 2:47am Sep 4, 2009 2:47am
Quoting pathumracled
helto there Tjpld :


hope you can share some of the properly functioning but non-performing EA's that you have encountered (via e-mail , PM or posting here). Let me have a chance at optimizing them and see where that leads.

Did not get what you mean when you said that's going to take a while.

Cheers any way you responded, so there is still a heartbeat in the thread
No sorry I only have profitable EAs. Damn it! I better drop them and look for losing EAs instead....
  • #45
  • Oct 13, 2009 4:41pm Oct 13, 2009 4:41pm
I know a lot of people think this idea is stupid, but I think it honestly has merit. The thing is if you reverse it, you need to reverse EVERYTHING. I have had several systems where I would have stop out after stop out of 5-10 pips, with an occasional winner of 25-50 pips. So, even after the spread, it should produce winners of 3-8 pips (with a 2 pip spread) and an occasional loser of 27-52 pips.

Because my system consistently lost money (and not just 1 or 2 pips, always 5+), and my stop loss was completely mechanical (move to breakeven at 5, +4 at 10, then trail 40% of the move after that), it would seem reasonable to do the opposite. So, whenever a position when negative against you, you trail the tp to BE, then start to trail it to lock in a loss. You exit at the same time you would if trading the other way (typically for just 5-10 pips profit), and most moves end up with some profit or at BE, with the occasional loser. This trail would lower your losses, as most times it returned to hit your stop loss, and rarely hit the overall tp (now stop loss) of +50 (now -50).

Haha, I'm not sure if this made any sense. But I think the idea has more merit than people want to give it. The thing is, we need a COMPLETELY mechanical system that fails consistently (not just a poor trader to reverse trade). And the system (including tp/sl/and any moves of these to be, lock in pips, etc.) needs to be completely reversed.

Please feel free to give any input. If anyone has a consistent losing system (with decent sized losses to cover the spread 10+ would be optimal) let me know. I'm NOT trying to make an EA.. trading with a robot is not intelligent; I just want a completely mechanical system that can be traded manually.
  • #46
  • Oct 13, 2009 4:48pm Oct 13, 2009 4:48pm
Laff. Markets aren't straight up and down, they wiggle. You can't just reverse and expect 1-1/N results.
  • #47
  • Oct 13, 2009 5:16pm Oct 13, 2009 5:16pm
Quoting Rabid
Laff. Markets aren't straight up and down, they wiggle. You can't just reverse and expect 1-1/N results.
You can if you account for double the spread, and reverse everything. I don't think you understand what I'm saying; I'm not saying "buy where you sell" and keep your tp and sl at the same place. The reason people lose is the exit, not the entry! If you reverse your stop loss/take profit management as well (which is counter-intuitive, and effectively changes the greed/fear dynamic if you think about it), it should yield the opposite results minus double the spread.

So if you have a system that is completely mechanical (not just entries, exits as well), which yields results like this:

-10
-15
-10
-12
-14
-8
+34

-59 pips + 34 pips
net: -25

It would end up like this:
+6
+11
+6
+8
+12
+4
-36

+ 47 pips - 36 pips
net: +11

This is with a 2 pip spread; could be much better on e/u with a 1 pip spread.

We just need to find a consistently poor performing completely mechanical system. I understand the math here, please take some time and consider what I said before responding with the typical "you can't do that" without doing any testing.
  • #48
  • Oct 13, 2009 5:19pm Oct 13, 2009 5:19pm
Quoting saint_berzer
This reminds me of the Seinfeld episode where George decides to do everything opposite....because his life has been a failure. Then everything happens his way....freaking hilarious.

Good idea!
That was a kickass episode. One of my favorite episodes. Freaking hilarious.
  • #49
  • Oct 13, 2009 5:42pm Oct 13, 2009 5:42pm
Quoting sindutzz
If we have a losing system why dont we just do a reverse trade ?
Or if we have a losing EA, why don't we just edit the EA so when the EA give a BUY signal then change it into SELL signal. And do same for SELL signal.

Has anyone ever consider this?

Sorry for this stupid question
Even a "losing" system with proper money management can be a "winning" system.
  • #50
  • Oct 13, 2009 8:47pm Oct 13, 2009 8:47pm
Quoting Cbh123
You can if you account for double the spread, and reverse everything. I don't think you understand what I'm saying; I'm not saying "buy where you sell" and keep your tp and sl at the same place. The reason people lose is the exit, not the entry! If you reverse your stop loss/take profit management as well (which is counter-intuitive, and effectively changes the greed/fear dynamic if you think about it), it should yield the opposite results minus double the spread.

So if you have a system that is completely mechanical (not just entries, exits...

#$# Reverse signalling or Reverse trading? The idea has more than merit. Just takes a lot of time to undestand your expert, unless your wrote it yourself. In which case you would not need to worry too much about the nay-sayers. We reverse because (mong other things) a large number of the freely available EA's, will loose more than the win. if yu cannot program then your chances of geting a system that has flipped is greater than our finding one (inexpensively) that will be on the extreme right of the bell curve of profitablility.

In Short, you can profit from reversing. So keep on track and see. There is some good work in this regard on FF just deligenlty look arround. You'll be surprised.

Just pay keen attention to What you acually have left at the end of the day. Account balnace (as seen from back tests) whist in open trades can be the greatest of the deceivers, so pay keen attention to the Forward tests and on the nature of your Free Margin. If you are running to larger a negative P/L, for twoo long, look out your lookig to make a tent or Roof with your reversed trades. That done, then go live with the smallest practicable live lotizes, that suit your system.

Most of it you'll need to prove to and for your self, after all who will benefit or loose from your exploit on this "hair-brained" idea, that so many have taken up and abandoned? Fortune favours those who venture.
In Forex we either Buy, Sell or Stand aside. But profit lies only to the extreme right of the chart, the part thay yields the next pip. Good luck and keep working on this idea.
  • #51
  • Oct 14, 2009 5:49am Oct 14, 2009 5:49am
Quote
You can if you account for double the spread, and reverse everything. I don't think you understand what I'm saying; I'm not saying "buy where you sell" and keep your tp and sl at the same place. The reason people lose is the exit, not the entry! If you reverse your stop loss/take profit management as well (which is counter-intuitive, and effectively changes the greed/fear dynamic if you think about it), it should yield the opposite results minus double the spread.

I understand what you're saying, it's been tried... doesn't work. Mostly because price doesn't move in a straight line. But also because a lot market movement is purely random.

Take an EA sometime and try it. Doesn't have to be a total loser, just any EA. If your theory is correct then it should turn any mildly losing EA into a mildly winning one. You can see the wiggle from backtests, and the randomness from forward testing.
  • #52
  • Oct 14, 2009 1:04pm Oct 14, 2009 1:04pm
Quoting Rabid
Doesn't have to be a total loser, just any EA. If your theory is correct then it should turn any mildly losing EA into a mildly winning one. You can see the wiggle from backtests, and the randomness from forward testing.
I appreciate your input Rabid. I think what your saying is correct; a mildly losing one wouldn't work. What I am saying is we need one that consistently loses, so much so that even after subtracting totaltrades*spread*2 from -1 * the loss, it still would have a positive expectancy. If the market is completely random than there aren't winning strategies either. It doesn't need to be an EA, just needs to be completely mechanical. Basically I'm trying to reverse poor trade management; not just poor entries.

Basically, if a EURUSD based strategy gives 10 trade signals in a session, it needs to result in losing MORE than 20 pips per session consistently. Because that is the negative difference flipping a strategy (including flipping all exits so you enter and exit at exactly the same time you would have otherwise) would cause.

The question isn't whether or not this would work. The question is whether there is a consistenly poor enough system (in different market conditions of course) to make it work.
  • #53
  • Oct 14, 2009 6:38pm Oct 14, 2009 6:38pm
Quote
The question isn't whether or not this would work. The question is whether there is a consistenly poor enough system (in different market conditions of course) to make it work.

Search for the unholy grail?
  • #54
  • Oct 14, 2009 6:42pm Oct 14, 2009 6:42pm
Quoting Rabid
Search for the unholy grail?
I love it. That is perfect. Hopefully it can prove more fruitful than the search for the holy grail.

You would think there would be a system that always loses.. who knows.. most people probably lose because of poor money management. But I know I've failed trying several systems, and I used good mm principles.(never risk more than 1% of account, etc.)

Could someone please give us a complete loser system?! Is that so much to ask?
  • #55
  • Oct 14, 2009 7:09pm Oct 14, 2009 7:09pm
Quoting GulfCoastPip
Even a "losing" system with proper money management can be a "winning" system.
This is not correct, its a myth. A negative expectancy system will always remain a negative expectancy system. Money Management cannot change that fact

Ironically, a positive expectancy system can very easily end up losing money through inappropriate money management.

How many times does this need pointing out, there's hardly a week goes by withous someone trotting out this same old line.
  • #56
  • Oct 15, 2009 8:41am Oct 15, 2009 8:41am
I suspect a consistently losing strategy will be as hard to find as a consistent winner.

As for good MM, good MM won't turn a losing strategy into a winner but bad MM can turn a winning strategy into a loser. There are a lot of systems out there that can be profitable with the judicious use of MM.
  • #57
  • Oct 15, 2009 9:04am Oct 15, 2009 9:04am
What is the horizon of the system you are trading and have you considered stretching it. By this I mean are you trading the system on a timeframe with a restricting SL that is causing losses based on the fluctuations of the market.

Also how indepth was you back testing and did/does it hold a viable edge? Have the results from backtesting replicated in any way with the forward testing phase?

Some models, while they test great over a period, have a period of inactivity and periods where you are out of the money and taking losses. Just because you are in one of these periods doesnt mean that it wont come out in the end producing profits, therefore the adage of just changing from a buy to sell to see if the results will change is very short sighted if you testing proved to be adequate.

Like in business there will be times when shit happens and it seems the whole world is falling around you. If your looking at trading as a long term prospect, as in years, then a short lean period is part of business and you have just better get used to it, IT will happen again and again.

If your foresight is shortsighted then you will probably become a victim of the ever changing world of emotional baggage, in which there is no model that will suffice because at moments of instability and losses (causing emotional pain) you will once again bring the prospect of "if only I bought all those times instead of sold, gee Id be rich".

Develope a plan, trade the plan and buckle yourself in for the long haul.
  • #58
  • Oct 15, 2009 9:06am Oct 15, 2009 9:06am
Quoting Rabid
I suspect a consistently losing strategy will be as hard to find as a consistent winner.

.
Quite true. This year in the football tippings (AFL for the aussies here) I tried to do the opposite and pick the loosers for every round. You wouldnt believe but seriously I averaged the same number as the guys trying to pick all the winners. Go figure
  • #59
  • Oct 16, 2009 12:06am Oct 16, 2009 12:06am
[quote=Rabid;3150141]

But bad MM can turn a winning strategy into a loser. [quote]

I think bad management of exits (sl, tp, whatever exit method) can make a system bad as well.

This is what I'm looking for. A system which has poor trade management (slightly different than money management), which makes it a consistent (big enough) loser (for example, moving the stop to breakeven to early, etc.)

If we can find a system like this, where the exit causes you to lose more than 2*spread per trade over a long period; that would make a winning system when reversed.

I think this way of thinking is counter-intuitive; but maybe that's why it could work. For example, maybe a system puts a trailing stop; and that ends up killing all the good trades to early and doesn't give them room to breath.

This would mean a reversal would have a trailing take profit; which after being set to breakeven, would lock in a bigger loss as the trade moves against you! Sounds crappy right? Not so fast; if the trade is 20 pips against you and normally you would exit, but this strategy has your trailing tp at -5, then most of the time it will hit your tp and save you 15 pips loss! This is because you know from losing with the same strategy that your trailing stop got hit too often, causing you to miss out on pips.
This reversal would end up reducing the size of your losses.

On the other hand you would have a stop loss based on your previous take profit, so if you were going for 50 before, you would risk losing 54. If your system was bad though; it hardly ever hit +50 for you, so you know the likelyhood of getting -54 is very low.

If your old system in most cases ended up exiting with -10 or so pips; you know most times you will exit (based on the same criteria; crossover or what have you), with +6 pips.

Basically this strategy would take greed/human emotion based position management (moving stops too early, etc.), and turn it upside down, letting the statistics work for us.

Has anyone had a strategy that consistenly lost pips so we can try this out?
  • #60
  • Oct 16, 2009 12:21am Oct 16, 2009 12:21am
[quote=Cbh123;3152262]
Quoting Rabid



Basically this strategy would take greed/human emotion based position management (moving stops too early, etc.), and turn it upside down, letting the statistics work for us.

Has anyone had a strategy that consistenly lost pips so we can try this out?
Basically this strategy would take greed/human emotion based position management (moving stops too early, etc.), and turn it upside down, letting the statistics work for us.

No strategy can automatically take out the element of human emotion, it is something that each trader needs to find for themselves.

Has anyone had a strategy that consistenly lost pips so we can try this out?

Do you believe that finding a constant loosing system, that is a system that even with sound MM and controls still looses, and simply changing the theory of reverse orders is going to be any easier than finding a winning strategy.

I do look forward to seeing this!
0 traders viewing now