Why inverting trades doesn't work? | Page 3 | Forex Factory

Options

Go to
Permalinks
Print Thread

Similar Threads

Why do indicators not work/work? Why is Forex not truly random? 129 replies

Inverting the chart 3 replies

Inverting currency pairs in mt4? 2 replies

Overlaying/Inverting Indicators? 3 replies

Why doesn't it work more recently? 17 replies

Attachments: Why inverting trades doesn't work?
Exit Attachments

Why inverting trades doesn't work?

  • #41
  • Aug 9, 2010 11:02am Aug 9, 2010 11:02am
Quoting Trad3r
You can see in the backtest reports how sell limit and buy limit orders of the reversed EA are every day at EXACTLY the same price than the buy stop and sell stop orders of the original EA respectively, but because the spread, orders are hit at different times, and this makes that in the end both EAs end up losing money, in the test the total amount of orders are different, but if you look order by order you can see the effect of the spread.

The spread when I have made tests is 1.1 pip in both versions, and this is a generous spread, imagine...
You could program your EA so that it doesn't set limit orders, but sets market orders when certain events happen. This way you avoid the "time-issue" you mention, and you will have a method that is both practically as well as theoretically a true inverse.
The nail that sticks out gets hammered back in
  • #42
  • Aug 9, 2010 11:12am Aug 9, 2010 11:12am
Quoting Noloqy
You could program your EA so that it doesn't set limit orders, but sets market orders when certain events happen. This way you avoid the "time-issue" you mention, and you will have a method that is both practically as well as theoretically a true inverse.
No, you can't, if you want to open trades at the same time, you have to open them at different prices, prices with the spread against you. This is the reality. Here we are talking about true reversing methods, with same prices, but you can modify my EA if you want to open trades at the same time with the price you get at those moments and you will see how you end up losing again.
  • #43
  • Aug 9, 2010 2:34pm Aug 9, 2010 2:34pm
Quoting Trad3r
This post is for you and for Craig and Razzle, here you have a simple EA coded by miself (you can edit it and fix mistakes, if they are) that is not random at all, a system with fixed entries, and the reversal method, it is a daily breakout EA:
You are failing to understand the simple fact that your EA is actually RANDOM. Your belief that your entry signal is non random just because its based on some arbirary rules is simply incorrect. Just because you have an arbitary "rule" does not make the entry non random.

If the system had a positive edge of sufficient magnitude to overcome transaction costs, then you'd have a rising equity curve. If it had a negative edge of sufficient magnitude then reversing it would lead to a rising equity curve.

I'm not suggesting for a moment that markets are random, because they are not, but your entry (and exit) almost certainly is.
  • #44
  • Aug 9, 2010 2:39pm Aug 9, 2010 2:39pm
Quoting McKeen
Thanks so much in advance to anyone who might shed some light to a dog trying to understand the laws of financial instruments.
I think most of it has already been said.

Spread is the silent enemy, whose effect is significantly underestimated by many folk. And as Razzle points out, a genuine edge is necessary, one that exploits a real market inefficiency, and would consequently be borne out by the test results. A system that's marginally profitable (even ignoring spread) would need likely need several hundred trades to generate a result that exceeds the boundaries of statistical error, and thereby prove the validity of the edge.

Are you inverting your exits (if so, how?), or merely your entries? If the latter, then (for example) there's a fair probability that a long and a short trade, entered at the same time/price, will both get stopped out, if the market zigzags sideways; I'm seem to be forever reading the generalization that forex moves sideways 70%-80% of the time. If you're making your TPs your SLs, and your SLs your TPs, then that's a 'valid' inversion. Using any kind of trailing SL, or complex scaling out, could be rather more difficult to invert.
  • #45
  • Aug 9, 2010 3:30pm Aug 9, 2010 3:30pm
Quoting Razzle
You are failing to understand the simple fact that your EA is actually RANDOM. Your belief that your entry signal is non random just because its based on some arbirary rules is simply incorrect. Just because you have an arbitary "rule" does not make the entry non random.

If the system had a positive edge of sufficient magnitude to overcome transaction costs, then you'd have a rising equity curve. If it had a negative edge of sufficient magnitude then reversing it would lead to a rising equity curve.

I'm not suggesting for a moment that markets...
OK, I'm not going to discuss with you the definition of randomness, just give me an EA with a true non-random system according to you and I will reverse it to continue arguing with proves, not with beliefs.
  • #46
  • Aug 9, 2010 4:01pm Aug 9, 2010 4:01pm
Are markets random? Yes and no.

NO, in that there's a rationale behind every order that's placed; but

YES, in that it's nonetheless impossible to know the rationale of, and action taken by, every participant, at a given point; but

NO, in that a higher number of these orders are invariably placed at or near certain key levels; but

YES, in the sense that if one does not understand where the majority of these levels lie, and/or the nature of the 'trends' that move prices between levels, then the patterns will likely appear to be random. Hence what one trader perceives as 'randomness', might appear as perfect order to a more proficient trader, as he views the data through his own acquired 'filters'.


Razzle touches on a great point, IMO. To win at any game, there must be both (1) real aspects of underlying non-randomness in the game itself, AND (2) one's skill and knowledge must be such, as to make suitable decisions that recognize and exploit them. Take horse racing as an analogy. If I'm knowledgeable about the horses' form relative to the track conditions, weight carried, barrier position, jockey's track record, etc (and at locating situations where I see an inequality between my estimated probability of the outcome, and the odds being offered), then I'm making a truly informed bet. But if I simply place my bet according to 'backtesting' that told me that the horse wearing saddle cloth #5 wins the fourth race more often than the third race on Tuesdays, then I'm effectively making a random bet, even if the 'real' underlying ingredients that determine the probabilities are far from random. So it is with the markets.

In this context, it's important to understand the difference between (1) using data to test a predetermined hypothesis about price behavior, and (2) merely using the data itself to create a hypothesis. The latter opens itself up to the possible dangers of curve fitting.
  • #47
  • Edited 4:52pm Aug 9, 2010 4:32pm | Edited 4:52pm
Quoting McKeen
My question to anyone is, what are the reasons that inverting trades also seem absolutely impossible, even on higher timeframes, wide tp´s & sl´s and over extended periods of time? Stops and TP´s average around 50-500 pips so once again, spreads just can´t be the whole reason for this. Or could it? What is it that I fail to see and understand?

Thanks so much in advance to anyone who might shed some light to a dog trying to understand the laws of financial instruments.

/ McKeen
I've been there too. The answers are, unfortunately, that :
- the spread eats you alive ;
- the leverage associated to the spread shallow you whole ;
- your entries do not have an edge ;
- the market is mainly noise/random/intentional noise ;
- invalid theory behind trading approach.

Point 1 : The spread
everything has been said and I can only recommend trading EURUSD on Oanda 0.9 spread with a daily volatility of 140 pips, or +/- a daily volatility of 155 spreads.
For example, GBPUSD with FXCM will have a 3 pips spread with a daily volatility of 163 pips, or +/- a daily volatility of 54 spreads.
That's 3 times worse than the first example and that's really really bad. And they are far worse cases. Like a 32 spreads daily volatility for AUDUSD on avafx.


Point 2 : the leverage
A 1 pip spread with a 50:1 leverage steals 0.25% of any position you create. A 4 pips spread with a 200:1 leverage steals 4.00% of any position you create.
And there are some 500:1 leverage brokers........


Point 3 : no edge
Edges are especially difficult to create because of the vastly different paths the market can take to go from point A to point B. There are no easily understood rule, no apparent design (or far too many...), no discernable edge. But if you really want one, try using a program called forex strategy tester. You are limited to 4 entries conditions and 2 exit conditions. Backtest 4 months of 1 min data to find an edge. The advantage here is that the program calculate the result or your strategy, with drawdowns and everything, instantly. Make a tiny change and tada ! There are some more advanced features but that's the dealbreaker.
You'll soon find out that an edge can be really hard to find but also really simple to create. But most strategies and their inverse will end up with a negative equity curve.


Point 4 : intentional noise
Good entry points are taken out extremely fast by robots and there are some very intentional oscillations that make it difficult to find a good entry point but make it easy to catch a stop order. That's the way it's intended and this false chaos-real grand theft will take out most the position traders or trend traders. Just because that's an easy way to milk profits out of the unsophisticated traders.
But there's also the case of true noise, when noone is interested in moving the market and your entry/exit edge becomes worthless. That's the time when your money management skills should kick in and save you from frustrating losses.

Point 5 : invalid preconceptions
To make it short, you might approach trading like having to make good trades while you might be better off doing sequences of trades... Your trading system would then start to look like "Potential entry point ; first trade ; reaction X noted ; second trade ; reaction Y noted ; third trade ; reaction X+Y noted ; mega huge trade ; wait ; close everything ; yipee ; call the callgirlS".
More seriously, hanover just gave you a hint you should also check out 7bit's thread or Ronald Raygun approach.
  • #48
  • Aug 9, 2010 4:38pm Aug 9, 2010 4:38pm
Quoting Trad3r
OK, I'm not going to discuss with you the definition of randomness, just give me an EA with a true non-random system according to you and I will reverse it to continue arguing with proves, not with beliefs.
I cant actually give you what your asking for because... well this is the bit that's going to screw with your head.... my entries are actually RANDOM, yes you read that correctly, the entries are random (but with a directional bias), but the systems have positive expectancy.

As I pointed out previously, I also trade the inverse of each of my systems in a demo account on an ongoing basis. These systems generally have rising equity curves, the inverse signals generally have falling equity curves. These observations are not based on belief, they're based on the results over 120,000 trades over a 6 year period, and probably close to 2 million trades generated from back tests.

If you stick at this trading lark, sooner or later you'll come to the same conclusions.
  • #49
  • Aug 9, 2010 4:57pm Aug 9, 2010 4:57pm
Quoting Trad3r
This post is for you and for Craig and Razzle, here you have a simple EA coded by miself (you can edit it and fix mistakes, if they are) that is not random at all, a system with fixed entries, and the reversal method, it is a daily breakout EA:

hour1 = the hour to begin to open the trades, if they have to be opened.
hour2 = the hour to close all previous trades, if they are still open.

I'm using hour1 = 1 (the beginning of the asian session + 1 minute) and hour2 = 22 (the closure of the WS session + 1 minute) in GMT, you can change...
all the variables in the system rules you outlined are stationary...whereas market processes are nonstationary...this fact alone guarantees independence of system results....
now you could invert the rules all you want, the results would still oscillate around the breakeven line (excluding slippage and transaction costs)...

the system needs to model the reality as closely as possible and not map just one part of the reality...
Hear the music before the song is over ...
  • #50
  • Aug 9, 2010 5:03pm Aug 9, 2010 5:03pm
Quoting Razzle
I cant actually give you what your asking for because... well this is the bit that's going to screw with your head.... my entries are actually RANDOM, yes you read that correctly, the entries are random (but with a directional bias), but the systems have positive expectancy.

As I pointed out previously, I also trade the inverse of...
Maybe few moths ago it screwed my head, but not now, I was discussing about reversing losing system, even with fixed entries and exits, and how spread makes the reversal lose too in the long term. But after one year trading I've figured out that the real "holy grail" is not about the entry, but how to manage the trade to exit as better as you can. First time I read the "coin toss system" of Ronald Raygun I thought that he was crazy and that was a complete waste of time, but with the time I've learned that it's impossible to predict the market, even if you base your prediction in a 99'9% accuracy analysis of past data. This is because you have to gain skills in managing trades if you want to became a successful trader. This is in what I'm working now, and it's harder than finding a good entry strategy, because psicology is involved in this affair.
  • #51
  • Aug 9, 2010 5:05pm Aug 9, 2010 5:05pm
Quoting GlobalMacro
the system needs to model the reality as closely as possible and not map just one part of the reality...
True and false. Your statement is ambiguous. A system needs to tend towards reality but just mapping the part you judge interesting and admitting that chaos and unknown link the parts you are interested into might also work. You'll tell me that modeling chaos or at least taking it in account is part of modeling reality...
  • #52
  • Aug 9, 2010 5:21pm Aug 9, 2010 5:21pm
Quoting Intensity
True and false. Your statement is ambiguous. A system needs to tend towards reality but just mapping the part you judge interesting and admitting that chaos and unknown link the parts you are interested into might also work. You'll tell me that modeling chaos or at least taking it in account is part of modeling reality...
YES...EXACTLY...OF COURSE!

we need to model and conquer ambiguity then we'd be one step closer...that's the real challenge...
Hear the music before the song is over ...
  • #53
  • Aug 9, 2010 7:06pm Aug 9, 2010 7:06pm
Quoting McKeen
It is an interesting idea especially since this is basically what most brokers do I hear, betting against the trader. I am going to create such a project in the future, however I seriously doubt it could be effective in practice. But if they can, why couldn't we?

McKeen
When Kasparov was beating the IBM computers people used to say that he was actually beating the programmers from IBM not the computer itself.

You can't be something that you are not, so if you consistently loose money your code will reflect it.

Now the catch

You don't loose money because you trade consistently wrong. If this was the case your inversion approach would be right.

You loose money because you trade imperfectly ( and you do it consistently) for the reasons I explained on the last post.

Your code reflects your "SELF" just like Deep Blue reflected the programmers behind it. An error in the "self" will be passed down in your code.

No free lunch here buddy.....you still have to do your home work and become a real trader before you code something that will really work .
  • #54
  • Aug 10, 2010 11:48am Aug 10, 2010 11:48am
Quoting Trad3r
I'm not agree with you, this is not the reason, I tested the same inverting the trades AND the SL and TP, if in the original the SL is 10 and the TP 20, in the inverted EA the TP is 10 and the SL 20, and results were the same, the total loss.

There's another reason why inverting trades don't work (inverting SL and TP too), the reason is that odd are against us, the typical example is that you need to win a 100% to recover from a loss of 50%: if I have in my account 1,000 dollars and I lose the 50%, I have now 500, so I need to double up my account...

While i do believe making this work is tough, try wider stops and bigger trades. at a 10 pip stop loss and 20 pip tp, spread is already 10% of your trade before you even begin. That will kill it. For the love of science try a 40 pip stop and a 100 pip TP for both the original system and the inversion, and see if it improves. Too many trades would cost you too much at 10 pip SL. see ya
  • #55
  • Aug 10, 2010 2:07pm Aug 10, 2010 2:07pm
Quoting Platypus
You loose money because you trade imperfectly ( and you do it consistently) for the reasons I explained on the last post........
No free lunch here buddy.....you still have to do your home work and become a real trader before you code something that will really work .
Sure, but I would make the distinction between strategy and execution. In your "last post", you talked about "natural inclination" and "discipline", but the OP was running EAs, which will execute flawlessly.
  • #56
  • Aug 10, 2010 8:06pm Aug 10, 2010 8:06pm
Quoting hanover
Sure, but I would make the distinction between strategy and execution. In your "last post", you talked about "natural inclination" and "discipline", but the OP was running EAs, which will execute flawlessly.
My point is that the code will necessarily reflect the coder.

Since the coder does not have the skills to trade efficiently (and consistently) it will be passed down into the code.

If he was able to trade efficiently ( and consistently) minus (-), he could flip his code and get the same results ( efficiently/consistently) but with a inverted sign, plus (+) in the case.
  • #57
  • Aug 11, 2010 1:37am Aug 11, 2010 1:37am
Quoting Platypus
If he was able to trade efficiently ( and consistently) minus (-), he could flip his code and get the same results ( efficiently/consistently) but with a inverted sign, plus (+) in the case.
That's so wrong... I'm speechless. Looks like you haven't read the thread
  • #58
  • Aug 11, 2010 9:14am Aug 11, 2010 9:14am
Quoting Intensity
That's so wrong... I'm speechless. Looks like you haven't read the thread
You are failing to understand the concept....the goal of the thread was to explain why a system that looses money will still loose money when traded inversely.

Try to swap the word speech to brain and maybe you will realise why you can't grasp the concept !
  • #59
  • Aug 11, 2010 9:22am Aug 11, 2010 9:22am
So how do you define trading "efficiently"?
The nail that sticks out gets hammered back in
  • #60
  • Aug 11, 2010 10:10am Aug 11, 2010 10:10am
Quoting Noloqy
So how do you define trading "efficiently"?
Oh boy....Ok....I'm gonna exercise my patience here and try to explain one more time.

Retail traders have a mental laziness that amazes me.....

Ok lets try to use the same analogy from an older post.

Take a piece of hard wood and try to put a nail on it, all the way down, with a screw driver. I would say that it is close to impossible.

Now try to turn the piece of wood around or try to hit it with the screw driver handle or what ever. It does not matter because you are doing things the wrong way. The screw driver is not "efficient" it is not the right tool, it is plain wrong to start with, it is an error in principle.

His system is the screw driver, it is wrong.... and I hate to use the word "wrong" here in this context. It is actually "inefficient" deficient right in the beginning, in the blue print so it is almost mathematically impossible to give a different outcome.

Take a look at the pics the guy posted here, they are almost a mirror of each other. His trades were wrong in one sense now they are wrong in the other sense but still wrong, they will keep going south.
0 traders viewing now