Breakout & RETURN: 97.5% accuracy | Forex Factory

Options

Go to
Permalinks
Print Thread

Similar Threads

ECN && STP && Scalping && Hedging - Broker Review - Oct 2009 24 replies

Intergrated MM with System Accuracy and Risk Control 2 replies

Accuracy vs. net profit 10 replies

historical data accuracy puzzlement. 1 reply

accuracy problems or stable market 0 replies

Breakout & RETURN: 97.5% accuracy

  • #1
  • First Post: Edited Jul 24, 2010 5:04am Jul 7, 2010 4:04pm | Edited Jul 24, 2010 5:04am


Makes me think...
  • #2
  • Jul 7, 2010 4:44pm Jul 7, 2010 4:44pm
It sounds like you are trading with no stoploss.

Run through your stats. For every winning trade, what is the highest floating loss?
  • #3
  • Jul 7, 2010 5:25pm Jul 7, 2010 5:25pm
Quoting Ronald Raygun
It sounds like you are trading with no stoploss.

Run through your stats. For every winning trade, what is the highest floating loss?

From all the data I have collected while analysing 5 minute EURUSD chart from may 2009 until march 2010 I have drown some conclusions. I thought that stoploss which is 200pips away from the breakout level is optimal. 2411 breakouts out of total 2470 did not reach 200pips. (2411/2470 = 0.976113 -> 97.5%)

When it comes to the highest floating loss I can't give you the exact number right now Ronald. I'm pretty sure that losses shown in backtesting are a little bit exagerrated because the EA does not delete orders that remained from the grid, after some number of them were closed with TP. This means that there is plenty of orders that are placed in a wrong place having SL close to the opening price and TP pretty far. I still don't know how to code the EA so it can delete them, but I keep learning so hopefully I'll overcome this problem soon.
  • #4
  • Jul 7, 2010 5:34pm Jul 7, 2010 5:34pm
However I can give you what Excel says:

When the first order is opened (8pips away from the breakout level) there is no floating loss. When the next one is opened (18pips away) the first order shows 10 pips loss. When the third one is opened (28pips away) the firtst one shows 20pips and the second one 10pips of loss totalling 30 pips.. and so on.

In case of the trade where the price broke through the breakout, did not reach 200 pips, and when it opens the last, 18th order the floating loss shows -1710 pips. (around £113 if we use 0.01 lot per order and GBP account).
  • #5
  • Jul 7, 2010 5:38pm Jul 7, 2010 5:38pm
I played bit with your excel data, found out the following.

> Enter a trade only if the breakout gets to 100 pip far from resistance/support. This means that yout TP is 100pip
>set SL 400 pip

With a 3pip spread this setup ends up to 8557pip profit:
-> wining trades:163 resulting in 15811pip profit
-> losing trades: 18 resulting in 7254 pip loss

The only factor that can be crucial is the number of consecutive loosing trades however I am not sure if your ranking is based on time. If yes then you have the first 17 trades all loosing ones!

That was a quick and dirty exercise and I might need to re-check the figure or someone can have fun with it.

Way2Freedom
  • #6
  • Jul 7, 2010 5:45pm Jul 7, 2010 5:45pm
Quoting Casper C.
What is ForEx? Chaos this is what it is. So is there any chance to predict how it will behave in the future and take advantage of it? Apparently no.

In theory however even in chaos you can find something that happens all the time. Something that always repeats itself.

Here I will prove that there is one interesting thing that always happens in this market.
The very definition of chaos is "a state lacking order or predictability". If you have found something that always repeats itself, then you have proven that the market is predictable.

Think on that for awhile.
  • #7
  • Jul 7, 2010 6:00pm Jul 7, 2010 6:00pm
Quoting way2Freedom
I played bit with your excel data, found out the following.

> Enter a trade only if the breakout gets to 100 pip far from resistance/support. This means that yout TP is 100pip
>set SL 400 pip

With a 3pip spread this setup ends up to 8557pip profit:
-> wining trades:163 resulting in 15811pip profit
-> losing trades: 18 resulting in 7254 pip loss

The only factor that can be crucial is the number of consecutive loosing trades however I am not sure if your ranking is based on time. If yes then you have the first 17 trades all loosing ones!

That...

I'll code your suggestion and test it as soon as get my laptop back. Thanks for another idea of how to use what we know.
  • #8
  • Jul 7, 2010 6:10pm Jul 7, 2010 6:10pm
Quoting Darkstar
The very definition of chaos is "a state lacking order or predictability". If you have found something that always repeats itself, then you have proven that the market is predictable.

Think on that for awhile.

I thought about it for a longer while. As far as I know the general chaos theory states that even "a state lacking order or predictability" potentially can be put in to mathematical formula, thus predicted. Which means that chaos is not really chaos.

I've found in a few places a statement that Forex chart is an example of perfect chaos, this is why I used it.

Anyways I didn't mean to create this thread to discuss words which definitions are contradictory to themselves, so please lets focus on breakouts and returns.
  • #9
  • Jul 7, 2010 6:47pm Jul 7, 2010 6:47pm
You put the horizontal line on the chart after the fact, not before. Therefore, you really don't know where the true top and bottom is until after it happens.
  • #10
  • Jul 7, 2010 6:49pm Jul 7, 2010 6:49pm
Quoting Casper C.
I thought about it for a longer while. As far as I know the general chaos theory states that even "a state lacking order or predictability" potentially can be put in to mathematical formula, thus predicted. Which means that chaos is not really chaos.

I've found in a few places a statement that Forex chart is an example of perfect chaos, this is why I used it.

Anyways I didn't mean to create this thread to discuss words which definitions are contradictory to themselves, so please lets focus on breakouts and returns.
Chaos theory utilizes a probabilistic model to define "potential" outcomes, but thats not the same as a consistently repeatable event. In any event, I agree it's debating semantics.

The point I wanted to make was that (as you have proven) the market IS predictable, to a reasonably high degree.

When you start with the idea that the market is random, it leads you to a number of conclusions. Starting with the assumption that it is predictable leads you somewhere else entirely.

Good luck in your development.
  • #11
  • Jul 7, 2010 7:05pm Jul 7, 2010 7:05pm
Quoting eagle4x
You put the horizontal line on the chart after the fact, not before. Therefore, you really don't know where the true top and bottom is until after it happens.
I don't really know what you meant by that. I decided to place a horizontal line (set resistance/support) when: the 4th bar from the open one is higher than two candles on its left and three candles on its right. That's all. Before the close of the third 3rd candle on its right nothing happens.
  • #12
  • Jul 11, 2010 7:04pm Jul 11, 2010 7:04pm
hey casper..... was thinkin about coding this up real quick but had a question.....

the high of the seventh bar [6] is higher or equal to the Low of the
fifth bar [4]

you've used high as part of the last condition for a bottom..... that does not follow the theme up until that point......

just to clarify, did you really intend it as the the high of the seventh bar [6] is higher or the low of the seventh bar [6] is higher .....

thanks......h
to trade and code, keep both simple... no call to impress....h
  • #13
  • Jul 12, 2010 6:06am Jul 12, 2010 6:06am
Yes, I meant low! Sorry for that, just a typing error.

Could you post your code here for me? I'm still beginner with mql4 so it's not that easy for me to code such thing. Thanks in advance!

Cheers
  • #14
  • Jul 12, 2010 10:10am Jul 12, 2010 10:10am
Quoting Casper C.
I'll code your suggestion and test it as soon as get my laptop back. Thanks for another idea of how to use what we know.
Here is another pattern in the data that can generate huge profit.
Anytime an order is trigred keep adding new orders on an 8 pip basis. In other word,
1st order: 8pip target
if TP not hit then add a second order at the level of 16pip TP
If 2nd TP not hit, add a third one at TP level of 24pip
.
.
.until you achieve a TP order of 300 pip (at this level, you will have 37 open orders where the first one has a TP of 8pip and last one with TP of 300)

The SL is alwyas 300 pip.

Based on the data, you will have 29 loosing order that will result in 332,808pip loss= (29 order* (300SL+3 spread)*37.87 opened order)

And a total of 1,625,555 pip gain. The net profit is 1,292,747 pip.

The list of profitable trades can be ranked in terms of TP
1523631174272552569007142182112929971041001154451270561389571411172151372516166401719941182365219277972032400213748522430762349197245587225631252670980277946128885922998397301089003112012533136224341492263516303037182595

you read it as follow:
1 order with TP of 8 (net TP is 5=8-3pip spread)
2 order with TP of 36pip
.
.
.
37 order with TP of 182,595 pip

This can be appleied on micro account and still lead huge profit.

The only risk is on very long term percentage of the loosing trade is getting high.

Can anyone check this strategy?


Way2freedom
  • #15
  • Jul 12, 2010 1:37pm Jul 12, 2010 1:37pm
hey casper ...... we're on the road today....... will post it this evening when we get back to motel room..... it's on my computer running there.....

ran some backtests from 1/1/09 on the 5 minute eurusd..... can verify your 97% rate..... as you have found out, high win rates are not equal to profitability in all cases.....

97% and higher win rates are easy to achieve.... the remaining 3% tends to whack us pretty good......

//------

my true interest here was in you comment number 3,

3. But we know it will return to the breakout level.

it almost stands to reason that can not be true...... backtesting verifies this..... as example, the 2nd and 3rd trades that were made in the first week of 09 never had the price return to the breakout level......

i only coded in the first trade at each signal..... my thoughts were if it did not return any subsequent added trades would not either.....

your choice of trades themselves was also of interest..... many would be reversing the those..... that is to say they would buy at the resistance break and sell at the support break..... where you are doing the opposite.....

both ideas can be combined into one....

will post it this evening.....h
to trade and code, keep both simple... no call to impress....h
  • #16
  • Jul 12, 2010 4:26pm Jul 12, 2010 4:26pm
Quoting way2Freedom
Here is another pattern in the data that can generate huge profit.
Anytime an order is trigred keep adding new orders on an 8 pip basis. In other word,
1st order: 8pip target
if TP not hit then add a second order at the level of 16pip TP
If 2nd TP not hit, add a third one at TP level of 24pip
.
.
.until you achieve a TP order of 300 pip (at this level, you will have 37 open orders where the first one has a TP of 8pip and last one with TP of 300)

The SL is alwyas 300 pip.
If I'm not mistaken this is exactly the way I was thinking. If you go back to my first post you will se that actually you are following the same rules as I pointed out.

The only difference is that you are placing more orders ( every 8 instead of 10 pips, with exception of the first order in my way ) and your SL is 300pips.

Quoting way2Freedom
Based on the data, you will have 29 loosing order that will result in 332,808pip loss= (29 order* (300SL+3 spread)*37.87 opened order)

And a total of 1,625,555 pip gain. The net profit is 1,292,747 pip.
Way2freedom
I'm not sure if I follow your calculations correctly. By 29 orders do you mean TRADES that I defined (event of placing series of pending orders) or does it reflect smaller orders that were placed as a grid? If it is the second one there may be a mistake. Each of these orders will end up with different loss. The first one will make 292 pips of loss, the second one 284, the third one another 8 pips less, etc. So if you meant TRADES the loss will be much bigger than your estimate.

Can you verify this one for me?
  • #17
  • Jul 12, 2010 4:59pm Jul 12, 2010 4:59pm
Quoting hayseed
//------

my true interest here was in you comment number 3,

3. But we know it will return to the breakout level.

it almost stands to reason that can not be true...... backtesting verifies this..... as example, the 2nd and 3rd trades that were made in the first week of 09 never had the price return to the breakout level......

i only coded in the first trade at each signal..... my thoughts were if it did not return any subsequent added trades would not either.....

your choice of trades themselves was also of interest........


The period I checked for breakouts and returns was almost random. At the moment I started my broker's server couldn't refresh more of five minutes chart than last ten months. This is why it's 10 instead of initially planned 12. I thought that it should be enough to make this data reliable considered that it is M5.

While I was dragging cursor from every breakout to to every peak/bottom I was firstly shocked that this thing happens so often and secondly I started to realise that one should have huuuge account if he doesn't want to have any losses (no SL at all). I calculated roughly how big the account should be in order to survive every single breakout no matter how far it went in the mentioned 10 months and I ended up with over 8 million pounds (assuming that we used only microlots throughout that period). Perhaps my statment that 'we know as a fact that the price will return' was too bold, but it's not the case.

There is a thing in the chart that happens extremely often. We know what it is. We have something that, according to Ronald Raygun, is way more important than the entry. We know where to exit and I can't imagine anything more transparent. The question is where should we enter to make the most of it?
  • #18
  • Edited 5:45pm Jul 12, 2010 5:08pm | Edited 5:45pm
scroll right, scroll left.
  • #19
  • Jul 13, 2010 7:18am Jul 13, 2010 7:18am
Quoting Casper C.
If I'm not mistaken this is exactly the way I was thinking. If you go back to my first post you will se that actually you are following the same rules as I pointed out.

The only difference is that you are placing more orders ( every 8 instead of 10 pips, with exception of the first order in my way ) and your SL is 300pips.



I'm not sure if I follow your calculations correctly. By 29 orders do you mean TRADES that I defined (event of placing series of pending orders) or does it reflect smaller orders that were placed as a grid? If it is the...
Yes, I tried to follow your rules for trading and find out the result. At this moment, what matters is to test that the low percentage of trade loss is not gonna increase to a high percentage in a different market dynamic. Based on your excel data, they will be 29 loss trade which is great result. We need to test if it stays at this level on a long period.

You are right for the loss calculation as it should be smaller than mine (300+292+284..+8) but even with my majoration the result are incredible!

Way2Freedom
  • #20
  • Jul 14, 2010 12:23am Jul 14, 2010 12:23am
Quoting hayseed
hey casper ...... we're on the road today....... will post it this evening when we get back to motel room..... it's on my computer running there.....

ran some backtests from 1/1/09 on the 5 minute eurusd..... can verify your 97% rate..... as you have found out, high win rates are not equal to profitability in all cases.....

97% and higher win rates are easy to achieve.... the remaining 3% tends to whack us pretty good......

//------

my true interest here was in you comment number 3,

[i]3. But we know it will return to the breakout...

Heyseed, how is your code developing? Do you have anything to be posted here? I'll appreciate that.

Cheers
C.
0 traders viewing now