EURUSD | Page 21395 | Forex Factory

Options

Go to
Permalinks
Print Thread

Related Instruments

EUR/USD, GBP/USD, USD/CHF

Similar Threads

MT4: how to change "EURUSD" to "#EURUSD"? 3 replies

Re: EurUsd short term 15 replies

did oanda just drop its spread for eurusd to 1 pip? 11 replies

EA for multiple lot limit order for EURUSD 0 replies

NFP nice bump up on EURUSD 2 replies

Attachments: EURUSD
Exit Attachments

EURUSD

  • #427,881
  • Apr 25, 2011 5:03pm Apr 25, 2011 5:03pm
Quoting Forexia
I know. I haven't found a profitable bot either. I admire those traders who are able to stay awake around the clock to trade.

Hope you had a good Easter!
Humans will always have an advantage over EAs, since rules to handle the market, most often based on its whole history without distinctions and being tuned in, are just too dumb to filter out the noise or become so complex to be known as curve-fitting, without adding any value since predicting the future is physically impossible (same goes for staying awake around the clock, I'm curious who does so and still is profitable).
I'm not talking generally here, I have made significant efforts in crafting EAs and indicators.
The only thing where machines are better off are with speed. High frequency Trading is obviously better and safer using reactive automats and general statistics, but it has no conscience either and does not really make any senseful decisions. We humans have the ability to suppose, EAs can only use whatever information is available from the past.
However, EAs are immensely purposeful for testing your edge in your strategies on a large scale of data without miscalculating since they don't get tired crunching numbers.
  • #427,882
  • Apr 25, 2011 5:15pm Apr 25, 2011 5:15pm
Quoting talentfx
Im back in long I think this still has a lot of upside potential
you indicators are not showing buy signals and you bought ..
  • #427,883
  • Apr 25, 2011 5:18pm Apr 25, 2011 5:18pm
Quoting frevd
Humans will always have an advantage over EAs, since rules to handle the market, most often based on its whole history without distinctions and being tuned in, are just too dumb to filter out the noise or become so complex to be known as curve-fitting, without adding any value since predicting the future is physically impossible (same goes for staying awake around the clock, I'm curious who does so and still is profitable).
I'm not talking generally here, I have made significant efforts in crafting EAs and indicators.
The only thing where machines...
Yup. I personally, can't think of any other way to design and program an EA other then using the lagging indicators.

The next best thing is using some combination of fractals and swing highs... and that has produced... interesting results. lol

It comes down to being able to find a strategy based on a combination of indicators, that are completely objective.

IMO something like the stochastics is a good place to start, just due to its nature of following candles and putting more weight into the closes.
  • #427,884
  • Apr 25, 2011 5:20pm Apr 25, 2011 5:20pm
Quoting frevd
Humans will always have an advantage over EAs, since rules to handle the market, most often based on its whole history without distinctions and being tuned in, are just too dumb to filter out the noise or become so complex to be known as curve-fitting, without adding any value since predicting the future is physically impossible (same goes for staying awake around the clock, I'm curious who does so and still is profitable).
I'm not talking generally here, I have made significant efforts in crafting EAs and indicators.
The only thing where machines...
they also don't suffer from emotions, thought it was worth a mention...

Here's a thing, all bias and overbought/sold conditions is a fallacy, there is only fair market value which is fluid and evolving as new participants come in and as new info gets priced in, not saying it's right or wrong, just kicking around ideas, been doing a lot of reading lately...
  • #427,885
  • Apr 25, 2011 5:24pm Apr 25, 2011 5:24pm
Quoting Venice
Yup. I personally, can't think of any other way to design and program an EA other then using the lagging indicators.

The next best thing is using some combination of fractals and swing highs... and that has produced... interesting results. lol

It comes down to being able to find a strategy based on a combination of indicators, that are completely objective.

IMO something like the stochastics is a good place to start, just due to its nature of following candles and putting more weight into the closes.
Everyone uses lagging indicators, at least when it comes to confirmation. Unfortunately there is too much noise whipsawing you, and if you smooth that out, you're always too late or even a target.
I personally dislike indicators for automated trading since they essentially only transform the price we already know. They can be of value to filter some characteristics though for manual trading, which they are used for mainly.

The other way is to predict the market and act as a countertrend-trader using R/S levels that you somehow calculate. The downside of that approach is that you'll be more often wrong than right, which is where you need to apply sound cash management and pyramiding to get some edge.

The best approach to avoid curve-fitting is the pragmatic one - find some very simple signal-giving somewhat (e.g. the usual MA crossing) and optimize its inputs for several markets (that is not one specific market).
Or roll out rocket science and find something nobody has ever found .
  • #427,886
  • Apr 25, 2011 5:28pm Apr 25, 2011 5:28pm
Quoting FibbinArchie
they also don't suffer from emotions, thought it was worth a mention...
That surely is another important point. My bad I forgot.
However, emotions play an important role. I wouldn't let an EA trade my strategy since it proved to be not as effective as manual trading.
No lunch for free I think.
  • #427,887
  • Apr 25, 2011 5:33pm Apr 25, 2011 5:33pm
Quoting FibbinArchie
Here's a thing, all bias and overbought/sold conditions is a fallacy, there is only fair market value which is fluid and evolving as new participants come in and as new info gets priced in, not saying it's right or wrong, just kicking around ideas, been doing a lot of reading lately...
I personally think the markets are impulse-driven, meaning you kick them with some information and they fly until a contrary information hits, like an object in space. Anyways, information if not news is also any kind of technical level, popular calculation etc., being expectations.
Thats kind of a selffulfilling prophecy, and that's why all those technical things mystically work.
  • #427,888
  • Apr 25, 2011 5:35pm Apr 25, 2011 5:35pm
Quoting frevd
That surely is another important point. My bad I forgot.
However, emotions play an important role. I wouldn't let an EA trade my strategy since it proved to be not as effective as manual trading.
No lunch for free I think.
well there you go, the proof of the pudding is in the eating...

and your point on indicators, how about a possible example of an application being measuring if directional momentum is increasing or decreasing, as I said, there aren't really any right or wrong answers, just trying to contribute to the discussion...
  • #427,889
  • Apr 25, 2011 5:39pm Apr 25, 2011 5:39pm
Quoting frevd
I personally think the markets are impulse-driven, meaning you kick them with some information and they fly until a contrary information hits, like an object in space. Anyways, information if not news is also any kind of technical level, popular calculation etc., being expectations.
Thats kind of a selffulfilling prophecy, and that's why all those technical things mystically work.
yep, totally agree on your second point, but there is an element of chaos in price movement, although even chaotic systems have underlying motivations and a degree of ordered structure driving them...

not quite sure where I'm going with this, having a bit of a brain dump...
  • #427,890
  • Apr 25, 2011 5:40pm Apr 25, 2011 5:40pm
Quoting FibbinArchie
well there you go, the proof of the pudding is in the eating...

and your point on indicators, how about a possible example of an application being measuring if directional momentum is increasing or decreasing, as I said, there aren't really any right or wrong answers, just trying to contribute to the discussion...
Yep, the universe is so full of information that you have to apply filters in order to make some sense (finally that's all what's sense is about, categorial structures and conditional connections, itself being just a human concept).

so if market behaves overall chaotically, the best thing to do is follow a strict filtering strategy that detects patterns and have a trial and error on the results. This also underlines why money management is way more important than the entry signals or exits. The Stoploss function is the best fit for minimizing the cost of participation by not minimizing the potential profits.
  • #427,891
  • Apr 25, 2011 5:45pm Apr 25, 2011 5:45pm
Quoting FibbinArchie
yep, totally agree on your second point, but there is an element of chaos in price movement, although even chaotic systems have underlying motivations and a degree of ordered structure driving them...
The definition of a chaotic system is that it is not random, but way too complex to be predictable and behaving like random (an minor change changes the whole outcome). Thus you can apply the catch-all phrase of statistics, e.g. the Stochastics. However, since the market consists of thinking individuals that know when they are being abused, market adapts and adjusts all the time to prevent that. Physicians found out that markets do not behave like normal distributions, i.e. they cannot be caught using a standard deviation around a median, they are somewhat intelligent.

Quoting FibbinArchie
not quite sure where I'm going with this, having a bit of a brain dump...
Haha, yeah, weird thoughts, but helps killing time and I find it interesting.
  • #427,892
  • Apr 25, 2011 5:46pm Apr 25, 2011 5:46pm
Quoting frevd
Yep, the universe is so full of information that you have to apply filters in order to make some sense (finally that's all what's sense is about, categorial structures and conditional connections, itself being just a human concept).

so if market behaves overall chaotically, the best thing to do is follow a strict filtering strategy that detects patterns and have a trial and error on the results. This also underlines why money management is way more important than the entry signals or exits. The Stoploss function is the best fit for minimizing...
define an edge, enter when said edge is present and realise that even though your edge is present your trade is still only based on probabilities...

doesn't even really matter what that edge is so long as it can be demonstrated over a decent sample size...
  • #427,893
  • Apr 25, 2011 5:48pm Apr 25, 2011 5:48pm
Quoting frevd
Everyone uses lagging indicators, at least when it comes to confirmation. Unfortunately there is too much noise whipsawing you, and if you smooth that out, you're always too late or even a target.
I personally dislike indicators for automated trading since they essentially only transform the price we already know. They can be of value to filter some characteristics though for manual trading, which they are used for mainly.

The other way is to predict the market and act as a countertrend-trader using R/S levels that you somehow calculate. The...
The real issue here is really how to manage orders when they are open. Just as with a lot of the things, I watch indicators be correct to some extend, in that you will at least have a chance and getting into profit.

The next road that leads me down... is identifying as many TP areas as possible.

Fractals, DPP, WPP, MPP, highs, lows. Give it any reason to exit and it'll do it. That's what I am currently working on... and in doing so, it even manages to survive range'ish conditions, because TP is at the top of the range thanks to the fractal.

I may not know rocket science, but I sure as hell know rocket surgery
  • #427,894
  • Apr 25, 2011 5:51pm Apr 25, 2011 5:51pm
Quoting FibbinArchie
define an edge, enter when said edge is present and realise that even though your edge is present your trade is still only based on probabilities...

doesn't even really matter what that edge is so long as it can be demonstrated over a decent sample size...
I could long not believe that. I would think that there was a holy grail. Then I made the test. I wrote a stupid EA, meaning one that did everything not only wrong in my opinion, but completely random.
Of course it was not better or worse than all my previous candidates.
That said, since EAs always are rule-based, I cannot tell whether that applies to humans as well.
  • #427,895
  • Apr 25, 2011 5:53pm Apr 25, 2011 5:53pm
Quoting Venice
The real issue here is really how to manage orders when they are open. Just as with a lot of the things, I watch indicators be correct to some extend, in that you will at least have a chance and getting into profit.

The next road that leads me down... is identifying as many TP areas as possible.

Fractals, DPP, WPP, MPP, highs, lows. Give it any reason to exit and it'll do it. That's what I am currently working on... and in doing so, it even manages to survive range'ish conditions, because TP is at the top of the range thanks to the fractal....
Hehee, you sure do.
Yep, seems it's all about experience with your particular market, and staying in tune, as markets do change from time to time whenever it is obvious that someone has an edge.
  • #427,896
  • Apr 25, 2011 5:53pm Apr 25, 2011 5:53pm
Quoting frevd
The definition of a chaotic system is that it is not random, but way too complex to be predictable and behaving like random (an minor change changes the whole outcome). Thus you can apply the catch-all phrase of statistics, e.g. the Stochastics. However, since the market consists of thinking individuals that know when they are being abused, market adapts and adjusts all the time to prevent that. Physicians found out that markets do not behave like normal distributions, i.e. they cannot be caught using a standard deviation around a median, they are...
hmm, yeah perhaps why trying to predict future price based on linear statistics is perhaps the worst thing you can do, since price action and human decision making is not linear in nature, a better model perhaps is to build a trading system that is flexible and adaptive...
  • #427,897
  • Apr 25, 2011 5:57pm Apr 25, 2011 5:57pm
Now as I keep saying Venice, tonight I really am going to get a good night's sleep so as I'm nice and fresh for the London open, have a good night guys...
  • #427,898
  • Apr 25, 2011 5:58pm Apr 25, 2011 5:58pm
Quoting FibbinArchie
hmm, yeah perhaps why trying to predict future price based on linear statistics is perhaps the worst thing you can do, since price action and human decision making is not linear in nature, a better model perhaps is to build a trading system that is flexible and adaptive...
I don't know actually. Human predictive potential might be only linear, only. You ever wondered why nobody can estimate exponential development, be it increases in debt or environmental change? I guess they only forecast linearly, that's hard-wired into our brains.
But the base changes irregularly, that's where you can possibly begin with nonlinear calculations, and that's where you soon end in chaos .
  • #427,899
  • Apr 25, 2011 5:59pm Apr 25, 2011 5:59pm
Quoting FibbinArchie
Now as I keep saying Venice, tonight I really am going to get a good night's sleep so as I'm nice and fresh for the London open, have a good night guys...
CU later.
  • #427,900
  • Apr 25, 2011 6:02pm Apr 25, 2011 6:02pm
Is that a first breakout attempt or are my eyes too tired already?
564 traders viewing now