Optimize Indicators ? - page 2
You are missing trading opportunities:
- Free trading apps
- Over 8,000 signals for copying
- Economic news for exploring financial markets
Registration
Log in
You agree to website policy and terms of use
If you do not have an account, please register
FourX:
Indicators ARE by and large what we use as entrance signals
Tip: The optimizer is unable to manipulate Boolean TRUE or FALSE settings. I'm going to make a 'translation' function as a work around
1. A matter of distinctions. What you are referring to is also what i am referring to: optimizing the settings for the indicator to get the optimal results: this is all the strategy tester does. Not modifying the indicator code itself WH, though this is of coarse a possibiliity, but certainly beyond my capacity at this point in time.
I've been reading an article posted here on not falling into the 'Optimization trap' which I have been doing: Only looking for the most profitable settings. But the author makes the valid point that if the settings are such that only a minor change in them results in drastically reduced results, then in operation when anything other than the exact conditions, then the performance is going to be terrible. Utilizing settings that give good results that cover a whole region of the the 2D graph will result in a much more robust EA performance. If in the same ball park, then will still be profitable, as opposed to having to hit the exact mark or get the golf ball into the very small hole.
2. I take it you are talking about being able to optimize Boolean True or False condition in this one? Again, what you are suggesting is what I am likewise doing. Basically just using the Integers 0 or 1 as false and true respectively. The integers are valid conditions for the Boolean false or true, but as we know, the Strategy tester won't allow one to select them as an optimization option, so just use the integers 0 or 1 themselves instead. Can either assign this to a Boolean False or True or just test for 0 or 1. 6 of one, half a dozen of the other. Their may be cases where it will be advantageous or necessary to utilize the actual Boolean. So whatever the situation dictates.
The only way to optimise entry conditions is to trade and find the best settings of the indicator one is using. Therefore the EA is the method of optimising your indicator. One method I use is to get in and out using the indicator and then find the best settings for stops and takes if the indicator entry and exits are not good enough. Indicator entry and exit is always slow because indicators are behind the market. If relying just on indicators the trick is to have them get you in when there is reasonable movement to get you straight in to profit but indicators are easily fooled.
1) My computer is always slow and 'behind' to start with, so this doesn't help matters. )< 8(
1) I haven't been using the Strategy Tester at all for quite some time now as I have found it not to really provide any realistic results. So have only been doing forward testing/Demos for quite some time. But this is a very slow method by which to analyse a whole range of numerous variables, so I'm giving it a go again. But one can 'optimize' the settings for the indicator just as we do the EA itself. Many EAs do contain external variables for settings for indicators, thus one can, and we often do, optimize indicators as well as was pointed out earlier in this thread.
2) Well indicators only utilize data from the 'past', so in that sense they always are always behind to start with. But they are trying to project into the future hence the term 'Indicator' Some do it much better than others. There are so many approaches to this, so there are so many indicators and EAs.
3) As you say:getting into a consolidated sideways market doesn't hold the prospect of immediate profits. So part of the task and challenge for 'static' FX trading code is to identify when there is enough movement and activity that the possibility of profits at least exists. I think that most people and EAs utilize numerous indicators to try and measure and judge this factor as they do with many other factors.
I'm starting to think that it is not very realistic to forecast the future activities of the market based purely on data from the past. That one must include some 'real world' fundamentals as well. Look at the current USDCHF charts. The farthest bask that I can get data for it is 1989: quite a ways. It is now considerably lower than any time in this lengthy period and is still dropping. What indicators can predict this ?
Many years ago when I was interested in the futures /commodity markets the, the Japanese Yen crashed and dropped by 2/3 in a VERY short period of time because the information that the whole stock market there was basically a rigged 'insider trading scam'. What indicator could of predicted this? However I knew enough about 'real world' economics that it would regain it's value and get back to at least very close to what it had been before this crash in a relatively short time: 1 ~ 2 years. Which it did. But This is not a time scale that many, if any at all, trade the ForEx on. if one has deep enough pockets and the willingness to do so, they could make a lot of money over an extended period of time riding the USDCHF down for the many years it has been dropping now.