THE IDEA EXCHANGE - page 7
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
If anyone is interested, I'm currently enjoying Fibonacci levels. At first I built extensions myself, and then I started using DinapoliTargets (available in CodeBase). But there are peculiarities. Here's what I found:
1. if price has passed the starting line, it will almost certainly reach the 61.8 level;
2. if the price has not passed the start line, a stop order can be placed for a point behind the line;
3. if the start line is already behind at the moment the indicator is switched, a limit order can be placed for it with a target at 61. 8;
4. if some levels coincide or almost coincide at different TFs, their significance is higher, and the price is likely to reach them
5. if levels coincide completely on two adjacent TFs, the price is likely to reach 161.8;
To confirm the decision, we may use any trend indicators or stochastics. But they often lie (least of all my DynamicRS_3CLines - also available there). It is better to wait for switching levels. The indicators are just for self-reassurance.
Also. I liked working on low TFs better. Especially on M5 and Yen. In this case the stops are the shortest, so you can open positions up to (free funds)/2000. If the target is closer than 10 pips, there will be problems with TP on the real. IMHO. So you will have to wait for the moment to manually close the order.
I am also dealing with the DiNapoli indicator. At first I scrupulously looked into the code and checked if the algorithm corresponds to the classical DiNapoli formulas. Then I started experimenting with the Expert Advisor proper according to the given methodology. (My observations are as follows.
The start signal often appears "retrospectively". Aim 1 and 2 are usually achieved with great consistency. Especially at low timeframes! At the moment the best results are obtained on mf15 and especially on m30. I have been testing this on GBPUSD.
The best thing about it is that you can optimize it almost by eye (parameters - stops) - with enough reliable (to my surprise) visual predictability!
The best results I've got (on m30) - if a small trailing stop is activated when the first target is reached.
One more thing. In addition to the indicator DiNapoli itself, the author of this indicator (Mishanya) has developed a whole system of working together with the indicator by the channel DiNapoli and oscillator DiNapoli. But there is no description of this system anywhere. I searched all the Internet on this subject - I found a couple of sites, but they say, that by the request of the author the material has been deleted!
If anyone is interested, I'm currently enjoying Fibonacci levels. At first I built extensions myself, and then I started using DinapoliTargets (available in CodeBase). But there are peculiarities. Here is what I found:
I'm doing the DiNapoli indicator as well. At first I scrupulously looked into the code and checked if the algorithm corresponds to classic DiNapoli formulas. Then I started experimenting with the Expert Advisor proper according to the given methodology. (without filters and other evidence so far) My observations are as follows.
Observations at least do not contradict mine. That's good :)
Now to channels and oscillators. The channels, and indeed the levels all around them, have been blatantly appropriated by DiNapoli. But that's not the point now. The methodology, yes. That's his. But everything else is purely Fibonacci. Channels are very well described by R. Fischer. I have two his books in electronic form, but I have problems with e-mail - I work via GPRS. So it's too expensive to send. One is called The New Fibonacci Trading Methods (The New Fibonacci Trader), and the second Fibonacci FACTORY: APPLICATIONS AND STRATEGIES FOR TRADERS.
I like the first one better. But both differ in that there are no calculations. It's a pity.
Now the practice. Channels are good. But they imply quite distant in time targets. That's why I haven't tried them. My principle is to take them as soon as possible. This is only possible on small TFs and with targets close to them. That is why only extensions have been implemented so far. But I can say that this method has given me some confidence for the first time in 10 years of my investigations in forex. Sometimes I cannot believe what is happening. For instance, last week on EURJPY I accidentally observed a perfect match between all levels on M5 and M15. After that, the price reached exactly 161.8 and reversed. After a couple of hours of flat the same thing happened again in the opposite direction. I entered successfully at the start line, price hesitated for a while and then rushed to 61.8. I barely managed to move the TP to the 100 level. In 5 minutes it was already going towards it. I moved it again. Now it moved to 161.8. Then the trend started to exhaust and I set a trawl of 25 pips, which was my big mistake. I try not to use trawls in such a way now - it's a stupid and harmful tool. I made a mistake of 5 points. The trawl was struck near the 100 level, and then the price easily reached 161.8! That is how it happens.
Now about the signal. The fact that it appears retrospectively is not a problem. I have already written - put a Limit order at its level. If the price comes back to it, it will definitely reach 61.8. But this applies to the case if the start is behind already at the time of switching levels. If it is ahead, it is not certain that it will be broken. Then the price will reverse. Very easy. But at first I also fell into this trap. I was in a hurry and got hit in the face by a forelock :)
In principle, the technology is quite mature. Manually it works somehow. It needs to be automated. That's why I've modified the indicator somewhat (made arrays instead of lines), but it works crookedly - I use zero bar head-on, so it only works in the real chart when getting quotes. Take a look. Maybe someone will upgrade. It's not convenient to mess with source code. I've commented out pieces of unused code (haven't deleted almost anything).
I also forgot to mention the DiNapoli oscillators.
I don't know, maybe I haven't figured it out yet, but in my opinion, they are only for self-assurance. They won't make you feel worse, but they won't make you feel better either. IMHO.
I have outlined the principles of my own system outline here: http://onix-trade.net/forum/index.php?showtopic=88&st=135, after that I have a couple more posts there explaining the picture. The system is based on a book by Miner. I have not yet seen a decent translation of it, I read it in the original. I have not read Fisher well either, I will have to have a look.
The essence of the implementation is in the nuances, which are not present in the general description on onyx.
Now a couple of words on the subject. Trawl is bullshit, it ruins everything. Oscilli are also rubbish, they show the average temperature in a hospital. There is either one - a discrete market model, or another - a continuous model. They cannot be combined together. The prospects for the discrete model are huge, but there are many obstacles.
In principle, the technology is quite mature. Manually it works somehow. It needs to be automated. That's why I've modified the indicator somewhat (made arrays instead of lines), but it works crookedly - I use zero bar head-on, so it only works in the real chart when getting quotes. Take a look. Maybe someone will upgrade. It's not convenient to mess with source code. I've commented out bits of unused code (haven't deleted almost anything).
I dare to suggest that those present dabble with one of the versions by DiNapoli, which at my request (as best I could) was produced by a specialist! You can find the source code in the download. All the indicator attributes are present in the visual mode - very clear!
Signals "retrospective" are skipped. After the first (second) target is reached - the stripe is moved to Breakeven. Then, if necessary, trawl may be activated. Or exit by Take Profit.
Right now - blindly from a torch, I've set it (without optimization) on m15 on GBP and ran it since Jan. 2007г.
GBPUSD symbol (Great Britain Pound vs US Dollar)
Period 15 Minutes (M15) 2007.01.02 00:00 - 2007.10.12 22:45 (
Model All ticks (based on all smallest available periods with fractal interpolation of each tick)
Parameters Lots=0.1; SigPoint=3; tral=54; stoploss=59; barn=100; Length=9;
Modeling quality 90.00% Initial deposit 10000.00
Net profit 84.42 Total profit 5512.63 Total loss -5428.21 Profitability 1.02 Expected payoff 0.39 Absolute drawdown 133.37 Maximum drawdown 920.36 (8.53%) Relative drawdown 8.53% (920.36)
Total trades 218
Short positions (% win) 121 (55.37%)
Long positions (% win) 97 (58.76%)
Profitable trades (% of all) 124 (56.88%) Loss trades (% of all) 94 (43.12%)
Largest profitable trade 191.42 losing trade -63.25
Average profitable trade 44.46 losing trade -57.75
In principle, the technology is quite mature. Manually it works somehow. It needs to be automated. That's why I've modified the indicator somewhat (made arrays instead of lines), but it works crookedly - I use zero bar head-on, so it only works in the real chart when getting quotes. Take a look. Maybe someone will upgrade. It's not convenient to mess with source code. I've commented out pieces of unused code (haven't deleted almost anything).
I dare to suggest that those present dabble with one of the versions by DiNapoli, which at my request (as best I could) was produced by a specialist! You can find the source code in the download. All the indicator attributes are present in the visual mode - very clear!
Signals "retrospective" are skipped. After the first (second) target is reached - the stripe is moved to Breakeven. Then, if necessary, trawl may be activated. Or exit by Take Profit.
Right now - blindly from a torch, I've set it (without optimization) on m15 on GBP and ran it since Jan. 2007г.
Rebus, it's pretty much the same as mine in principle. My system is still unfinished; automating it is quite a serious undertaking.
I stated the principles of my own system outline here: http://onix-trade.net/forum/index.php?showtopic=88&st=135, after that I have a couple more posts there explaining the picture. The system is based on a book by Miner. I have not seen a decent translation of it yet, I read it in the original. I have not read Fisher well either, I will have to have a look.
The essence of the implementation is in the nuances, which are not present in the general description on onyx.
Now a couple of words on the subject. Trawl is bullshit, it ruins everything. Oscilli are also rubbish, they show the average temperature in a hospital. There is either one - a discrete market model, or another - a continuous model. They cannot be combined together. The prospects for a discrete model are huge, but there are many obstacles.
You don't have to read Fisher. He's got a lot and it's all off-topic really. It's like a high school encyclopaedia. But that's where I saw the Fibo channels. The calculations are all just DiNapoli's. By the way, I got a little excited about DiNapoli (he's a smart guy, I respect him a lot :)). His channels are different from traditional Fibo channels after all.
About it working, I can feel it in my spinal cord. But there are a lot of external conditions. I cannot fully understand it yet. Found a few, will try to check them programmatically. But that's all the time.
I emailed you, but at the address I had (at mail.ru). It would be nice to exchange ideas. I myself am in the process. Right now I'm picking up a demo. But every time I raise it too much, I dump almost everything I've gained in yet another experiment. I'll have to try and do a clean week sometime. Right now the balance is a little over 4,000 with a start of 3,000. As soon as I double it and master my technique I will open small real account and check if there is still a problem. But for now this is the first mechanism that allows me to steadily increase my deposit. I have read your link. I don't get the fucking idea. Stupid, I guess :) I will read it later at my leisure. Maybe I'll get it.
Good afternoon, everyone! I've been thinking about this.
It is becoming more and more evident to me every day that the "grail" formula (without any irony) is as follows:
A good profitable Expert Advisor must contain (not one or two, but) at least half a dozen different versions, oriented towards different reasons! For example - for an Up-trend, for a Down-trend, for an active flat, for working with models, for working with channels, etc. ......
In the most ideal case - existence of a "semaphore" for switching versions, - but this is not an obligatory condition....
Also - multicurrency, timeframes...
Good afternoon, everyone! I've been thinking about this.
It becomes more and more evident to me every day that the "grail" formula (without any irony) is as follows:
A good profitable Expert Advisor must contain (not one or two, but) at least half a dozen different versions, oriented towards different reasons! For example - for an Up-trend, for a Down-trend, for an active flat, for working with models, for working with channels, etc. ......
In the most ideal case - existence of a "semaphore" for switching versions, - but this is not an obligatory condition....
Also - multicurrency, timeframes...
A dozen more conditions could be added. But really, it's either there or it's not. I think the second one. Which means you need experts for different conditions. And at the same time, the profitability of the worker at the moment exceeded that of the others.