a trading strategy based on Elliott Wave Theory - page 244
![MQL5 - Language of trade strategies built-in the MetaTrader 5 client terminal](https://c.mql5.com/i/registerlandings/logo-2.png)
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 such a noticeable difference in results in the above pictures can only be explained by the "predilections of the brokerage company management", then we might as well put everything else down to it after a real trading failure! For example, the villains brokers have changed the methodology of filtering or just changed the provider of streaming data for objective reasons. Of course, I also can't guarantee anything - I'm just giving my opinion... What if you manage to find something, which will be the same for quotes from any broker (of course within some reasonable and understandable limits, but certainly not as in the pictures above)? Well here I can give you an example. If we take several brokers, whose servers have the same server time, then it is an absolutely known fact that the prices of O-H-L-C bars of the D1 period will differ by a negligibly small value. And all calculations made on this data will also have similar results, the difference between which can only be really explained by "management bias". That is, once you have calculated the system at one broker, you can work without recalculation at another broker with approximately the same results. There will probably be more consistency in this than in the stability of the MF of the chosen broker...
Yes, the coincidence is not bad. That's good.
And concerning the difference of data from different brokers I would like to say my IMHO. :-)
1. If the strategy depends on the quote provider, then it's better to abandon it. At the same time the data can really differ and this should be taken into account by the trading program. So if a strategy works differently on different data streams and in both cases successfully, then that is a real strategy and the required robustness.
2. The difference in the results of kagi and renko constructions on data from two different brokers may have very different reasons. Sergey, are you sure that there are no gaps and associated gaps in your broker's data?
If they exist, the zigzags will differ not only in form, but also in statistical structure. And this will inevitably affect H-volatility.
3. No one has said anything about this so far, and it's important. Renko construction does not have the unambiguity that is quite inherent in kagi. As a result, for an H=10 parameter value, for example, 10 renko constructions can be made. And they will differ from each other. I have not seen anything about this in Pastuhov, but I do not understand all the details there, maybe I missed something. Or maybe that is what he has left up his sleeve. This ambiguity does not change the validity of the theorems he proved, but it introduces another degree of freedom and should be studied at least to some extent before drawing conclusions about the suitability of renko for trading. If I have time I will post my results today as well.
4. solandr, I understand your concern that Neutron has entered a slippery slope. :-)
However, I think your concern is unfounded. We are all here looking for reasons and insights, not parameter fitting schemes. That's why this discussion has such a quasi-scientific form, so distasteful to many. Participate constructively. The example you gave with diaries is obvious: any integration of ticks into one bar erases the differences. And the bigger the bar interval. The differences, by the way, are the same, but relative to the size of the bar they are insignificant.
H-volatility, which Pastukhov introduced, is a non-trivial characteristic of the market. And if it is indeed a fundamental characteristic, as stated by the author, it should be slightly sensitive to the changes of the quote provider and sufficiently stable in long time intervals. This is what we are trying to find out - is it true? Aren't you interested in the answer to this question?
Of course I am. That's why I try to participate in the discussion only constructively, expressing some of my doubts, inspired by some of my experience in trading and expert writing. And many other forumers, who don't take part in the discussion but nevertheless receive some useful information, are also interested in the answer to this question. In fact, everyone has been waiting a long time for some foreseeable results of on-line demo testing. I hope they are not too far off?
PS: It will probably be a very unusual event - the fact that the information in the open thesis will be enough to create something workable and not only in MathCad. It's just that I know well enough from my own personal experience what a PhD thesis is and so I have no great illusions about the thesis under discussion. The main task of a thesis, which must be necessarily fulfilled, is to meet the formal requirements of VAK and ONLY the SECOND thing is the possibility to actually (practically) use the results obtained in it! In WAC (and firstly in the Academic Council) no one will require from you any test-statments of the idea in real life. There more than enough of numerical simulation, which is the main proof of the possibility of practical use of the results of the thesis in practice.
Although it is possible that I am overly sceptical about it! Well you will be able to show it. All the more so, as everything is shaping up quite successfully at the current moment in time.
As far as I have guessed, it's about comparing two values with different dimensions (points and ticks, respectively), but you can't compare them like that... or am I just not sure what you mean?
Sergey, maybe, I don't understand something, but to compare meters with meters or minutes with minutes will hardly be interesting, if not to take into account absolutely simple comparisons. For example, "to be 15 minutes late is less than to be 30 minutes late". Usually people make comparisons and look for regularities between different metrics.
to Solandr
It's just that I know well enough from my personal experience what a PhD thesis defense is and therefore I don't have great illusions about the thesis under discussion.
I share your point of view and have no illusions about the thesis under discussion. The opinion is purely personal and is formed on results of reading.
I cannot say anything about Alexey Denisov and the Fibonacci timeframes he uses, because I am not familiar with him and his method. If you are interested in whether there will be a pullback in the pound to the 1.99 level, I believe there will be. )I'm expecting a mark much higher, possibly even above 2.1000. Gotta watch and count...(.
The signs are probably right here http://awakening1.narod.ru/foto.htm : )))
and hieroglyphics, yes, quite possible...
I'm just "at the beginning of my journey" and much is not yet clear to me, but I am confident that movements (trends) can be "calculated" and predicted with 99% probability out of 100. And the more I "dive into the analysis", the more my confidence in this fact increases :)))
There are no coincidences in the market, ALL movements are logical!
To write a really working strategy, you need to understand all the details, not to mention the fundamentals. The principles inherent in my strategy, which I "understand and apply by hand", are very difficult to explain to a programmer in words. I've already experienced with this, so it's probably faster to learn the programming language yourself :))).
As for stop levels, it's not that simple... In my opinion, one cannot use the same pattern for all currency pairs, moreover, I'm sure that for different timeframes the "stop loss scale" should be different.
To be honest, I don't see the point in it now... :)
:о)))
It will be fun to watch the "OKA" vs "BMW M6" race :-)))
[img]Interbank FX, LLC
A/C No: 4947 Name: 2006 February 3, 21:01 (local time)
Closed Transactions:
Ticket Open Time Type Lots Item Price S / L T / P Close Time Price Commission R/O Swap Trade P/L
9958 2006/12 00:12 balance Transfer from #4947-5 14187.60
11400 2006/12/21 13:17 buy 1.09 usdjpy 118.35 0.00 0.00 2006/12/21 15:29 118.46 0.00 0.00 101.21
11694 2006/12/27 11:18 buy 0.02 usdjpy 118.66 0.00 0.00 2006/12/28 12:56 118.93 0.00 0.81 4.54
11728 2006/12/27 16:41 buy 1.31 usdjpy 118.81 0.00 0.00 2006/12/28 15:18 118.92 0.00 53.06 121.17
12454 2007/01/03 15:07 buy 1.42 usdjpy 119.46 0.00 0.00 2007/01/03 17:40 119.61 0.00 0.00 178.07
12915 2007/01/09 14:21 buy 2.07 usdjpy 119.49 0.00 0.00 2007/01/10 15:22 119.74 0.00 27.95 432.18
14146 2007/01/17 14:51 buy 2.03 gbpusd 1.9696 0.0000 0.0000 2007/01/17 15:32 1.9706 0.00 0.00 203.00
14429 2007/01/18 14:35 buy 2.02 eurusd 1.2959 0.0000 0.0000 2007/01/18 19:49 1.2961 0.00 0.00 40.40
14872 2007/01/23 17:47 sell 2.04 eurusd 1.3017 0.0000 0.0000 2007/01/24 08:05 1.3004 0.00 16.42 265.20
14932 2007/01/24 11:12 buy 1.02 usdchf 1.2476 0.0000 0.0000 2007/01/24 15:45 1.2486 0.00 0.00 81.69
16028 2007/01/30 09:14 buy 0.97 eurusd 1.2957 0.0000 0.0000 2007/01/31 05:21 1.2961 0.00 -8.49 38.80
16201 2007/01/31 06:24 buy 1.91 eurusd 1.2962 0.0000 0.0000 2007/01/31 06:34 1.2964 0.00 0.00 38.20
15119 2007/01/25 18:44 buy 1.40 eurusd 1.2966 0.0000 0.0000 2007/01/31 15:06 1.2975 0.00 -49.00 126.00
16329 2007/01/31 15:19 buy 1.92 gbpusd 1.9554 0.0000 0.0000 2007/01/31 15:22 1.9561 0.00 0.00 134.40
16370 2007/01/31 15:30 sell 1.92 usdchf 1.2471 0.0000 0.0000 2007/01/31 15:44 1.2460 0.00 0.00 169.50
16659 2007/02/01 07:56 sell 1.92 gbpusd 1.9654 0.0000 0.0000 2007/02/01 08:00 1.9641 0.00 0.00 249.60
0.00 40.74 2183.96
Deposit/Withdrawal: 13020.02 Credit Facility: 0.00 Closed Trade P/L: 2224.70
Open Trades:
Ticket Open Time Type Lots Item Price S / L T / P Price Commission R/O Swap Trade P/L
16481 2007/01/31 15:58 sell 1.92 gbpusd 1.9578 0.0000 0.0000 1.9675 0.00 30.24 -1862.40
0.00 30.24 -1862.40
Floating P/L: -1832.16
Working Orders:
Ticket Open Time Type Lots Item Price S / L T / P Market Price
No Transactions
A/C Summary:
Closed Trade P/L: 2224.70 Floating P/L: -1832.16
Deposit/Withdrawal: 13020.02 Total Credit Facility: 0.00
Balance: 15244.72 Equity: 13412.56
Margin Requirement: 1920.00 Available Margin: 11492.56
[/img]
I am particularly interested in the open trade? does it have the potential for positive realization, of course, within theory??
always thanks...
SarkeeV no offence, but I can't tell you whether these trades fit into Elliot wave theory or not. To give an unambiguous answer, you need to analyze and get to the bottom of it.
Unfortunately, I deal with Forex only in my spare time, which I have very little.
My work occupies most of my time. To my great regret, it has nothing to do with the Forex market.
P.S. I have a question - is gsb your nickname? A couple of days ago I got an e-mail from St. Petersburg.
If it's from you, the offer is VERY INTERESTING, but I will not be able to take part physically.
If not, sorry and don't be offended, it means I've got the wrong address :)
:о)))
grasn it will be fun to watch the race "OKA" against "BMW" :-))
So, one can put an engine from Messerschmitt on "OKA" too. :о) It was just an honest joke, without any hints, rather than fanatical assertion of any directions (I'm far from fanaticism by nature). Besides, I don't really have Elliott Wave Theory and not exactly in its entirety. Only a reduced basic model with boundary conditions (rules and conditions are taken from the program winwave32) plus my own research in this direction (I just imagine the Elliott wave in the form of signals with the subsequent DSP). The wave itself, propagating through a "reliable" channel has a fractal structure inherited from the source. Well, there is something else interesting...
Besides, calculation of the forecast already takes 3-7 hours, depending on the specific structure of the series itself. And this is quite a lot...
And these are my results of kagi and renko yield calculations for EURUSD, all ticks 2006 and the Wiener series of 2200000 ticks you presented earlier. In comparison.
For kagi you can see the difference between the statistical structure of real and model tick constructions.
For renko this difference is if not problematic, then not so significant. Nevertheless just for renko it is possible to arbitrage at rather large H. At H=49 and 46 it is possible to earn even on a random process. :-)))
I think, that jerkiness of profitability for Renko is caused by that ambiguity of constructions, about which I wrote in the previous post. As confirmation, I give H-volatility data for 20 variants of reneko constructions at H=20 for euro and model ticks.
You can see that for the model ticks the H-volatility is not very different for different variants. For several of them it lies exactly on 2. The same cannot be said about real ticks. The difference is significant here. Depending on the construction method it is possible to obtain both a non arbitrage variant Hvol=2, and a significant difference from 2, that is the possibility to make profit.
But the price series is one and the same ! So is there an opportunity for arbitrage or not ?
I wonder if anyone can answer this question ?
You, Yura, be careful with your jokes. I'm sure there will be references to this joke of yours as proof of the possibility of building a miracle TS that earns under any conditions and even, on a random process. As for the possibility of arbitrage on large H, I would be extremely careful here. After all, the lack of statistics may present any surprises.
I have divided a number of my EURUSD ticks of the last year into two equal pieces and estimated stability of the total profit in points, at one point spread, depending on H:
Two islands of stability for renko-building in the area of 15 and 33 point renko-breaks attract my attention. From the very beginning of my work with reneko partitions I wondered about the dependence of returns on the starting point, the uncertainty you mentioned, and got reassuring results:
One can note the very low sensitivity of the Renco-partitioning method to initial conditions! Probably for this reason, Pastukhov did not pay any attention to this point in his work. I think this is the answer to your question:
If you are an optimist, I can say that there is an arbitrage opportunity! And it is not a fact that EURUSD is the optimal pair for it.