For the buyers and customers of Expert Advisors (trading robots), this is dedicated.
Well? ))))))))))
Does anyone else have any doubts about the need for complex mathematical analysis of trading systems?
Secure yourself against possible mistakes before you put some serious real money on your trading account!
I do not use complex calculus - I have an eighth-grade education ))))
I trade without maths. The topic is interesting, go on (no irony, honestly).
...
Any logical decision-making rules, dependencies, indicators can be described in the form of an accurate mathematical model, and once this model is created, it can be tested on as large an array of historical data as desired and the verification results evaluated.
Of course, the formation of an accurate mathematical model, writing additional verification programs and various checks take considerable time. Sometimes one week, sometimes several - that's a lot of work. But...
1. Subsequent checking of the ready model on all available history takes, unlike an Expert Advisor, from fractions of a second to several seconds.
2. Further on, you can do everything with the model. Change the settings, rules, sets of rules, etc., while immediately getting visual charts of results of such changes, which will allow you to write a correct profitable Expert Advisor, which will not need to be changed later.
3. Unlike advisor, mathematical models allow you to get very fast, accurate and comprehensive information, the result of checking the feasibility of any trading hypotheses, assumptions, systems.
>>>So the buyer or customer gets a "cat in a poke" and cannot check the rules written into it <<<<
--- How so? That's what the terms of reference are written for. Point by point. Each point can be checked to see if it works in the program.
If all the clauses work, the programmer has done his job. The programmer has programmed what was described in the requirements specification.
If the logic of making a profit in the order was incorrect - the programmer had nothing to do with it. The blame of the customer should not be shifted onto the implementer.
Figuratively speaking - the customer ordered a "crooked" couch carpenter, when receiving the couch fit the drawings, but it is impossible to sleep on it. Is the carpenter to blame?
I don't use complex calculus - I have an eighth-grade education ))))
I trade without maths. The topic is interesting, keep it up ( no irony, honestly ).
Thank you for your interest in the topic .) I will certainly continue.
What is the difference between this "mathematical model" and the EA? It seems that someone is reinventing the wheel. You can change everything in an EA: parameters, rules, and everything else. How fast will depend on how you write it all.
This article is not aimed at programmers. I personally don't write trading robots. It's not my bread and butter. But you probably program it? Are you familiar with the MQL programming language? Are you familiar with it? You're right, the Expert Advisor can change the rules, parameters and everything else. But not everyone can change the code of the Expert Advisor. And ordering a new program every time is at least long and costly. A ready-made matrix can be rotated like a kaleidoscope, and the analysis results will be obtained in seconds.
Put yourself in the customer's shoes. How long does it take just to agree the TOR, correspondence, etc.? And how much time will it take for the Advisor to check a large historical interval? You know the answer. And if something is wrong and the programme has to be rewritten? What will he do if the program is written correctly, but the rules do not work? And this is while the customer cannot do without the services of a programmer!
My task is different. Namely, the programmer must provide the client with a clear set of rules, the use of which will give him a given financial result. He has written (for example, with your help) the Expert Advisor and began to work quietly.
I am not 'reinventing the wheel'. In the teams where I have worked and am working, this has always been the case. Everyone does their own thing: programmers do their programming, analysts do their research and verification.
If all the points are working - the programmer has done his job. He has programmed what was described in the TOR.
If the logic for making a profit in the order was flawed, the programmer has nothing to do with it. You should not shift the blame from the customer to the implementer.
It's very interesting to see how different people perceive the subject). I already answered this question above and I will repeat it. My subject is not in spite of programmers. Moreover, I don't want to offend anyone. If a programmer did his/her job correctly - respect to him/her. But it is a case where there was a customer and there was a contractor. If a person understands what he/she is buying - great! And how many people buy Expert Advisors (trading robots) simply for the sake of advertising, having the vaguest idea of what they are paying for?
My task is only to ensure that the client comes to the programmer with a clear set of proven rules, the use of which will give him a given financial result. So the client will not put the money into the trading account and lose it, not understanding how it happened. For such people I am talking about the possibilities of mathematical models for trading in FOREX.
It's pretty stupid to run an EA on ALL available history. What worked great in the 70's does not work at all in the 2000's.
The starting post is a very good chart which shows that the TS - SHOULD NOT be tested on "all" history. Any TS is built for a certain market behaviour, and should be used exactly when the market has that behaviour. Of course, in each TS there should be a "stop-cock", which clearly indicates that the system has stopped working and needs at least a revision, or even needs to be replaced.
As for the purchase of robots - I do not see anything wrong with it. The trader builds the TS, and in order not to perform the routine actions of its usage, orders an Expert Advisor that will precisely follow its rules. By the way, the Expert Advisor must also monitor the "stop crane" itself and give a signal to the trader about changes in the market behavior.
yes, can you tell me more about how or what calculates the mathematical model, by µl or other means?
Another attempt to invent the Grail?
What is the math model if we are all acting as the other side of the INSIDER deal?
There is one law in the marketplace: "A CROWD MUST LOSE". And the operation of this law does not depend on the presence (or absence) of your matrix. :)
When starting to trade on a real account, the trader must clearly understand that sooner or later he will lose this deposit. It's an axiom! He will lose it because he is a TOTAL.
The main task of a trader is to try to earn as much as possible before losing his deposit! That's what robots are written for. These robots are NOT GRAILY, they just do routine work. These robots have strategies sewn into them that, at a given stage, allow them to make money. But that doesn't mean at all that they will earn forever! Remember, we are all a CROWD and the crowd has to lose!
That's why when I read about testing robots over a long period of time (3-10 years) and the author admires the test results, it makes me personally smile. The author of this super robot is just very far from the real events on the real market: you're trading against a cheat that's much stronger than you! :)
Want a free Grail? - "Team up with a loser trader and do the opposite! And you'll always be in the black!" ))))))))))
But seriously, let's leave the creation of Grails to someone who is actually capable of performing miracles.
The mathematical model is just a good way to evaluate the rules which the trader puts into the Robot. Robots are needed! I am not calling for them to be abandoned in favour of something else, quite the opposite.
As for "we are all a CROWD", not every trader considers himself a representative of a "CROWD", rather the opposite - an individual! And such a person is right.
You write: "We are trading against a cheat, which is much stronger than you are. Perhaps. That is his ("cheat") strength and his weakness. He cannot react to you personally! He can react to cumulative actions. In the market there are always those who win and those who lose. You're not going to argue with that? Well then. Comprehensive analysis of your own algorithms of trading decisions, which the mathematical model can provide you with, will allow you to be more statistically among those who profit than vice versa. That's all.
I know enough people and financial institutions around the world who just belong to the category of those who are more likely to be on the plus side. They don't shout about their Grail, they don't need big words and they don't offer anything to anyone. They simply do not need it. Why should they offer and sell their rules to anyone when they already work perfectly well?
- Free trading apps
- Over 8,000 signals for copying
- Economic news for exploring financial markets
You agree to website policy and terms of use
Any Expert Advisor is an individual set of clear rules for making a trading decision to open or close an order, to enter or exit the market. The totality of these failures forms the TS (trading strategy or a trading system) embedded into the program. However,
First, most people who buy or order EAs do not know the language of EAs, i.e., MQL. Therefore, the buyer or the customer gets a "pig in a poke" and cannot check the rules written in it, not to mention more complex systems, such as those using artificial intelligence. Then he wonders why the trading robot shows completely different results.
Secondly, the customer (the buyer), if necessary, cannot change these rules without assistance.
Thirdly, he cannot change the set of rules himself.
Fourthly. Checking of the robot on large time intervals takes too much time, so such checks are done by very few traders. And we have to check every Expert Advisor on all available history. Something wrong, the Expert Advisor shows a loss, we order a rework or change the EA and run the test again, but it is long, inconvenient, and it costs time and money again. Those who have tried it understand what I'm talking about.
Fifth, the profit diagram of most trading systems resembles a saw-toothed pattern, i.e. if tested on a certain interval, the EA may show good results (but not over the whole interval) and this depends not only and not so much on the trend direction and type, but on other reasons. See the example below.
Finally, most customers do not have a clear, tested TS. In this case, they can only rely on the advertising and authority of the Expert Advisor creator.
A proven trading strategy is a strategy tested on historical data for 10(!) or more years.
A proven profitable trading strategy is a proven trading strategy in which you:
What is the exit?
Any logical decision rules, dependencies, indicators, can be described in the form of an exact mathematical model, and by creating this model you can test it on an arbitrary large array of historical data and evaluate the results of the test.
Of course, forming an accurate mathematical model, writing additional verification programs and various checks take considerable time. Sometimes one week, sometimes several - it is a lot of work. But.
1. Subsequent checking of the ready model on all available history takes, unlike an Expert Advisor, from fractions of a second to several seconds.
2. Further on, you can do everything with the model. Change the settings, rules, sets of rules, etc., while immediately getting visual charts of results of such changes, which will allow you to write a correct profitable Expert Advisor, which will not need to be changed later.
3. In contrast to the Expert Advisor, mathematical models allow you to get very quick, accurate and complete information, the result of checking the feasibility of any trading hypotheses, assumptions, systems.
Bottom line.
According to statistics 98% of all traders lose their investments. They are driven by excitement, not cold calculation. When deciding to invest, for example in FOREX, even a few thousand dollars, and use a trading system or expert advisor, you must understand in exact figures the risk that you are taking. Testing any trading system with mathematical models will give a huge number of people this understanding and, at the very least, an opportunity not to lose their money. And for those who are already working and making profit, this check will allow them to get the most out of their TS.
In conclusion, I will give you an old example. I made this mathematical model back in 2000. The time horizon - 1999. - Present time horizon - 1999.
Currency pair EURUSD, time frame H1.
Here's such a funny graph ... Pay attention, a trading system, which showed excellent results from 2000 to 2004, and brought its customers (already then, well-to-do people from one known capitalist country) in 4 years of work a little less than 50 000 pt, became unprofitable.
The reversal from a stable profit to a net loss didn't happen in a month or a week... or even in a day. It happened in 1 hour! It wasn't an abrupt change in trend type and direction. No. It was an instant rule change. It is impossible to trace such a change visually on a candlestick chart. Only... let's call it a "market architect" can change the rules this way. And you, can you track it?
Well? ))))))))))
Does anyone else have any doubts about the need for complex mathematical analysis of trading systems?
Protect yourself from possible mistakes before you put some serious real money into your trading account!
18.02.2015.
Dear friends!
Firstly, I consider this information to be personal and confidential.I can't answer everyone, otherwise I would have to do no work, just this forum. I will be grouping questions and answering as much as possible.
With specific questions and serious suggestions on the subject, please contact me in person. You can find the email address in my profile.
Part 1.
Judging by your responses and those emails that I received just a day (!) after I had started the topic, I had no idea how big a gap in most traders' perception of methods of information analysis and decision-making from the methods of such analysis used in decent financial structures. This is probably why financial structures make profits on a regular basis, while the "crowd", as one of my opponents put it, makes regular losses. This kind of analysis is a common practice in a serious approach to trading, and I only wanted to bring this experience to the masses.
You can sneer at what I am writing and try to compare this type of analysis with a trading program. Understand that one does not cancel the other! You need an analysis, and the obtained rules can and should be integrated in a trading robot. As I wrote above, the analysis is my part of the overall work. I have always worked in a team. Why? Because it is much more effective than when one person is a "healer and reaper and maker of sticks" - and more effective in any area - in the army, in sports, in work...
Requests to show the amount of my money, my account, as well as "to provide a key to the flat, where the money is kept" will always be ignored.
Because:
Secondly. The results of my accounts are my version of the return/risk balance. My regular income is 20-25% per month. Maybe these figures won't impress some, but personally I'm quite happy with them. I myself have tested the rules with which my money works and I am happy with the result.
Thirdly. I am an advocate of "portfolio investments". This approach diversifies risks. It gives me more peace of mind.
Fourthly. I don't try to "knock on closed doors", forcibly convince someone that I am right, etc.
Fifthly, I do not ask anybody for money, neither for management, nor in any other way. I don't need it.
Part 2. "The mathematical model
What kind of a thing is a "mathematical model"? A mathematical model should not be understood as a "one size fits all". Your personal mathematical model is individual. It is a set of rules for decision-making specifically by you (!) described in the exact language of numbers and formulas.
When I opened this thread, I thought that the concepts of complex analysis and mathematical model will be close to everyone and clear, but perhaps it is necessary to supplement the main block of my topic by putting some examples, so that people can have a clear idea of the subject. I don't promise to do it urgently. It's a lot of work, but I'll do it for sure.
Good luck and profit, everyone!