Some signs of the right TCs - page 31

 
Mihail Marchukajtes:
I see such equities, especially when the author calls them interesting I really feel sorry for such people, because the main thing in the market is not to deceive HIMSELF. What's so interesting about it? All the same overstaying with a drawdown of the balance. You may think that the real drawdown will be bigger by the deposit size. Law of meanness. Do not deceive yourself with synthetic tools as well. I agree that they may be very good, but one has to trade on real quotes anyway. IMHO naturally!!!!
The meaning of this particular algorithm is the most suitable to the subject. It is a self-adapting robot. It does not imply optimization at all. You may use it on any symbol with equal settings (I checked 28 pairs, 28 stocks plus 5 cryptocurrencies) and it does not shed. And it works on minutes, on history of any duration. Even if there are gaps in the history, it does not matter, it will digest them. That is why this result was interesting to me. I assumed that if we add a static coefficient it should not affect operation, I thought it would not affect reverse quote too, but it turned out that there are some problems, because on a reverse quote trades should be identical to the direct one.
 
Valeriy Yastremskiy:

...

And the predictability of multifactor functions depends on the number of factors and their predictability. In general, a function can become unpredictable if there are a large number of predetermined factors. And if the factors are predetermined but unaccountable, the problem is even more difficult.

Well, everything becomes clear at this point. All we have to do is to gather known factors into functions, and in consequence of their unconditional limitedness (we cannot gather ALL factors) our functions will automatically become predictable.) In fact, traders do just that, forgetting, however, how many real factors they have.

Try to extrapolate in your imagination your "formulaic" approach to price to ALL traders and see how the mathematical "looping" of their TS multiplies Random in the Market exponentially.
 
Реter Konow:
Well, everything becomes clear at this point. It remains to gather the known factors into functions, and due to their unconditional limitations (we cannot gather ALL the factors) our functions will automatically become predictable. As a matter of fact, traders do just that, forgetting, however, how many real factors they have.

I'm just talking about the opposite. Molecularly, we can only calculate their motion up to a certain number of molecules. We can't account for all the jumping of electrons and the action of neighbouring molecules. Brownian motion does not stop, it starts only with a certain number of molecules, i.e. one does not give a Brownian motion and stops, two too, but a thousand may give, I don't know exactly. And I believe that as soon as Brownian motion begins, the possibility of an exact calculation ends, it becomes probabilistic. We have probabilistic miscalculations and price movement analysis.

 
Valeriy Yastremskiy:

I'm just talking about the opposite. Molecularly, we can only calculate their motion up to a certain number of molecules. We can't account for all the jumping of electrons and the action of neighbouring molecules. Brownian motion does not stop, it starts only with a certain number of molecules, i.e. one does not give a Brownian motion and stops, two too, but a thousand may give, I don't know exactly. And I believe that as soon as Brownian motion begins, the possibility of an exact calculation ends, it becomes probabilistic. We have probabilistic miscalculations and price movement analysis.

There used to be no Brownian motion in the Market. Traders flat out interpreted local and global events and decided to buy/sell. The less scientific the process was, the more predictable the market was. But, computers have changed everything. Now there is meaningless Brownian motion everywhere. Thanks to them and mathematical analysis).

But, since the trend cannot be stopped, it has to continue.
 
Реter Konow:
There used to be no Brownian motion in the Market. Traders flatly interpreted local and global events and decided to buy/sell. The less scientific the process was, the more predictable the market was. But, computers have changed everything. Now there is meaningless Brownian motion everywhere. Thanks to them and mathematical analysis).

What to do, that's the way it is. There were no scalpers before) and you couldn't do neural network algorithm optimization before either. But life is changing.

 
multiplicator:

so these are lagged quotes. that doesn't count.

Let me remind you, it's about whether it's possible to build an algorithm that meets 4 requirements

Nikolai Semko:
I would add the main one:

  • it has no period-dependent parameters.
  • the TS operation does not depend on the current chart timeframe
  • the operation of TS does not depend on the symbol-tool
  • the entire setup of TS is only the setting of risk management (the size of the deposit used)

some people think it's impossible,Dmitry Fedoseev:"Utopian fantasy". Actually it's arbitrage, I wouldn't say regular arbitrage, requiring a difference of 2 spreads, but it's spatial arbitrage nevertheless. And it is so possible and even frequent in reality, that many brokerage companies prohibited its use in their client agreements. At the same time they were proud to use it in their advertisements: "the execution policy: the best price available on the market is chosen".

As for the lag, I agree at 20%. That is how long the markets of the 2.5 days in the above statement are panicking due to the fall in the price of oil. Both for all of us and for the brokerage companies panic is not a common occurrence, everyone is not prepared for it. It is hard to quote and brokerage companies' algorithms are not ready for it. The trading algorithm that has generated this quote is ready.

There was no panic during the previous 2 days, but the algorithm increased the deposit 10 times https://www.mql5.com/ru/forum/333746/page29#comment_15344956.

As a matter of fact, I never set the goal differently. The first 3 features were natural from the beginning, the fourth appeared while programming and implementing the trade. Actually, 4 points of requirements: not only the size of the part of the deposit to be traded, but also the value of the opening threshold in spreads and the minimum size of the expected profit itself specified as a commission. Expected profit itself is simply defined as the size of rate deviation of this brokerage company from the average rate of the covered brokerage companies.

The questions, why I do it, have already been asked here. I will tell you 3 motives: interesting, one; adds knowledge, two; if you need to raise money - you can go to demo accounts contest with real prizes, three.

Dmitry Fedoseev
Dmitry Fedoseev
  • www.mql5.com
Добавил опрос Как у вас с электропитанием? Добавил тему Круглые скобочки против квадратных скобочек Соревнование перегрузки оператора [] и вызова обычного метода с параметрами. 1. Один параметр: class C1{    protected:    int sum;    public:    void C1(){       sum=0; Выставил продукт Библиотека для расчета формул. Формула задается...
 

As if to confirm what I said above

"As for the lag, I agree at 20%. That is how long the markets have been panicking out of the 2.5 days in the above statement, caused by a fall in oil prices. And for all of us, and for the brokerage companies, panic is not a common occurrence, everyone is not prepared for it. It is hard to quote and brokerage companies' algorithms are not ready for it. The trading algorithm that has generated this situation is ready.

- MT5 terminal has just been updated to build 2361. Without any prior announcements, all of a sudden. Plus the announced update to build 2360 happened just recently, this morning. Seems to be a reaction to the panic. I wonder what will be written about this build.

 
Vladimir:

As if to confirm what I said above

"As for the lag, I agree at 20%. That is how long the markets have been panicking out of the 2.5 days in the above statement, caused by a fall in oil prices. And for all of us, and for the brokerage companies, panic is not a common occurrence, everyone is not prepared for it. It is hard to quote and brokerage companies' algorithms are not ready for it. The trading algorithm that has generated this situation is ready.

- MT5 terminal has just been updated to build 2361. Without any prior announcements, all of a sudden. Plus the announced update to build 2360 happened just recently, this morning. Seems to be a reaction to the panic. I wonder what will be written about this build.

Don't panic, the ichthyosaurs will be ours. Even if they have updated the terminal.

 
Алексей Тарабанов:

Don't panic, the ichthyosaurs will be ours. Even if they have upgraded the terminal.

This thread is not about how people panic. It's about algorithms that are resistant to various changes in the properties of incoming rates. Now I looked, the deposit is already 2617884. Three days out of 10k on the back of unexpected chatter in these flows.

 
Vladimir:

This thread is not about how people panic. It's about algorithms that are resistant to various changes in the properties of the incoming rate stream. Now I looked at it, the deposit is already 2617884. Three days out of 10k on the back of unexpected chatter in these flows.

Just a shitty feed on the broker's demo. There is no point in this activity. On the contrary, when you make a normal system, you have to cut out the shitty places in the feed, because they will distort the real result (read "achievable in reality").