Probability, how do you turn it into a pattern ...? - page 10

 
Neveteran >>:

Логика существует в виде программы уже около года.
Вокруг данной модели поведения в рынке существует инвест фонд с приличным капиталом и постоянно открытым конкурсом управляющих.
Цель данной ветки, поиск людей готовых отойти от стереотипных подходов в управлении капиталом, готовых принять участие в семинаре FXYalta сентябрь - октябрь 2010

P.S. я действительно плохо владею русским, по причине что начал изучать его в возрасте 20 лет.

But can your strategy work with a net position?

because essentially any profitable strategy should not depend on a position accounting system.

If there is a statistical advantage, it does not matter how you account for the position, it (statistical advantage) will manifest itself anyway.

 
Neveteran >>:

Давайте отбросим количество инструментов, на уровне понимания логики вцелом, это совершенно непринципиально.

А теперь примитивно, повторюсь по раскрутке первого цикла. (замечу что я сейчас опишу половину логики)
...

О! Well, that's a different story!

The idea is new, everything seems great and it may work...but there is one thing: the correlation of pairs, as a result of which the rise and fall of different pairs is not

It's not so easy to operate with independent events, and hence with probabilities.

Another thought: remove locks from the system - just close profitable positions. Why give extra money for the spread? You will free up more margin.

And one more thing: during the second cycle the price may go far enough and you will get 0 for 15 instead of 7 for 8 and you will be stuck with it for half a year, unless of course Kolya Morzhov

comes sooner...

 
It is not clear what statistical feature of a series of quotes is being used. If you can't identify it, then for you time series are akin to a random process on which it is impossible to make money, no matter how you convert it.
 

I think it's just another load of nonsense. :) Alas :) How much is it? If it's not too expensive, you should take it.

 
SProgrammer >>:

Помоему это какая-то очередная ахинея. :) Увы :) Надо узнать сколько стоит? Если не дорого то надо брать.


It's like an American movie.
A client looks through a window where a girl is stripping, when only her pants are left, the window closes and an inscription appears.
The seminar will reveal the other secrets of brilliant trading will be held in Yalta, in September. The cost is blah-blah-blah.
 
This is an interesting idea.
I think that for its implementation we take quite a specific number of instruments, depending on the number of currencies used. For example, for 4 currencies we take 6 pairs. Now, we face the question of calculating an initial position for each symbol in such a way as to obtain the equality to the volume in the deposit currency or equality to the point value in the deposit currency.
How this problem is solved in your case?
 

Ticket Open Time Type Size Item Price S / L T / P Price Commission Taxes Swap Profit
69769109 2010.03.22 15:14 buy 0.10 gbpaud 1.64812 0.00000 0.00000 1.64691 0.00 0.00 0.00 -11.07
69769114 2010.03.22 15:14 buy 0.10 eurnzd 1.91937 0.00000 0.00000 1.92012 0.00 0.00 0.00 5.28
69769118 2010.03.22 15:14 buy 0.10 audcad 0.93233 0.00000 0.00000 0.93295 0.00 0.00 0.00 6.08
69769133 2010.03.22 15:14 buy 0.10 gbpusd 1.50128 0.00000 0.00000 1.50730 0.00 0.00 0.00 60.20
69769135 2010.03.22 15:14 buy 0.10 gbpcad 1.53618 0.00000 0.00000 1.53687 0.00 0.00 0.00 6.77
69769140 2010.03.22 15:14 buy 0.10 euraud 1.48046 0.00000 0.00000 1.47864 0.00 0.00 0.00 -16.65
69769142 2010.03.22 15:14 buy 0.10 nzdcad 0.71916 0.00000 0.00000 0.71821 0.00 0.00 0.00 -9.32
69769146 2010.03.22 15:14 buy 0.10 gbpjpy 135.247 0.000 0.000 135.743 0.00 0.00 0.00 55.07
69769152 2010.03.22 15:15 buy 0.10 audusd 0.91111 0.00000 0.00000 0.91501 0.00 0.00 0.00 39.00
69769191 2010.03.22 15:15 buy 0.10 eurcad 1.37989 0.00000 0.00000 1.37980 0.00 0.00 0.00 -0.88
69769205 2010.03.22 15:15 buy 0.10 audchf 0.96831 0.00000 0.00000 0.96787 0.00 0.00 0.00 -4.16
69769214 2010.03.22 15:15 buy 0.10 eurusd 1.34847 0.00000 0.00000 1.35328 0.00 0.00 0.00 48.10
69769220 2010.03.22 15:15 buy 0.10 audnzd 1.29777 0.00000 0.00000 1.29816 0.00 0.00 0.00 2.75
69769225 2010.03.22 15:15 buy 0.10 eurjpy 121.494 0.000 0.000 121.875 0.00 0.00 0.00 42.30
69769236 2010.03.22 15:15 buy 0.10 cadjpy 88.082 0.000 0.000 88.299 0.00 0.00 0.00 24.09
69769244 2010.03.22 15:15 buy 0.10 nzdjpy 63.309 0.000 0.000 63.438 0.00 0.00 0.00 14.32
69769247 2010.03.22 15:15 buy 0.10 usdjpy 90.105 0.000 0.000 90.061 0.00 0.00 0.00 -4.89
69769259 2010.03.22 15:15 buy 0.10 eurgbp 0.89834 0.00000 0.00000 0.89767 0.00 0.00 0.00 -10.10
69769263 2010.03.22 15:15 buy 0.10 audjpy 82.129 0.000 0.000 82.403 0.00 0.00 0.00 30.43
69769271 2010.03.22 15:15 buy 0.10 chfjpy 84.832 0.000 0.000 85.115 0.00 0.00 0.00 31.43
69769273 2010.03.22 15:16 buy 0.10 nzdusd 0.70267 0.00000 0.00000 0.70449 0.00 0.00 0.00 18.20
69769287 2010.03.22 15:16 buy 0.10 cadchf 1.03904 0.00000 0.00000 1.03721 0.00 0.00 0.00 -17.30
69769297 2010.03.22 15:16 buy 0.10 usdcad 1.02290 0.00000 0.00000 1.01969 0.00 0.00 0.00 -31.48
69769304 2010.03.22 15:16 buy 0.10 nzdchf 0.74688 0.00000 0.00000 0.74507 0.00 0.00 0.00 -17.11
69769308 2010.03.22 15:16 buy 0.10 gbpchf 1.59556 0.00000 0.00000 1.59448 0.00 0.00 0.00 -10.21
69769312 2010.03.22 15:16 buy 0.10 usdchf 1.06262 0.00000 0.00000 1.05784 0.00 0.00 0.00 -45.19
0.00 0.00 0.00 205.66
Floating P/L: 205.66

Just opened buy on 26 pairs. I see that losing/profitable ones are almost 50/50. If we close the profitable ones and average half of the unprofitable ones and reverse the other half...

 

Let me ask the author a question - where in nature does the normal distribution come from?

And in order not to overcomplicate with continuous random variables what will be the distribution of the "magic" (with 32000 faces) cube?

 
Neveteran >>:

.............
Дальше разжовывать нехочу.

That is not required. Just as it was not required by this post repeating your previous one, and mine, where I named the main points of your system as follows:

1. opening both ways, or, same thing, opening on x number of pairs.

2. lock. (ha! for the sake of the balance curve, which no one needs for shit - only to get your mind blown).

3. Averaging/martin on loss. (Very new - prize in the studio!).

4. Walking out on hopelessness fatigue. (How I understand you...)

And don't talk about cycle memory for the hundred and fifth time! It's not children here, for God's sake. You can go on screaming that it exists - no one will argue with you. Who knows what you mean by that. Maybe the state of the account - he remembers, no doubt about it. No hypothesis, no proof. It's the only thing you can cling to in order to justify the profitability of the TS. What is this?

Without it, there is no point in going further. Except that it seems to me that ..... OK, what if?))

===

I want to ask Alexei again:

What, including spitting on TA, did you see new there? There was no such thing as entering the market, locking and averaging from the ground up?

If you discard all the verbal crap, what remains is what remains.

===

Alright then. Suppose flies get turned on by the heat, and with sleight of hand you can make a deterministic process out of a random one. Can the author give the ALGORITHM of the Expert Advisor? Only without any homebrew concepts and definitions - this is not a network marketing meeting. Just an algorithm. And the reasoning (there is no reasoning here, except for butter, and, judging by the approach, is not expected) somehow come later.

 
kharko >>:
Идея интересная.
Думаю, что для ее реализации берем вполне конкретное число инструментов, в зависимости от числа используемых валют. Например, для 4 валют берем 6 пар. Теперь возникает вопрос вычисления первоначальной позиции по каждому инструменту так, чтобы было равенство по объему в валюте депозита или равенство по стоимости пункта в валюте депозита.
Как решается эта задача в вашем случаи?

The weighting problem is solved as a system of linear equations.
If we assume that the number of available trading instruments is one, then the topikstarter's idea is pure martin.

Simultaneous operation on a portfolio of trading instruments is almost operation on one synthetic (as a portfolio of all trading instruments) trading instrument. I.e. it is almost a martin.
Reason: