Let's build a mini grail!? - page 9

New comment
 

this is what a grail should look like.

The question is, of course, that I don't know how or what it's based on.

it just works, that's all...

the issue is, of course, that i don't know what it is based on. it just works, that's all... and it allows me to cut my losses.

as i am a registered user, i don't know why it works, it just works and when it happens, i simply stop the system and restart it without losing even 10% of my profits.

if i knew why it worked, i would have put it on permanent)

i increased the risks 15! times to see what happens when it still loses...if you can call it that...a chart...ahem...a month and a half of profits...

If you see a "glow thread" as I call it, grab luck by the tail ... and if you want to tell me if you've figured out how it works ...

for me personally it goes like this:

...bacchanalia...bacchanalia...orders adjust to the right place for strong market movements...profit...profit...profit...bacchanalia...bacchanalia...profit etc.XD

why bacchanalia... because the system has no logic whatsoever... it just knows how to open orders when they are closed with a SL and TP...

and that's it...and the market does all the other logic for the robot...

This system, in my opinion, is by no means a grail. but as an interesting fun toy, it's a very good one)
 
Martin Cheguevara:

this is what a grail should look like.

The question is, of course, that I don't know how or what it's based on.

it just works, that's all...

the issue is, of course, that i don't know what it is based on. it just works, that's all... and it allows me to cut my losses.

as i am a seasoned trader, i don't know why it works and i just don't know if it works, but it works and when it does, i simply recognize it and stop the system and restart it without losing 10% of my profits.

if i knew why it worked, i would have put it on permanent)

i increased the risks 15! times to see what happens when it still loses...if you can call it that...a chart...ahem...a month and a half of profits...

If you see a "glow thread" as I call it, grab luck by the tail ... and if you want to tell me if you've understood how it works ...

for me personally it goes like this:

...bacchanalia...bacchanalia...orders adjust to the right place for strong market movements...profit...profit...profit...bacchanalia...bacchanalia...profit, etc.XD

Describe how it works, you can think about where the profit comes from and whether it is a profit.
 
Maxim Romanov:
Describe how it works, you may think where the profit comes from and whether it is a profit.
I'm telling you I don't know the logic of the market...
I can only tell you how I arrived at it...
I divide all market movements into three categories:
Trend, flat, flat-trend.
Each category has a 33% probability... more or less.
I thought I would take quanta of profit in two of three categories: trend, flat and trend. Quants, i.e. partial profit on trends, roughly speaking, by locking part of profit by the trading session the system locks part of profit simultaneously trading along the trend.
The third category of flat is essentially the same trends, only they are not directed to the edges of the "imaginary price corridor", but directed to the centre.
That is, there is a trend trade from the edge of the corridor for the pullback.
But it appears to be a paradox to trade from the edge of the corridor and to the edge of the corridor using two mutually opposite order systems ... giving an absolute Lok ...

 
That's the problem ... one cannot combine the unconnected ... although everything is possible of course ...
Almost everything...
If it works, then it will be a real grail that makes money on absolutely everything.


And what about order systems....the difference from trading with indicators is incommensurable here because when an indicator, even a very good one, detects something the order system is already trading it and making money on it.
 
We need some kind of uncertainty criterion and if we put it into flat order systems as well as into trend and trend-flat systems...
Then they will not be mutually exclusive...
But here is the problem again: the criterion is a number which is not relative, and everything in the market which is not relative leads to a loss...
The uncertainty criterion should even be determined by the market itself:₽ that's some kind of nonsenseXD.
 
But then again, all these criteria, etc., i.e., i gather the truth from my head ... The market should not be like this... The system should switch from flat to trend and vice versa...
Because according to the probability distribution(see my comments about leptokurtosis...) there is practically no time interval between the market change from flat to trend.
Either a long flat, or a very fast and very strong trend, or a sharp breakthrough of the minimum or the maximum, that's all.
 
Martin Cheguevara:
I'm telling you I don't know the logic of the market...
I can only tell you how I arrived at it...
I divide all market movements into three categories:
Trend, Flat, Flat-Trend.
Each category has a 33% probability... more or less.
I thought about taking quanta of profit in two of the three categories: trend, flat and trend. Quanta, i.e. partial profit on trends, roughly speaking, the system locks part of profit by the trading session and at the same time trades along the trend.
The third category of the flat trend is in fact the same trends, only they are not directed to the edges of the "imaginary price corridor", but directed to the centre.
That is, there is a trend trade from the edge of the corridor for the pullback.
But it appears to be a paradox to trade from the corridor edge and to the edge of the corridor, as two mutually opposite order systems... giving an absolute Lok...

If I understand correctly, I have implemented such a mechanism in its essence and it gives positive expected payoff. The essence of the following: the system trades bounce from trends, but if there is a long ubiquitous trend, then under certain conditions new systems start to create inside, trading smaller trends and their bounce. Then the profit obtained from the inside series is used to compensate the loss that has accumulated on the open position in a long trend. This mechanism itself leads the system from the negative to the positive expected payoff. Apparently, it is done in the same way, only with the help of lots. Where the profit comes from I cannot say scientifically, so I will say it in very simple words. The market has an excessive trendiness in general and the value of this trendiness is fluctuating within small limits, but if there is a trend on some scale, then on small scales fluxes start to appear so that the general trendiness remains approximately stable. We all know how to make profit in the flat market (not all of course, but I think many do), and that is where the switch from trend trade to flat trade appears. Due to these fluxes we manage to earn more than the price passed along the vertical line during the time of the large trend. That is, the price moved 100 pips vertically and incurred a loss of 100 pips, but inside it made a swing of 110 pips. As a result 100 compensate for the loss and 10 make a profit.

Now I have 80% of all profits going to compensate for losing positions, but I am working on a significant reduction of this percentage

 
Martin Cheguevara:
But then again, all these criteria and so on, i.e., i just want to believe what i have in my head, and i am the absolute truth... The market should not be like this... The system should switch from flat to trend by itself and vice versa...
Because according to the probability distribution (see a bunch of my comments on this leptocurtosis stuff...) ) there is virtually no time interval between market transitions from a flat to a trend.
Either a long flat, or a very fast and very strong trend, or an outburst, which has happened recently in the pound, or a constant relentless breaking through the lows or the highs, that's all.

that's how it will go from trending to flat. If there is a big trend, there will be a flat inside.

 
Maxim Romanov:

If I understand correctly, I have implemented a similar mechanism in its essence and it gives positive expected payoff. The point is: the system trades reversals of trends, but if there is a long trend without a reversal, then under certain conditions new systems that trade smaller trends and their reversals start to be created internally. Then the profit obtained from the inside series is used to compensate the loss that has accumulated on the open position in a long trend. This mechanism itself leads the system from the negative to the positive expected payoff. Apparently, it is done in the same way, only with the help of lots. Where the profit comes from I cannot say scientifically, so I will explain it in very simple words. The market has an excessive trendiness in general and the value of this trendiness is fluctuating within small limits, but if there is a trend on some scale, then on small scales fluxes start to appear so that the general trendiness remains approximately stable. We all know how to make profit in the flat we all know (not all of course, but I think many do), and that is where the switch from trend trade to flat trade appears. Due to these fluxes we manage to earn more than the price passed along the vertical line during the time of the large trend. That is, the price moved 100 pips vertically and incurred a loss of 100 pips, but inside it made a swing of 110 pips. As a result 100 compensate for the loss and 10 make a profit.

At the moment I have 80% of all profits going to compensate for losing positions, but I'm working on a significant reduction of this percentage

let me guess - new systems -> magic_order_buy++;magic_order_sell++; =)

I've already done this kind of thing, but not now.

But still 20% of profit almost guaranteed from losses that are guaranteed to be covered is already pretty good)

 
Maxim Romanov:

If I understand correctly, I have implemented a similar mechanism in its essence and it gives positive expected payoff. The essence is this: the system trades bounce from trends, but if a long non-flat trend appears, then under certain conditions, new systems that trade smaller trends and their bounce start to appear inside .

And if the conditions are practically nonexistent for more than half a year and the word "inside" is simply irrelevant)?

12345678910111213141516...36
New comment