If profitable EAs ? - page 8
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
I can write, but there is a psychological moment, and there will be a lot of text (just a general hint) because it is not within candlestick patterns, timeframes, etc.
I'll think about it....
I can write, but there is a psychological moment, and there will be a lot of text (just a general hint) because it is not within candlestick patterns, timeframes, etc.
I'll think about it....
Any TC idea can be described in two or three sentences.
I can tell you're lying, for it reveals a simple chain of logic:
1. you write about"a stupidly straightforward idea, but very beautiful".
2. You forget what you wrote in the last post, "and then Ostap carried" - began to rave about the heavy text, otherwise you can not describe.
p.s On the subject - there is.
Don't bullshit the nitwits, eh.
Any TC idea can be described in two or three sentences.
I can tell you're lying, because a simple chain of logic reveals it:
1. you write about"a stupidly straightforward idea, but very beautiful".
2. You forget about what you wrote in the last post, "and then Ostap carried" - began to rave about the heavy text, otherwise you can not describe.
p.s On the sabotage - there is.
OK, I'll respond to the comment.
For me with my practical experience the idea is beautiful (at the same time the mathematical model is extremely complicated), stupidly straightforward and simple, but to make it clear to the masses, we need to go over basic concepts (Level 2 nuances) and explain a lot of things.
A year ago, we had a long time ago a system developed for the modeling where order opening was based on point, local signals characterizing the market condition at the current time and in a small vicinity near the current time. In the real market a loss was accumulating when working with market orders.
I had to change the approach in the simulation, to give the future states of the market depths (for the order opening module, forcing the entry and exit points) in which in the interval of 1 second (any price can be set) the worst prices were taken with the largest volumes (Ask\Bid) plus the commission added and at these Ask\Bid prices the strategies were considered for market entry/exit.
The hint is that we should think and solve the question how to remove slippage at Market Execution when we work with market orders, because the work is intraday and the charts are not synthetic, but based on updates (without filters on the server side) of Level 2 market conditions. I.e. to create a system based on.......
This is what HA was thinking of explaining.
The model uses many components (neural network committees, spectral analysis, deterministic chaos, sau, for capital management dynamic programming and some other topics)
From my experience - there is nothing complicated - you just need to present and explain it correctly
And what is the basis of the system?
OK, I'll respond to the comment.
For me with my practical experience the idea is beautiful (at the same time the mathematical model is extremely complicated), stupidly straightforward and simple, but to make it clear to the masses, we need to go over basic concepts (Level 2 nuances) and explain a lot of things.
A year ago, we had a long time ago a system developed for the modeling where order opening was based on point, local signals characterizing the market condition at the current time and in a small vicinity near the current time. In the real market a loss was accumulating when working with market orders.
I had to change the approach in the simulation, to give the future states of the market depths (for the order opening module, forcing the entry and exit points) in which in the interval of 1 second (any price can be set) the worst prices were taken with the largest volumes (Ask\Bid) plus the commission added and at these Ask\Bid prices the strategies were considered for market entry/exit.
The hint is that we should think and solve the question how to remove slippage at Market Execution when we work with market orders, because the work is intraday and the charts are not synthetic, but based on updates (without filters on the server side) of Level 2 market conditions. I.e. to create a system based on.......
This is what HA was thinking of explaining.
The model uses many components (neural network committees, spectral analysis, deterministic chaos, sau, for capital management dynamic programming and some other topics)
p.s. let me clarify my position: in my opinion, this requires no more than general school level maths.
I was talking above about trading via FIX using L2 data.
If I'm in the mood and time I'll briefly describe what it's for, neural network committees, spectral analysis, deterministic chaos, sau, dynamic programming ...... all included in one model. Maybe someone will be stimulated to dig into L2. Real big money doesn't come easy.
p.s. no arbitrage strategies are used, it is strictly FX market with indicative price and analysis inside liquid symbols.
The question remains unanswered, where is the proof that the application of these methods is not pseudo-practical lah-buh-buh-buh?
The question remains unanswered, where is the proufe that the application of these methods is not pseudo-practical la-boom?
Daniil Stolnikov:
Интересно - как это может красивая, тупорыло прямолинейная и простая идея быть чрезвычайно математически сложной?
Fermat's theorem
Actually the hint is that we need to think and solve the question, how to work with market orders (Market Execution) to remove moments with slippage, based on the fact that the work is intraday, charts are not synthetic, and are built on updates (without filters on the server side) states of the Level 2 market. I.e. to create a system based on.......
This is what HA was thinking of explaining.
What is the problem? Especially if the normal exchange of why do market places?
In short, it is solvable. Or it is not critical. Or the problem is not.
How does this relate to the model?
I am interested in another question - did the development costs pay off?
Of course it is good not to fuck up with everything (research directions), otherwise it may happen that years and resources have been wasted, so I understand you about the references. Of course I can pull up software archives and take videos of ready-made solutions and get people fired up, but so much has been made up that I'm too lazy to make straws for others.
I think the self-made GUI pictures are just a distraction... well, to show at least some results.