Developing a stable trading robot - page 7
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.e. can you at odds 50/50? 0)
Or do we understand "chance" differently?
Or is it trolling again?
The mathematical term is chance from probability or frequency:
odds = p / (1 - p)
Chances, frequencies and probabilities have nothing to do with expectation. I.e. 90% of profitable trades may be profitable, i.e. p = 0.9, but at the same time, 10% of losing trades will exceed the profit and the expected payoff will be negative.
MO = profit * p - loss * (1 - p)
Accordingly, profitable TS with strict or average profit and loss should meet the following condition:
profit * odds > loss.
The only drawback is that it needs 1l cents, or 10k, which I don't have yet, to work consistently with a 0.1 lot. The only disadvantage of this robot is its stable operation with a lot of 0.1 cents, or 10k, which I have so far. I am waiting for participants willing to help check the performance of the robot in real life. I only have 1k.
You're confused... I think he was talking about a cent account, which requires 1,000,000 cents ($10,000). He didn't publish the availability of cents first for nothing, he's a millionaire! :))) Comrade Yusuf has 100,000,000 cents ($1,000). But Roman correctly noted, if you use the minimum lot 0.01 (I myself use such a micro-real - throw 1 dollar and have fun!!!), then the money is enough to start testing the Expert Advisor on the real account. But something is wrong here, if he wants to share risks! :))))
You need hard evidence to make that claim.
This implies that you have managed to build a TS with expected payoff >0. Maybe you should apply other methods of money management, they give more profit with less risk of ruin compared to martin. And the pictures you gave don't prove anything, you need 3-4 hundred trades on a forward period.
You're confused about something... I think....
I've got an ancient model of my own lying around somewhere and I still can't get to it. In this model it is possible to get out of the flat, even if the external information inflow is zero. I.e. there should be a reason, but it is not necessarily an inflow of new information.
Alexei! Anything is possible in the market - you know that. But the model is built on some assumptions about the "physical" nature of the process. Since L. Bachillier, random (now fractal) wandering has been discussed. One cannot make money on it (if one is in the market all the time) - it is proven and developed in options. )
Your model pretends to analyze the external information - which channels, which market participants receive it? Does Vasya Pupkin receive it in time and in full? Nothing has been voiced. Maybe there are mistakes? Information is not news - information is already processed news, practically - the field of existence of a new "fair" price. The market eventually settles on it. All the more so because the volumes of liquidity involved in the adjustment are also news, which the market continuously takes into account.
So the question is, why are you not discussing your brilliant conjectures, which are stuffed in a drawer and allegedly you don't get around to sorting them out?
Open your own forum - let's discuss it together. :)
And the fact that "tangential integrals" :) will appear - the readers will have a reason to look through the old notes.
also skipped maths and high school?
There is nothing brilliant there. And that ancient model, in general, has nothing to do with alexeymosc's information research. And it is not developed to the state of extracting valuable information... I am just saying this by the way, to let you know that sometimes it seems to me that movement is not always conditioned by incoming information.
Information is not news - information is already processed news, practically - the area of existence of a new "fair" price. The market eventually settles it. Moreover, the volumes of liquidity involved in the adjustment are also news, which the market continuously takes into account.
Here in the feature selection thread it turns out that the information (in general, as an abstract statistical concept, not a specific piece of it, which is usually called news) for a move on the zero bar is in the distant past. It has not even been fully internalised in the past. That is the inefficiency.
We have investigated the phenomenon using different methods - chi-square test (me) and information theory(alexeymosc). In fact it is almost the same, if you look closely. But the results in the form of my namesake become more convex, sort of.
Since L.Bachillier the random (now already fractal) wandering has been mused on. One cannot make money on it (if one is in the market all the time) - it is proven and developed in options. )
Bachelier was certainly cool for his time. Shall we recall some even more ancient model?
keep87:
You may find it very difficult to give up, but I'll console you, I spent half a year writing various grids (like Ilan), calculated all the moves and combinations, as you have already understood, I was convinced of the complete incompatibility of such a system with any income, "-half a year". I did it too, I share my experience, although you probably won't listen, I recommend to get acquainted with relevant books on probability theory.
P.S. Until you clearly identified a pattern you want to beat, there is no difference between the price movement on the chart from random (like roulette) for your system. I suggest you to find a pattern so you know what you're working with.