Auto or manual - page 11
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
1. Nobel)
2. It is not feasible in forex, more so for stocks. for crypto too, as long as there is liquidity)
3. this is the selection into a portfolio of instruments without links to each other. But there is a global BUT that cannot be taken into account, an external factor affecting several instruments at once. The goal is correct.
Why do we need a Nobel Prize) the market will pay for everything
In forex, there is also asymmetry, but it is not so obvious and more difficult to understand. In short, in equities, as the price increases, so does the amplitude, but there is only one asset. In Forex it works in both directions. The amplitude grows but not so obvious and it is the equivalent of 2 assets traded back and forth. But it allows you to earn less. The point is that the amplitude of fluctuations increases and in addition currency pairs, in contrast to the shares, have a flat character, but not all of course. If you gain profit in dollars and pounds separately and then sum it and translate into dollars, you can earn. In the screenshot I showed GBPUSD chart and what I got after conversion.
order type and set mathematics and logic as the logic. Simple functions such as multiply, add, divide, subtract, if, and, or. Let him choose which data to analyse and which mathematical functions to apply. Generate a million individuals, allocate a conditional dollar to each and make them all trade. Next, who has earned on breeding, let them breed with mutations and so on.
In the beginning it will be a wild randomness, over time, the system should learn something and start to identify some patterns. But this is very difficult and not cheap.
The idea is that this is almost the case with me.
Any TC "lives" in the system only as long as it has not exhibited "unacceptable" behaviour. As soon as such behaviour is detected, the system is immediately "eaten up". It is replaced by a new system of the same type, re-tested on the history.
It needs to happen in real time, through gene transfer from successful individuals with mutations
can i hear more about the strategy generator? do you mean selecting a random number generator to trade from a pool of plummer flatcore ts-oks :-) a sharable one that might turn out to be profitable when traded on the real?
---
I've seen a strategy generator here somewhere before... I think so, you set the conditions - and it generates stratagems...
Can you elaborate on the generator you have in mind?
I wrote above, have a look
You have to make it happen in real time, through gene transfer from successful individuals with mutations
Yeah, that's in the classic genetic algorithm. I don't have that. For the simple reason that there is only one TC of each species. And to organise a classical genetic selection, each of the League's 700 TCs must be represented by a set of TCs with different parameters, as soon as the next one shows unacceptable behaviour - it is removed from bidding, and a TC whose parameters are created by "crossing" the parameters of active TCs with mutations is put in its place.
But, all this is too complicated, we need to completely rework the League code.
Plus there are also appearing constraints of brokerage companies, like limitation of number of pending orders. For example because of this we had to divide the League into three divisions.
So for each TS there is only one instance, and when it fails, it is immediately re-optimized, and then it starts working again.
Why do we need a Nobel) the market will pay for everything
In forex there is also asymmetry, but it is not so obvious and harder to understand. Briefly, in equities, as the price increases, the amplitude also increases, but there is only one asset. In Forex it works in both directions. The amplitude grows but not so obvious and it is the equivalent of 2 assets traded back and forth. But it allows you to earn less. I will show you the bottom line a bit later.
I put off the subject of the logic generator in 2015. It's complicated, but not so complicated that I couldn't solve it. I wanted not to prevent the computer from generating any strategies, and to do this I started with mathematics. That is, to allow it to generate absolutely any strategy. I will lay down the selection of an instrument, selection of order type and set mathematics and logic as the logic. Simple functions such as multiply, add, divide, subtract, if, and, or. Let him choose which data to analyse and which mathematical functions to apply. Generate a million individuals, allocate a conditional dollar to each and make them all trade. Next, who has earned on breeding, let them breed with mutations and so on.
In the beginning it will be a wild randomness, over time, the system should learn something and start to identify some patterns. But it's not very easy and it's not cheap.
I wrote above, have a look.
hmmm... interesting...
I think I saw something similar here, there was something like a mess of indicators, interpretations of their readings and values, their weights, etc. And the optimizer in the tester chose variants that worked in profit with SL, TP...
Probably even in an article in MQL5 Wizard... Exactly here somewhere... I wasn't dreaming, anyway... :-)
Why do we need a Nobel) the market will pay for everything
In forex there is also asymmetry, but it is not so obvious and harder to understand. In a nutshell, in equities, as the price increases, the amplitude also increases, but there is only one asset. In Forex it works in both directions. The amplitude grows but not so obvious and it is the equivalent of 2 assets traded back and forth. But it allows you to earn less. I will show you the bottom line a bit later.
I put off the subject of the logic generator in 2015. It's complicated, but not so complicated that I couldn't solve it. I wanted not to prevent the computer from generating any strategies, and to do so I started with mathematics. That is, to allow it to generate absolutely any strategy. I will lay down the selection of an instrument, selection of order type and set mathematics and logic as the logic. Simple functions such as multiply, add, divide, subtract, if, and, or. Let him choose which data to analyse and which mathematical functions to apply. Generate a million individuals, allocate a conditional dollar to each and make them all trade. Next, who has earned on breeding, let them breed with mutations and so on.
In the beginning it will be a wild randomness, over time, the system should learn something and start to identify some patterns. But it's not very easy and it's not cheap.
It's expensive without optimization... Although the power is growing and the logics are really described by Boolean and mathematics. Come to that)
Your theses:
1) all experts periodically earn, but lose more than
2) to earn, it is necessary to rotate experts in time, switching off those who currently lose
3) no criterion of future failure, so rotation is guessing and belated, because time is needed to state the period of losing or earning
The absence of profit at a distance is the result of the absence of some regularity in the Expert Advisor logic, which, by definition, allows to predict the future state of the process with the probability more than 0.5. Since it is absent, why should we imitate market analysis inside an Expert Advisor using indicators, if there is no valuable forecasting information coming from such "analysis"? You can just trade a coin or a set of pennies and lose the spread in the same way.
How is it possible to conclude from the fact that "all EAs are losing" that one can earn on a distance with such a set? This is absurd. If the probability of a profitable outcome is obviously less than 0.5, the amount of its realizations is a sure loss. How can one, having no methods of analysis, no regularities, i.e. no criteria for intelligent decision making, pull the set of knowingly losing balance trajectories to the area above zero? You want to add negative numbers to get their positive sum, well, that's just impossible.
what do locs have to do with it? There are no lots, it's a picture in the terminal. There is only an open and a closed position.
Here is an example on 28 stocks without leverage, netting, with commissions
30% a month is for dreamers. There are no algorithms that give such returns. You just have to forget about such figures forever! For those in the know, it is obvious that the yield of any algorithm is limited either by liquidity or by the capabilities of the theoretical model. If you take HFT and arbitrage, you can essentially provide any yield there, even 1000% per month, but the question is the amount you can wrap. Ultimately, in these algorithms, the returns are hinged on liquidity.
If you use other algorithms, you won't get the same returns. To start with, we need a normal theoretical model and build on it.
Yes, I have achieved 20-25% annual return in foreign currency on American equities and I am working on reducing resource consumption, simplifying the algorithm, improving the yield to 30-35% per annum and smoothing out the yield curve. And this is an excellent yield, it is enough for me stably with the risks tending towards zero.
30% per annum with minimal risk - that's the grail.
No well this is exactly my favourite topic, now I'm sprinkling you with questions:
You are sitting on an MQL forum and not at a large hedge fund investor meeting, only forex traders reside here,
My trading robot has a realistic 30% mark, you can reach it, it's not about stability, but 99.999% of people have enough liquidity to use forex.
You say that all algorithms are limited by liquidity. The question is how much money are we talking about?
in forex as I know the maximum guaranteed volume of up to 200 lots in the major operators, and that's a second 1 pip = $ 2,000, with 5 leverage it would be about $ 4 million of capital required, which allows you to theoretically make the forex market 100% annually on effective TS, on any algorithm without a lack of liquidity.
I'm not even talking about CME, where you can generally trade huge volumes.
is such a sum not enough for you and for the dreamers of this forum?
Or do you want to manage like Buffett with 10 digits?
Where is the bottom line where everything stops working, liquidity will be lacking, I don't understand, this is what amount in your mind?
No well this is exactly my favourite topic, now I'm sprinkling you with questions:
You sit on the MQL forum and not in the meetings of investors of a large hedge fund, only forex traders live here,
My trading robot has a realistic 30% mark, you can reach it, it's not about stability, but forex liquidity is enough for 99.999% of people.
You say that all algorithms are limited by liquidity. The question is how much money are we talking about?
in forex as I know the maximum guaranteed volume of up to 200 lots in the major operators, and that's a second 1 pip = $2000, with 5 leverage it would be about $ 4 million of capital required, which allows the theoretical to make the forex market 100% annually on an effective TS, on any algorithm without a lack of liquidity.
I'm not even talking about CME, where you can generally trade huge volumes.
is such a sum not enough for you and for the dreamers of this forum?
Or do you want to manage like Buffett with 10 digits?
Where is the bottom line where everything stops working, liquidity will be lacking, I don't understand, this is what amount in your mind?
A realistic figure of 30% per month? Well, with expectation=0 yes, it is real, why not? Today it's +30, tomorrow it's minus 40.
I was talking about liquidity when I wrote about HFT and arbitrage strategies. Are you going to run HFT algorithms on forex using mt5 terminal? With such algorithms there is competition for time. Yes, you can theoretically wrap large sums on CME and forex with these algorithms, but you also have to compete for time. I think very few of the local public can do that.
I didn't say that any algorithm rests on liquidity, only those on which you can make high percentages with proven efficiency, those are quite specific algorithms. And where they don't hinge on liquidity, they hinge on competition for time.
And with the algorithms that have been discussed here, based on price analysis, making 30% per month is very unlikely. You can do it, but it's random, not profitable. I myself made 150% a month in 2009, a few months in a row, and took out a profit. There were other stories when I made 50% a month. But it all makes no sense, if expected payoff = 0.
The essence of what I'm writing is that you need to learn how to make at least some stable profits with proven efficiency. Find a pattern that consistently yields at least 1% annually on top of commissions in forex. I see trading as a job, in the context of gambling, yes I am wrong, in the context of stable returns I am right. I myself will bring money to the person who will show me how he makes 30% a month stably and with minimal risk and explain why it works. And there will be no money problems at all, the sums will be normal. I would completely step back from development and research and would do nothing but raise money for this. But it won't happen.
And it makes no difference which forum to sit on, it's the same everywhere.