The future of automated trading - page 2
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
Interesting situation:
On forums where they trade manually - they write, automatic trading is rubbish and few people use it.
And on forums where they trade automatically - they say that soon they will not trade manually...
Everyone is pulling the blanket over his side, that's all.
In fact, most EAs are just partial human substitutes trading on principles laid down by the programmer.
And a small percentage of systems analyze the market and think by themselves what to do.
Even a praised neural network is like a little child that won't grow, but just learns to perform some actions partially.
It takes a lot of time to compose algorithms for network training, selection of input parameters, activation functions, normalization.
During this time, we can stack up a bunch of ordinary linear EAs and even merge them into one.
And let's not forget: One person's destiny doesn't always fit another person's destiny. There are people who trade manually, and there are people who are traded by experts.
Automated trading may be rubbish, but the percentage of trades performed on the exchange by automated systems is steadily growing and already exceeds the number of trades performed by humans. Whether it suits someone or not, but it's an objective fact. Mass conveyor production and robotic assembly plants replaced manual handmade production long ago. Algorithms don't go after algorithms for the good of life. Market efficiency is growing, which means fewer and fewer fish in this pond, analyzing the market is already very difficult to make something.
A computer has an undeniable advantage over humans in the amount of information processed. Let a man try to look through several thousand charts in a day to see if there is a trend on them. In practice, it is practically impossible. A computer, on the contrary, will need dozens, or even hundreds of times less time to accomplish this task. This is one of the aspects of using computers in trading. For example, you have a trend following system with a good performance. The task: to find the instrument on which a trend movement is currently occurring. Open it and make a profit. Now the large funds analyze all markets: Asian, European, American. They analyze futures, currencies, commodities, etc. on these markets. This kind of work can only be done by a computer. By the way, I recently read an article about an American company where only robots trade. Well, the income of this corporation is measured in billions of dollars a year. So you have to wonder which is better: hand trading or robot trading. Just in our country this whole industry is at an antebellum level. And with dealing centers with requotes at 0.01 lots you cannot even think about good trading.
To a large extent it is a moot point. On the one hand, yes - computers carry out a certain type of analysis much faster than an average trader. On the other hand any existing algorithm is so far inferior to the human brain in its ability to understand and evaluate (at least theoretically) all the factors and processes existing in the market.
The situation is as simple as a pig's ear, let me give you an example:
Many people play chess, some even well. But just how many people in the world play at the level of, say, KARPOV or KASPAROV? Many of those who play can try to play against the computer (they even try to do so). But who among those who play chess will be able to put up a decent fight against a supercomputer, and not against an ordinary desktop PC, that has been honed for chess?
In my opinion there are two trends in this market:
1. People start using mechanics and automation to make their work easier. First, they use indicators and scripts that help to automate some analysis and trading processes. Then traders' appetites and demands grow, and they think about full-featured mechanics. At least the maximum possible automation of the trading process.
2. Against this background we can see that a certain number of traders reject GLOBAL and ALLOWED automation. It is caused by the fact that many traders, for various reasons, being disappointed in automation of trading processes, return to manual trading.
In any case, one thing is unquestionable - mechanical trading is and will be used by most traders. This is due to the fact that there will always be more people willing to pass the responsibility on to a no-nonsense robot.
Just with time, the trader begins to understand the disadvantages and advantages of automation, begins to take a different view of the processof trading operations and the possibility of market analysis with the help of AI.
PS
Ultimately machines remain REAL in markets where a lot of technical analysis is required, for example in the stock market. They are also in demand in trading processes where a lot of trading is done.
Some people trade with robots, others do it by hand. Ultimately, it is everyone's choice. But what must be done either way is to test your ideas and strategies on history. Doing it with "pencil and paper, by hand", as Dr Elder advises, is simply impossible. After the fact, we always know what will be on the right side of the chart - this will inevitably make adjustments in manual testing. But the robot does not know this. Therefore, the only possibility to check if your thoughts are correct is to write a trading robot and run it on history. Many may argue that the process of trading is too multifaceted, too complex. Many strategies are impossible to formalize. That's true, and I'm ready to admit that robots, due to this limitation, trade worse than humans. But the strategy has to be so effective, so "wooden", that even a robot can earn on it. If a robot can do it, a human can do it.
The process of speculation is a monotonous repetition of the same actions a hundred, a thousand times over many years. A man is not an assembly line; it is difficult for him to do the same job for years on end. It is better for him to keep inventing, changing, "improving" something. This is the main reason why many systems that are profitable on history stop being profitable in real conditions (I'm not considering pseudo-profitable over-optimized systems). Even trading robots is difficult to maintain uninterrupted process. In my practice there have been cases of power failure, operating system errors, and provider's maintenance. These are the days when the market offered an opportunity to make a good profit, and it was missed. It is difficult to stick to any system while trading manually. Traveling, physical impossibility to be at the monitor 24 hours a day, other work, free time, sleep. All this inevitably takes up most of your life. All you have is a few hours a day to devote to trading. To trade during this time, just because it is free, is a sure way to the graveyard of traders.
Some people trade with robots, others do it by hand. Ultimately, it is everyone's choice. But what must be done either way is to test your ideas and strategies on history. Doing it with "pencil and paper, by hand", as Dr Elder advises, is simply impossible. After the fact, we always know what will be on the right side of the chart - this will inevitably make adjustments in manual testing. But the robot does not know this. Therefore, the only possibility to check if your thoughts are correct is to write a trading robot and run it on history. Many may argue that the process of trading is too multifaceted, too complex. Many strategies are impossible to formalize. That's true, and I'm ready to admit that robots, due to this limitation, trade worse than humans. But the strategy has to be so effective, so "wooden", that even a robot can earn on it. If a robot can do it, a human can do it.
The process of speculation is a monotonous repetition of the same actions a hundred, a thousand times over many years. A man is not an assembly line; it is difficult for him to do the same job for years on end. It is better for him to keep inventing, changing, "improving" something. This is the main reason why many systems that are profitable on history stop being profitable in real conditions (I'm not considering pseudo-profitable over-optimized systems). Even trading robots is difficult to maintain uninterrupted process. In my practice there have been cases of power failure, operating system errors, and provider's maintenance. These are the days when the market offered an opportunity to make a good profit, and it was missed. It is difficult to stick to any system while trading manually. Traveling, physical impossibility to be at the monitor 24 hours a day, other work, free time, sleep. All this inevitably takes up most of your life. All you have is a few hours a day to devote to trading. Trading during this time, just because it's free, is a sure way to the graveyard of traders.
To sum it up:
Man is a lazy creature.
In the villages we used to work monotonously from 6am to 10pm, now we don't want to bang on the buttons to eat afterwards :D
What are you talking about? I'm not talking about forex kitchens, I'm talking about real exchanges and real trading for real money.
Who would doubt it, where else but here I can meet the "real Livermore".
I don't know about the Livermores, but there are plenty of idiots here who have no information but an opinion on any subject.
One cent is the minimum change in the share price, i.e. you cannot earn less than one cent. And it is for this one cent that computer algorithms are actively fighting.
Many people play chess, some even well. But how many people in the world play at the level of, say, Karpov or Kasparov? Many of those who play can try to play against the computer (they even try to do so). But who among those who play chess will be able to put up a decent fight against a supercomputer, rather than an ordinary homebrew PC?
A good analogy. A few big players put themselves chess supercomputers in the stock market colocation, and what can pioneers with metatraderies do against them?
It's not the supercomputer that decides, but the person behind it.
So you're ready to wrestle in chess against Deep Blue and even hope to win? Kasparov failed to do so.
Besides, behind the supercomputers of the big companies are the best mathematicians and economists. And what can pioneers with metatraderies do against them?
So you're ready to wrestle in chess against Deep Blue and even hope to win? Kasparov failed to do so.
Besides, behind the supercomputers of the big companies are the best mathematicians and economists. And what can pioneers with metatraderies do against them?
Good analogy. Some big players are putting chess supercomputers in colocation on stock exchange, and what can pioneers with metatrader do against them?
Good analogy. A few big players put themselves chess supercomputers in a colloquy on the stock exchange, and what can metatrader pioneers do against them?
It's not about what kind of computer one has. Ordinary terminals can be combined into a network and get a supercomputer (if you want and with enough skill).
All the matter is that whatever the client terminal may be, it does not allow to implement all the required (and at the same time it works quickly and error-free).
And if you consider the fact that any server software is in a better position in this case, it becomes clear - It is not possible to beat the DC (market) all the time, at least if you use one strategy.