Machine learning in trading: theory, models, practice and algo-trading - page 466
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
Do you think banks, HF go bankrupt? Go bankrupt.
So why, according to your market model, do you think that big money can do everything?
If they could do everything, they wouldn't go bankrupt.
Once the price (based on what you write and the fact that the banks and the CF go bankrupt) is unpredictable, even for the same banks and the CF, as a consequence one can assume that you can apply a trading model that will bring profit over time.
With respect.
It's not my model of the market, I just know how the big players trade.
So you yourself answered your own question why they go bankrupt - they ran out of money.) Of course this money needs to be managed, but in any case, the winner is the one who has more money.
Let me make it clear: you can't just go to the exchange and buy over 1000000 contracts, nobody is going to sell you that much. The point is to buy these contracts without anyone noticing.
The same thing about the price, the price never just goes anywhere. Why is there a flat in Asia? Because no one trades, Europe and America are always struggling.
You're right, you said "for a while", absolutely all strategies that are in the open access, only some are lost in a week, and some live for 2-3 years while there is some kind of a flat.
Well, you yourself have answered your own question why they go broke - the money ran out). Of course this money should be managed, but in any case the one who has more money wins.
When you say that every deal has a counter agent, 1 out of 2 wins. This is understandable, the flow of money passes from one hand to another. But the same way a large order is split into smaller ones and the market goes against it, the same players can split the bank position, like fishermen with nets against a seiner.
Ivan Rubtsov:
Let me make it clear: you can't just go to the exchange and buy over 1000000 contracts, nobody will sell you so much. The idea is to buy these contracts without anybody noticing.
I know that a big player splits his position into many smaller ones and can accumulate it for months.
As for the price, the price never just goes anywhere. Why is there a flat in Asia? Because no one trades, there is always a struggle in Europe and America.
Price is a derivative of the actions of players. as a consequence, if the volume of buyers goes up if the volume of sellers goes down.
Right you said "some time", absolutely all of the plum strategies that are in the public domain, only some are lost in a week, and some live 2-3 years while there is some kind of a flat.
There is nothing static in this world, so every object has a beginning and an end.
Ivan Rubtsov: In this world, nothing is static and every object has a beginning and an end.
It's not my model of the market, I just know how the big players trade.
So you yourself answered your own question why they go bankrupt - they ran out of money.) Of course this money needs to be managed, but in any case, the winner is the one who has more money.
Let me make it clear: you can't just go to the exchange and buy over 1000000 contracts, nobody is going to sell you that much. The point is to buy these contracts without anyone noticing.
As for the price, the price never just goes anywhere. Why is there a flat in Asia? Because no one is trading, Europe and America are always fighting.
You're right, you said "for a while", absolutely all strategies that are in the open access, only some are lost in a week, and some live for 2-3 years while there is some kind of a flat.
Almost like the great Paukas said, only he had another "and some people die rich before they are drained") )
Almost like the great Paukas said, only he also had "and some people die rich before they live to drain. )
Every person has a time when alzheimer's occurs, every trader has a time when he lost zero, or even a big loss, the size of the GDP of a small country, but not everyone lives till then, it is important that a man decides for himself what and how in his life will happen, They say that a trader is unlikely to work as a clerk in a bank or as a manager, this is unacceptable.
Almost like the great Paukas said, only he also had "and some people die rich before they live to drain. )
And he's right.
)))
The point is that we have several types of participants in the market, one of them is some Gennady, who sits and waits for a candle to bounce off the price and there is an uncle who has 100500 billion and a Depth of Market in front of him and he takes a position in a certain direction and he does not care about the price, indicators, analyst calculations and what else is there).
Put yourself in his place and everything will become clear at once, for example you want to make a good profit and push the price with marques and such and no here according to the rules now the flag/triangle/head shoulders/ - no now nothing will work =))
That's why it's IMPOSSIBLE to predict the price, you can't get inside that uncle's head and predict his intentions of what he's going to want tomorrow.
And all these figures/indicators/algorithms/candles make no sense, if they worked everyone would already be living in Miami and drinking Coke.
Right, I didn't argue.
So, what are the options?
went into philosophy...
It's a useful thing to do.
Good luck
Another interesting topic is the application of fuzzy logic and fuzzy neural networks to market forecasting. There is a huge scope for research and system improvement.
Methodologically very interesting article - combines different models
I think there is too much with models, I saw similar examples solved on one arima (+seasonal). And they have both stl, and then arima, and then forest.
And the forest itself is trained explicitly with overrides - in the pictures you can see how they adjust parameters so that the error in the training data was as small as possible. And the test data is not so perfect. The data is just very simple, that is why the model does not lose, and if it were Forex the balance would go to zero quickly.