Machine learning in trading: theory, models, practice and algo-trading - page 2150
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
The influence of the news background does not necessarily have a momentary impact on price behavior. It depends on the significance. It depends on the weight of players' wallets. But any news that comes out will affect the behavior of players. It plays the role of TA. A strong player can wait for the best moment when liquidity gathers in the market.
I don't see any obstacles!!! (contradictions) The task is just to determine which news at what moment become significant, and what it may depend on)
Quiet everyone, I'm hurt pretty bad.
Another bottle I broke before I left the store?)
You broke another bottle before you left the store?)
No, that was before the fun always, now it's different.
I think that unambiguous synchronization is impossible to achieve, based on various factors and conditions of their preservation and issuance. But I do not see a problem in this. The time of the news within an hour before and after the time of close attention. The task is to understand what changes introduce a particular news in the behavior (model) of BP. What does it depend on. From the day of the week, or hour of the day, or day of the month or lunar calendar. Or all the same from the forecast and the weight of the news. and the correlation of the forecast weight and the action. and difficult, the combination of all inputs and the result.
Got carried away with another task while at it. Described earlier. It's like an equilibrium problem on new data. that's how we keep the equilibrium. and don't fall. Here from ticks to a year fractally and rules can be described. Basically a balance strategy on a slide of bumps))))
I see. I also think that it is impossible to get a guaranteed synchronization - I will have to be content with what is) Model is simple - SB with drift. Just want to see how variance and drift rate change around the news. I'm thinking of using MO to study the dependence of these two parameters (or at least their ratio) on news parameters.
The concept of equilibrium is a very important one for current economic science. Its achievability, sustainability (efficiency), etc. Before Keynes it was thought that equilibrium was always achieved by itself - the "invisible hand of the market") Now it is believed that the economy cannot be in effective equilibrium by itself and this leads to a constant increase in the role of the state in it.
It is clear that you mean stability in a different sense, but I believe that the two are connected - the stability in quotes comes from the state's attempts to make the economy stable.
I don't see any obstacles!!! (contradictions) the task is just to identify what news at what point becomes significant and what it may depend on)
It's hard for me to judge how you factor the news background into your analysis of the markets. Maybe I missed something when I missed it).
I see. I also think that it is impossible to get a guaranteed synchronization - I will have to be satisfied with what we have) Model is simple - SB with drift. I just want to see how variance and drift rate change in the vicinity of the news. I'm thinking of using MO to study the dependence of these two parameters (or at least their ratio) on news parameters.
The concept of equilibrium is a very important one for current economic science. Its achievability, sustainability (efficiency), etc. Before Keynes it was thought that an equilibrium was always achieved by itself - the "invisible hand of the market") Now it is believed that the economy cannot be in effective equilibrium by itself and this leads to a continual increase in the role of the state in it.
It is clear that you mean sustainability in a different sense, but I suppose they are related - sustainability in quotes comes from the government's attempts to make the economy sustainable.
The reaction to the news can be short-term, or it can be blurred over time. It can be played out before it is announced. Synchronization is hard to achieve. It still hinges on the TA. Important news is most likely played by the big players in manual mode. And then the robots come into play. The work of robots is easier to trace.
I see. I also think that it is impossible to get a guaranteed synchronization - I will have to be satisfied with what we have) Model is simple - SB with drift. I just want to see how variance and drift rate change in the vicinity of the news. I'm thinking of using MO to study the dependence of these two parameters (or at least their ratio) on news parameters.
Clearly you mean stability in a different sense, but I suppose the two are related - stability in quotes comes from the government's attempts to get the economy to be sustainable.
I agree, the same way I compare states, variance or width and velocity. Also the ratio of the variance of the section under study to the longer one. And by the ratio to the point. I use for a single estimate and write an algorithm to start the window from that point and return and drop the point. Such a manual filter.
I like Keynes, but that's for later.
I simplify the problem to just equilibrium without conditions on the new data. We have historical data, we have a real point where to keep the equilibrium or choose the direction of movement and at each new given to make a decision. If the price does not go beyond the limits, the equilibrium is stable, and if it does and stays there for a long time, it passes the filter, then we should change the equilibrium disposition.
It's hard for me to judge how you take the news background into account in your analysis of the markets. Maybe I missed something when I missed it).
By changing the speed of the price channel and its width.
I agree, the same way I compare states, dispersion or width and velocity. Also the ratio of the dispersion of the investigated section to the longer one. And the ratio to the point. I use for a single estimate and write an algorithm to start the window from that point and return and drop the point. Such a manual filter.
I like Keynes, but that's for later.
I simplify the problem to just equilibrium without conditions on the new data. We have historical data, we have a real point where to keep the equilibrium or choose the direction of movement and at each new given to make a decision. The problem is simplified by the fact that there are borders of equilibrium and it is reduced to the definition of these borders. If the price does not go beyond the borders then the equilibrium is stable, and if it does and stays long, i.e. passes the filter then we have to change the disposition of the equilibrium.
One of the standard approaches is the same CUSUM. It only applies to errors of the model used instead of price increments.
Someone with a knowledge of game theory
this package might help.
https://cran.r-project.org/web/packages/EvolutionaryGames/vignettes/UsingEvolutionaryGames.html