Bayesian regression - Has anyone made an EA using this algorithm? - page 10
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
Formulas cannot be copied into a txt file.
One of the great physicists said: "Nothing is more practical than a good theory". :)
This is the theory that you have to build on empirical observations. Then introduce a level of significance and test the theory with new data.
Well, I don't swing that far... I use theories that came before me.)
How about this? I think you mean something else. I mean theories of price behaviour that can be deduced and deduced.
The price behaviour is the behaviour of the totality of different groups of market participants, the ratio and value of their open positions changing dynamically and stochastically. :)
To my mind, it's interesting (for a researcher), but not the monetary one.
Earlier I formulated two possible strategies - trend-finding and pips-finding. The first strategy is reduced to detecting the beginning and the end of a trend.
The second one is catching of small (almost noisy) price movements. I think both strategies should be formulated in terms of merely statistical characteristics of price and/or price increments at corresponding timeframe.
Recently I have come across an interesting phrase - statistical arbitrage. I am studying it now. :)
Recently I came across an interesting phrase - statistical arbitrage. I am studying it ... :)
There is a much more interesting word "cointegration" - widely used in professional trading. The author, Granger, even got a noob.
The meaning is as follows.
If you take a time series and add them together in a certain way, you get a stationary result. And if so, then a deviation from the average of this stationary result will necessarily cause a return to this average. Corresponding models show that 90% of the profit has a value of 2% to 5% of pips at 4 digits.
There is a much more interesting word "cointegration" - widely used in professional trading. The author, Granger, even got a noob.
The meaning is as follows.
If you take a time series and add them together in a certain way, you get a stationary result. And if so, then a deviation from the average of this stationary result will necessarily cause a return to this average. Corresponding models show that 90% of profits have a value of 2% to 5% of pips at the 4-digit level.
... Classic use of cointegration is spread trading and arbitrage as one of its varieties....
https://forum.mql4.com/ru/46066/page9#584962
Totally confused me... :)
And also in the same thread :Forward test for a TS whose balance isnot stationary is self-defeating...
Stationarity of what do you mean ?
Thanks for the information. Do you have a link to it? :)
Found this:
... Classic use of cointegration is spread trading and arbitrage as one of its varieties....
https://forum.mql4.com/ru/46066/page9#584962
Totally confused me... :)
And also in the same thread :Forward test for a TS whose balance isnot stationary is self-defeating...
Stationarity of what do you mean ?
You found the branch correctly. It was attended by very qualified people.
Co-integration is the term behind an algorithm that has a stationary output.
Spread trading and arbitrage is trader jargon, which usually, but not necessarily, has no basis in the requirement of stationarity of the RESULT of the addition of two series.
At the time of the thread you mentioned I had access to paid Evews and examples on it, but then I switched to R and have almost valuable results on R.
Again.
Co-integration is mainstream in professional trading. For individuals it is what is called arbitrage or spread trading here, if the requirement of stationarity of the result is met. For trading teams it is portfolio trading.
However, there is an extremely important note.
Co-integration, and the trading techniques based on it, are not always possible. Approximately every 10 years the markets collapse and everything goes to hell, including cointegration. This mess lasts about a year, and then you can use cointegration again. That's why I switched to a tool called machine learning.
... Every 10 years or so, markets collapse and everything goes to hell, including cointegration. This mess lasts about a year, and then you can use cointegration again. That is why I switched to a tool called machine learning.
Stationarity of what do you mean ?
Does "machine learning" help in collapsing markets ? :)