Econometrics: State-space model forecasting - page 13
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
idd, ACF, bootstrapping, risampling.............
all you have to do is run it through the tester!
3. A run in the tester won't do anything. 2. The tester won't solve anything. 1. Don't look at the tester as a one-size-fits-all solution to all problems.
Don't get too rowdy. You have finally motivated the econometrician to program the strategy in MQL, and you are demotivating him/her to the utmost. Why?
You are not writing the truth.
1. No one looks at the tester as a 'one-size-fits-all problem solver'.
2. Yes, it will. It will instantly show potential profitability/loss.
Yes, he will. See point 2.
A tester run won't do anything. The tester will not solve anything. Don't look at the tester as a universal solution to all problems.
Running it in a tester won't do anything. The tester won't solve anything. You shouldn't look at the tester as a universal solution to all problems.
The word 'model' is an empty word to people. We stick it in the tester - and believe, believe, believe the result .....
Unfortunately faa's advice on how to turn a point forecast into a trend forecast has proved quite difficult for me. It's going to take some time. But things are moving....
..and the model will redraw...although I may be wrong...
Could you clarify 'redrawing'.
When each bar arrives, not only are the parameters of the model recalculated, but a specific one is selected from the 18 used. If you draw charts on history using the newly obtained model, they are very likely to be different. Is it re-drawing? The graph I posted is not redrawn. The value of t+1 stays the same, the point t is drawn, which is not re-drawn when a new bar comes, although it has a different value in the new model.
All models have four parameters that are identical in meaning. Here is a graph of the first one for a 100 bar run of 1036.
With this parameter whizzing around, what does the word "redraw" mean?
ZS. The others also change a lot.
Could you clarify 'redrawing'.
When each bar arrives, not only are the parameters of the model recalculated, but a specific one is selected from the 18 used. If you draw charts on history using the newly obtained model, they are very likely to be different. Is it re-drawing? The graph I posted is not redrawn. The t+1 value remains the old value, a t-point is drawn which is not redrawn when a new bar arrives, although it has a different value in the new model.
Don't get too rowdy. You've finally motivated the econometrician to program the strategy in MQL, and you are demotivating him/her to the utmost. Why?
You are not writing the truth.
1. No one looks at the tester as a 'one-size-fits-all problem solver'.
2. Yes, it will. It will instantly show potential profitability/loss.
Yes, he will. See point 2.
;)) OK. Instead of saying "A tester run won't do anything", I'll put it this way: "A tester run will do something."
Although this "something" will not be the solution of the task. But it may show the way forward.
Could you clarify 'redrawing'.
When each bar arrives, not only are the parameters of the model recalculated, but a specific one is selected from the 18 used. If you draw charts on history using the newly obtained model, they are very likely to be different. Is it re-drawing? The graph I posted is not redrawn. The value of t+1 remains the same as it was before and the point t is drawn; it is not redrawn when a new bar arrives although it has a different value in the new model.
as we are interested only in 1 point (prediction for 1 step), you should watch its change... i.e. you ran the model - wrote the forecast point... gave a new data line... ran it again - wrote it and so on... you should use a simple first model, not using Anasable18... you can use your hands you can use scripts....in this way you made 20-30 steps for example and with the same model you made as on 1 screen....then compare both results and the actual value....I think it will draw black....and is not suitable for trading...