Econometrics: one step ahead forecast - page 18
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
I'll make another prediction.
The model is unchanged. Opening price at 00:00 on Monday.
Forecast Monday 14 November
The only way to test the model for adequacy is to run it in a tester or on a demo. Anything else is a waste of time.
Especially in this table it is not clear where the "right/wrong" comes from.
The only way to test the model for adequacy is to run it in a tester or on a demo. Everything else is bogus.
Predictions are available, anyone can make them on the demo.
But both are not that easy. The article contains an Expert Advisor without MM. If we use it, we can trust it.
Moreover, we can't see where the "right/wrong" comes from.
The two columns to the left of the output. If the actual price increment coincides with the predicted increment (the sign of increment coincides), it is correct.
Attached to the article is an EA with no MM. If you run it through, are we to believe?
Yes, with an eye. You gave a test of the Expert Advisor. Spread is down. Why the forecasts?
Two columns to the left of the output. If the actual price increment coincides with the forecasted increment (the sign of increment coincides), then it is correct.
Add a column with start value, very inconvenient.
Better yet, look for another model, instead of pissing this one off.
Yes, with an eye. You did give a test of the EA. Spread was lost. Why the forecasts?
Add a starting value column, very inconvenient.
Better yet, look for a different model instead of pissing this one off.
Why do you need forecasts?
The Expert Advisor is built on forecasts. There are plenty of forecasts on the net. I'm trying to get a forecast that can only be trusted based on its internal structure. And then run it through the EA.
Add a column with a starting value, very inconvenient.
This is the second column.
Better yet, look for another model instead of trying to make this one work.
Suggest it. When I opened the thread, I hoped that the team would take part. After all, I need a model at the level of a formula or at least a wish of WHAT to model. So far, the plan is as follows: we take the model, if the forecast does not suit us, we look for problems in the model and try to eliminate these problems. At the moment we have model no 2. We got a very good quality SS, corresponding to 97% of the quote
This is the second column.
How is the state space? I'm ready to support. I should warn you straight away that I haven't managed to figure it out, so only support.
This is a very productive method of research, which requires some effort to understand.
And it seems to me that the state-space method you're using has only borrowed its name (figuratively speaking) from the miracle program you're using.
avtomat:
And it seems to me that the state-space method is just the name (figuratively speaking) borrowed from the miracle program you're using.
Actually, EViews is one of the most advanced packages in econometrics. Reading additional literature on state space has not revealed anything that is not in the package.
My offer to collaborate on the state space, still stands.
An example application is in the attachment. But this is not a guide to the program. Find the package, it's with the documentation. Each chapter of the documentation has links to books that outline the relevant algorithms.
And start your own branch. I think there will be at least as many people interested.
He has a branch. Very close models. I am not just referring to him for a reason. Of all the forum members I know, avtomat is the closest to state-space models.