Econometrics: one step ahead forecast - page 31
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
Finally. There is a forecast, but there is no forecast estimate. This begs the question: Should I take only those that are bigger than spread, stop levelling? Or bigger than forecast error? What to do?
I do not know what this error is.
If it's standard deviation (s.c.o.), and the prediction value itself is a normally distributed value with exactly that s.c.o., then it would be a good idea to ignore all predictions that are less modulo than at least two s.c.o. Then if the modulus of the forecast is less than two s.c.o. (somewhere around 118 points) there's about 95% probability we won't make a mistake attributing the forecast value to zero.
It turns out that a forecast whose modulo value is less than 2 s.c.o. should be considered uninteresting (it is a zero movement forecast).
There is an indicator for forecasting: hodrick_prescott_indicator. It makes it so (I deliberately put arrows on it to compare EURUSD, 4H):
I'm already answering myself (it hasn't been an hour, but only 15 minutes have passed), although I had BUY)):
And you for tomorrow ))))
I don't know what this error is.
If it's standard deviation (s.c.o.) and the forecast value itself is a normally distributed value with exactly such s.c.o., then it would be nice to ignore all forecasts that are less modulo than at least two s.c.o. Then if the modulus of the forecast is less than two s.c.o. (somewhere around 118 points) there's about 95% probability we won't make a mistake attributing the forecast value to zero.
It turns out that a forecast whose modulo value is less than 2 s.c.o. should be considered uninteresting (it is a zero movement forecast).
Error = s.c.o. + systematic error.
Here is a graph of that error:
And here are the statistics:
Range = 5 pips, s.c.o. = 1 pip! It can be considered a constant over the whole interval. The forecast itself can hardly be called normal
But here's his error = constant, and that's something. I think it's possible to accept a channel of two errors. We'll have to think about it, though.
I'm already answering myself (it hasn't been an hour, it's only been 15 minutes), even though I had BUY)):
And you for tomorrow ))))
I have shown the price of trend lines in my article "Indicator Analysis". All this art is close to Repin or Aivazovsky
No of course, probably some intricate patterns. A trend line is just a trend line. On M1, when price follows the trend line, it bounces off the line with an accuracy of one pip. What is it? - Just buy from this trendline - it could be an hour, it could be 5 minutes.
No of course, probably some intricate patterns. A trend line is just a trend line, on M1, when price follows the trend line, it bounces off the line to within a point. What is it? - Just buy from this trend line - it may be one hour, it may be 5 minutes.
Right now I have a forecast until 00:00, and in the morning I will make a forecast from 00:00, i.e. with a lag.
According to my indices, the trend change in EURUSD from rising to falling or flat is starting right now.
What about you?
According to my indices, the trend change on EURUSD from rising to falling or flat is about to start.
What about you?
(If I have 80% profitable trades - they are everywhere and always - that's statistics!) Good luck with your predictions!
Out of 10 forecasts 7 are correct - that's the kind of statistics you need. What is a justification? - Formulas, rhombuses, dice.
Not enough coffee grounds, a very effective forecasting tool. Lots of people are not doing badly.
(If I have 80% of profitable trades - they are everywhere and always - that's statistics!)
Are you a millionaire? You have a profit factor = 4! What's the problem? What are you doing on the forum?
к 00:00 1.3529. I have a prediction for the day with a fixed prediction time, after all. Really question: I don't know, because I can start forecasting when I get Open at 00:00, but what to do 2 hours before that?
Tantrik:
Wait, you are messing with my head. What timeframe do you work with? How can something happen right now?