Econometrics: one step ahead forecast - page 33

New comment
 
faa1947:

New forecast

Forecasting long. Very large difference in the value of the forecast using EURUSD and DX lags.

Can I make it clearer? Are you making forecasts on the basis of XP? The trend graph does seem ))) to be true after all. Yesterday you predicted a slight dip in the rate at midnight. Yes - it was.
 
new-rena:
Can I make it clearer? Are you making predictions based on KP?

A regular component is extracted from the quotient using HP and the oxide (noise) is added to it. Here is the formula to do this:

KOTIR = -11.2430621824*HP1(-1) + 35.6424734363*HP1(-2) - 34.0025338417*HP1(-3) + 10.602618658*HP1(-4) - 0.63687458837*HP1_D(-1) - 0.971411444769*HP1_D(-2)

You take the last 4 bars of HP and two bars of noise.

 
faa1947:

A regular component is extracted from the quotient using HP, and an oxide (noise) is added to it. Here is the formula to do this:

KOTIR = -11.2430621824*HP1(-1) + 35.6424734363*HP1(-2) - 34.0025338417*HP1(-3) + 10.602618658*HP1(-4) - 0.63687458837*HP1_D(-1) - 0.971411444769*HP1_D(-2)

Takes the last 4 bars of HP and two bars of noise.


I had that thought once, but haven't tried it myself. And also - what is NR??? Is it a normal probability distribution?
 
new-rena:
I had that thought once, but haven't tried it myself. And also - what is NR??? Is it a normal probability distribution?
NR is a Hodrick-Prescott filter. I've explained this many times in this thread, as it's not about the filter at all.
 
faa1947:
HP is a Hodrick-Prescott filter. I have explained this many times in this thread, as it's not about the filter at all.
That's what I'm saying about what the hodrick_prescott_indicator showed yesterday. Could you add a little more to it?
 
new-rena:
That's what I'm saying, that's what the hodrick_prescott_indicator showed yesterday. Could you add a little more to it?
You have to write the formula given above. HP is a part of this formula. For it you know quality of reflection of source cotier = 97%. You will have an indicator for which the quality is known.
 

New forecast

Date Value Change Forecast Forecast Error Forecast Error Change Change Forecast Forecast
of fact Open prices to based on in pips based on in pips forecast forecast on EURUSD on DX
eurusd DX on eurusd on DX matched? matched?
2011.11.08 23:59 1,383
2011.11.09 23:59 1,3524 -0,0306 2011.11.09 23:59 1,3798 56 1,3663 67 -0,0032 -0,0167 Yes Yes
2011.11.10 23:59 1,361 0,0086 2011.11.10 23:59 1,3613 60 1,3742 70 0,0089 0,0218 Yes Yes
2011.11.11 23:59 1,3778 0,0168 2011.11.11 23:59 1,3541 59 1,3766 71 -0,0069 0,0156 No Yes
2011.11.14 23:59 1,3624 -0,0154 2011.11.14 23:59 1,3676 59 1,3673 69 -0,0102 -0,0105 Yes Yes
2011.11.15 23:59 1,3525 -0,0099 2011.11.15 23:59 1,3650 59 1,3634 69 0,0026 0,0010 No No
2011.11.16 23:59 1,3455 -0,0070 2011.11.16 23:59 1,3529 57 1,3627 69 0,0004 0,0102 No No
2011.11.17 23:59 1,3468 0,0013 2011.11.17 23:59 1,3446 57 1,3521 70 -0,0009 0,0066 No Yes
2011.11.18 23:59 N/A N/A 2011.11.18 23:59 1,3422 55 1,3479 70 -0,0046 0,0011 N/A N/A



The outlook for the dollar index is clearly better.

 
faa1947:

The forecast for the dollar index is clearly better.

Are you judging this by five events? You seem to have a good psychic ability, which is not surprising, since you are an econometrician and they are all like that. Well, it's pretty cool, we just applied a filter (which, by the way, doesn't fit to the f*cking crap) to the quotes and now we are econometricians. "Oh, give me two" (C).

PS: You'll see, it will end up 50/50 for both indices. I won't distract you - have fun with it.

 
Farnsworth:

Are you judging this by five events? You seem to have well-developed psychic abilities, which is not surprising, since you are an econometrician, and they are all like that. Well, it's cool, you apply a filter (which by the way doesn't fucking fit) to the quotes and now we are econometricians. "Oh, give me two" (C).

PS: You'll see, it will end up 50/50 for both indices. I won't distract you - have fun with it.

I'm not judging by 8 (not 5) trades, but enticing people who understand at least something to discuss the model itself, which is not a filter, or suggestions for improving the model's performance, test results, etc. Only a few worthwhile thoughts on 33 pages so far.
 
Farnsworth:

Are you judging this on the basis of five events?

Show the class in your branch.
1...262728293031323334353637383940...139
New comment
Reason: