I question the well known alpha=2/(n+1) for a one pole EMA, as the assumption is that prices rise at one unit per bar.
Since when do prices do this?
I defer to the formula on page 161 of Rocket for determining cutoff, which is all that matters - not trying to compare with SMA, however noble that may appear to someone maybe not as pedantic as myself. As it turns out roughly OK within loose bounds - if near enough is good enough.
I have seen slightly different formulations of alpha somewhere which may go someway to correcting this, but I have forgotten.
An example.
One pole EMA cutoff = 12, say. Alpha is about 0.4.
Now, 2/(4+1) = 0.4. Cutoff of 4 bar SMA is about 8. Is 8 close to 12? Yes and no. But at these high "frequencies" I think it does matter a lot.
Agat: Prompt me please someone have a color version of this indicator?
Agat, sorry for the late answer, do you mean a color version of the fisher transform line(sorry if it seems like a dumb question, just want to be sure)?
mrtools: Agat, sorry for the late answer, do you mean a color version of the fisher transform line(sorry if it seems like a dumb question, just want to be sure)?
Agat: Yes, exactly.The signal line is not needed. Also, I noticed that this indicator gives an error "Incorreсt start position for ArrayMaximum(Minimum) function".
That is a benign error (happens only on a first run or when time frame or symbol is changed, and does not affect the result. Anyway, here is a version without that message : ehlers_fisher_transform.mq4
Allow the use of cookies to log in to the MQL5.com website.
Please enable the necessary setting in your browser, otherwise you will not be able to log in.
You are signed in. Two-factor authentication has been disabled. Set it up again and scan a new QR code, or continue with two-factor authentication disabled.
I question the well known alpha=2/(n+1) for a one pole EMA, as the assumption is that prices rise at one unit per bar.
Since when do prices do this?
I defer to the formula on page 161 of Rocket for determining cutoff, which is all that matters - not trying to compare with SMA, however noble that may appear to someone maybe not as pedantic as myself. As it turns out roughly OK within loose bounds - if near enough is good enough.
I have seen slightly different formulations of alpha somewhere which may go someway to correcting this, but I have forgotten.
An example.
One pole EMA cutoff = 12, say. Alpha is about 0.4.
Now, 2/(4+1) = 0.4. Cutoff of 4 bar SMA is about 8. Is 8 close to 12? Yes and no. But at these high "frequencies" I think it does matter a lot.
End rant!
I have come a cross the term called '' Portmanteau Test(autocorrelation and variance ratio tests)'' which looks interesting: Portmanteau Test | the Statistics Glossary .Some interersting articles about the portmanteau test Using R in Algorithmic Trading: Testing whether an instrument follows a random walk | Mechanical Forex
Indian Stock Market: An Empirical Analysis of Informational Efficiency - Gourishankar S. Hiremath
Ehlers filter by Alexander Gettinger
ef.mq4
nef.mq4
Can someone convert this : https://www.mql5.com/en/forum to metatrader, please?
Prompt me please someone have a color version of this indicator?
Prompt me please someone have a color version of this indicator?
Agat, sorry for the late answer, do you mean a color version of the fisher transform line(sorry if it seems like a dumb question, just want to be sure)?
Guess this would be a combination of Ehlers and Williams, this is an ift of adaptive super smoother wpr with mtf.
Guess this would be a combination of Ehlers and Williams, this is an ift of adaptive super smoother wpr with mtf.
Nice. Thanks
Agat, sorry for the late answer, do you mean a color version of the fisher transform line(sorry if it seems like a dumb question, just want to be sure)?
Yes, exactly.The signal line is not needed.
Also, I noticed that this indicator gives an error "Incorreсt start position for ArrayMaximum(Minimum) function".
Yes, exactly.The signal line is not needed. Also, I noticed that this indicator gives an error "Incorreсt start position for ArrayMaximum(Minimum) function".
That is a benign error (happens only on a first run or when time frame or symbol is changed, and does not affect the result. Anyway, here is a version without that message : ehlers_fisher_transform.mq4