|
PureBytes Links
Trading Reference Links
|
Trey Johnson wrote:
>Hello Alex,
> Could you give another example, one that's open source possibly?
I don't know of any, other than crude ones like Kaufman's Adaptive
Moving Average.
I've written some myself, and they are a real chore to tune, to get
the right balance between overshoot, lag, and smoothness. After
trying it myself, I can understand how Mark Jurik spent so long
developing his JMA.
John Ehlers might have something in one of his books.
The basic algorithm involves some measurement of noise, comparing
each price movement to that noise, and adjusting parameters
controlling velocity and position of the underlying smoothing
function (such as an infinite-impulse-response Kalman filter) based
on the price movement compared to the noise.
-Alex
>-----Original Message-----
>From: unicorn@xxxxxxxxxxxxxxx [mailto:unicorn@xxxxxxxxxxxxxxx] On Behalf Of
>Alex Matulich
>Sent: Monday, November 13, 2006 7:23 PM
>To: omega-list@xxxxxxxxxx
>Subject: Re: Hull Moving average
>
>>The Hull Moving Average solves the age old dilemma of making a
>>moving average more responsive to current price activity whilst
>>maintaining curve smoothness.
>
>The "problem" as stated is a straw man. The problem isn't a
>dilemma. It's more of a trilemma, involving THREE things, not two:
>
>1. maintain smoothness.
>2. minimize lag.
>3. minimize overshoot.
>
>Unfortunately, the Hull Moving Average suffers from overshoot, which
>is typical of low-lag smoothers based purely on impulse response
>functions.
>
>Statistical smoothers, on the other hand, DO balance these three
>features. Jurik's moving average (JMA) is one example.
>
>-Alex
|