looking for a smooth macd

Discussion in 'Trading' started by divergence, Dec 29, 2003.

  1. Does anyone know of a cheap or free indicator or code for a smooth macd.i like the jurik one but i dont want to spend $200 dollars to test it for my system.does anyone know of a comparable indicator.thanks:)
     
  2. tiocsti

    tiocsti

    Smooth is pretty easy to get, smooth with no delay cost -- much harder. I don't have any suggestions on something better, or even as good, as Jurrik's stuff (or atleast, my impression of it based on his web page) but there are a few possibilities for smoothness:

    Patrick Mulloy developed the TEMA and DEMA moving averages, which uses a modified form of exponential smoothing. You can get some basic information at: http://www.paritech.com/education/technical/indicators/trend/tema.asp

    You could use linear regression which has some good smoothness characteristics as well as tracking the price very closely. This is not, however, a moving average.

    Kaufman Adaptive Moving Average is annother possibility -- it changes its speed based on the volatility in the price series.
     
  3. Thanks Tiocsti for the help:)
     

  4. There is a "zero lag MACD" available for Amibroker which was based on ZeroLag EMA - see Technical Analysis of Stocks and Commodities April 2000
     
  5. abogdan

    abogdan

    Try FIR filter for TS (you can get the code for free at their forum). Its as good as it gets. And, by-the-way, there is no "0" delay smooth interpolators by definition.
     
  6. Well yes...Could he use a weighted macd to get his desired results?

    If so.....Investor R/T has this in their expensive charting program...

    www.linnsoft.com

    Michael B.



     
  7. abogdan

    abogdan

    Well, it depends on acceptable lag. Search the NET there is tone of free staff you can get. Remember, the best things in life are free.

    P.S. I have tested them all, the best but very complicated is IIR filter. FIR filter is easier but "harrier"
     
  8. Thanks everybody for your help.:)