The perfect moving average

Discussion in 'Technical Analysis' started by aphexcoil, Sep 6, 2002.

  1. Gordon,

    Yes, I would have already been short, and that spike may have come close to taking my stop out -- I cannot average up or down when I begin trading one contract, but if I make it to trading two contracts, I will average up or down BUT STILL KEEP A STOP.

    I basically just trade off the 4MA and wait for it to change direction and immediately after a one bar confirmation, make a trade. Here are the results for a given day with that strategy (much faster than a crossover strategy).

    I also only trade with the major moving trends.

    Here's an example:

    09/06/2002 5 Minute chart
    #31     Sep 8, 2002
  2. this was posted by super_ego:

    "One more time.....Stocks/futures are, on a daily basis, trying to MOVE INTO and MOVE AWAY FROM the 5 period moving average ON A DAILY CHART.

    It is massive movements away from the 5 period moving average that are going to make you lots of money. Because when your stock or futures contract diverges away from the 5 period moving average (5MA) it will have a tendency to "snap" back to the 5MA."

    i totally agree with this concept. in short, prices move in a direction, but extreme moves will revert to the mean.
    #32     Sep 8, 2002
  3. wdbaker


    I stand corrected, I see what you mean, putting down the pipe now :D

    Maybe for you its not an issue but for me at this point cost is an issue so this ma is about as close as I am going to get to perfection for now, how much does it cost to use the JMA???

    well I was close enough for hand grenades:)

    #33     Sep 8, 2002
  4. yes, for the most part, crossover systems suck.
    #34     Sep 8, 2002
  5. aphie,

    i'm gonna take your chart and show you how i want to play it. but i don't know how to REALLY do it yet because i can't figure out how to calculate what i want to know. hehe...

    anyway, it's basically what i just talked about. you find the extremes for the current volatility and fade them. i will reupload your chart in a few min.
    #35     Sep 8, 2002
  6. ok here's aphie's chart that i added to.
    #36     Sep 8, 2002
  7. echo


    It sounds to me like you are talking about Bollinger Bands. No? They are a moving average with bands that expand during increased volatility and collapse as volatility diminishes. Or are you thinking of something else?
    #37     Sep 8, 2002
  8. wdbaker


    Here is the picture, I still say it is not bad for a free MA

    #38     Sep 8, 2002
  9. Gordon,

    I don't think you'll ever find a way to capture extremes. You may come up with a probability distribution system that only captures trades that are a high divergence from the mean, but capturing the very extreme would be tough.

    However, if you took my system and averaged down while still keeping your stop, you could bring your total cost a little closer to the extremes.

    Superego's assessment of the 5MA, in my opinion, is not totally correct. The 5MA is simply telling you where prices are currently trending on a very small level. Of course prices will snap around it -- because it is such a small MA. Either the prices will snap back to it, or the 5MA will meet up with the prices. However, imagine a very strong downward or upward trend where the trader was expecting snapbacks but never got one -- that would hurt.

    However, if you traded with the 5 MA as a guide to where prices are likely to trend on the next bar or two AND found a way to adjust an MA to not give as many false signals (so you could stay with a trade longer), you would do super.

    My system will eventually use MA's, scaling into and out of positions and averaging up and down. I'm breaking a trading rule on purpose because that is exactly what helps my system in the long run in terms of probabilities.

    #39     Sep 8, 2002
  10. wdbaker


    Could something like this work???

    Check the spread sheet, it contains sample data and chart.
    Not saying this is the answer but hopefully moving in the right direction, if I am totally wrong then you can point that out for me so that I don't continue to delude myself with fantasies of grandeur and the HG :D :D

    Would probably need to be modified for real data, just trying to use the same sample as previous which showed the JMA, DEMA etc...

    #40     Sep 8, 2002