Some way to find trend

Discussion in 'Technical Analysis' started by SEALAND, Jan 5, 2015.

  1. expiated

    expiated

    I made two modifications to the NPP Trading System this morning.
    ScreenHunter_7606 Apr. 24 16.00.jpg
    I believe that once I have my forecast model down pat, my success rate should increase to 100% on most days.
    ScreenHunter_7607 Apr. 24 16.02.jpg
     
    #31     Apr 24, 2018
  2. expiated

    expiated

    I’m continuing to refine my forecast model in an effort to maximize consistency in terms of executing successful trades…

    ScreenHunter_7612 Apr. 25 08.52.jpg

    I also returned to a style of entering positions primarily following short-term pullbacks as opposed to intermediate-term reversals...

    ScreenHunter_7613 Apr. 25 08.59.jpg

    For example, with respect to the above image, I would attempt to enter short positions as “price” bounces off the upper band of the tan-colored envelope and lock in profits as the candlesticks make contact with the lower band of the tan-colored envelope (which replaced my dynamic support/resistance Darvas box-looking indicator, thereby conserving precious memory).

    The two intermediate trend lines were essentially replaced by the “faster” orange- and green-banded simple moving average envelope, and the red simple moving average and brown proprietary moving average have taken over the role of the former key preeminent shorter-term intraday moving average. (The colors are “off” in the above image.)

    In terms of finding the trend: Since generating a line chart which plots every single data point results in a jagged graph that is virtually useless in terms of figuring out if price is ultimately headed up are down, I regard simple moving averages as the most ideal solution in that they can reduce the number of haphazard saw-shaped “invalid signals.”

    From my perspective, logic would seem to suggest the possibility of arriving at a single line (moving average) within a given time frame that removes the maximum number of false positives/head fakes while tracking the overall direction of price closely before it begins to lag so far behind price’s actual trajectory as to render it statistically inaccurate.

    These are the lines I have sought to discover through systematic analysis and I have based my Numerical Price Prediction strategy on this premise, having identified such moving averages in the longer-term (see the black-dashed moving average in the above image), intermediate (see the orange- and green-banded envelope), and shorter-term (see the brown proprietary moving average) time frames.

    Of course, the lines must be converted accordingly when moving from a chart of one time period to a chart of another, and this strategy is useful ONLY for carrying out a guerrilla-style approach to intraday DAY trading, since there are intrinsic predictability limitations that lead to error growth with time, making the application of this methodology to swing or position trading totally out of the question.
     
    #32     Apr 25, 2018
    birdman and Turveyd like this.
  3. birdman

    birdman

    @expiated ... how many MA's do you use? Are they SMA or EMA? What are the numerical values for each? Thanks in advance!
     
    #33     Apr 25, 2018
  4. birdman

    birdman

    @expiated - I worked up a crude version of your plan for stocks and ETFs using 5SMA, 10SMA, 15EMA, 20SMA and it looks much better than what i was using. I will test it out tomorrow. Thanks!
     
    #34     Apr 25, 2018
  5. Let's assume market has gone down for the day and made a LL. After this occurs, market makes a HH. Until we have a HL, we are still in a downtrend. Once we have a HL we are now in an uptrend for the time frame you are watching to place trades in. A 1 min TF is noise and should not be used.
     
    #35     Apr 26, 2018
    birdman likes this.