reverse engineering MACD-R2

Discussion in 'Strategy Building' started by nukethewhales31, Nov 1, 2008.

  1. needed a place to store my data/testing/ideas figured why not on here.
     
  2. tested original MACD indicator what was referred to as MACD1

    parameters

    -buy signal
    -enter long this bar @ close
    -when MACD crosses above signal line
    -sell signal
    -exit long this bar @ close
    - when MACD crosses below the signal line

    results tested DOW 100

    3470 trades

    # of profitable trades - 1268
    # of unprofitable trades - 2202

    36.541% profit rate
    63.459% loss rate
     
  3. Is there an overall profit?

    Can you actually trade "this bar" at the closing price?

    What time interval are you using? Is this daily index value data?
     
  4. wow i think im still drunk from last night dunno how much work im gonna get done today cause im sure a throbbing hang over is right around the corner...
     
  5. overall with the low success rate only long trades entered further testing with short included will be coming. but I did see a modest profit just from using this entry exit system. I contribute that to the fact i was testing long exit/entries in a bull market more info will come

    also looking down to individual plays some securities that where -30% with this system showed modest profits with only 20% of trades being winners..


    At closing price of this bar is also at opening of next bar as well. Yes you can trade close of bar and next bar open as if you trade the bar before it closes you receive more false signals.

    yes this seems to matter as well there are i find different statistics for different time periods I tested multiple time periods. why? i see different time periods as layers im sure there are others that think that larger layers of the market high time intervals = fundamental layers as you decrease in layers or time periods the more distortion comes into play..

    I posted the time with the highest win prob rate which was daily / hourly by 1.5%
     
  6. ok finished testing refined MACD what is referred to as MACD-R1

    Parameters

    BUY Signal
    - MACD-R1 crosses signal line upward
    - wait 3
    - If no cross has occured enter long

    Sell Signal
    - MACD-R1 Crosses signal line downward
    - -OR-
    - Specific Target Price Reached

    Gauge OPCMWR%%P RATE at Intersection FOR largest overall gain.

    Results

    AVG SUCCESS RATE FOR:

    3% - 78.47% profitable trade rate
    - AVG profit per trade about=5.5%
    5% -73.78% profitable trade rate
    - AVG Profit per trade about = 7.2%

    AVG HOLD TIME
    3.3 bars
     
  7. 4982 trades made - forgot to put that in there
     
  8. Started work on MACD-R2
     
  9. tested what is referred to as MACD-R2

    parameters

    CLEVEL-F-.5%-3.5%
    MACD CROSS
    CLEVEL-T-3.5%-.5%
    EMA ADJUSTMENT

    ON MACD CROSS
    ENTER CLEVEL-F-.5%-3.5%
    ADJUSTMENT LEVEL HOLD
    EXIT CLEVEL-T-3.5% - .5%

    results

    better then mentioned out of 3,437 trades

    Profit Rate 96.3772%
    avg profit per trade 6.27%
    @ 2-2.5-34,74,26
     
  10. 96% rate is better then expected because as in most system profit rate doesnt directly mean more profits if there is no management... this system has management almost built into it.... weaknesses tho are loss off profits on fakeouts caused by pullbacks.

    its mentioned that profits can be maximized by use of leveraging in options... I modified for FX...

    EMA adjustment found that 34,77,25 wasnt optimal for setup

    Found 34,74,26 was

    accuracy increased at cost of signals....
     
    #10     Nov 3, 2008