HLHB Version 3.0 – Still No Good!

After two weeks of testing, the result of the HLHB Version 3.0 backtest is finally in. If you read the title carefully, you’d know that it’s no good!

Without further ado, here are the details:

The HLHB System Backtest Results (2009 to 2010)

• Total Gain/Loss: -3.86%
• Total Number of Trades: 90
• Number of Winning trades: 29
• Number of Losing Trades: 61
• Win %: 32.22%
• Loss %: 67.78%
• Largest winning trade (pips): 819
• Largest losing trade (pips): -300
• Longest winning streak: 3
• Longest losing streak: 8

What I like:

The 20-pip filter is FREAKING AMAZING! It helped me avoid fakeouts and kept me in A LOT of losing trades.

The system also generated a lot of signals, which allows the law of averages actually happen. That’s a good thing if the system is actually profitable.

When the system wins, it wins a lot. According to my backtesting, it can bring in 600-800 pips if you are able to catch a true trend reversal.

What I don’t like:

The RSI sucks. Not only did it prevent me from taking winning trades, it also did not help keeping me out of trend reversal fakeouts! If I didn’t use it, the system would actually have been profitable. Heck! It made me miss a 1000-pip move in 2010!

Also, the amount of losing trades is too big and the losing streaks are long. It confirms that the trade catches a lot of fakeouts.

Final Verdict:

The system is still far from being the Holy Grail of mechanical systems, but I think it has potential. I just need to tweak a few things and then backtest it again. I’ll do a bit more research on possible indicators that I can use so I can up my winning percentage and actually make the system profitable.

You guys have suggestions on what indicators I should use instead? Hit me up on Twitter and Facebook or write ‘em in the comment box below. Thank you!

  • fxsableng

    Hi Huck,
    In my newbie opinion, there’s no indicators suck!
    They suck only if you don’t understand what they are and how to use them properly, otherwise they’re all (well, maybe most of them) really good.
    The major flaw in your HLHB system is you depend on MA crossover as the basis (you’re trying to catch the trend, right?), and that’s the first fatal mistake!
    The first thing you should understand is how the MA calculated and works, and how the price moves or reacts around the MA (you can call this price action, or I personally call it rhytm), not how to catch a trend.
    And the second fatal mistake is you set the other indicator (whether it’s RSI or stoch in the previous version) in a completely wrong way, I’m sure it will invalidate the MA crossovers most of the time although sometimes it helps you avoiding fakeouts.
    Well, it’s just my newbie opinion though, just trying to help you out, hehehe. Good luck with your system :)
    By the way, if you’re making good result with your other system, why not stick to it?

  • edkhoo168

    Try add Stochastic to your system.

  • edkhoo168

    Try add Stochastic to your system.

  • sergeyrar

    Hi Huck !
    From my personal experience its better to let the machine find what works best ! its way faster and easier
    what I mean by that is
    You should optimize your system to see what really works !!
    first you need to change the MA, Period, RSI parameters to external variables to have something to “play” with.
    next run an optimization lets say on 2000-2001 history data of your pair (better be 90%+ modeling quality, otherwise its no good) take the profitable setups and keep testing them through 2001-2002 , 2002-2003 … 2010-2011
    If you remain with a consistently profitable setup you may have something here !
    secondly, you may want to find how well your system really operates:
    1- find the standard deviation of returns and the mean return of your system
    2- evaluate the risk of ruin ( the chance to lose more than x% of your account balance )
    3- If you feel you don’t have enough history data you may want to run a monte carlos simulation on the price to predict some future movements and check the performance on that data as well

    These tools are really helpful , without them its really like throwing darts at the dark !

    for more info see – link to forum.mql4.com
    u can also contact me on facebook “sergey an”

    GOOD LUCK !

  • fxsableng

    Hi Huck,
    In my newbie opinion, there’s no indicators suck!
    They suck only if you don’t understand what they are and how to use them properly, otherwise they’re all (well, maybe most of them) really good.
    The major flaw in your HLHB system is you depend on MA crossover as the basis (you’re trying to catch the trend, right?), and that’s the first fatal mistake!
    The first thing you should understand is how the MA calculated and works, and how the price moves or reacts around the MA (you can call this price action, or I personally call it rhytm), not how to catch a trend.
    And the second fatal mistake is you set the other indicator (whether it’s RSI or stoch in the previous version) in a completely wrong way, I’m sure it will invalidate the MA crossovers most of the time although sometimes it helps you avoiding fakeouts.
    Well, it’s just my newbie opinion though, just trying to help you out, hehehe. Good luck with your system :)
    By the way, if you’re making good result with your other system, why not stick to it?

  • edkhoo168

    Try add Stochastic to your system.

  • edkhoo168

    Try add Stochastic to your system.

  • sergeyrar

    Hi Huck !
    From my personal experience its better to let the machine find what works best ! its way faster and easier
    what I mean by that is
    You should optimize your system to see what really works !!
    first you need to change the MA, Period, RSI parameters to external variables to have something to “play” with.
    next run an optimization lets say on 2000-2001 history data of your pair (better be 90%+ modeling quality, otherwise its no good) take the profitable setups and keep testing them through 2001-2002 , 2002-2003 … 2010-2011
    If you remain with a consistently profitable setup you may have something here !
    secondly, you may want to find how well your system really operates:
    1- find the standard deviation of returns and the mean return of your system
    2- evaluate the risk of ruin ( the chance to lose more than x% of your account balance )
    3- If you feel you don’t have enough history data you may want to run a monte carlos simulation on the price to predict some future movements and check the performance on that data as well

    These tools are really helpful , without them its really like throwing darts at the dark !

    for more info see – link to forum.mql4.com
    u can also contact me on facebook “sergey an”

    GOOD LUCK !

  • o990l6mh

    Hey Huck!

    I often read your blog posts and in them I recognize a bit of what I’ve been up to in the past.

    I do wish I had comforting words or good ideas on indicators, but in my experience mechanical systems simply do not work very well. If you tweak them to backtest well, then chances are they will perform even worse in live trading. Optimizing and playing with indicators got me nowhere, but of course that’s just my personal experience.

    My suggestion to you is that you instead start looking really close at the supply/demand concept. That is the only thing that has worked for me.

    I’d love to see you move in that direction, but if you were actually able to figure out a mechanical system that works consistently, that would be even more impressive!

  • o990l6mh

    Hey Huck!

    I often read your blog posts and in them I recognize a bit of what I’ve been up to in the past.

    I do wish I had comforting words or good ideas on indicators, but in my experience mechanical systems simply do not work very well. If you tweak them to backtest well, then chances are they will perform even worse in live trading. Optimizing and playing with indicators got me nowhere, but of course that’s just my personal experience.

    My suggestion to you is that you instead start looking really close at the supply/demand concept. That is the only thing that has worked for me.

    I’d love to see you move in that direction, but if you were actually able to figure out a mechanical system that works consistently, that would be even more impressive!

  • sergeyrar

    well yeah… over optimization is BAD !!
    that’s why you do it on one period only, and check if it survives all other periods as well – as is
    If it survives 10 years.. chances are its a good system

  • sergeyrar

    well yeah… over optimization is BAD !!
    that’s why you do it on one period only, and check if it survives all other periods as well – as is
    If it survives 10 years.. chances are its a good system

  • antorise

    hi you really have to try to combine ADX 14 with your basic system. i think it could really help !!!

    remember ADX 14-period. and back test it. dont forget to tell me the results. thanks

  • Huck

    @fxsableng You’re right. I guess I got carried away by my emotions. I’m sorry, the RSI doesn’t suck :( I’m just exploring different kinds of systems. Don’t worry, I only use my HLHB on demo. It’s my discretionary system that I trade real money with. Thanks a lot for your insights!

    @edkhoo168 I had Stochastic on before, it wasn’t working out as well as it used to :(

    @sergeyrar Thanks! I’ll have a look at this! :)

    @o990l6mh I find enough comfort in your kind words :) I think I’ll give it more time, maybe I’ll be able to develop a profitable system. Don’t worry I pinky swear that I’ll use my mechanical system strictly on demo. I’ll be sure to keep in mind the supply and demand concept when I trade with my discretionary system. Thank you! :)

    @antorise I tried it before, but it didn’t work out well :( You can view the backtest results here: /blogs/loonie-adventures/hlhb-version-2-results.html

  • antorise

    hi you really have to try to combine ADX 14 with your basic system. i think it could really help !!!

    remember ADX 14-period. and back test it. dont forget to tell me the results. thanks

  • Huck

    @fxsableng You’re right. I guess I got carried away by my emotions. I’m sorry, the RSI doesn’t suck :( I’m just exploring different kinds of systems. Don’t worry, I only use my HLHB on demo. It’s my discretionary system that I trade real money with. Thanks a lot for your insights!

    @edkhoo168 I had Stochastic on before, it wasn’t working out as well as it used to :(

    @sergeyrar Thanks! I’ll have a look at this! :)

    @o990l6mh I find enough comfort in your kind words :) I think I’ll give it more time, maybe I’ll be able to develop a profitable system. Don’t worry I pinky swear that I’ll use my mechanical system strictly on demo. I’ll be sure to keep in mind the supply and demand concept when I trade with my discretionary system. Thank you! :)

    @antorise I tried it before, but it didn’t work out well :( You can view the backtest results here: link to babypips.com