Trend following on the short side

Discussion in 'Automated Trading' started by HuggieBear, Feb 13, 2011.

  1. fjpenney

    fjpenney

    #11     Feb 13, 2011
  2. good results...you will soon be a billionaire if you can annualize that :)

    what are your predicted drawdowns?
     
    #12     Feb 13, 2011
  3. fjpenney

    fjpenney

    I would be a fool to expect to have that return annualized and have it every year. I designed the system to have a high MAR (Calmar ratio) and therefore included ETF's covering a wide range of asset classes. Also, I don't trade for a living and therefore had to design a system that has holding periods of weeks rather than minutes.

    The system as traded on C2 uses 1:1 leverage (i.e. 100%). My expectation for Max DD is 15% to 20%. Given that some of the ETF's I use are relatively new, it is impossible to perform a 15 year backtest, for example, to get a very good feel for the Max DD to be expected.
     
    #13     Feb 13, 2011
  4. curious, what is your expectation for cagr?
     
    #14     Feb 13, 2011
  5. fjpenney

    fjpenney

    The return generated by a trend trading system depends on the strength and duration of the trends of the underlying securities plus the amount of leverage that is used. C2 limits the amount of leverage at 1:1 which I agree with for the site but I use a slightly higher leverage personally.

    The backtested CAGR is north of 50% but that is based on a somewhat optimized system. In this case, the results are somewhat optimized because of the ETF's selected and the parameter settings in my algorithms. However, all ETF's used are equally weighted so I am not using the benefit of hindsight to optimally weight my holdings.

    The best hedge funds average 30%. If you check the long term results as per the Hennessee Group, you will find that the average hedge fund return is just over 12%/year.

    Based on the above, I expect an average return between 25% and 35%. If all the asset classes trend sideways it will be impossible to generate high returns. The system as presented on C2 holds a maximum of 25 ETF's but I monitor more than that. This way, I have a higher likelihood than I would otherwise have of being fully invested at any time.
     
    #15     Feb 13, 2011
  6. You right...I just wanted to demonstrate short trend performance. This is how the system has performed 2002 - 2010:

    [​IMG]
     
    #16     Feb 13, 2011
  7. What software are you all using to backtest?
     
    #17     Feb 13, 2011
  8. NoDoji

    NoDoji

    The markets have been in an uptrend for nearly 2 years. Did you test your intermediate/long term mechanical short trend following systems during the decline from October 2007 through March 2009? That would be where an intermediate/longer term short trading system would make sense.

    I'm not sure what your definition of intermediate/long term is, though. If it's trading weekly/monthly swings off trend/channels, then short and long strategies should work similarly, with the longs running longer and the shorts (pullbacks in the long term uptrend) running shorter, time-wise.
     
    #18     Feb 13, 2011
  9. ehsmama

    ehsmama

    I think you can try to use 2 things
    1. Use a Filter to enter Shorts..so Bull Markets keep your short system mostly on sidelines fo
    2. Exit Short when Profit Target hits...Not on pullback
    I think short + Long system will always perform better than a Long only system..But combination of 2 is important..Not Just long only or Short only
    I think using Tradingblox as a development platform will help you see things in new light
    RM
     
    #19     Feb 13, 2011
  10. I use Metastock and Amibroker. The reports I posted came fram an older version of Metastock. I have a lot of systems there, around 150 or so, and I have no time to move them.
     
    #20     Feb 14, 2011