stable fully mechanical system for 2000-06?

Discussion in 'Trading' started by shortie, Jan 3, 2007.

  1. is anybody familiar with a purely mechanical system that has been successful in the last 7 years? i would define success as ~15% annual return with a max drawdown of 25%. the system has to be consistent year-over-year (no losing years, for example). the system is for trading stocks using an account of ~100K.

    the system results should be from live trading (not from backtesting). in other words, the data from years prior to 2000 could have gone into the system development. after that the newer data could have been used to modify the system as it trades forward. but at any given moment the system should have been always trading live even as it is being modified.
     
  2. Systems dont make money, traders make money.

    However to answer your question, a simple system like a 45 day channel breakout system (long only) for the DOW 30 portfolio with a 3 times 10 day ATR trailing stop, risking 0.5% of your equity per trade migt match the kind of returns/drawdowns you are looking for.

    The win rate for such a system is around 45%.

    This should be farely easy to mechanise.

    However the system will trade about 100 times a year which means some years you might make 0% and some years 40%, but it will average out at 20% or so.

    Its unlikely to have a losing year, but it isnt out of the question.
     
  3. Vishnu

    Vishnu

    Check out the Active Trader section of stockpickr.com.

    Full disclosure: I have an ownership stake there.

    The Active Trader section has about 6 or 7 trading systems backtested since 2000. More coming in the next few weeks.
     
  4. i think i did not ask the question i wanted to ask. i wanted to know if anybody actually traded a mechanical system that worked well during the last 7 years. but i see that this is also a wrong way to ask a meaninful question. maybe i don't know what i want to know :)

    with your example, i would guess that depending on parameters one chooses the system may or may not work well during the last 7 years. so, one may have 10 breakout systems that fail in one way or another and 1 that works well. now if i ask 11 people who traded one of the variant of the system there will be one who will say that he made good $$ with his system. but this still does not tell us whether this guy was just lucky or good.
     
  5. Exactly true.

    No mechanical system produces great returns for one simple reason:

    A trading system manned by an experienced, PRO trader...
    Who uses discretion perhaps 50% of the time...
    Will CRUSH any purely mechanical system.

    So in a Zero Sum Game...
    PRO traders manning computerized systems and using discretion...
    Are consistently taking money away from people simplistically running purely mechanical systems.

    Mechanical systems are about as real as the Fountain of Youth...
    As are basically a fantasy of people unwilling to do the Hard Work to become PRO traders...
    Or a scam to sell and fleece the same lazy slugs.