Example Trading System

Discussion in 'Strategy Building' started by sidinuk, Jan 17, 2005.

  1. Who needs charts? You are really devoted to providing a service. Thanks.


     
    #141     Feb 24, 2005
  2. Well, some of us still need them for trading, but obviously I cannot speak for all ET members. :D :D

    I agree, good job sidinuk! Now, if we could also get the control sheets for other markets similar to the control sheet for YM, that would really be great. Don't forget that this is the lazy trader board, so my request is fully justified and you may even be stripped of your membership if you do not comply with it... :D :D :D

    Considering that this system was pronounced dead on arrival (or 'going forward') by some theoreticians here, it's doing remarkably well this month...

    I guess the best approach to the markets is 'shut up and trade' or at least a similar approach ('shut up and calculate') works really well in quantum mechanics according to a friend of mine, a theoretical physicist. But what do I know...
     
    #142     Feb 25, 2005
  3. No one said it would never make money. Just that it will likely lose over long periods, unless another stock market bubble improbably arises.
     
    #143     Feb 25, 2005
  4. Actually, what does it mean a 'long period of time'? How long? Conveniently long that you can at some point declare that you are right? Right now I can say that you are wrong (the system is making money), but when the system is in a drawdown you might be saying that the system is losing, so we both can be right, but at different times. If you really want to make a prediction that makes sense in this business, state how long you believe it should take for the system to be declared a loser. I believe that one should give it at least a year and keep checking its performance every year to see if it keeps making money. And it will as long as volatility is there. Volatility is cyclical so in some months the system will do worse, in others better, but on the whole its should do well.

    Another stock bubble? Who needs it? All we need for this system to work is volatility and this can be high during a deep bear market as well as was the case in 2002.

    If you check the performance of this system for ER2 and ES, you will see that the former market did much better than the latter over the same period of time. Why? Because of the higher volatility of ER2. So why would we need a stock bubble for this system to work? Beats me, but I am a simpleton and it works for me... no matter whether it's bull or bear. All I need is volatility...

    So let's see a year from now how the system will have performed in YM and ER2 as these are the markets that are worth following. And keep checking it on an annual basis. That's my idea of 'a long period of time', precise enough to be subjected to some objective test. My prediction is that as long as we do not experience a further deterioration of volatility we should do well. Obviously, it's impossible to predict volatility, so we may do worse than in the past too, but that's the risk everyone who trades this kind of systems has to take. Or anyone who trades anything, for that matter. Simply saying that the system will do worse unless we have a bubble is not justified on any grounds in my opinion.
     
    #144     Feb 26, 2005
  5. Let the predictions to astrologers....

    From a probability point of view, you need an infinite time series to determine the average of a brownian motion with bias when you only have one realization of the process. That's why a lot of people in trading can make money with a system...which actually doesn't work...

    What you can do is determine the average of the distribution with a confidence interval. That's very easy :

    1) Compute the average of returns : r
    2) Compute the volatility of returns : sigma
    3) Compute r +/- 1.96*sigma/sqrt(N) where N is the number of observations. With this you get a 95% confidence interval....

    If you get a 5% +/- 35%, it's quite difficult to say something. If you get 15% +/- 2%, then yes, your system is very good.

    Of course all of this is based on the traditional assumption of stationnarity, which in trading is well known to be a difficult quality to find...
     
    #145     Feb 26, 2005
  6. Time only matters in figuring the yield. Drawdown matters only to the trader and his/her personality.

    This system may not be for everybody....


    But....

    We don't need no fricken Charts!

    Thanks Sidinuk
    Michael B.
     
    #146     Feb 26, 2005
  7. I don't want to get into a pissing match over this system. I haven't backtestested it. and really, the guy who started this thread hasn't backtested it either. As I recall his results were based on a month or two which is certainly not adequate. Of course, he didn't claim they were adequate, so I'm not criticizing him.

    I have backtested similar systems years ago and they were losers in the stock market, particularly if you closed them EOD. If you look at the systems evaluated at Futures Truth you will find very few S&P daytrading systems that ever make money, and I am pretty sure that most of the ones that have survived, like the R-Breaker series of systems, are pullback systems or faders, not vol breakouts.

    I am a big fan of Gary Smith of Trading for a Living fame (not the RM guy) and I know he traded a 60 minute B/o system, but he used a lot of discretion and also had the biggest bull market in history at his back. He also traded long only.

    I really wish someone could prove me wrong with hard data, so I could kick back and start trading this system, but I don't see that happening.
     
    #147     Feb 27, 2005
  8. sidinuk

    sidinuk

    Despite missing out on Thursday the system still produced 2 winners out of 2 for the Dow last week providing 152 points profit. In fact across the 5 tested contracts the system produced 10 winners out of 10 trades for a total profit of $3,452.50 before commission.
     
    #148     Feb 28, 2005
  9. Correction. I think Gary smith's book was called Living the Dream or something similar.
     
    #149     Feb 28, 2005
  10. has anyone back tested this system on the ESTX50/DAX, N225/HSI, ive just finished eyeballing about 4-5 months worth of charts and it appears to lose money??????
     
    #150     Mar 3, 2005