logo Satuki @ Home

Trade to win

Sign up for my latest trading ideas. Free!

Day 2 Experiment: Trading Automation of SRS and FAZ

Posted by Satuki On April - 21 - 2009

Today is the second day of the experiment. The computer closed both of them.  The FAZ trade was up 30% and the SRS trade lost 5%. 

 

When I woke up this morning, FAZ was up 15% in pre market.  And the computer sent a sell order right after the opening bell rang.  It kept checking the gain/loss of these 2 positions tick by tick.  In terms of the FAZ position, if the gain was = or > 25% and a sell order would be executed. Due to the gap up, it was 5% more than the targeted 25%.  That was a bonus.  Of course, that would be extra risk if it were in the different direction than my trade. For the SRS position, the gain never reached 25%, and it hit the stop loss, 5% and it was out.

 

The FAZ exit was perfect.  It was almost the high of the day.   Question is whether it should be considered luck or discipline.  The computer knew it captured a giant return (30% in 2 days) and it was neither afraid nor greedy.   Greed and fear are the biggest enemies of traders.  The computer has none of those issues.  So It held onto the FAZ position and closed it according to the plan.  It never flinched in any order executions.  I have no problem closing a position when it hits its target.  If you have been reading my blog for while, you know that.  But I still have issues pulling the trigger when there is entry signals  The computer also showed rock solid discipline in the execution of the SRS trade because it followed the trading plan religiously.  The experiment will carry on with SRS.   Let’s see if SRS will show us the darkest side of machine trading.  That is 5 losers in a row. 

 

Here are the 2 trades closed today.  Click on the pictures to have a better view

 

What do you think?

Related Posts


Would you like to receive the latest trading ideas or other updates from my blog? Then join hundreds of others on my mailing list below. It is totally free and easy to opt-out.

Enter your email address:

Related Posts

Swing Trading, Trading
  • Mike

    Too bad the system isn’t set up to go short. (or long FAS)

    impressive results!

    OT – how much further up are we going and where will the roll over take us? I’ve heard some wild numbers (900 -> 450 SP500). I’ll guess 870 -> 760 -> 840 -> 680. : )
    I’d love to get your guess.

    • http://www.momdaytrader.com/blog/ Satuki (Trader Mom)

      I doubt that the market is going to drop that much. I thnk we have hit a big bottom here. I would flip to the bull side(swing trading wise) when it hits 750. The reason I keep shorting the market is because I do not see much upside. I am just expecting a quick small drop to 750 for some type of healthy corrections. The market has been acting much stronger than I think. It has been chopping hard.

    • Daniel

      I also use tradestation and I’ve been testing a machine algorithm on the platform for weeks. Have you done optimizations and can you print out and display the backtesting results? You do this by going to “View: Strategy Performance Report”. Then you can click the “save” icon on the performance report and save to an “mht” file. This file is in html format and is then all ready for display on your website. It works great and is the best way to review performance.

      I trade FAS with my system, long and short. I backtested from Oct 2008 which is as far back as FAS goes. Net profit would have been $33,000 in that time investing a max of $10,000 at any one time. The biggest single losing trade was $1,200 (factored into the “net” gain). I am eager to find out if there is a better performing system with risk/reward considered. Note that I’ve just started testing mine as well as I’ve been modifying it, so this is paper gains so far.

      • Daniel

        One correction to my post. The $1,200 was not a single trade but the worst run of losing trades. I looked at it and said, ‘hum, that would mean my stop loss was at %12!” It was instead a series of trades in the red. My worst single trade was $700 but that was an exception as my stop is normally closer to %5 or %6.

      • http://www.momdaytrader.com/blog/ Satuki (Trader Mom)

        I did not do optimizations since it is curve-fitting.

        • Daniel

          Regardless, if you do not do optimizations for fear of curve fitting, backtesting is still invaluable. If you have an algorithm that is not curve-fitted then it should do great in a backtest, even better than all the curve-fitted algorithms.

          Also, if you optimize over a 6 month period, and then backtest over a 2 year period and get the same results regardless of time-frame, then you are strongly countering the curve fitting argument. Basically, if your 6 months of optimized data works for 2 years you must have found some “first principals” by which your system is operating. I also would think you would tell from the nature of the algorithm if it is curve-fitting. It is MACD types of systems and RSI ones that tend to be susceptible to curve-fitting because they are statistical and not absolute. How can a doji reversal be curve-fit if it follows the programming rules?

          • http://www.momdaytrader.com/blog/ Satuki (Trader Mom)

            Yeah, I did a lot of back testing. It was all good on paper. It did not work out in real world as expected.

            As for optimizations, if you are referring to the “optimization” button in TradeStation, I do not think it works since it tries to give you the best returns for the period of time that you are testing by manipulating your parameters.

            Things will be different when you put real money on the line. I used to very zealous about machine trading. But I am not a big fan of it anymore.

            Of course, do not let me discourage you. I might not have found the trick. LOL

          • Daniel

            Ok, well I guess if you find the time I’d really appreciate seeing your backtest results. That is all I’m after, because I want to compare from others who are trying to do machine trading. I am not trying to convince you of the virtues of machine trading at all, or convince anyone that it is working for me. I’m simply wanting to compare my own machine trading results to other algorithms. That way I can better judge if what I’m doing is worth all the effort.

            I actually do believe that I have something useful because I’ve seen so many of my attempts fail in backtesting. Dozens of attempts failed this way without spending a dime in a trade. One algorithm has worked out, but I’m not taking that for granted. I will use it with great caution until it proves itself in forward testing. That is, real trading with small position sizes.

  • Mike

    Too bad the system isn’t set up to go short. (or long FAS)

    impressive results!

    OT – how much further up are we going and where will the roll over take us? I’ve heard some wild numbers (900 -> 450 SP500). I’ll guess 870 -> 760 -> 840 -> 680. : )
    I’d love to get your guess.

    • http://www.momdaytrader.com Trader Mom

      I doubt that the market is going to drop that much. I thnk we have hit a big bottom here. I would flip to the bull side(swing trading wise) when it hits 750. The reason I keep shorting the market is because I do not see much upside. I am just expecting a quick small drop to 750 for some type of healthy corrections. The market has been acting much stronger than I think. It has been chopping hard.

    • Daniel

      I also use tradestation and I’ve been testing a machine algorithm on the platform for weeks. Have you done optimizations and can you print out and display the backtesting results? You do this by going to “View: Strategy Performance Report”. Then you can click the “save” icon on the performance report and save to an “mht” file. This file is in html format and is then all ready for display on your website. It works great and is the best way to review performance.

      I trade FAS with my system, long and short. I backtested from Oct 2008 which is as far back as FAS goes. Net profit would have been $33,000 in that time investing a max of $10,000 at any one time. The biggest single losing trade was $1,200 (factored into the “net” gain). I am eager to find out if there is a better performing system with risk/reward considered. Note that I’ve just started testing mine as well as I’ve been modifying it, so this is paper gains so far.

      • Daniel

        One correction to my post. The $1,200 was not a single trade but the worst run of losing trades. I looked at it and said, ‘hum, that would mean my stop loss was at %12!” It was instead a series of trades in the red. My worst single trade was $700 but that was an exception as my stop is normally closer to %5 or %6.

      • http://www.momdaytrader.com Trader Mom

        I did not do optimizations since it is curve-fitting.

        • Daniel

          Regardless, if you do not do optimizations for fear of curve fitting, backtesting is still invaluable. If you have an algorithm that is not curve-fitted then it should do great in a backtest, even better than all the curve-fitted algorithms.

          Also, if you optimize over a 6 month period, and then backtest over a 2 year period and get the same results regardless of time-frame, then you are strongly countering the curve fitting argument. Basically, if your 6 months of optimized data works for 2 years you must have found some “first principals” by which your system is operating. I also would think you would tell from the nature of the algorithm if it is curve-fitting. It is MACD types of systems and RSI ones that tend to be susceptible to curve-fitting because they are statistical and not absolute. How can a doji reversal be curve-fit if it follows the programming rules?

          • http://www.momdaytrader.com Trader Mom

            Yeah, I did a lot of back testing. It was all good on paper. It did not work out in real world as expected.

            As for optimizations, if you are referring to the “optimization” button in TradeStation, I do not think it works since it tries to give you the best returns for the period of time that you are testing by manipulating your parameters.

            Things will be different when you put real money on the line. I used to very zealous about machine trading. But I am not a big fan of it anymore.

            Of course, do not let me discourage you. I might not have found the trick. LOL

          • Daniel

            Ok, well I guess if you find the time I’d really appreciate seeing your backtest results. That is all I’m after, because I want to compare from others who are trying to do machine trading. I am not trying to convince you of the virtues of machine trading at all, or convince anyone that it is working for me. I’m simply wanting to compare my own machine trading results to other algorithms. That way I can better judge if what I’m doing is worth all the effort.

            I actually do believe that I have something useful because I’ve seen so many of my attempts fail in backtesting. Dozens of attempts failed this way without spending a dime in a trade. One algorithm has worked out, but I’m not taking that for granted. I will use it with great caution until it proves itself in forward testing. That is, real trading with small position sizes.

  • Mark

    Trader Mom: thanks for sharing your experiment. You use TradeStation. What kind of learning curve does that have? I’m with TDAmeritrade now and considering a move to Tradestation. Will continue to watch your experiment!

    Mark

    • http://www.momdaytrader.com/blog/ Satuki (Trader Mom)

      Mark, I will write a post comparing TradeStation, Interactive Brokers and ThinkorSwim, which most full time traders use. I will keep you posted

  • Mark

    Trader Mom: thanks for sharing your experiment. You use TradeStation. What kind of learning curve does that have? I’m with TDAmeritrade now and considering a move to Tradestation. Will continue to watch your experiment!

    Mark

    • http://www.momdaytrader.com Trader Mom

      Mark, I will write a post comparing TradeStation, Interactive Brokers and ThinkorSwim, which most full time traders use. I will keep you posted

  • Michelle

    Good job!

  • Michelle

    Good job!

  • Michelle

    Satuki

    I forwarded your web site address to my other friends. You will more friends here. :))

    • http://www.momdaytrader.com/blog/ Satuki (Trader Mom)

      Thanks Mchelle. I am glad that you like my blog and trading.

  • Michelle

    Satuki

    I forwarded your web site address to my other friends. You will more friends here. :))

    • http://www.momdaytrader.com Trader Mom

      Thanks Mchelle. I am glad that you like my blog and trading.

  • http://none.com/ J

    TraderMom,

    Does today’s last SRS candle look like an engulfing bar?

    Thanks

  • http://none.com J

    TraderMom,

    Does today’s last SRS candle look like an engulfing bar?

    Thanks

Trading Ideas

Receive Updates via Email Delivered by Google FeedBurner

Enter your email address:

Popular

    USER LOGIN


    Twitter: Follow Trader Mom
    RSS Feed: Subscribe in a reader




    Page optimized by WP Minify WordPress Plugin