Frosty's trading bot goes live part 2

Discussion in 'Automated Trading' started by frostengine, Jun 18, 2007.

  1. jimmyrey,

    Yes both slippage and commision is arleady figured in. Loss due to tech problems however was not. This is no longer as much of a problem, since after the last ISP outage I moved the bot over to a data center where I work.
     
    #191     Jul 7, 2007
  2. jonnyman

    jonnyman

    Honestly common technical problems are solvable and should not impact trading except in extreme situations. You have relocated to a data center, which solves most problems however for those who have not:

    a) buy a dual-wan router and subscribe to two isps

    b) buy a battery backup

    Unless both isps go down simultaneously you should be fine. If you are not around to watch the bot (very dangerous in my opinion but that's neither here nor there) buy an advanced backup system that will tell the computer when the battery power is engaged, and when it is low - have the bot stop and flatten in this case. Those are the two main causes of heartache and they've saved our systems more than once.
     
    #192     Jul 8, 2007
  3. Another day in the books, and the road to recovery is still continuing. Today +$250.4

    Granted, still a long hard road and the past week's winnings doesn't even get me back to above where I was before the last big loss I took... but at least we moving in the right direction.

    Today +$250.40
    Week to Date: +250.40
    Since going live: -$2662.8
    Total days: 13
    Avg per day: -$204.83

    Up Days: 7
    Down Days: 6
     
    #193     Jul 9, 2007
  4. Shawn B.

    Shawn B.

    Frosty, could you please tell me what software and broker you are using to backtest and actually run the bot? Thx
     
    #194     Jul 10, 2007
  5. OK

    Good

    Now, before it happens again:

    1. Pull up charts of the past 13 days.

    2. Study price action on the days when you made money, AND (more importantly, because it is your bot's weakness) on the days when you didn't make money.

    3. Then, once you see the picture, devise strategies to curtail your losses on the days when your strategy proves to be ineffective.

    Good trading,

    JJ

    P.S. I took a very small loss today, and now I'm just sitting here looking at the slop. That tells you something about the strateg(ies) I use to trade the markets, what my Risk:Reward ratios are and what happens for me on those trend days that are the bane of your bot.
     
    #195     Jul 10, 2007
  6. nkhoi

    nkhoi

    today was kind of ranging in the morning however 20 EMA was below 34 EMA for the whole time, hm...
     
    • 2ma.gif
      File size:
      31.7 KB
      Views:
      125
    #196     Jul 10, 2007
  7. Worthless advice. Sample set too small to devise robust strategy.
     
    #197     Jul 10, 2007
  8. Another day, this one not so great but still up for the week.....

    Today -$229.20
    Week to Date: +$21.2
    Since going live: -$2892
    Total days: 14
    Avg per day: -$206.57

    Up Days: 7
    Down Days: 7
     
    #198     Jul 10, 2007
  9. maxpi

    maxpi

    Actually, I learned in statistics 101 that 30 is a large sample size. I have found that it is good enough for noodling and getting ideas, especially in the recent market where you were attuned more to it than the distant past.

    People go to large data sets and think they proved something because of the size of the set. They toss out good strategies that just need filtering to fit the market conditions.
     
    #199     Jul 10, 2007
  10. Yep, sounds just like 100 level stats at a mediocre community college.
     
    #200     Jul 10, 2007