A path towards profitability

Discussion in 'Professional Trading' started by garachen, Feb 7, 2012.

  1. garachen

    garachen

    Per person we do about 3000 contracts/day in maybe 300-400 trades.

    I can't really advise on that particular method. Although it's one I'm familiar with it's not one I employ and I've never had great success trading ES. There are just too many participants for my taste unless it's overnight.



     
    #11     Feb 8, 2012
  2. which instruments are you trading with frequency so high?

     
    #12     Feb 8, 2012
  3. garachen

    garachen

     
    #13     Feb 8, 2012
  4. What are you trading in 1 sec timeframe, what is your size and expectancy?
     
    #14     Feb 8, 2012
  5. garachen

    garachen

    For fast time frames it will trade things that move fast:

    metals and energy

    size depends. between 10 and 100 usually depending on available liquidity

    expectancy. A few increments. Algo specific.
     
    #15     Feb 8, 2012
  6. For your algo work, what framework do you use for data storage and analysis? What data sources?

    I'm guessing a lot may be custom but I always try to find the best tools for the job.
     
    #16     Feb 8, 2012
  7. garachen

    garachen

    This isn't anything that can be done by the average person or firm. That's why I focused on manual trading in this post instead.

    I collect every tick off almost every futures exchange. There are some in Asia I do not have. And I store that full book as MANY terabytes of data very highly compressed. So compressed that I have to worry about cosmic rays and bit rot making it unusable.

    Then we run the algo through our simulator. We are able to simulate our latency, market congestion and exchange fill algorithms. Hit that with a genetic optimizer for about a week then test resulting parameter sets for stability.

    Data is bought from a variety of vendors that cover different markets. All the other software we wrote ourselves - including the database.
     
    #17     Feb 8, 2012
  8. You claimed to be an algo trader, but you need "ability," "intuition," "obedience," "calmness,"

    You are contradicting yourself.

    You will never make it as a discretionary trader. Your posts have nothing remotely related to being a successful discretionary trader.

    99% traders lose, they just lose.
     
    #18     Feb 8, 2012
  9. Understood and appreciate the insight. Too high-end for a garage quant likely to compete.

    But, I've been a programmer long enough to know that it isn't all quite so otherworldly when you break it down. And, with services like Amazon EC2, we all have on-demand clusters -and the future for this type of work looks interesting.
     
    #19     Feb 8, 2012
  10. garachen

    garachen

    I'm not sure why you are so angry. I admit, the post deviates from the original intent which was a pathway for manual trading. I still do manually trade some and my employees manually trade but I now spend much of my time dealing with adding new products/exchanges to the algo system.

    My goal is not to become a successful discretionary trader. This year my goals are to finish learning 'chasse neige' by Liszt and to fly my RC helicopter upside down. I'm pretty tired of manual trading. I find it tedious and boring and am glad I've been able to outsource it to other people successfully. I suspect it might get more interesting when we connect to Indian Multi Commodity Exchange later this year. Crazy new exchanges usually make trading interesting again.
     
    #20     Feb 8, 2012
    jtrader33 likes this.