multichart vs amibroker ?

Discussion in 'Trading Software' started by mcgene4xpro, May 21, 2010.

  1. LeeD

    LeeD

    mcgene4xpro, given your interest in 100-millisecond bars expresses in a different thread, it's quite interesting that you narrowed the platform short-list to the 2 platforms with the most concise language but the least tools for handling orders and executions at a low level.

    I can't comment regarding charting but AmiBroker has amazingly fast back-tester for bar-based trading systems. Its language allows converting backtests into matrix calculations. The drawback is this works on bars only and you need to be very much aware that not minding a few simple rules may lead to forward-looking bias.

    Regarding speed, for trading systems of simple to moderate complexity on a dozen of instruments pretty much any platform will be fine. Computers are getting faster and you need a trading platform that will do what you need and will make you the most productive, not something that's marginally faster.

    Obviously, some heavy calculations like multidimentional optimisation on every tick or calculations done on hundreds of symbols can stretch modren computers... but 1) it's your strategy code that will be the bottleneck; 2) If you do need hundreds of symbols, make sure the platform can handle this load at all.

    A few things to watch out for:
    1) There are a few platforms out there with a history of freezing when market activity on liquid instrumenst like S&P 500 E-minis spikes. Some platforms may also develop a few-second lag and don't catch up after activity in the instrument slows. This freezing can happen both in data handling and when updating charts. I don't want to call any platform names hear but earge you to do your own research.
    2) Make sure the platform can backtest the same exact code that is going to be used for trading. Having separate "trade execution" code carries potential for difiicult-to-catch errors.

    You can get in one platform pretty much every essencial feature except millisecond precision and backtesting on level 2 (market depth) data. So, besides saving a few hundred dollars there is no reson to compromise.
     
    #11     May 22, 2010

  2. my response inside quote.. thanks man
     
    #12     May 22, 2010
  3. ET99

    ET99

    in that case, don't make initial cost a criteria.

    look at the total cost... that includes your time investment in learning the package and getting your strategy up and running.

    Amibroker is fast and powerful...
    but would take more effort to learn.

    MultiCharts uses EasyLanguage.
    There are thousands of ready made codes and resources on the internet.
     
    #13     May 22, 2010
  4. ET99

    ET99

    there's an old thread here with a hot debate between Amibroker and MultiCharts. Lots of fireworks, accusations and innuendos.
    don't remember the title tho.
     
    #14     May 22, 2010
  5. thank you for sharing
     
    #15     May 22, 2010
  6. LeeD

    LeeD

    You want the platform to allow your strategies to track individual orders in the market, see partial executions, cancel open orders. Ideally, you want all this mechanics to work in backtests... including partial executions.

    I looked at AmiBroker and Multicharts a couple of years ago. So, feel free to correct me if my info isn't up to date.

    I am sending you a PM....


    You are welcome!
     
    #16     May 23, 2010
  7. ET99

    ET99

    Your info aren't up-to-date.
     
    #17     May 23, 2010
  8. pspr

    pspr

    AmiBroker used to be called the poor man's Metastock. Since Metastock has serious limitations and doesn't seem to have added important features that other platforms now use, I wonder if AmiBroker has passed Metastock in functionability even though Metastock RT still costs around $1600.
     
    #18     May 23, 2010
  9. So, it is perfectly suits me :)

    thanks
     
    #19     May 23, 2010
  10. Mcgene -- here's your PhD teacher:

    1. it should be hobby, not hoppy
    2. "even with harsh too intellegent words" - requires serious revision, MS Word will tell you.
    3. it should be intelligent, not intellegent

    Back to trading, if indeed you are going for your PhD, stick to it, 80% people on this board haven't made any money for the last 4+ years, 20% made less than 50k a year and the remaining 20% it's still a puzzle...

    2c$
     
    #20     May 23, 2010