collaborations

Discussion in 'Strategy Building' started by ssrrkk, Nov 7, 2011.

  1. You know your system better than I, and you could probably answer that question better based on the specifics : mainly how important resolution (or loss thereof) and noise/accuracy are to your signal generation. Also, with the IB tick data, you won't get much out of doing that. Maybe with some better feeds.

    Hey, we're almost collaborating over here! I got some ideas from you, you got some ideas from me, that's kinda rare on these forums. Cool.
     
    #41     Nov 9, 2011
  2. ssrrkk

    ssrrkk

    I appreciate this discussion -- it's like a breath of fresh air compared to wasting hours and hours arguing whether statistics and probability theory should be even considered at all when devising strategies.

    I agree with the problems of IBs tick data. And it's not only their ticks that are problematic -- their day open high low close values are often off by a significant amount Of course, the open is off almost every day by a huge huge amount compared to other data sources, even yahoo and marketwatch. So it is very tricky to use these extreme values in your signaling (e.g., pivots will often be off by a lot). I am seriously debating about getting IQFeed.
     
    #42     Nov 9, 2011
  3. kut2k2

    kut2k2

    #43     Nov 11, 2011
  4. ssrrkk

    ssrrkk

    #44     Nov 11, 2011