HFT Myths

Discussion in 'Automated Trading' started by hft, May 3, 2013.

  1. Hello hft,

    You previously mentioned that the big banks e.g. GS, JPM were best for HFT in both equities and futures. Are you truly comfortable executing with firms like GS & JPM. Are there not risks with firms that engage in their own proprietary trading to monitor your proprietary trading activity with ulterior motives.
     
    #121     May 13, 2013
  2. Term

    Term

    What do you think is in store for the future of HFT industry?
     
    #122     May 13, 2013
  3. #123     May 13, 2013
  4. hft

    hft

    Maybe I'm being naive but it doesn't concern me. They certainly can't step in front of my order or fill/reaction since I'm connected directly to the exchange with no intermediaries. In theory they could push certain positions against me (like a index vs. futures spread) but a) it's probably not worth their time and b) it would take a heck of a lot of balls and capital to push those types of spreads around.
     
    #124     May 13, 2013
  5. hft

    hft

    Direct from exchange, period. Filter out what you need for live trading, record everything, keep anything you need often easily accessible and archive the rest for future use.
     
    #125     May 13, 2013
  6. onelot

    onelot

    just repeating the above for another look, thanks...
     
    #126     May 13, 2013
  7. hft

    hft

    1)
    Inexperienced: CS/Math undergrad and/or post-grad from top school and unanimous passing of panel interviews. Sucks but that's just the way it is. There's so many qualified applicants it's just terribly hard to stand out otherwise. If you had something extraordinary technological achievement that stuck out that could strike a chord too.
    Experienced: All boils down to how much money you can make and what resources would be required to allow you to make that money. Then all the intangibles too (can we believe you - the industry is small so it's not that hard to verify someone's track record, personality fit/etc.)

    2) Most HFT isn't high-risk and easy to flesh out with 1-lots, so there's little reason not to start there and ramp up quickly if it works. The bigger allocation is the resources spent on getting you up and running, which goes back to how good your track record is. A rule of thumb I go by is 1 man-month per million in estimated yearly PNL. So if I think you can bring in $6M a year I'll give you 6 man-months of resources to get you going. It's a better pitch if you bring in your own team and the firm is purely paying you salaries for X time as in investment.
     
    #127     May 13, 2013
  8. onelot

    onelot

    ok, that's interesting, thanks. is that mid-7fig target a typical minimum these firms want to see?

    you know of any firms similar to an HTG on the equity side?
     
    #128     May 13, 2013
  9. Thank you for the thread. A lot of useful information.

    In a HFT firm, do quants trade their own model (do everything) or they just develop the model, let software engineer develop the trading engine/GUI and trading is handled by trader?

    Are trader mainly looking at the screen to make sure everything works fine and adjust some parameters only? Trader might
    also need to find the opportunity? e. g. if strategy is pair trading,
    trader need to find a good pair?

    How profit is shared among quants/software developer/trader?

    Whose contribution is the most? quants or trader?
     
    #129     May 13, 2013
  10. hft

    hft

    Yeah I'd say 5M is a typical minimum for most larger firms to start taking someone seriously, maybe a little less if their resource requirements are minimal or line up cleanly with what's already available.

    A lot of the bare-bones firms like HTG are really hesitant to delve into the HFT equities space. I'm also not that familiar with those types of firms anyway so can't help you there. I don't know of any particular one that's really strong in equities off the top of my head.
     
    #130     May 13, 2013