is statistical arbitrage viable for individuals ?

Discussion in 'Automated Trading' started by jjk2, Dec 30, 2007.

  1. jjk2

    jjk2

    i'm just starting to read a few books on this...still need to figure out the math side of it. do you suggest any good books ?

    i know iknow, its tactically not arbitrage

    however, for those of you who are doing this, are the opportunities lot less frequent (ie. hedge fund arbs)

    am i wasting time by trying to expand into this territory ?

    cheers,

    i've so far enjoyed the level of maturity in this forum (no abusive mods). :)
     
  2. a 50/50 odds with 53/47 payout would be a good example of stat arb ( roulette , with double zeros for a house). What tools will you use to evaluate trading position the same way ? User defined "edge" ?
     
  3. 1). Give it time, the abuse might not emanate from our mods, but will most certainly orginate from arrogant snots like myself.

    2). There are no good books on Stab Art (are you serious?). There might be a few papers that you might find useful though - but nothing that will make you a trillionaire or get Jim Simons agitated enough to send his hired goons over to your place for an attitude adjustment. Google to the rescue.

    3). Define Stab Art. This can be anything from High Frequency sheah to bullshit pair trading to longer term long/short multi-factor models, universal portfolios, etc, etc, etc...

    4). ET is probably not the best forum for such questions..., oh wait... abusive mods? Hmm.. let me guess, you have already been "blended" at nuclearphynance?
     
    • arb.jpg
      File size:
      15.5 KB
      Views:
      4,394
  4. rosy2

    rosy2

    you need to define statistical arbitrage first. it is so broad that almost everything is considered stat arb
     
  5. jjk2

    jjk2

    im not that much active in nuclearphynance but i do read the posts, very smart people there....i think just too smart sometimes

    you are right, theres no book that will reveal the ultimate guide to millions, i never expect that, actually from any book.

    i just want to know how to do practical things on excel and such, so if im asked, do you know stat arb, i can explain it to someone at some length without embarassing myself.
     
  6. jjk2

    jjk2

    yes i should be more specific.

    im interested in the type of stat arbs thats employed usually in trading desks. namely pairs-trading. finding high correlation derived from statistical infernce of historical data valid ? and then taking advantage of any price discrepancies.

    (http://books.google.ca/books?id=FTF...oi=print&ct=title&cad=one-book-with-thumbnail)

    thats the one of book im going through.

    more specifically, i wonder if it can be applied to the currency market...i'll leave out the derivatives for later....much more complex i think.

    i guess i was just looking for ppl with actual experience in this field.
     
  7. I'm trading very simple stat arb strategies myself and yes, I think they're viable for individuals. You have to differentiate between high frequency strategies (e.g. index arbitrage which buys/sells index futures vs. the underlying baskets of stocks) and medium/longer term arbitrage strategies.

    I think the success of many high frequency strategies (<1s, 1s, 1m) is more and more determined by low latency and low transaction costs. I would say as an individual you could still make money here but you wrestle with institutions that have access to faster connections, hardware and lower transactions cost. However, I never traded in this space so maybe someone else can elaborate.

    The strategies I employ trade correlated baskets of equities over time spaces of days and weeks. Works out well for me since transaction costs and latency don't play as big a role as security selection and money management rules.
     
  8. Jumping pretty quickly from trading FX to statistical arbitrage aren't we?

    The mods didn't abuse you at FF, they banned you for posting 6 straight months of verbal diarrhea. Grow up.
     
    MiniDowTrader likes this.
  9. Obviously it's tough for publishers to find highly-qualified statarb traders willing to write tell-all books. AFAIK the only recent book on statarb per se is http://www.amazon.com/Statistical-Arbitrage-Algorithmic-Insights-Techniques/dp/0470138440/ which got mixed reviews on Amazon, and which, having read it last weekend, I can't recommend with much enthusiasm either.
     
    #10     Jan 1, 2008
    MiniDowTrader likes this.