HFT robots seeing orders before the exchange ?

Discussion in 'Order Execution' started by softdown, Dec 12, 2012.


  1. Ummm ... first, I always think it's a great question to ask "Why should anyone listen to [me]?" ... but I don't think it's appropriate to phrase it as a rhetorical question as DeeDeeTwo seems to have done here.

    Now, regarding your statement about spreads ... I'm not sure where you are coming from, but I'm guessing it has something to do with my comment about my average spread being about $0.10 ... in case there is some misunderstanding, please note that this is my average profit spread inclusive of commissions, not the average spread of the stocks I'm trading ... I trade hundreds of stocks with average market spreads of $0.02 - $0.05 ... in 2012 I traded something like 1000 stocks and profited on all except 24 (just ran those numbers).

    In any case, I'm not intending to brag about my p&l spreads in isolation from the other details of my trading program ... if DeeDeeTwo makes a better ROC, with equal or lesser risk, in earning his/her $0.01 spread, than I do earning my $0.10 spread then DeeDeeTwo has the better trading program and I wish I was trading like DeeDeeTwo.
     
    #41     Jan 3, 2013
  2. Just a gentle question here guys, are you paying any spreads to re-balance yourselves?

    Hard to imagine you're muddling through, delta neutral all the time.

    p.s. Like the banter .. cool heads hey.
     
    #42     Jan 3, 2013
  3. Good question.

    The only time you might penny-jump...
    Or hit the NBBO and "pay the spread"...
    Is when you are getting imbalanced...
    And need to move closer to market neutral.

    Staying market neutral is not remotely trivial.

    It requires sophisticated infrastructure (usually custom)...
    A rock solid understanding of the interrelationships you trade...
    And, as you point out, is a direct ongoing expense.

    As for Trader442...
    His statements about volatility's effect on spreads...
    And claim that "predatory" robots are somehow more prevalent in low volume stocks...
    Are just plain wrong.
     
    #43     Jan 3, 2013

  4. Well, I'm never "balanced" (always net long a giant book of crap stocks heading for bankruptcy), so "rebalancing" is not an applicable concept ... but sure, I take liquidity all the time to close positions out ... just not a lot on a net basis. I would guess very roughly that excepting days when I'm on strong defense, about 3-5% of my executions take liquidity. This is way more than in the good old days when taking liquidity was seldom a good idea on a strict risk/reward basis. These days, with all the "enemies" out in force, taking liquidity while it's there, before somebody else eats it, is frequently a good idea in hindsight.

    Of course, when I'm in strong risk-reduction mode I'll suck up liquidity like an old lady with a $10,000 AmeriTrade account.
     
    #44     Jan 3, 2013

  5. I'm confused here about the first statement, DeeDeeTwo ... can you please clarify precisely which of my statements are included when you refer to my "statements about volatility's effect on spreads", so I can defend my dubious honor and perhaps edify one or more readers of this thread?

    Regarding my supposed "claim that 'predatory' robots are somehow more prevalent in low volume stocks" ... please correct me if I'm wrong, but I think I only claimed that the "enemies" are very active in low volume stocks. If I said or implied anything more then I was wrong. I would guess that all other things being equal, prevalence of activity generally correlates to profits to be had ... and certainly very low-volume stocks are not generally very profitable compared to higher-volume stocks, even when the larger spreads and greater proportions of uninformed order flow are considered.

    However: When the "enemies" happen to be jumping all over a very-low-volume stock while I'm trading it, as they frequently are, there is certainly no mistaking it.
     
    #45     Jan 3, 2013
  6. Whatever, I'm strung out on mescaline...
    And cannot unravel what on earth this is all about...
    But you seem to be good at whatever it is that you do.

    Carry on.
     
    #46     Jan 3, 2013
  7. As I see it, a smart quoting algo is absolutely necessary to play the game, especially to layer & to back away fast .. running off (let's call it) a basket pricer.. Re-balance can be manual, but based on the same pricing, this time as a risk measure.

    My 2 ticks.

    Thoughts welcome..
     
    #47     Jan 3, 2013
  8. Everyone talks about "algos"...
    Like you are gonna fake out people with 8 figure tech budgets...
    Plus your broker who is frontrunning you...
    (Or, at best, cherry picking customer orders)...
    Plus every slimeball ECN hustler...
    Plus every insider on the Street.

    I got a news flash... no you are not.

    What I deal in in Automation...
    The ability to send 1000s of Limit Orders and auto-respond to fills...
    While keeping a 100+ position Portfolio market neutral...
    And relying on volume to overcome all the snags.

    When you have 100 positions or more...
    You are not terribly concerned about what happens with one of them...
    Which makes things like a "smart quoting algo" a marginal luxury at best.

    Some things are doable. Most are not.

    Every day I see businesses in my neighborhood...
    That will 100% be gone in one year.

    Game selection is the key to all Zero Sum game success.

    Most people here pick the sexiest, toughest, highest profile game...
    Then build gambling systems and delude themselves for years that they are "trading".
     
    #48     Jan 4, 2013
  9. Ah .. I meant within an index arb book, inside a broker, working around the waste-of-time tech kingdom.
     
    #49     Jan 4, 2013
  10. HFT robots hijack thread.
     
    #50     Jan 8, 2013