Looking for a collaboration

Discussion in 'Automated Trading' started by Леончек, Jul 28, 2025.

  1. Yes! I prepared a better illustration, please find attached
     
    #11     Jul 28, 2025
  2. Not an IQ test lol. We are trying to maximize profit by perfecting our execution. We know the edge for every trade with a great precision but we are struggling with estimation of our impact costs and what is the optimal amount to buy based on the order-book for every situation. Please look at the pdf above if you have time
     
    #12     Jul 28, 2025
  3. Q.E.D.

    Q.E.D.

    Sounds like an impressive group -- but in case some readers are not aware of the history of Long Term Capital, a hedge fund that had a super impressive group of people, including 2 future Nobel Prize winners, and despite number of years of excellent returns, almost took the entire financial system down:

    https://en.wikipedia.org/wiki/Long-Term_Capital_Management
     
    #13     Jul 28, 2025
    Bugsy likes this.
  4. Very educational story, thank you
     
    #14     Jul 28, 2025
  5. qlai

    qlai

    If you are already trading, you should have an idea of the impact. The book is not static, as you know, so just because orders are displayed, doesn’t mean orders may not be cancelled.
     
    #15     Jul 29, 2025
    comagnum likes this.
  6. MarkBrown

    MarkBrown

    i think you have an interesting question and i commend you for stimulating my brain with a question that is unique and i never would have thought to ask.

    however after thinking about it, i think the question has a fallacy in the assumption that what you see on the dom has anything to do with reality. the dom is real until it isn't, ask yourself who would show their cards in a poker game? why do you think the dom exist in the first place, for your benefit?

    as smart as your team is, you are down a rabbit hole chasing something that is vapor and has no substance. it actually pains me to know what you need, but i will not make it easy for you and drop it in your lap. actually i will the tool you need has been posted by me in a thread on this forum already.

    since your already hunting you might as well look for something that is 4x faster < (clue)
     
    #16     Jul 29, 2025
  7. Bugsy

    Bugsy

    Phil Ivey? Daniel Negreanu?
     
    #17     Jul 29, 2025
  8. trismes

    trismes

    Rough rules of thumb: use proxies to get a true TOB and slice at a max of 20% TOB or less, repeated at X interval, where X is based on TWAP/VWAP/secret sauce. Cum. size <1% ADV fill your boots. 1-10% ADV can be modelled, very roughly, with quadratic decay. Beyond that, shit show.

    That's the noddy version I can understand. more advanced path starts here: https://www.smallake.kr/wp-content/uploads/2016/03/optliq.pdf
     
    #18     Jul 30, 2025
  9. kbs

    kbs

    Interesting study. It is build for large portfolios, that properly trade a lot of regulated exchange traded products.

    Jim Rickards a former lawyer and prem dealer, who lost his life savings in the Long Team Capital Mangement, now book writter. Has written about false lividity in some crypto exhanges.


    I do not trade crypto. And I do not know, to what extent, or if false lividity exist.

    The study also exclude the discretionary trader. That according to the studie go for fast gain and think less about impact.

    I do not know if you example with buying 100* 1000K a 100.000K position falls under that if selling after 2 houres.


    But I whish your good luck
     
    Last edited: Jul 30, 2025
    #19     Jul 30, 2025
  10. demoncore

    demoncore


    You have no slippage at $1K per, and if you do than the market you've identified has no capacity. Perfecting execution would be trading inside the spread and keeping costs to a minimum. I assume with perps there is a ticket fee so this sounds a lot like a hobby/voyeur thing.

    You're stuff is too small to generate any revenues. Crypto transaction costs are still ridiculous.
     
    #20     Jul 30, 2025
    comagnum likes this.