Fully Automated Stocks Trading

Discussion in 'Journals' started by ValeryN, Jun 14, 2020.

Thread Status:
Not open for further replies.
  1. I doubt you have a real execution edge in vanilla equities unless you are running some sort of a DMA system. My sense is that your tricks are likely to be just part of your retail-size edge (i.e. you are more sophisticated than an average retail size trader, so you get away with being undetected).
     
    #301     Dec 15, 2020
  2. ValeryN

    ValeryN

    It is an edge. If I execute differently I will predictably make less money.

    Val
     
    #302     Dec 15, 2020
  3. That could be because either (a) you actually make money on execution (i.e. you have an edge, like the market makers do) or (b) you lose less money on execution (no edge, you just optimized your execution process to lose less). Which one do you think it is?

    PS. A more analytical approach would be to check - if you were to get done exactly at the arrival price, would you make money money or less money?
     
    #303     Dec 15, 2020
  4. ValeryN

    ValeryN

    Strictly speaking it is both, with vast majority coming from (b).

    Sorry I can't give more details. Tried to describe it vaguely just now but either way I put it - looks like too much.

    From a practical perspective of retail trader, even if I would NOT have (a), I'd still call executing at lesser costs comparing to pretty much any other retail trader including ones running successful system - an edge.

    Val
     
    #304     Dec 15, 2020
  5. traider

    traider

    Have you tried any of the brokers execution algos? Are they smart enough to add a tiny bit of edge to execution?
     
    #305     Dec 16, 2020
  6. ValeryN

    ValeryN

    I didn't do any formal comparison. IB is the only one for Canadians with API. Alternatives require forming LLC in US which I am still hesitant to do. I am now eying Alpaca though, as they started private beta testing for Canadians with 100k+ accounts and I have 1 system that could particularly benefit. So, might eventually get a good comparison done.

    Having said that, I do rely on general decency in IB executions. Overall they do reasonably good job. Unless you do something plain stupid like sending a directly routed MKT order to illiquid market or trade without LMT after hours. Not even sure they allow that these days. Doing those would be basically like opening a window and throwing your money away, then blaming strangers for picking them up. IB and OTC* is kinda like a wild west, but I would imagine that's with any broker.

    Val
     
    #306     Dec 16, 2020
  7. ValeryN

    ValeryN

    This Sunday at 9am PST, I'll do a my process explainer and Q&A for a fresh batch of friends who asked.

    If anyone interested PM me, I'll send you a Zoom link to join within a next couple of days.

    Draft agenda:
    - Example of what a systematic strategy consist of
    - What realistic single strategy performance looks like
    - What trading it manually looks like
    - What automation looks like
    - Q&A

    Plan is to keep it ~40-60 mins. But I miserably failed at that last time I tried. Maybe this time will be different.

    Val
     
    #307     Dec 16, 2020
    shuraver likes this.
  8. It's all right, we are having a general discussion. My main point is that most of alpha for a retail trader comes from looking at capacity constrained opportunities, not from building execution infrastructure or finding some unusual alternative data sets. That's why I was so surprised that you felt that you had an execution edge of any sort (hint - unless you are rolling some serious tech most probably you do not).
     
    #308     Dec 16, 2020
    d08 likes this.
  9. mhparker

    mhparker

    Even if the true source of retail alpha is capacity constraint, a somewhat-high-frequency trader still needs a half-decent execution infrastructure to extract it. Is a carpenter's edge the nail or the hammer?
     
    #309     Dec 16, 2020
    ValeryN and Nathan07 like this.
  10. Craig66

    Craig66

    My 2 cents on the "execution edge" thing, I don't think it's a binary outcome whereby you either have or don't have "an edge". There are many components to execution: latency, venue selection, price, order splitting, timing (potentially among multiple instruments). Of course, any retail trader is going to be orders of magnitude slower to market than anybody with serious DMA access. But intelligent placement and routing can, at least, help to win rebates on a good proportion of orders, I wouldn't call that a disadvantage. As already discussed in this thread, speed matters less in lower volume equities.
     
    #310     Dec 17, 2020
    ValeryN likes this.
Thread Status:
Not open for further replies.