Midpoint match and pegged to midpoint orders

Discussion in 'Order Execution' started by luisHK, Sep 26, 2013.

  1. luisHK

    luisHK

    Anyone using those and willing to share feedback ?

    I hold positions in stocks showing over 1% spread during much of the trading day, several % around opening and might want to trade such stocks more often. Midpoint orders look useful but what about stocks with low liquidity ? On a midpoint sell order for instance, can't the traders at the bid just drop the bid so it move down significantly, so the new midpoint becomes the former bid, where they can now buy , in which case I would have actually sold on the bid rather than midpoint ?
     
  2. luisHK

    luisHK

    I would also use such orders for spreads smaller than 1% btw, but am not familiar at all with those.
     
  3. I've tried using every variety of pegged orders I could come up with in an effort to improve my fills on illiquid stocks but the results ranged from no improvement to horrible. I'm not a blame everything on the HFT boogeyman guy, but these types of orders seemed to be particularly susceptible to those types of algos.
     
  4. luisHK

    luisHK

    Thanks for your reply, feedback is especially useful as it seems useless to try those orders through papertrading.

    I will wait a bit longer.
     
  5. gmst

    gmst

    i think only real way out of this conundrum is to ask experienced traders or to basically have a bankroll of 50,000 usd as test bankroll which you use to test all kinds of market impact costs.
     
  6. luisHK

    luisHK

    uh ?!? I hope I was asking experienced traders and apparently got a reply from one of them already. Those orders don't seem very popular so I doubt there will be many newbies replying.

    As of 50k, I'd like to waste less in execution experiments but indeed it might be a sensible move to check the waters with small midpoint orders. Can't say the first reply I got was very encouraging though, especially that I suspected those orders could be played against, as mentioned in the OP.
     
  7. gmst

    gmst

    My above comment is general in nature. Few months back, I was wondering about the impact cost, liquidity, execution, trade size etc. on different stocks - on nasdaq, vs nyse etc. Same question can be asked for different futures, all the spread combinations, option spreads/flies, other countries equities, options, bonds, spreads etc.

    So, when i said 50k to test all kinds of impact costs, I covered a very wide range of universe. I wasn't really replying to your specific question - I just made a general comment. I apologise if my comment actually caused any confusion :eek:

    EDIT: For more clarification, I have no experience myself with pegged to mid kind of orders, their execution success or the way they distort the market or affects your fills. But I assume if you want to seriously experiment on this aspect, it will cost at least 5k to test different conditions and document your findings. Similarly, I wonder how much time/effort/dollars it will cost to compare execution quality/fill rate/ impact cost on different dark pools and other exchanges - while we have so many exchanges here in US.