Anyone Have a VWAP Program for CQG or TT

Discussion in 'App Development' started by msiper, Apr 8, 2013.

  1. msiper

    msiper

    Hello,

    I am looking for a sound VWAP program to be used within either TT or CQG. Would like to match or beat settlement price. Off-the-shelf already built program is preferred but recognize this isn't a sophisticated strategy so would be interested in someone who could build one quickly as well.
    Thanks
     
  2. hftvol

    hftvol

    sell-side banks would kill to get their hands on a VWAP algorithm that beats the theoretical vwap price most of the time (keep in mind many vwap algos kick in during trading sessions not beforehand as a pure function of when clients send their vwap orders to the DMA team of a broker). Most such desks lose money on vwap orders but they still do it because they can attract large volume at times and run the positions themselves given this is a desk with such mandate and the ability to manage the risk on their own books.

    It gets even trickier to beat the theoretical vwap for teams that get paid couple more basis points in exchange for guaranteeing vwap to clients because the type of assets they have to run their vwap algo over are much "trickier" than average (such as low liquidity, recent public offering and start of trading, small number of shareholders, large insider transactions...).

    So you are asking for someone to give you such algorithm? Good luck!



     
  3. hftvol

    hftvol

    that just shows how to calculate the vwap price, it is not a vwap algo itself that determines how much to trade at any time in order to match/beat the theoretical vwap price. Actually beating vwap is technically close to impossible, unless one creates a strategy overlay which trades more or less at any given time as a function of other non volume related metrics and may increase risk but also expected return. The reason for that is that a) vwap algos by definition are reactive, thus trade when trades already printed, often at less favorable prices, b) costs of executions need to be factored in while the theoretical vwap assumes fills at traded prices that already occurred, and c) most traditional vwap algos are heavily gamed by predatory algorithms (such as causing vwap orders to "chase" short-term momentum). Despite the simple math behind it , creating a good vwap algorithm is anything but trivial and can get as complex as thoroughly analyzing micro market dynamics and queue priorities. Just a heads up and reiterating that OP should not expect anyone to hand him the silver platter, I won't for sure....