IB Slippage: paper vs live

Discussion in 'Order Execution' started by SPYAlgoTrader, Dec 17, 2018.

  1. so anybody trading amzn like this on ib ? actual live slippage?

    qlai: good points
     
    Last edited: Dec 17, 2018
    #11     Dec 17, 2018
  2. That is total nonsense and not categorically true. It totally depends on how the paper trading algorithm at your broker simulates fills. For example, there is a demo account at IB where you can login even without a paper trading account. Fills there can take minutes which is obviously much worse than fills in a live trading account. Again, it depends on the simulated fill algorithm. Specifically at IB I often had unreasonably delayed fills in my paper trading account when I tested changes in my API infrastructure. Correct answer: one cannot tell, best solution is to make conservative fill assumptions for live trading.

     
    #12     Dec 18, 2018
    Overnight likes this.
  3. dholliday

    dholliday

    I have found that IB paper trading is very realistic.
    I have tested running the same system on live and paper, stocks and ETFs with wide range of liquidity, at the same time. All market orders. Over several hundred trades they averaged out to no difference, though they do very on individual trades. IB must be running an algorithm which takes into account of where you would be on the order book. On low liquidity stocks I have seen the paper trades take many minutes to trade just like the live trades.
     
    #13     Dec 26, 2018
    Overnight, qlai and SPYAlgoTrader like this.
  4. I doubt they have such sophisticated fill simulation algorithms. They most likely just fill you at best bid/offer for market orders and limit orders as well as long as those limit orders are "marketable". You will see that IB's fills on illiquid stocks will be very different from live fills. Pick some stocks that trade in Singapore or HK that trade across 2-3 price levels the entire trading session and you will see what I mean.

     
    #14     Dec 26, 2018