Wednesday, May 26, 2010

A question about CME order execution

After reading Chasing the Same Signals I started thinking about order execution, about which I know next to nothing. So I went to the CME site to educate myself. Alas, I’m still ignorant.

In a study dating from March 2009 the CME analyzed immediately executable orders--that is, orders that can be at least partially executed at the time they reach the central limit order book. One finding puzzled me: “order quantities between six and 49 contracts are being executed with lower market impact than orders of five or fewer contracts.” Market impact means “the difference between the middle of the market at the time of the order’s arrival and the order’s execution price, or the average execution price in the event of fills at different prices.”

The fact that the small trader often doesn’t get the best price is not a function of speed of execution. Orders to buy or sell between one and five contracts were filled within 30 milliseconds after their arrival at CME Globex 86% of the time and within 50 milliseconds 91% of the time. Orders for six to ten contracts were filled within 30 milliseconds 85% of the time and within 50 milliseconds 93% of the time. As order size increased so did average fill time.

Why do the smallest traders not get the best price? Perhaps they’re simply addicted to market orders and by definition always give up the spread. Perhaps CME’s matching algorithms tilt in favor of the larger trader, with FIFO being the default for small size. The upshot is that I don’t know the answer. I’m sure that many of my readers are more knowledgeable on this score than I am; if so, please share.


  1. Brenda,

    I was waiting for somebody who *knew* to provide a real answer, but since there are none so far, here is a guess: larger trades only trade at times of great liquidity, whereas smaller traders trade basically 24/5, so on average, they should get worse fills because they trade at times with reduced liquidity and larger spreads.

    I'm not sure that market orders are relevant here, since all trades are market orders lifting book orders (ok, this is not entirely true - think limit orders at bid/ask price, but for the purpose of the discussion I think it applies).

    Best trading,


  2. Jorge,

    It seems that the CME confined its study to floor hours—that is, 8:30-3:00 Chicago time. Two graphs speak to the question at hand. Exhibit 1 shows that the smallest trader does indeed account for a higher percentage of filled orders when there is reduced liquidity (mid-day) than when there is ample liquidity (open and close). But Exhibit 5 shows that he experiences the highest market impact cost of all traders on the open and second only to the 51 to 100-lot trader both mid-day and on the close. The CME study states that “while many variables can influence an order’s impact on the market, the depth of book, size of the order and the client’s urgency will have a significant effect.” Which takes us back to the two variables—whether the order falls under the category of liquidity provider or liquidity taker (since I don’t know any way to express urgency other than with a market order) and whether CME’s order execution algorithms favor the strong.

    A still confused