jpb Veteran   Posts: 168 Joined: 5/11/2005 Location: Brown Deer, WI  User Profile | After Steve Leurman's initial post in March, I converted the strategies to MOC but haven't traded them. Instead, I took an account I was trading M&M in and adjusted AutoTrade to run its analysis before market close (around 2:20pm CDT). It finishes somewhere between 2:30 and 2:45 CDT. On occasions it runs a little longer which is ok since it submits market orders. It has been running for about 3 months. June was roughly flat. July and into August were down. After mid-August things started to turn around. That account is now down about $500 on an account size of about $15,000. A few weeks ago, it was down $1,200. Unlike Bill, my account saw an August draw down. I have another AutoTrade instance running using the MOC version of the strategies with an Inactivity Stop. I start this one around 2:12 pm CDT so it completes before 2:45 pm (the approximate cutoff for MOC orders on GXT/IB). On occasions it doesn't complete until 2:50. Between 2:45 and 2:50, it is hit and miss if IB will accept the order. I also trade this account with other systems so there is a lack of funds to execute all the orders so I can't accurately assess the performance. What I can say is that when I spot checked between the 2 approaches when both instances were able to submit orders for the same symbol, the MOC orders tended to be slightly more favorable. While I haven't done a detailed analysis yet, I'd say I've seen a slight edge go to the true MOC orders instead of issuing orders in the last 15 - 30 minutes of the market. I don't have a pure MOO version of ATM M&M running to compare with for real world results so I'm trusting the simulation is a close approximation to MOO vs MOC. | |
