It's based on whatever your chart's "Session begins" and "Session ends" times are because, yes, whatever bars you have before the open will effect the Keltner's position at the open. The chart I was showing was just an illustrative sample, not meant to be an optimal scenario.
The following user says Thank You to shodson for this post:
I've been receiving NT7 messages since yesterday that NT7b18/Zen-Fire would not connect and that the NT tech guys are working on it; and it didn't really connect; I'll try to connect to check if the connection is now ok for me.
You can actually get quite good mean reversion trades leading up to the report (till 10:20 AM). After that it becomes choppy as the DOM starts to thin out in to the report. CL tends to follow equity markets after 9:30AM especially if there is no other news catalyst.
Last edited by aviat72; July 21st, 2010 at 12:20 PM.
I decided to focus only on this strategy today, although I did make one discretionary trade for a -6 loss . The first trade was the outside bar at 6:45. I originally didn't place the order because both the Dow and CL looked like they could bounce, but after the price tried to push up and failed then I felt good that the short would hit at least 8 ticks so I entered. The second trade was my crap discretionary trade, which I dumped out at a small loss after it failed to push up. I should have never entered so close to the 78.00 handle and the news (oh yeah, also I haven't been trading multiple lots with CL yet and I forgot to change my ATM back to 1 unit on this trade, doubling my loss). The third trade was the inside bar after the news. The 2nd profit target was 2 ticks below the low for the day and I figured that if the price dropped to 2 ticks below the low that it would probably keep going, so I moved my proft target down and made an extra 10 ticks. This somewhat made up for my stupid trade earlier. I set it for audible alert and plan to keep trading it as long as it continues to work.
That said, the obvious concern with this stategy are the occasional excessively large stops. I am not sure I would be willing to take a 50+ tick loser. For example, on the outside bar trade above, I was going to stop out at the high of the current bar, which would have been a 42 tick loser, figuring if it pushed up that far again that the trade was most likely a dud. That said, in the long run it may possibly be better to simply filter out large stop trades. I am including the stop amount as a stat in my testing so I can analyze performance by stop amount. But I just started and because there are so few losers, I don't have enough data at this point. Is anyone else considering this in their testing and have any results to share?
Also, I don't know what other people are finding, but so far my limited testing has found that trading after lunch seems to be a waste of time. I need more data but are you guys finding the same thing?
I figured out they finally fixed NT so that you can now see your trades on the chart when trading CL with Interactive Brokers, so here it is. Good luck the rest of the day, it's nap time for me.