One of our room members has put a lot of attention to the stretch pattern of Toby Crable. Thi sinspired me to look into this pattern in some more detail. In the process, I ran into another pattern that Crable refers to as the bull bear and the bull hook. I decided to code these patterns up and do a little research and then I saw that Ned wrote about a similar situation in his Introduction to Market Profile article when referring to excess. As I looked at his example, I could not help but notice it was the same pattern (it could be different, but in this case it was the same.
Initially I ran a test to see if there was a bias on these patterns over near term future bars on 30 minute bars. I found in fact that there was but the margin was relatively small (about 10 ticks on average with a fairly large variance). But having seen this pattern over and over at key turning points, I realized there must be something more to it so I continued my research and investigation. There is no doubt about it, this pattern is one you want to know and it is pretty simple. Here is how it works:
For an up pattern you will have a new low and a close that is higher than the last close. For a down pattern you will have a new high and a close that is lower than the last bar.
Both ways this is a form of an encompassing pattern around the previous close and this pattern is also called a key reversal. Here is a chart with dots to show the pattern:
Note how this pattern can identify good trade areas. However, in testing it is less than random. So I decided to try to identify various contexts where this pattern would work with controlled risk and found it works better in various counter-trend situations. Of course this is exactly where Ned was suggesting to use such a pattern in his article- at excess.
What I found interesting was I was able to actually code a profitable mechanical system off this pattern in a short time based on the above ideas. Here were the results:
Not the greatest system in the world but it does show a bias with about 53% winners and an average trade of about $90. Over the test period of about 4 years this system gained a lot more on the long side by taking advantage of fast bull rallies using the pattern off new low areas but where the market has leveled off. I did not present this system to convince you of a good system but only to show there is a long term beneficial bias for this patterning.
During my research I also imagined another pattern that would capture missed patterns of the above specific case but I have not yet tested it fully. In this case, you look for situations where bars have not closed in the middle of their range.
The pattern here is the previous bar closed in the bottom third of its range, made a new low and then closed higher to buy. For a sell, the bar closes in the upper third of its range, makes a new high and then closes lower on the next bar. Why did I come up with this pattern? because sometimes the Bull and Bear hook pattern misses the move if it does not take out the low or high. The disadvantage there is markets that are in a strong bear or bull mode will often fail to to a high or low retest and this version of the pattern can actually trigger on an inside bar:
Thirty minute lower and higher closes in the right context are successful trading patterns. Give some thought to this and share your findings in the Oil Trading Room.
That's all for now...