To date, I have read enough on the forums to not trust the NT Strategy Analyzer's results (largely because NT7 lacks the "look inside bar" capability that TS apparently uses). Therefore, to date, I've elected to test my ideas by building indicators with entries, trade management, reporting and summaries built in via output window and screen prints. As you might expect, although this approach allows me to be in better control of how my entries and exits are managed (through coding in a 1 minute bar series--which still isn't 100% but is certainly better than hourly, daily, etc resolution), it is decidedly not a high throughput factory--and walkforward testing would be like a bludgeon to the head. Therefore, instead of brute force indicator coding to validate my trading ideas, it's time for me to CORRECTLY learn the NT7 strategy and walkforward capabilities--especially if I am going to scale up my strategy development factory (ala kevinkdog).
For my first foray into NT strategies, I first downloaded the NT7 strategies that sam028 and baruchs coded (from here (). Again, thanks for Sam028 and Baruchs for their willingness to code this up.
Second, I went and found as much CL data as I could, which was from this post in the Elite Downloads section ( and the CL historical tick thread (. So this is the data that I used to run my analysis.
I then opened up the NT Strategy Analyzer window and loaded the nexusfi.com (formerly BMT)Baruch2 strategy in the Walk Forward Optimization with the following settings:
136243
I then ran my walkforward test, and lo and behold I received 11 walkforward periods with the following corresponding optimized parameters:
136244
For completeness, here are the WF parameters that Kevin had Baruchs include in his strategy (I am assuming that Kevin generated these WF parameters via his own WF analysis):
A summary look at my WF results show that the resulting parameters are fairly similar to Kevin's WF parameters, however the run I did was just my initial run and I am interested more in process than I am in specific results at this point. For instance, I could have messed around with the dates to get the walkforward periods to line up exactly, but each run took my computer about an hour to do, so I'm not worried about that at this point. What I am worried about is making sure that I am using the NT Strategy Analyzer's Walk Forward features correctly--and that I understand what is happening well enough to have confidence to use the WF tool in my …