Note: Big Mike I tried to attach the Bloodhound template, but since it's a .bht file it was considered invalid.
For any of those wishing to have this Bloodhound template, PM me and I could send it to you by email...
I’ve been a nexusfi.com (formerly BMT) member since December 2011, but it wasn't until around April of this year that I really began using the forum. I had initially downloaded some indicators, but other than that I couldn't care any less for spending time on forums; I was always partial to books. But after having read through some of the countless threads and having watched many of the most excellent webinars, it’s now obvious to me that I've allowed such a precious resource to pass by virtually unused for over a year. I had previously thought of forums as nothing more than “time-robbers”, but now that I have benefited so much from the endless treasure trove of content here over the past several months, my feelings toward forums, and especially this forum, have dramatically changed. And so now I think it’s time I finally started being a contributor rather than solely a user. I had considered starting a journal, and I’m sure I will eventually do that, but I would first like to share a bit about some of the things on which I've been working. I have only just begun to learn NinjaScript, but I am quite proficient with the visual programming software—Bloodhound. So if I were to contribute anything of any real value here (for now, at least), it would have to be concerning my work with Bloodhound.
I recently watched Linda Raschke’s webinar on Volume/Price Relationships, (which is brilliant, as is the follow up webinar . She inspired me to try and solve within Bloodhound for this volume divergence concept on a 1 minute chart. I, like many of you, have been using this concept for years, but I used it mainly on Daily and Weekly charts (a perfect example would be the March 2009 lows on the S&P500 weekly chart), never on 1 minute bars. What I have come up with may not be the greatest way of doing this, but it could serve as a good starting point on which to build.
Let me try to sum up the logic here as simply as I know how. The initial event that I solved for was the Volume Spike. I know Linda talked about using a standard deviation of 4 or 4.5, but here I’m using a standard dev. of 3. I have built a custom session template in which I eliminate the first minute of the session and the last minute before 4pm EST. So the template runs from 9:31am EST to 3:59pm EST. This eliminates the giant outliers caused by nothing more than the fact that it was the first minute and last minute of the day. It also eliminates the very low volume bars between 4:00-4:15pm EST, and provides for a somewhat smoother moving average (SMA,100) and StdDev. To qualify, this Volume Spike must also be making either new highs or lows within the 60 period Donchian Channel, which simply means that High[0] or Low[0] will be equal to the Upper or Lower plots, respectively. I then extend this signal forward 60 bars with a Reset Condition on Opposite Signal. While many of the signals come in within 15 to 30 minutes of the initial Volume Outlier event, I chose 60 minutes (as this is designed around a 1 minute chart) as some of the signals do come in around 40 to 60 minutes after the initial event. I then eliminate the first 10 bars of this signal, since I want there to be at least 10 bars between the initial Volume Climax and the Volume Divergence Signal, otherwise there would be many false signals. This signal is then Reset if Volume[0] > the maximum Volume of the previous 60 bars.
For the Volume of the actual Divergence signal, I am looking for any of the following events within the anaBetterVolume indicator(which can be found here: : Breakout or Climax, Churn, Climax Churn, Weak Breakout or Climax, Weak Churn, or Weak Climax or Churn; OR if Volume[0] > 2 times Volume[1]. AND Volume[0] must be > the VolSMA(100). These volume events I then extend forward 2 more bars, to be sure …