I thought. I pondered. I looked at the true formula (at least in it’s beta form). I can’t figure it out. If we bend the “purity” for a second, there becomes a good reason to to knock off 3 more. Anything that involves the open. I assume the logic would follow something like saying we don’t care where price begins (present), we care about where it’s going (future). Thus the only candidates left are:
2H-2C and 2C-2L. Together these form the complete bar with the Close as the anchor. Coincidence? Quite possible. The “actual” VP is made from the latter, 2C-2L. I created my own variable, Vega Sec (VS) which is the supplementary value of VP.
VS=2H-2C (or 2-VP)
Next we have a value that apparently has huge implications called the Distinct Vega, which is equal to the average of the moving average of the Omega, multiplied by the VP. Hm. I don’t quite understand the multiplication aspect of it..
Nevertheless, I threw in the value and started looking at it. played around with the data a little bit, until I noticed something a little weird.
I took the row of values from the average of moving averages (so I guess it’s really representing the moving average of the moving average of the range H-L??). I averaged it. Next to it, I subtracted the current value from the previous value. And took the average of that. Answer? 0. ??? I don’t quite understand why. I’ve never been great at math so I’m very likely to just be missing something, but it was an interesting find. It can be done with any data set, as far as the rand() function is concerned. I would think that with a set of random numbers, the difference between each number (cell2-cell1) would be random as well. As a result, the average of the differences would be random. Not the case. It’s pretty much always close to 0. With this in mind, I took a look at those values, and graphed them alongside the MA.
On the top is the AMA. On the bottom is the difference between each period in the AMA. The curved style is showing some minor points of interest in where the bounds of the chart are. The 0 line is clearly the popular zone, and any points heading to .0004 or .0006 are rejected and shoot back for the 0 line. This is someone helpful in the vacuum that I have a better idea of when the AMA will turn, meaning I have a slightly better idea of when range will contract/expand. The downside is that these numbers are quite small ( I think forced due to 20 periods). Perhaps adjusting some things will allow me to better predict when range will contract or expand? This is just 1 part of many pieces to form pieces to form a model, but every piece must be perfected if the model is to be of the same quality.