Thanks Ehorn, and I hope I didn't come across as defensive or facetious. Far from it, I really did find those charts very instructive, especially with the added illumination provided by Bi9foot's recent post. I'd seen them before, but more or less glossed over them, whereas this time, I felt I had a few more hooks on which to hang the concepts they presented. So the relationship between volatility and volume underpins the basis for dynamically shifting PACE values. The next mental hurdle I'm trying to vault is it's role in Analysis, and how that melds with the information provided by Overlap to enable "grooving on adjacent-bar knowledge acquisition". Your posts are not a distraction, but rather informative sign-posts on the excellent route Jack has laid out before me. I've had some outstanding guides and teachers over the years, and I fully agree with you that as a pedagogue, Jack is right up there.
Relatively un-busy period at work, so good time to think out loud. Volatility and Overlap, quasi-analytical tools that enable grooving on adjacent bar knowledge acquisition. Initially, I thought it was all about degree of overlap, until I was given the following clues: Overlap is a TIMING issue⦠What is important is what is NOT overlapped⦠⦠which I misinterpreted and starting barking down a wrong tree. The vocabulary was later provided in the form of seven symbols. Volatility, is measured in tics and intimately connected with PACE. Both of these are displayed in the look-up table. So overlap, volatility and PACE provide information that is crucial when hunting traverse FTTs, but which is extraneous when trading just channel FTTs. So at the moment, Iâm reading these three tools, collectively, as markers of progress, where a Failure to Traverse is synonymous with a Failure to Progress, and a deviation from an expected and repeating pattern manifests as a notable absence, or What Wasnât That. Price accelerates from point 1 to point 2, shuffles across to point 3, and takes off until it runs out of juice, then the cycle replays. So volume increases as price moves from point 1 to 2; there should be a concomitant increase in volatility, and decrease in overlap. Volume dwindles as price moves to point 3, characterised by decreasing volatility and increasing overlap. And at point 3, volume surges, price takes off, volatility increases and overlap once again decreases. A resistance level approaches, at which volume, price, volatility and overlap can be used to gauge whether a range expansion will take place. If so, they can then be used to determine if price will reach the LTL. So, if the fuel tanks run dry while price is still moving, it should invariably stop dead in itâs tracks. Likewise, if itâs full steam ahead but volatility uncharacteristically drops, or overlap unexpectedly increases, price could be signalizing an intention to reverse. But hang on, I remember reading somewhere (I think the Channels document) that noise increases as price approaches a trendline. Increasing noise should correspond to decreasing volume, and vice versa. Clearly, more thought is required...
Left to right on the scale would be increasing overlap in points, if I'm not mistaken. As you move down the chart it to lower volume levels it shows more overlap days. The negative numbers I would think are gap days.
Another conundrum, that recently resurfaced, was representing dynamic variables with a static log. This was most apparent during early bars, which often compelled me to complete two rows per bar to reflect the varying intrabar landscape. Bar 49, the (?double) IBGS OB was also a stirling (and costly) example, yet one which in the right hands could have turned a handsome profit. So figures canât be used. Comparative relationships need to be described, hence I-R, D-R etc and the various overlap symbols. Therefore, volatility canât be expressed as a fixed percentage because it is a dynamic value. But on the other hand, surely just increasing and decreasing canât capture all the relevant information. Deducing the complete role of volatility should reveal the requisite vocabularyâ¦
Presumably two reasons for this: 1) the increased sensitivity of price to volume on the traverse level and 2) the need for greater precision. After the first several bars, I switched to recording M values at the 3/4 mark of each bar, in tandem with the colour change of the bar countdown timer. But perhaps this is part of the meaning behind Overlap being a timing issue; i.e. that once the level of overlap is unlikely to increase further intrabar, the trader records M and then moves to A. Speaking of MADA, a couple of the unnecessary losses were due to doing M<strike>AD</strike>A, like a lizard in the headlights.
As I watched the prv surge and then fall intrabar, I often wondered at what point I could record the final result. Several times an initial burst of volume would spike the prv to over 200%, only for it to progressively fall, and the final volume be less than the prior bar. The quoted statement would suggest that the reverse does not happen, though perhaps that's just one specific context. For certainly, I've seen many bars that begin with paltry prv, only for it to explode to new heights as price breaks out of a formation, or makes a HH/LL intrabar. May seem like I'm stalling on doing the meat and potatoes of the debrief, but I'm just trying to crystallize a few thoughts, and had two staggered shifts today, with more of the same tomorrow.
Annotation is crucial to every other process, and while I'm annotating more thoroughly, I'm still not annotating correctly on a consistent basis. As far as traverse construction goes, I'm riding the pendulum, rapidly going from paucity to surfeit. The first four bars of the market day were amongst those that I missed... had I seen them, I would have annotated thus... <img src=http://www.elitetrader.com/vb/attachment.php?s=&postid=2156579>
I've been reprimanded by a well-wishing observer for inventing. I sincerely hope none of the above posts are classed in that category, being simply a record of my thinking process and an informal prelude to the next stage of the debrief. Due to scheduling/time constraints, I've not had a sufficient span of time in which to complete the debrief. As I'm only working for eight hours tomorrow, I will have ample time.
For those with an interest, here is a bar overlap chart using 5-min bars based on 5 days of data late October. The last few bars of each day were eliminated due to the high volume spikes at the EOD from closing of positions. The pace levels were a bit off (5 days for the chart, but the pace levels are calc'ed over a longer period), so I added another extreme level to even the spread rather than re-sort all the numbers. - EZ