Also just a detail in case the NQ looked more spotty. It is due to the tick change that occurred within the dataset, for some of the more keener folks who may have wanted to jump on my case. The 0 volatility bar is also correct. Where else would that bar occur except in VDU... I'm sure there are still many lurkers who will argue the other way around and say price leads... Well do the projection and see that the PRV would have told you to stay out since it is the ultimate risk territory. Then again you could have just held until volume kicked in and then analyze whether you were on the right side. In any event, I know there are alot of lurking entry/exit type of folks and thus the tables were presented to gear towards a get in circa the open and get out at the end of the bar... However, it is not that difficult to comprehend the continuous form of this discrete table... Regards, MAK
Mak sundays are your most productive days I notice - you tend to emerge from your lab with findings and post totally thought provoking trains of thought - very much appreciate your detailed comments. multiple points 1. thanks for the grid of the 5 min ES probability/projection - you referred to this in the past and I actually did a lot of analysis back then but this time you added some key additional comments that identified the odds of a decent move in price vs the lower ended scale summarized by the 50/50 comment. The FOMC days now make a lot more sense and at the risk of rates going up another .25 point - I look forward to the next FOMC meeting with new keen eyes in place. On that note - I am also watching the NQ most days so yes. for the record somebody is interested. 2. In ref to my logs - all your comments are helpful - in particular - Failure to STR/SQU prior to a spike adds a lot of clarity and calls for re-calibration. I had to read the post a few times to make sure I was clear and I believe I am now. The chart Grob kindly posted on the pairs/str/squ way back when is very familiar to me - I have it printed for some time now and I assumed you had derived your findings from that very chart ! Interesting to learn it was new to you. I was watching the DOM this week based on that and your description but it all ties in like roads to Rome. Failure to STR/SQU is exactly how I need to look at this live - I was not looking at it from this perspective but it makes sense as an orderly chain of events to observe. All of these are pieces to the overall puzzle. In a nutshell, my recommendation is to conditionally (if statements) capture the {SQU/NEUT/NO STR}=>{LONG SPIKE} and the {STR/NEUT/NO SQU}=>{SHORT SPIKE} state. These two conditions trigger your recenter short enough sentence but potent and complete - thanks 3. no point three for now .. much to consider and a whole new xls to play with in the morning sincere thanks again for on going sharing and feedback - very much appreciated IRL
Elegant solution. Well done... Tiny technicality, for you IB folks, make sure when you do the above replacement that you use B19 with the "=" sign. The other two should be without the "=" sign. The rest should just pipe along... If there is concern about synch, the atomic clock solution is overkill. If you want to eek out ahead, set your system clocks to be 5 seconds ahead of your datafeed's clock. Perfect synch with the datafeed is unnecessary. It might actually be more beneficial to have slight offset. Actually now that I think about it, a 90 sec offset might actually synch the noise with the actual beginning of the actual 5M bar... You can do this due to the continuity of markets despite have discrete price increments... Perhaps this is too much detail... MAK
Mak, About nononsense's comments being misguided, have a crack at the following SCT excerpt posted by your great guru master: Trades continue on the short term trading level by comparing the progress of the current bar with the prior bar as a trending reference bar. See IF 1, IF 2 Report. When IF 2 is invoked a reversal is made. It is possible that a subsequent wash and reversal may be required. When this comes up, the wash and reversal is done and the monitoring is done for the remainder of that bar on the 1 min chart using the IF 1, IF 2 strategy as applied to the 1min bars. This converts a 5min bar strategy to an intra-bar strategy. At the end of the remaining portion of the APA invoked on a 5min bar, the monitoring returns to the 5 min bar thus ending the intra-bar strategy. The fractal change is based upon statistics that predicate that any loop filibration on one fractal can be terminated on the next faster fractal using the same rules. This is an abstract statistical method based on Alexanderâs Method for least connecting systems for optimum system operation. A concurrent theorem on the migration of the market operating point and what the controlling facet is, is what establishes the market as a system of interrelated cells and how that connectedness may be optimized. One of the dimensions of the operating point matrix is the set of fractals. After you're done with this exercise in mental sanity, we'll talk further. nononsense
I'm unable to get this to work with Ireland's changes. I'm seeing the volume come into the spreadsheet, but the actual volume, and prv volume and fractal (cells B3, C3, and C4) will flash with a value briefly and then return to 0. I've rechecked the changes Ireland recommends, and they seem to be correct. I've also tried Maks syntax with the same results. Did other IB users get this working ok?
I just got it to work also and have the same results as you. It seems as though the time in cell C4 keeps resetting to 0.
Try this change to cell C4 =(NOW()-TODAY()-OFFSET(DATA!A1,DATA!E1-1,0)) That should get the fractal setting correctly. however, there is something going on with the volume calcs which I haven't figured out yet.