Yes, his numbers differ from mine too and I try to match MP with VP. I think he's deriving his numbers from a multi-day profile since the mkt has been trading in this range for quite awhile.
Covert, It's possible I grabbed numbers from the wrong column I guess. Which market are you looking at specifically? b
Hi Bolter, Great thread. I started studying MP last Nov so your thread has been of great interest to me. One question I have is regarding neutral days (defined as range extensions on both sides of the IB). Do you attach any significance to them. The CBOT manual maintains they frequently signal market turning points while Dalton says to watch the close on a neutral day relative to the VA for an indication of strength of either the bulls or bears if the close is at one extreme, or continued neutrality if the close is in the VA. I have seen others maintain a neutral day is usually followed by another neutral day. Do you have any particular sentiments regarding neutral days? Thanks again for a great discussion.
Bolter... the value areas represented here are obviously not based on Friday's profile. How far back did you combine your profiles to get these levels? I'm trying to follow your logic. Thanks.
Student, UVA, POC and LVA always come from previous days MP - but bear in mind I use volume not TPO's so they will differ from standard MPs. I use multi-day VP's only for identifying S/R and vaccuums. Later
Possibly because my calculations are incorrect I've only put together the code in the last few days. I cross checked against Bolters and some Esignal charts that were posted (which are also not exactly the same) and my figures were close but not the same. The levels I posted were for the whole DAX session. I've just calculated them for the cash market (9:00 - 17:30) and they are much closer to the CQG figures: POC: 5731 UVA: 5748 LVA: 5720 I'm just using 70% rule for UVA/LVA - 70% of TPOs lie in value area. Distribute the rest equally above and below. Should I use one standard deviation. or something else ? IB might well be wrong. I will check this. One other possible source of discrepancy is using the close of one minute bars. Clearly the distribution of prices will be different in a fast market with long 1 min bars as compared to tick data. I was going to use shorter bar size for the TWS backfill, but I couldn't get it to work. Despite the doco, it seems that they just don't work. And TWS/IB data may be different, but not THAT different unless there is some problem somewhere. Anyway, I would like to get this right, so any comments appreciated.
Not sure what the problem is. I know that when I tested eSignal last year, their data was often incorrect. They often lumped size together on the tape (I use T&S heavily in my trading).