DislikedFlaws to every approach Hmmmmmm I am not going to try and say Murrey as hit upon the Holy Grail, far from it, but if you take a little time to think about it, he is always right. Why do I say this, easy , he does a chart shift when he is wrong, as Parisboy would say think Russian Dolls, so in my humble opinion you must look at more than one Octave setting when deciding to trade from a smaller time frame OB/OS area, in short, chart shifts are telling you if a trend is in place. I keep reading Buy low Sell high, Buy support Sell resistance, sounds...Ignored
A 2000 discussion on a "Murrey Math software forum" about the programmation of a "chart shift " within Murrey's EOD software. Hope it will help :
Brian, At the seminar in Beverly Hills THM talked about how he determined the high and low of the frame .......he said for a 16 day frame he used the range of the last 32 days...and if prices closed past 10/8 or -2/8 he would adjust the frame. He also mentioned something about the prices going past 8/8 and 0/8 for several days to adjust the frame. Steve ...the computer guy was in the front row. I know he is with this group or the other. Also try agent moulder he's good at details.Ron
Yes... Murrey says 2X. That's not what the s/w does though. I don't know if Murrey is intentionally misleading the audience or if Werling has done something different than Murrey thinks he has. Bruce
I have the feeling that the true answer lies somewhere in between. I think that Murrey's software looks back up to two times the time period chosen and select only the first significant high and low that matter. If there is another significant high and low before the one that it picked, still within the two times the period that was selected, it ignores it.
If I had more details I would not say "I have the feeling", but I would say it is. Anyway, my feeling is:
- say you set the frame to 16. The software should look back 32 and see what significant highs and lows are present. Based on an algorithm it would then determine what is "significant", and then choose the closest "significant" to the present and drop the one before it, even if the one before is more "significant". Murrey keeps saying that the last high and low are the ones that count. But in fact they have to be significant highs and lows. It does not really matter where they are located, as long as they are the LAST significant high and low.
Brian's software now looks at the high and low within the designated period and adds lines based on the highest high and lowest low present in that period. I don't think that is the correct way of doing it. Adding a range finder and then computing the "significant" highs and lows, then choosing the last "significant" high and low are the way to go, IMHO. One should be easily able to compare this thought by comparing the daily charts from Murrey's software and the daily charts with Brian's software. You will probably have to go through fifty or more charts, and see what settings in Brian's software will equate with Murrey's daily chart set, say, at 16 periods. Do not change the periods in Murrey, but change them in Brian's to obtain the same results as in Murrey's. Use charts that have good price excursion.
I am planning to do just this as soon as Murrey sends me the latest software that will read data after last December. I hope this explains my "feelings". Cheers. Bijan
In recent versions I added the 1.5x & 2.0x multipliers (eg if using a 64 frame, then 2.0x would use the extreme high/low in the last 128 days).
A recent post by THM lends credence to using ONLY "natural" or 1x settings though...
The public post was dated 2/1/2k and titled "MMTS: OEX Support off current lows:
Downside almost over: Reverse soon" he shows the OEX in a 64 frame (8 days per 1/8th time interval).
To make XChartMM "match" his graph, the 64 frame setting is used along with the 1x multiplier - eg, NO multiplier.
I've heard this from Murrey before. The issue is that his s/w doesn't work that way. Remember that Kristof Werling is the one who created the code and what he's done may not match THM's desires. What the s/w does is 1.5*N+1, even in 64-bar frames.
That aside, you have all the bases covered, so it's not a problem. Bruce
Wait - what I'm saying is that if Murrey's prog really used 1.5n+1, then the output of his recent OEX graph *should* be different, because the price extremes were larger at 1.5n than at 1.0n.
In other words, Murrey's OEX graph contradicts the 1.5n+1 scenario for all intents and purposes
Guys, I think this may prove my point that regardless what time frame you put yourself in, the software looks for the most recent "significant" high and low to calculate the frame size.
Another chart which implies that 1.5x is the one to use as a general rule (this one is frame size 16).
On 1x, the high and low in 16 days do not compute the frame to match THMs output.
I agree, Brian. The only thing I've seen besides 1.5X+1 is a shorter display that seems to be reacting to a rule they have about significant lows/highs. The ABTXQ chart does that after a whipsaw in the rally that precedes the displayed portion of the chart. I have to assume the Murrey s/w looks for reversals off intermediate highs and if they're greater than X%, he says it's significant. I don't recall reading about that in the book, but I think it's in the s/w.
Bruce
Murrey's EOD s/w currently uses a greater number of bars than the frame size to test for significant highs/lows. The number of bars tested is 1.5*N+1 where N is the frame size. So for a 16-bar frame, the number of bars tested for significant high/low is1.5*16+1 = 25 bars. So a 16-bar frame, if allowed to display unassisted, will usually show 25 bars of data, the 32-baar frame will display 49 bars and the 64-bar frame will display 97 bars.
Murrey's s/w also has a whipsaw detector (of sorts) that tests to see if a whipsaw is sufficient to constitute a significant high/low. If the detector gets activated, the number of bars displayed will be less than 1.5*N+1. Bruce
Pretty simple really. With the frame settings of 4, 8, 16, 32 & 64; the program finds the highest high/lowest low in the last X bars, where X is:
on "1x", or natural, frame setting x 1, eg 64 is bars
on "1.5x", it is frame setting x 1.5, eg 64 is 64 * 1.5, or 96 bars
on "2x", it is frame setting x 2.0, eg 64 is 64 * 2, or 128 bars
I personally use the 2x setting, this seems to match Murrey's program output in the majority of cases. Brian
Glenn, Don't worry, I've hardly got pretensions about setting up a new trading religion - I'm just groping around for knowledge and
understanding.
Regarding violation of math rules, I thought it was one of Murrey's rules (underlying premises might be better) that the frame is set by the fractal - the fractal is set by the "relevant" inflection high/low - not highest and/or lowest in the look back period even though this is how the s/w does it. If my understanding and description of the rules is correct (and please correct me if I'm wrong, I'm just new at this game) then its Murrey's s/w that breaks its own rules and premises. Would you like to comment on this? Thanks, Henry.
I have been using MM as one of a number of indicators day trading the NAZ e-mini. On balance I find it very useful for exits, not as good for entries. Regards Dimitri
With the frame settings of 4, 8, 16, 32 & 64; the program finds the highest high and lowest low in the last X bars to build the frame.
For example, using a frame setting of 64, X number of bars back is:
- 64 bars on "1x" (or natural) frame setting
- 96 bars on "1.5x", (64 * 1.5 = 96 bars)
- 128 bars on "2x", (64 * 2 = 128 bars)
It is up for some debate as to which one is the correct one to use. Several members have commented (summary of all) "I've never seen any chart from Murrey's s/w that has 2*N bars... only 1.5*N+1 or fewer. Murrey SAYS to use 2*N in his classes, but that's apparently not what his s/w does. You'd have to use the search mode to get that"
I personally use the 1.5 setting and it has matched the majority (95% +) of the charts I've seen as output from Murrey's own program. Hope this helps, Brian
1
