Registered User
Joined: 4/18/2005
Posts: 4,090
|
How can I convert this PCF to do the same thing but for the 20.40, and 50 Exp ma's?
((ABS(XAVGC3 - XAVGC8) + ABS(XAVGC5 - XAVGC8) + ABS(XAVGC10 - XAVGC8) + ABS(XAVGC12 - XAVGC8) + ABS(XAVGC15 - XAVGC8)) / XAVGC8) * 100
I would think it wouldn't be as long a pcf right??
|
|
Registered User
Joined: 3/21/2006
Posts: 4,308
|
Scott this should work for you. I have had to amend this formula a few times myself..
((ABS(XAVGC20-XAVGC40)+ABS(XAVGC50-XAVGC40))/XAVGC40)*100
|
|
Registered User
Joined: 4/18/2005
Posts: 4,090
|
thanks..
I responded to your "where's Scott" questino on another post...
|
|
Registered User
Joined: 12/21/2004
Posts: 902
|
I think that the formula is different from that. In the original formala, he uses 8 as a proxy for the midpoint between the different series, then he just sets up the difference between each ma and the "8"
midpoint.
So using Scott's parameters, the new formula would have 35 as the new mid-point proxy and would be ((ABS(XAVGC20 - XAVGC35) + ABS(XAVGC40 - XAVGC35) + ABS(XAVGC50 - XAVGC35)) / XAVGC35) * 100.
Alternatively, the 8 was half the 15 (the largest ma), so we would use 25 instead of 35 in the formula.
Or maybe I'm wrong....
|
|
Registered User
Joined: 3/21/2006
Posts: 4,308
|
Scott can you share with me your views on the significance of these three PMA's converging.
I did a little bit of testing and do not see anything yet...
I read the thread you wrote. I had no idea that you were such a craftsmen...
|
|
Registered User
Joined: 4/18/2005
Posts: 4,090
|
Well I had no real clues to anything... other than that on the HNC template bigger moves seem to come after convergences of the longer ma's. Also possibly a way to find pre-springboards. the consolidation needs to be gonig on for a while.
I'm using the 20 ma allot now to keep a sense of value... and when price is between 20 and 50 I start looking for the entry signal in the direction of teh original trend. But I have seen some times that trend reversals and bigger moves can come from their convergence. Which makes sense as the 20 crosses down through the 40 and 50. The 20,40,50 are actually martha settings that she suggets using as s double check for a stop loss... that allot of the Crowd tends to use ma's as stop loss areas. To be hones in some cases they seem to work.
On allot of charts I see visually that price can bounce along the 20 sometimes diping below it. I intended to play with the convergence of these fewer and longer mas in connection with a longer ma for trend confirmation and check out it's potential for catching longer market swings. So the idea is kinda like the HNC method but "Zoomed out" a bit. I have tended toover trade a bit and it has gotten me nowhere I'm trying to trade less and capitalize more signifficantly on slightly larger swings, trends, rallies what ever. I also some times add an even shorter ma, with stocks that have been moving with momentum.
It seems that most of my sucess in 06 was with following trends. It would be nice if I could poast my trade history here via attachment....but I can't so i'll spare the itemized detail. I did catch a couple of bottoms but most of my trades were with already trending stocks. I think I have learned enough to peg probably bottoms and throw them on a watch list. Begin trading when the 20>50 and they fit Dicemans criteria for recovery, accumulation, or bullish phases. Less focus on the recovery. I may still dabble in the ocasional divergent bottom .. but for now I'll be looking for them in intermediate correctinos to the longer term trend.
One example of what I THINK i'm looking for is RTN. while yes it looks nice on the HNC template I already have the three ma's on a hot key template that is less cluttered. So for example add a 100 ma and in late jan you have the faster mas close to each other and price nestled down in there as well, in a value zone... then a decent move. Also NOvember and April. Its just a bit less cluttered and the same concept on a larger scale. Of course other indicators are used to try to determine the direction of the break.
SCT was one that I found that met the criteria. I originally found it on the HNC scan I created but found that my F9 hot key showed a similar thing.. and all mas were converging on the HNC template making it difficult to see price. I still use the HNC and think I will use it heavily but I tend to focus more on candles and I'm trying to figure rules for the context in which I want to find my favorite candle entries.
This was one I wan't to play with.
|
|
Registered User
Joined: 3/21/2006
Posts: 4,308
|
I like Hohandy's idea of using a MACD-H 12,24,4 In the Middle window and on top of that I use a Stochastics 9,3,6
Look at the set-up for RTN on February 1rst and April 3rd (you do not see that every day)
|
|
Registered User
Joined: 3/21/2006
Posts: 4,308
|
|
|
Registered User
Joined: 12/21/2004
Posts: 902
|
QUOTE (apsll) I like Hohandy's idea of using a MACD-H 12,24,4 In the Middle window and on top of that I use a Stochastics 9,3,6
Look at the set-up for RTN on February 1rst and April 3rd (you do not see that every day)
I don't think that's my suggestion Apsll - doesn't sound like any MACD format that I use. I do have Tobydad's TSV26bb13 setup (which I really like) in my lower window though. But I don't want to steal credit for something good away from someone else 
|
|
Registered User
Joined: 3/21/2006
Posts: 4,308
|
I thought that it was you that suggested using a MACD-H with the HNC indicator? You are right about the settings those were mine, and the Stochastics were my idea also...
|
|
Registered User
Joined: 3/21/2006
Posts: 4,308
|
Scott - I've been doing a lot with the HNC/Diceman decision point setup over the weekend - and your PDS showed up as at the decision point. And this morning it's up nicely. Did you ever go in?
I've been thinking about combining the MACD-H analysis with the decision point.
Yes it was you, I quess I took you out of context. So I will take credit for the idea, but you inspired me...
|
|
Registered User
Joined: 4/18/2005
Posts: 4,090
|
APSLL.. what is the custom indicator in the bottom window?
|
|
Registered User
Joined: 3/21/2006
Posts: 4,308
|
I have just designed a system that I am going to test but I will share here first.
Sort any watchlist of your own design by the Diceman indicator and flag all the cadidates that are at the decision point. put these flaged stocks in a seperate watchlist and un-flag all.
Now sort this list by a 121 day linear regresion line (Visual Slope) and discard any with a negative slope..
Next in the middle window use a MACD-H 12,26,4 (I have an indicator that will tell you when MACD-H has just crossed up over zero (Percent true) -
XAVGC12 - XAVGC24 - XAVG(XAVGC12,4) + XAVG(XAVGC24,4) > 0 AND XAVGC12.1 - XAVGC24.1 - XAVG(XAVGC12.1,4) + XAVG(XAVGC24.1,4) <= 0
Now sort by this indicator (actual value) and you will be left with a handfull of stocks worthy of the example I used above RTN
I have this example as well NCZ (have fun, I know I will) -
|
|
Registered User
Joined: 12/21/2004
Posts: 902
|
wow - that does look like fun - OK - I'll take the credit
this will be fun to play with - let me ask - why is the MACD setting to 4 rather than 9?
and going back to my theory (which I will rejoin on the other thread in a minute) I wouldn't necessarily count this as meeting the price/MACD-H high criteria either - the high of mid-April was retested the next day so that MACD-H, IMO, is done for predictive purposes.
And sure, I am willing to listen to someone say otherwise - I thought long and hard about this, because it seems to negate a bunch of good candidates and it would be good to see examples of how some of these slight variations work (considering that the MACD-H of the mid-April high isn't actually higher than the one preceding it at a lower high price point). Maybe it will turn out that the better signal is the MACD-H crossing the 0 line rather than the price/MACD-H signal.
IS the bottom screen the 2 day RSI from the Stoop and Scoop setup?
|
|
Registered User
Joined: 3/21/2006
Posts: 4,308
|
The bottom screen is first Dicemans indicator -
((ABS(XAVGC3-XAVGC8)+ABS(XAVGC5-XAVGC8)+ABS(XAVGC10-XAVGC8)+ABS(XAVGC12-XAVGC8)+ABS(XAVGC15-XAVGC8))/XAVGC8)*100
And second my indicator for MACD-H crossing zero (Wich Bruce-L shared with me, thank you Bruce) -
XAVGC12 - XAVGC24 - XAVG(XAVGC12,4) + XAVG(XAVGC24,4) > 0 AND XAVGC12.1 - XAVGC24.1 - XAVG(XAVGC12.1,4) + XAVG(XAVGC24.1,4) <= 0
Hohandy, I like the (4) because it reacts quicker than the (9)
We will both get credit for this one (first a lot of testing will be needed)
|
|
Registered User
Joined: 12/21/2004
Posts: 902
|
Question for you about your indicator that you got from Bruce-L - if the MACD-H settings are 12,26,4 - I see the 12 and 4 in the forumla, but I see 24 instead of 26 - is that right? And it's just a visual thing in there that spikes up when MACD-H crosses zero from below?
|
|
Registered User
Joined: 3/21/2006
Posts: 4,308
|
Good eye Hohandy you are correct. That is a MACD-H 12,24,4
The visual thing that spikes up when MACD-H crosses the zero is the result of the percent true indicator.
|
|
Registered User
Joined: 12/21/2004
Posts: 902
|
Got another question. I missed out on the original HNC/Diceman decision point thread - I've since read through it and enjoy playing with it. But maybe this was discussed in later on in that thread that I missed, or in a subsequent thread about the decision point:
When I sort by the Diceman formula, I sort by visual value - it assigns a 0-100 number and those at 0 are at the decision point (obviously). 100 equate to a very confirmed trend (obviously).
Is there any disagreement as to how low a value assigned has to be in order to be "at the decision point"? Does it have to equal 0? Less than 10? etc?
Is there a consensus?
|
|
Registered User
Joined: 1/28/2005
Posts: 6,049
|
The original intent was to sort by "actual value".
The lower that value the tighter the moving averages
are.
Ultimately a "decision point" is an area of low volatility
but has no exact definition and will probably change
with different stocks.
(that was why I did not give an exact definition)
Thanks
diceman
|
|
Registered User
Joined: 12/21/2004
Posts: 902
|
QUOTE (diceman) The original intent was to sort by "actual value".
The lower that value the tighter the moving averages
are.
Ultimately a "decision point" is an area of low volatility
but has no exact definition and will probably change
with different stocks.
(that was why I did not give an exact definition)
Thanks
diceman
Something's not right here Diceman. I sort by actual value, but since each stock isn't in the same scale, a "2" reading may be near the top and a "5" reading may be near the bottom. Doing it by visual value, it sorts by distance from the bottom regardless of what the actual value is.
So is this a scaling problem - should I be scaling my screen a certain way to get the desired results from actual value?
|
|
Registered User
Joined: 1/28/2005
Posts: 6,049
|
Not sure exactly what you mean by the 2 and 5 near
top and bottom.
The equation takes the sum of the differences of the
MAVs (in absolute terms) and divides by the 8
mav (to adjust for different priced stocks).
So the value is the percent difference of the MAV's.
(the lower the value the tighter the averages)
The scale you are viewing the price chart on or the
price should not matter.
Realize that averages can be tight at tops (cycle tops)
bottoms (cycle bottoms) and at any "flat" price trend.
Thanks
diceman
|
|
Registered User
Joined: 12/21/2004
Posts: 902
|
Ok - I gotcha. The log scaling of the price chart make the prices look close together whenever the Diceman indicator line was at the bottom regardless of its value, so it looked like relative location of the line was more more important than the actual value of the indicator (because the log causes the scaling for the price chart to change each time). But when I set the price chart to all have the same scale, it becomes apparent that it's the actual indicator value and not the relative placement of the line that's important. Thanks, Diceman!
|
|
Registered User
Joined: 4/18/2005
Posts: 4,090
|
Wow this thread traveled fast. Just got back for prenatal coaching classes. Apsll its late but I printed out your method and am looking foreward to playing with it tomorow. Thanks for shareing. NCZ is a beautiful chart.. I love this type of set up.
Question.. do you keep the Diceman indicator on screen to sort by as a final step to bring the tightest decision points to the top first? I use it in a scan and select those that fit my other trading criteria. it seems like having an indicator that tell you the ma's are close while having the Ma's on screen as well is kinda redundant..... or did I miss something?
|
|
Registered User
Joined: 9/22/2005
Posts: 849
|
I am having trouble with both the diceman and the HNC indicators.
Will you please re-post both of them?
|
|
Registered User
Joined: 12/21/2004
Posts: 902
|
Bcraig - they can be found in this thread:
http://www.worden.com/training/default.aspx?g=posts&t=18942
|
|
Registered User
Joined: 4/18/2005
Posts: 4,090
|
I think they are the same... or variations of the same. I have this.. which I think works well.
((ABS(XAVGC3 - XAVGC8) + ABS(XAVGC5 - XAVGC8) + ABS(XAVGC10 - XAVGC8) + ABS(XAVGC12 - XAVGC8) + ABS(XAVGC15 - XAVGC8)) / XAVGC8) * 100
it measures the distance between the Ma's. you can plot it as a custom indicator to use for sorts or plop it in a scan and then select stocks from teh scan to be filtered with other conditions, and put the multiple moving averages on the chart. This is what I do. If i'm managing a large list then I'll plot the custom indicator and do a sort by it.. actual value. But the scan works pretty well... I get more options than I can trade. My scan:
All Stocks
HNC's Ma convergence (above formula) set to rank 11 - 65
BOP 20 (formula is BOP 20)
optionable - true (helps with liquidity)
Volume 1-Day set to rank 34-77 (might try vol surge today, instead)
Price pershare... your favorite I have it set 2-86 dollars
today the scan gives 182 possibilities.. most need further refining...but i'd rather sift through allot fo charts than automatically miss a great one becasue for some mathematical reason it was eliminated before I saw it.... so I like to leave some criteria kinda loose. It only takes a moment to look at it and eliminate it. Feel free to play with it. It is still a work in progress for me. some times I and in and remove features just playing to see what it brings up.
|
|
Registered User
Joined: 3/21/2006
Posts: 4,308
|
Scott, I did you a disservice by giving you an inferior formula for your original question asked of Diceman.
Hohandy had the better choice to use for your needs.
ABS(XAVGC40 - XAVGC35) + ABS(XAVGC50 - XAVGC35)) / XAVGC35) * 100.
This indicator is what you should use to meet your needs...
To answer your question of me (Question.. do you keep the Diceman indicator on screen to sort by as a final step to bring the tightest decision points to the top first)?
This is the first step in filtering down my watchlist to the final list for good cadidates. I only want stocks that are at the decision point when continuing the other stages of "elimination" Remember what Techni-Trader taught us in this process (you want to find every reson to eliminate a stock from consideration so you are only left with the cream of the crop).
|
|
Registered User
Joined: 12/21/2004
Posts: 902
|
Morning Apsll - why aren't you including the 25ma in your formula?
I tried playing with your new setup last night - I was able to work your MACD-H above 0 formula that you got from Bruce L into a PCF - but I was unable to make it into a custom chart indicator such as you have on your charts. I was getting an error message about sayinmg there was a problem because the answer is being expressed in boolean terms or something along that line. Is there something you need to tweak the formula with to get it into a chart?
I went back and reread Elder last night. I noticed at one place that he qualifies his MACD-H rules by talking about them applying to situations where the price and the MACD-H are at "multi-month" highs (or lows) and I'm wondering if some of the issues that we're having with MACD-H is that we're trying to apply it and make it work in too many inappropriate situations - that maybe we should be more judicious
and only apply them to major highs and lows rather than also including minor ones.
|
|
Registered User
Joined: 3/21/2006
Posts: 4,308
|
About the formula, The construct of these formulas is not my strong suit (I leave that stuff to better mathmaticians than I) My strength I think is in puting them to good use. Finding creative ways to apply them is fun and rewarding.
Remember, I have never read the works of Elder and so cannot contribut any insights to his theories..
I am intriqued however as to what makes MACD-H tick and that is an ongoing case study for me.(It appears that for this Elder gentleman as well). One of the problems I am considering, is the lagging nature of MACD. Price in some cases just moves to quick for the moving averages to catch up with it. That is why I like Stochastics better.
|
|
Registered User
Joined: 3/21/2006
Posts: 4,308
|
I forgot to answer your question. (I was unable to make it into a custom chart indicator such as you have on your charts).
You need to -
Add Indicator
Custom
Percent true
Then paste the formula in the white box
XAVGC12 - XAVGC24 - XAVG(XAVGC12,4) + XAVG(XAVGC24,4) > 0 AND XAVGC12.1 - XAVGC24.1 - XAVG(XAVGC12.1,4) + XAVG(XAVGC24.1,4) <= 0
|
|
Registered User
Joined: 12/21/2004
Posts: 902
|
ahhhh... Percent true... that's the key. Thanks, Apsll!
|
|
Registered User
Joined: 1/28/2005
Posts: 6,049
|
I don't necessarily like the "boxyness" of percent true indicators and like to plot "price based" indicators.
An example of that is shown here:
Indicator Color, Visuals, and MACD.
Thanks
diceman
|
|
Registered User
Joined: 4/18/2005
Posts: 4,090
|
Apsll
"Price in some cases just moves to quick for the moving averages to catch up with it. That is why I like Stochastics better. "
I'm not certain there is much that can be done about it... Macd is based on moving average crosovers and is a trend following system... by nature will be a bit late... tightening MacD to tight like most indicators will get you in faster but also give you more false signals... I think the first is better than the last. That is why it is important to use indicators that don't measure the same thing. If you pick say RSI and STO as your min indicators then you are likely to get mutually confirming signals... while that may feel good it one isn't telling you something the other isn't....unless they are set differently. Those may be a bad example but I think my point is clear. That was why I couldn't understand both MacdH and the Diceman indicator on the same screen .... unless you were at some stage sorting by it. by looking at macdh you'll see that it is crossing zero
Stochastics is considerably different in it's calculation so gives different information and ther for tends to have more predictive power and earlier signals. I don't recal where but I know that Macdh can be viewed as giving a buy signal when the bars start becoming shorter...or trending up from below zero... the safer enty is on the cross up, which is when MacD lines have crossed. Plot a macd and macdH with the same settings and 2 ma's withe the same settings ad observe the relationship. elementally there is little mystery in macd. I thinkthe supurfulous mystery comes in trying to interpret macdh divergences which are touted by Elder as "One of the most powerful indicator signals" They work for him.. I'm beginning to like macd as I get comfortable with it...but I don't know that they are magical. I think it can help you visualize price action in a way that is not necessarily self evident unless you are really good at reading pure price.
|
|
Registered User
Joined: 3/21/2006
Posts: 4,308
|
Yes I am aware of these things, but I disagree with your point about Stochastics vs MACD. I have always considered them in the same class of indicator (Trend), with MACD being used as a slower moving (trend gauge) and Stochastics as more of a short term trend indicator. In other words a short term trader might be more inclined to use Stochastics and an intermediate to long term trader might use a MACD.
I think that the HNC Template & MACD-H is a good marriage because when you are at the decision point and looking for a good entry the MACD-H is a good tool for that.(IMO)
And I do not consider any indicator magical, they are simply tools, it is what you do with these tools is that makes the difference...
|
|
Registered User
Joined: 11/11/2006
Posts: 359
|
Apsll: Of interest, MACD is a trend following system. However, by virtue of the fact that it has a "O" line it crosses, it is also an oscillator. The old saw about using oscillators in choppy markets and indicators in trending markets would make MACD usefull in both. Stochastics is generally thought of as an oscillator. I ran across, in my reading, one guru suggested using MACD lines crossing well below "O" as a buy signal. And SELLING when "O" is reached. Fascinating.
Mammon
|
|
Registered User
Joined: 11/11/2006
Posts: 359
|
I think you are on the right track in your thinking with your current system. Colin Alexander in his "Timing the Stock Market" had a page on believing in MACD. When the price and MACD are in accord, the odds stongly favor you. He liked it to an airline pilot and his instruments. There may be bumps along the way, but the instruments are still true. When both price and MACD start to weaken, then leave.
Mammon
|
|
Registered User
Joined: 4/18/2005
Posts: 4,090
|
Apsll
this is true.. but stochastics and MacD aer measureing different things... or the same thing different ways. Whatever .. it seems like you will get earlier signals from stochastics is my point. I agree that that stochastics may fit the short term trader better. I only recenly began to use it more becasue I had difficulty reading it beyond over bought and over sold... I didn't catch divergences and in cases where it got "stuck" I found frustrating.
Mammon: I've read that as well. dosen't make sense to me.. but I also heard about gurus suggesting buying when RSI crossed 50% line in it's indicator window..... My best luck with RSI is when it is coming up from over sold or crossing up through a slower RSI period.
|
|
Registered User
Joined: 3/21/2006
Posts: 4,308
|
Mammon, thank you for sharing that. I have looked at a few stocks were this has tested possitive. Look at CRWN in December 2006 and WFII in early april 2007. By the way I am using a MACD 15,30,9 regular not histogram.
This is worth looking into deeper.
I wish that when I was just starting out with my trading that there was a forum like this one, I frequented a few in those days but none were like this (you who are still new to trading are very lucky to have such a learning tool as this site). If I had, then I would have made my first million by now...
Apsll...
|
|
Registered User
Joined: 11/11/2006
Posts: 359
|
Apsll: I think the most worthwhile thing about Worden Bros, two actually , Ask a Trainer and this forum. Like you, I have seen several, mostly by very small traders who really didn't know diddly. I find my self going back to the books virtually every day and researching things from this forum. This has a real tendency to keep you mind active. I may not actively use everything, but it gives me more tools to evaluate. I have incorporated several things from the likes of you, Diceman, HNC and others on these posts and I find myself anticipating each days discussions. Trading is a lonely task. I have nobody around who even understands what I do, much less discuss things with. Sometimes this is not all bad. With few exceptions, differences can be discussed on this forum without rancor or disrespect.
You and I did not completely agree on Swoop, but I enjoyed the discussion. I learned early on that ALL of my ideas are not good ones. New traders are indeed fortunate to have stumbled on this.
Mammon
|
|
Registered User
Joined: 12/21/2004
Posts: 902
|
QUOTE (mammon) I ran across, in my reading, one guru suggested using MACD lines crossing well below "O" as a buy signal. And SELLING when "O" is reached. Fascinating.
Mammon
Mammon - I went back to my books last night and saw pretty much the same thing - In The Master Swing Trader I think it was. They even said that only the "beginner's books" teach you to buy when it crosses "0"
I'm trying to track down Gerald Appel's original book where he introduced MACD just to see what he had to say about it. It's out of print but I'm looking...
|
|
|
Guest-1 |