Registered User
Joined: 9/22/2005
Posts: 849
|
I see many references to MACDH and MACD-H and I have some questions.
What is it?
Is there a PCF to find it?
How is it interpreted?
|
|
Worden Trainer
Joined: 10/7/2004
Posts: 65,138
|
MACDH and MACD-H stand for Moving Average Convergence Divergence Histogram. You may wish to review the following:
Understanding Moving Averages - Part I, The Basics
Understanding Moving Averages - Part II, Using Multiple Averages
Understanding MACD
I'm sure your fellow traders will be able to shed far more light on its interpretation than I can since trainers cannot give settings, interpretation or investment advice.
-Bruce
Personal Criteria Formulas
TC2000 Support Articles
|
|
Registered User
Joined: 9/22/2005
Posts: 849
|
Thanks, Bruce.
I am familiar with MACD, but I have an impression that MACD-H is some function of MACD.
|
|
Worden Trainer
Joined: 10/7/2004
Posts: 65,138
|
The reason MACDH or MACD-H is used is to differentiate it from the non-Histogram version.
-Bruce
Personal Criteria Formulas
TC2000 Support Articles
|
|
Registered User
Joined: 9/22/2005
Posts: 849
|
OK. Thank you. That clarifies it for me.
|
|
Worden Trainer
Joined: 10/7/2004
Posts: 65,138
|
You're welcome.
-Bruce
Personal Criteria Formulas
TC2000 Support Articles
|
|
Registered User
Joined: 12/21/2004
Posts: 902
|
BCraig - I've discussed MACD-H quite a bit with Jcfla the past couple of days in the TNH and the MTZ threads (maybe where you saw the refernce) and during the first couple of weeks pf May in several threads with Scottnlena and Apsll. It's currently my favorite indicator and I guess I am guilty of talking it up on here.
While it is derived from the MACD, it functions quite separately from it (I rarely use MACD itself, in fact) and I find it quite useful in evaluating trend strength and entry. If there's anything about MACD-H I can help you with, feel free to ask. 
|
|
Registered User
Joined: 9/22/2005
Posts: 849
|
hohandy
I have seen the difference in the plots of MACD and MACDH and it would be helpful if you would post some specifics ss to how you use and interpret MACD-H with some examples.
|
|
Registered User
Joined: 9/21/2005
Posts: 566
|
I also have a couple questions for you HH,
When using bar can you explain the difference to me between the top 100 vs. the top 1%? There seems to be some overlap but not sure I understand the methodology.
Also, you said before you like really strong MACDH signals to base your decisions on. I have seen some stocks with strong MACDH on the bullish side, but then the stocks retrace and you get equally strong MACDH in the other direction. How do you analyze those types of stocks?
|
|
Registered User
Joined: 9/22/2005
Posts: 849
|
jcfla7
I'm glad you asked those questions. I don't know enough to be able to ask specific questions.
|
|
Registered User
Joined: 1/28/2005
Posts: 6,049
|
bcraig
In your lower window plot:
XAVGC12-XAVGC26
Select center zero line.
Notice that this "looks" like price but is smoother.
If we want to "time" price we can use this smoothed
version to do it. Instead of applying a moving average
to price. We can apply an MAV to this indicator.
Click on the indicator and select "add indicator".
Apply a 9 period exp moving average.
(this is MACD 12,26,9)
In your middle window plot the built in MACD.
Select the histogram plot.
Notice when ever it is above zero. The indicator
in your lower window is above the 9mav.
Take a look at the weekly chart of LEN.
Notice the period from 7/24/06 to 2/13/07.
There is a rally in price. It is jagged and somewhat "noisy".
Notice the "rally" in the indicator is smooth.
Hopefully this will allow us to ride trends longer and not get
whipsawed in our trading.
So at its core. The concept of MACD is to create a smoothed
price indicator. Then "trade" that indicator using a moving average.
The "histogram" is simply a method to see the relationship
between the indicator and its moving average.
Thanks
diceman
|
|
Registered User
Joined: 12/21/2004
Posts: 902
|
QUOTE (jcfla7) I also have a couple questions for you HH,
When using bar can you explain the difference to me between the top 100 vs. the top 1%? There seems to be some overlap but not sure I understand the methodology.
Also, you said before you like really strong MACDH signals to base your decisions on. I have seen some stocks with strong MACDH on the bullish side, but then the stocks retrace and you get equally strong MACDH in the other direction. How do you analyze those types of stocks?
Barchart is telling you two different things and they don't necessarily relate to each other. They assign a "Weighted Alpha" value (I think in another post I called it "Beta" - I apologize if i cnfused anybody) based upon a propreitary formula that considers price movement and volume change during the past 6 months. They assign a "Weighted Alpha" to every stock and also to every industry group - and then post a ranking of the Top 100 "Weighted Alpha" stocks (They also post a Bottom 100 of the stocks with the lowest "Weighted Alpha" values.) If you click onto every industry group in Barchart, it will list all the stocks in each group and will put them in descending order by "Weighted Alpha" value.
Other parts of the Barchart website show "Signals" - and look for those stocks which exhibit the strongest of various technical indicators. They list the current (as of today) strongest stocks in terms of immediate technical indicators and list the top 1% for the day both in terms of strength and direction. Some stocks with very high "Weighted Alpha values also will show very high technical indicators on any given day and so will get listed in the "Top 1%", but others in the Top 1% may have very low "Weighted Alpha" values but are making an important technical move. Top 100 looks at last 6 months and looks at price and volume. Top 1% is just the immediate day and is a combination of several technical analyses.
Top 100 stocks are listed because of status high Weighted Alpha values regardless of what their current chart looks like. Top 1% will generally have a good immediate chart setup regardless of whether the stock has been strong or not in the previous several months.
Now, about MACD-H. Go back to the analogy that MACD-H is all about a giant tug-of-war between the bears and the bulls. The further away the bar rises or falls from the 0 line, the more the bulls or the bears are in control, the stronger the trend. When the slope of the MACD-H bars starts going up (regardless of whether the MACD-H bars are above or below 0), the bulls are gaining control at the expense of the bears. When the slope of the bars is declining, the bears are gaining at the expense of the bulls. One more thing - generally there is a "cycle" (I think of them as "Islands") for a period where the MACD-H is below 0, followed by a period where the cycle (or island) is above 0. Think of the islands in terms of how much area they cover - how many bars each island encompasses, and how high or low it goes from 0.
Say I'm looking at a stock to go long. What I want to look at are the past several island/cycles. What I want to see is that the islands below 0 are steadily diminishing in size - goibg down less far at the lowest point in each cycle. Covering less area in each cycle. I want to see the islands above 0 as the exact opposite - islands getting bigger with higher highs each succeeding cycle. Also, at a minimum, I want the total "area' of the last up island to be larger (hopefully much larger) than the last down island. Remember, we're thinking mutual strength of the bulls vs the bears - the larger the up island in area, the higher the highest MACD-H in the island, the greater the bullish strength.
If a stock is going up, but is not pulling the MACD-H with it, such that the island size isn't bigger than it's last down cycle, or the highest MACD-H isn't making new highs, then it isn't a strong trend. I want the bullish strength to be kicking the bearish strength's butt - otherwise I'll go find another stock where that's happening.
So JC - to go to your example - if a stock makes a high MACD-H, that indicates bullish strength vis-a-vis the bears. But, if the stock then retraces and forms an equally strong MACD-H to the downside, that means the bears are now stronger than the bulls. Basically, there is no strength remaining in the uptrend - there's no reason to consider the stock going long if the bears are equal to the bulls - the stock isn't going to go anywhere. You only want to stick with a stock if the up island remains bigger than the down one - if it isn't bigger, there's no reason to be in the stock long unless you're in the for the long term.
Make sense?
|
|
Registered User
Joined: 9/21/2005
Posts: 566
|
Got it, so if your eyeing a stock that has made a nice run, and during the HNC/Diceman phase it forms a MACDH comparable in size to the MACDH formed during the prior up move, you would take a pass on the HNC/Diceman.
|
|
Registered User
Joined: 12/21/2004
Posts: 902
|
In essence, yes. Use the MACD-H as a gauge of bulls-vs-bears strength. Unless it's overwhelmingly in favor of the bulls, take a pass. If the MACD-H to the down side is as strong as it was to the upside, that means the bulls aren't the dominant as the were on the run-up and can't be expected to be as strong going forward - there are always other good stocks out there that you shouldn't have to compromise.
Bcraig - I haven't directly addressed your question, but I have answered Jcfla's questions with you in mind, describing how I use MACD-H and the theory/framework behind it. Basically you're looking at momentum caused by inbalances of the strength of the forces that are causing the stock to go up vs. the the forces that hold the stock back as measured by the differences between the fast line and the slower line of the MACD (which essentially is what the MACD-H measures.) If buying is so dominant that it overwhelms selling to create forward momentum, such that, for example, resistance lines are ignored, the fast MACD line will pull away from the slow MACD line and the MACD-H bar will be that much greater in size. As selling power increases, the bar will shrink again and it indicates that forward momentum is not as strong as before.
Look at a stock I was interested in this week - MTZ - on a 3 day chart. When I first saw the stock on Tuesday night (one 3 day bar ago), I was impressed by its MACD-H strength. One concern (raised by Jcfla) was that it was close to its overhead resistance from last year. But I was confident, given it's high MACD-H, that it would be powerful enough to go through the resistance - and it did, rather easily. If the MACD-H wasn't as high, I would not have been as confident.
I use the MACD-H in connection with the HNC/Diceman setup and I describe how I use this, and some of my results in the TNH, MTZ, and Hohandy threads. Hope this helps in your understanding of MACD-H - if not, ask away
|
|
Registered User
Joined: 5/31/2006
Posts: 47
|
Hi All
I am really enjoying the discussions here on the HNC/Diceman setup as well as the MACD-H indicator. I'm almost convinced to give it a try myself but I am attempting to be a swing trader as I only wish to hold for a few days to maybe a couple of weeks. Anyway, just thought I'd suggest that if anyone likes to look at a squiggly line instead of the histogram (me) just add a 1 period ma to the MACD-H and then uncheck visable. Works for me. Thanks.
Mark
|
|
Registered User
Joined: 12/21/2004
Posts: 902
|
Mark - I think MACD-H used in conjunction with HNC/Diceman is perfectly suited for swing trading.
It consists of 2 basic components:
!) the HNC/Diceman setup identifies a "decision point" when a series of price moving averages converge as a method of identifying entry. Price is going to move out of the decision point, the only question is, in which direction?
2) In the MTZ thread, I copied a quote from from Dr Alexander Elder's book Come into my Trading Room where he talks about the momentum created with a high MACD-H - basically if there is a multi-month price high accompanied by a multi-month MACD-H high, there is a enough upward momentum created that, in the event of a pullback, it is likely that the previous price high will be retested.
So combine this with the HNC/Diceman decision point. If you had a stock with a multi-month price and MACD-H high, that then pulls back to the decision point, the odds are now in your favor that from the decision point it will go back and at least retest the previous high. It may take a couple of days, it may take a couple weeks - but that sounds to me awfully like swing trading, doesn't it?
.
|
|
Registered User
Joined: 9/22/2005
Posts: 849
|
hohandy
If you had noticed MTZ n 04/27 when it had begun its climb, what would haae been your attitude?
I have the HNC indicator involving the long and short MAVs. What do I need to add to make the HNC/Diceman indicator?
|
|
Registered User
Joined: 9/13/2006
Posts: 143
|
Hohandy:
I know you have been very kind to explain every question, one more please!
Do you sort watchlist by macd 12,26,9 indicator? and on the RSH on 1/8/ I see MACD-H aound 70 BUT on May 9th I clearly see it at 100. Am I missing the point here?Appreciate your time and patience
Moe
|
|
Registered User
Joined: 12/21/2004
Posts: 902
|
I would have liked it Bcraig. I don't know that I would have entered at the point as volume indicators such as MS and TSV weren't all that great yet. But, it would have caught my eye. It wouldn't have made a MACD-H=100 watchlist candidate yet, because, even though the MACD-H reached new highs on 4/13, 15-17, price hadn't made multi-month highs yet. BUT, consider the level of the MACD-H during those 3 days compared to the level of the MACD-H during the high point in Feb. If forward momentum was this much stronger than it was a few months ago when the price was higher, that's a positive divergence that indicates that good things were coming and that previous price high would be taken out. So it definitely would have made an impression on me. The fact that in the couple of weeks before the April 12 price rise was a horizonatl consolidation where the MACD-H didn't go below 0 also would have made an impression on me. The bears had every opportunity to bring the price back down, and didn't and showed no strength at all. It was just a matter of waiting for the bulls to flex their strength again. I'm a big fan of horizontal price consolidations.
But the thing that really impresses me (and why I went in this week) on MTZ is the weekly chart. Look at the long horizontal movement from August 06 through April 07. Look at how deep the MACD-H was during August and September - and as the horizontal movement went on, the bears lost their strength, until the next time they were able to force the weekly MACD-H below 0, it was just a minor 4 bar blip - nothing compared to it's last below 0 trip the previous summer, or to the above 0 island the cycle before in Dec-Feb The fact that the bears couldn't do anything with that trip below 0 in March-April was a signal that showed that the bulls wouldn't get any resistance when they reappeared at the end of April. And there's been nothing to stop the price on the weekly chart since.
Also, look at it's very strong MACD-H profile on the weekly chart at this time.
|
|
Registered User
Joined: 12/21/2004
Posts: 902
|
oh, forgot - look at the Thread "A simple profitable method for the newbie, or not so newbie" for discussion of and how to create the Diceman indicator
|
|
Registered User
Joined: 12/21/2004
Posts: 902
|
QUOTE (moakhavi) Hohandy:
I know you have been very kind to explain every question, one more please!
Do you sort watchlist by macd 12,26,9 indicator? and on the RSH on 1/8/ I see MACD-H aound 70 BUT on May 9th I clearly see it at 100. Am I missing the point here?Appreciate your time and patience
Moe
Moe - I'm not sure that I understand your question. My MACD settings are 12,26,9 - but when I do sort, I sort by the MACD-H (the histogram) and I sort my Visual sort on a zoom=6. I don't care so much what the actual numerical value of the MACD-H is, just that it is at it highest point in the past several months - and the visual sort will show that.
Not sure what you're saying about the Jan and May MACD-H. You are correct in what you're seeing, but I'm not sure what it is that the point is that you think you're missing. If you could explain exaclty what you think you're thinking I'll be glad to try and help.
|
|
Registered User
Joined: 12/21/2004
Posts: 902
|
Moe - this is what I think you're saying - if you use a zoom=6 chart, you get roughly 3 months worth of data. Now, notice as you scroll through time, the values on MACD-H value bar change. It resembles an ocillator, but it contracts and expands as the values of the MACD-H go closer and further away from 0. And since the zoom=6 shows 3 months worth of data, it will expand enough to show the values of every MACD-H bar in the previous six months. So on Jan 8, the MACD-H settings reflected the deepness of a very low MACD-H bar back at the end of October (it had a value of 0.33) - so as long as that October bar is on the chart, the MACD-H axis will be expanded to accomodate that bar. And since they keep 0 in the middle, it also had to go to at least 0.33 above 0 to keep it symmetrical. Now since the MACD-H value of the Jan 8 bar wasn't as high as that October bar was low, on a visual sort, it would have shown the visual sort value of Jan 8 as being only 70 - it only rose 70% up the bar at that setting. Make sense?
For purposes of what we're looking for, we want an extreme high MACD-H value (visual sort = 100) so that means that at any time in the previous 3 months (at zoom=6), in order for MACD-H to = 100, there would have have to have been no other MACD-H's with a greater postive value (otherwise the axis would have expanded to accomodate it) OR no other MACD-Hs with a greater negative value (because otherwise the axis would have expanded to accomodate that).
I *think* that explains what your question is.
|
|
Registered User
Joined: 12/21/2004
Posts: 902
|
QUOTE (hohandy) And since the zoom=6 shows 3 months worth of data, it will expand enough to show the values of every MACD-H bar in the previous six months.
Of course I meant to say "it will exapnd enought to show the values of every MACD-H bar in the previous THREE months."
Sorry if I caused confusion
|
|
Registered User
Joined: 9/13/2006
Posts: 143
|
I think I got it. RSH was recommended by one of the news letters I have been subscribed in Nov.06 as L.term buy , but I have lost so much money following them that I stopped playing their game and decided to create my own game with the help of generous participants of this forum. Thanks again
|
|
Registered User
Joined: 9/13/2006
Posts: 143
|
Hohandy:
These are what I got sorting Barchart 100, since am not subscriber to IBD.
JNC,DK,GTLS,AZC
these are MACD-H=100 both on daily and wekly.
What do you think? did these show up on your scan too?
Moe
|
|
Registered User
Joined: 12/21/2004
Posts: 902
|
In retrospect that would have been good advice to follow re RSH. I don't know why you picked Jan 8 per se, but that would have been a great day to get in.
I don't use MACD-H exclusively for my entry/exit points. Once you understand MACD-H, you will find that it is useful to confirm other signals and indicators regardless of whether you're using the HNC/Diceman or MACD=100 setup.
For example, I screen for breakouts (specifically up day on high volume) - and Jan 8 was most definitely a breakout day for RSH coming at the end of a 2 month consolidation after a downtrend.
Notice the MACD-H pattern in this time too - after 11/14 there was very little bearish strength left in the stock - it was just a matter of waiting for the bulls to come back - which they did on Jan 8. And once the bulls took over, there was very little bearish strength to hold them back. Pretty much like I described for MTZ a few posts ago.
Hopefully in the future you'll be able to use MACD-H to help evaluate these kind of situations.
|
|
Registered User
Joined: 5/31/2006
Posts: 47
|
Hi hohandy
I just typed a lenghty reply and it didn't post. I guess I wrongly assumed that by looking at monthly MACD-h, you were looking to hold longer term. You are taking the same approach I am, buying dips in uptrends, we are just doing it a little different(for now). I may have to become a hohandy and elder student if you are really doing good with this method and I hope you do. With this original thinking and unselfish sharing, somebody should nominate you for knighthood. Attention Don Worden! Thanks.
Mark
|
|
Registered User
Joined: 12/21/2004
Posts: 902
|
QUOTE (moakhavi) Hohandy:
These are what I got sorting Barchart 100, since am not subscriber to IBD.
JNC,DK,GTLS,AZC
these are MACD-H=100 both on daily and wekly.
What do you think? did these show up on your scan too?
Moe
You got it Moe. Of those I like DK and AZC best primarily because of their weekly charts and the industries they are in.
On Barcahrt, you also may wish to incorporate the "Signals" stocks - the Top 100 Signals list as well as Top 1% strength and Top 1% direction. These are stocks that are currently scoring highest in immediate technical indicators according to Barchart.
|
|
Registered User
Joined: 9/13/2006
Posts: 143
|
Thank you for your hard work.
|
|
Registered User
Joined: 12/21/2004
Posts: 902
|
Thanks for the kind words guys. I've learned an awful lot myself on this board from some pretty impressive and unselfish people. So I don't mind making my own contribution. And a lot of my thinking about how to use the MACD-H came about just by discussing it in a bunch of posts with Scottnlena and then going back to the books and doing a lot of thinking in response to our discussions. To me, that's what I see as the value of boards like this - so don't hesitate to jump into discussions and ask questions and offer your viewpoint and perspective - you never know if it will jog something in your understanding or someone else's understanding and then we all learn from that. It's not a zero-sum game - if you profit from my understanding of something doesn't mean that I profit by it any less - and vice versa - and we all gain.
|
|
Registered User
Joined: 9/21/2005
Posts: 566
|
Well said HH, you should get a knighthood for all your help. Diceman should too.
|
|
Registered User
Joined: 5/31/2006
Posts: 47
|
True jc
I didn't mean to leave out diceman tobydad and others too. Everybody here is great. What a good place to get away from the BS and try to learn how to trade. Thanks.
Mark
|
|
Registered User
Joined: 12/21/2004
Posts: 902
|
QUOTE (moakhavi) Hohandy:
These are what I got sorting Barchart 100, since am not subscriber to IBD.
JNC,DK,GTLS,AZC
these are MACD-H=100 both on daily and wekly.
What do you think? did these show up on your scan too?
Moe
Moe - I need to amend my reply to you about these stocks, now that I've actually sat down and down my weekly drill and gave them all a good study.
These stocks are all, in fact, MACDH=100, HOWEVER I would not consider GLTS and JNC as proper candidates for the HNC/Diceman setup from their daily charts. The key to the MACD-H part of the setup is Dr Elder's observation that a multi-month price high combined with a multi-month MACH-H high will generate momentum to revisit the price high in the event of a pullback.
Both GLTS and JNC have the multi-month MACD-H, but neither has a multi-month high to accompany that particular MACD-H bar. They may work in the HNC/Diceman, or they may not, but I'm sticking strictly with Elder's conditions.
|
|
Registered User
Joined: 9/13/2006
Posts: 143
|
I agree Thank you.
|
|
Registered User
Joined: 2/2/2005
Posts: 27
|
How do I copy the barchart 100 into telechart?
|
|
Registered User
Joined: 12/21/2004
Posts: 902
|
The way I do it is to copy it into Excel, delete all the colums except for the symbols, and then save as a "text - tab delimited" file. That gives it a .txt root which then can be imported into telechart
|
|
Registered User
Joined: 9/22/2005
Posts: 849
|
hohandy
Thanks for your insightful response to these questions. When I have digested this information I will have another arrow in my quiver of indicators,
One of the great advantages of Telechart is the help available from those who are more experienced through these threads. I have learned many things which have helped yo crystalize my thinking about many things from money management (stops and risk analysis) to several scans and chart set-ups'
hohandy, Diceman, Tobydad, fptry,and HaveNoCents have been especially forthcoming with detailed responses to questions. In addition, I have gotten several hints about specific stocks to which I then subjected my own indicators. (I'm not above a little plagarism.)
|
|
Registered User
Joined: 12/21/2004
Posts: 902
|
Thanks Bcraig. It really is a group effort around here, and I don't hold a candle next to those other guys.
|
|
Registered User
Joined: 2/2/2005
Posts: 27
|
OK, I am in the slow group. I have tried importing barchart with no luck. Any other ideas?
|
|
Registered User
Joined: 12/21/2004
Posts: 902
|
ok - I'll walk slowly through the drill and see if you did it like this:
Goto barchart Top 100 page. I start at the very bottom of the data in the right hand corner (the page I'm on in the "Main View" page - the right hand column is 52week low - the data is 9.77 for stock MFRI).
Using my mouse I click on just to the left of the 9.77 - hold the mouse button down, sweep left to the stock name and then up to the top of the chart - stopping when the uppermost stock LPHI - is highlight. At this point, all data on the sheet, but nothing else is highlighted. I right-click and the little menu comes up (I'm in Windows, I assume you are too) I select copy.
I go to an Excel sheet and paste into cell AI - everything should copy into it's own cell and the first column should be in blue because it's formatted as HTML (doesn't matter). Goto the column labels beginning a B, click onto that, holding mouse button down, sweep across to column J - everything in columns B through J are highlighted, right click, when the nenu comes up select delete.
Goto the File "Save As" Menu - when the menu box comes up, type in the File Name in the File Name box, in the box below that ("Save as type", hit the little arrow and scroll down until I see "Text (Tab delimited)(*.txt)" Hit save. My excel a box comes up telling me some features are incompatible, blah blah blah, I hit "Yes". I also take note of what file directory my file is in.
Goto Telechart. Create New Watchlist, Put in new file name, hit the check next to "Import from File". A box opens up that says "Import Symbols From List". Make sure "List Type" is set to "Text File". It defaults to the C drive (on mine it does, anyway) go into that and find the directory that your file is in, and it should list all *.txt files in that directory and select your file.
Provided you did everything right, it should be a very routine drill.
|
|
|
Guest-1 |