Download software Tutorial videos
Subscription & data-feed pricing Class schedule


New account application Trading resources
Margin rates Stock & option commissions

Attention: Discussion forums are read-only for extended maintenance until further notice.
Welcome Guest, please sign in to participate in a discussion. Search | Active Topics |

the power of a divergence? Topic Rating:
Previous Topic · Next Topic Watch this topic · Print this topic ·
scottnlena
Posted : Thursday, March 8, 2007 6:13:22 PM

Registered User
Joined: 4/18/2005
Posts: 4,090
I'd like ot knwo if divergences rank verry highly with you guys.. between price and TSV, Price and MONEYSTREAM, etc etc. Is it something that is a nice little extra or an additional excuse to stay away if it's agains the intended trade direction? Or is it something to take verry serriously? Would you call them powerful indicators of price action and quality or sometimes a bit of a clue and some times ananomoly? ?
Apsll
Posted : Friday, March 9, 2007 8:19:42 AM

Registered User
Joined: 3/21/2006
Posts: 4,308
Scott:

I take them very seriously, you only need to look at the major Indexes to see the divergence between price and TSV starting in November 2006

If you zoom out on the charts you can see it better - TSV is in a down trend along with Stochastics as the price is still moving up..

Another good example is AKAM remember I mentioned the divergence on that one and you said that it was a good call..

The last example I will give will be INTC back in early December, I warned that the stock was topping because of divergent Stochastics (I did not mention at the time, but TSV was also divergent)

Here is the INTC thread -

INTC

Candlesticks and divergences are my two most reliable indicators...

(IMO)

Apsll...
Apsll
Posted : Friday, March 9, 2007 8:24:26 AM

Registered User
Joined: 3/21/2006
Posts: 4,308
Correction I guess I did mention the TSV divergence in the above thread..

I must be getting old, my memory is not as sharp...
diceman
Posted : Friday, March 9, 2007 8:47:10 AM
Registered User
Joined: 1/28/2005
Posts: 6,049
One thing that should be remembered is that oscillators
because of there calculation method. Will diverge at some
point. (it is impossible for an oscillator to follow price
up for an extended period of time.)

At some point oscillators have to switch to the mode:
Above mid-point=positive/below mid-point= negative.
This is true of things like TSV and MACD histogram.

My guess is that because of this factor. Divergence in
MS would be "more valid". (because it can continue to
rise.

Thanks
diceman
Apsll
Posted : Friday, March 9, 2007 9:04:23 AM

Registered User
Joined: 3/21/2006
Posts: 4,308
Thanks for that Diceman, I just learned somthing new..

If I may elaborate -

When price is running up (and there is no end in sight) Stochastics will climb into the nether regions and float sideways (this is called a momentum pattern for stochastics, I read that somwhere), Yet you can still have diverging stochastics peaks..

How does that principle apply to TSV?? Are you saying that as price reaches the pivot point that you allude to in your post above that TSV will cease to function?

I am just trying to learn, If you could explain for me a little deeper I would appreciate that..

Thanks Apsll...
bustermu
Posted : Friday, March 9, 2007 10:52:10 AM
Registered User
Joined: 1/1/2005
Posts: 2,645
QUOTE (apsll)
How does that principle apply to TSV?? Are you saying that as price reaches the pivot point that you allude to in your post above that TSV will cease to function?


apsll,

The following example may help you.

Suppose we observe a plot of TSV38 Simple on Zoom 8. I chose the period 38 because Zoom 8 has fewer than 38 bars. What does the observed behavior of TSV38 Simple tell us about Share Accumulation Distribution as measured by TSV1 over the observation interval. The answer is "absolutely nothing".

What I have stated is a well-known fact about simple moving averages and has nothing to do with TSV per se. It should serve as a warning about using increases and decreases in TSV Simple to determine divergences.

Thanks,
Jim Murphy
scottnlena
Posted : Friday, March 9, 2007 10:56:40 AM

Registered User
Joined: 4/18/2005
Posts: 4,090
Thanks guys this was great...made use of some trading down time to consider my strategies and lay some ground rules. Also learned allot more about the capacities of TC 2007 to find specific things. One video ofparticular interest was "sorting by linear regression lines to find divergences". Basically sort any list by puttina linear regression on price and an indicator. Sort price anf flag all with the pride direction you are looking for.. then re sort by theindicator and unflag all those that dont apply, the remainders have good chances of being in divergence. I've learned so much inthe last two weeks, I really think 07 will be agreat year for us! Alot of that credit goes to you guys. Thanks again.
Apsll
Posted : Friday, March 9, 2007 11:31:12 AM

Registered User
Joined: 3/21/2006
Posts: 4,308
Bustermu, I am not sure that I fully understand your point. I applied a TSV 38 simple and a TSV 6 simple in the same window and while veiwing the chart for the Nasdaq Index I came to the same conclusion, that TSV (38 & 6) was divergent from price, from November 2006 to late January 2007

If you could explain in laymans terms (for dummies like myself) I would appreciate that..

Thanks Apsll...
dcostello
Posted : Saturday, March 10, 2007 7:51:53 AM
Registered User
Joined: 11/3/2004
Posts: 39
One thing I've observed about divergences is the longer they've been divergent, the better the signal and usually the stronger the move. Personally, the best trading I've done with divergences has been price/MACD Histo.
scottnlena
Posted : Saturday, March 10, 2007 10:14:12 AM

Registered User
Joined: 4/18/2005
Posts: 4,090
dcostello

can you give an example of a MACD divergence?
diceman
Posted : Saturday, March 10, 2007 10:53:45 AM
Registered User
Joined: 1/28/2005
Posts: 6,049
If you look at a MACD histogram (12,26,9) on a strong
stock. It will typically rise as the stock moves up.
(due to the spread of the 12,26 averages)

However if the stock then just stays flat. The moving
averages will approach price. The moving averages
approaching the "same" value will make the
histogram "fall". (the spread of the 12.26 is moving
toward zero)

This will be viewed as divergence but I'm not necessarily
sure this is a bad thing. (this is a case where indicator
calculation can create divergence)

------------------------------------------------------------------------------

I think a part of what bustermu is alluding to is. Information
is lost when you average something that goes above and
below zero.

If you have a 4 period smoothing of TSV and the values are:

+100
-100
+100
-100

it will equal zero.

If you have:

+1000
+4000
+10000
-15000

it will also equal zero. However I don't think traders would
view this as the "same" type of movement.
--------------------------------------------------------------------------------------

If you look at the symbol GS with a TSV of 28
(look at both simple and exp)

You will see a divergence that forms after OCT 2006.

However if you apply OBV to the price chart. It looks
OK. (OBV is also a "price volume indicator" but one
that we know how its calculated and one that can continue
to go up)

With this information it is difficult to tell what part of
TSV divergence is being created by its smoothing,
the fact that it is an oscillator, that fact that it falls above and
below zero, or its calculation, and what part is real.


Thanks
diceman
Apsll
Posted : Saturday, March 10, 2007 11:03:20 AM

Registered User
Joined: 3/21/2006
Posts: 4,308
Very clear and concise thank you Diceman & Bustermu

I did some testing on successful momentum runs with TSV vs Money stream and you were correct, even though price was sky rocketing along with money stream, TSV leveled out at some point and even droped, creating the allusion of a divergence..

That is what I love about this forum, every day can be a learning experience...
rmr1976
Posted : Saturday, March 10, 2007 11:05:58 AM
Registered User
Joined: 12/19/2004
Posts: 457
I had posted about divergences just over a month ago:

http://www.worden.com/training/default.aspx?g=posts&t=19337

I think momentum divergences are the second most important concept in technical analysis, with the most important being trend, and the third being volatility (or noise), and the fourth being cycle.

In order to trade divergences, you have to know what the trend is first.

To do this, I prefer to use moving averages that provide support during uptrends, and resistance during downtrends. The moving average acts just like an automated trendline, with more touches making the moving average a more reliable description of the trend.

Changes in the trend and momentum can be spotted when the moving average changes direction.

I also prefer to find parameters on these moving averages that work on daily as well as weekly data.

The periods 21 and 55 work well on daily data and weekly data (using simple moving averages). They also reduce cyclical fluctuations on the weekly data being they are pretty close to the 1/2 and 1 year time frames (26 weeks is 1/2 year, and (52 weeks is 1 year).

After you define your trend, you need to determine the volatility around that trend. I currently use Bollinger Bands based on the 21 bar SMA, although envelope channels based on a very long term average (ie. 200 day) could work very well, and I'm looking at that also.

Divergences work best when the closing price is far above or below the dominant trend.

To confirm or filter out likely false signals, a comparison to volume is very helpful. A stock that is making new highs on low volume, while also showing a negative divergence, is likely to be a better short, or sell long positions, than a stock still breaking out on good volume from an intermediate term trend trade.

For short term swing trades, divergences with volume spikes make good short term trades to bet on a counter trend move back to the moving average.


jcfla7
Posted : Saturday, March 10, 2007 1:14:35 PM
Registered User
Joined: 9/21/2005
Posts: 566
RMR, that was a good post. I agree with your observations.
Apsll
Posted : Saturday, March 10, 2007 1:42:00 PM

Registered User
Joined: 3/21/2006
Posts: 4,308
Rmr1976:

Diceman brings up a very good example of a major divergence in TSV & Stochastics for the ticker symbol GS that started in October of 2006. If I were long this stock then I would have sold my shares during the consolidation of December 2006 (because of the divergence between TSV and price), and would have missed the last leg to the top in February 2007

Should we chalk this example up to the fact that divergence is not always reliable, or do you see somthing that we are all missing? (or I am missing)

I did enjoy your thread pertaining to the subject matter, and myself have relied heavily on the reliability of divergent indicators.. But then there is Dicemans example to contend with...
bustermu
Posted : Saturday, March 10, 2007 2:08:07 PM
Registered User
Joined: 1/1/2005
Posts: 2,645
QUOTE (apsll)
... TSV leveled out at some point and even droped, creating the allusion of a divergence..


apsll,

I do not know whether TSV has any predictive value or not. The illusions to which you refer are caused by the filters placed on TSV1 and not by TSV.

Let's put the blame where it belongs.

Thanks,
Jim Murphy
rmr1976
Posted : Saturday, March 10, 2007 4:34:04 PM
Registered User
Joined: 12/19/2004
Posts: 457
Much like failed chart patterns, failed divergences give you significant info about the strength of the trend, compared to the strength of the cyclic component of price movement. I'll go into detail in a minute.

If you have a chance, you should pick up Welles Wilder's New Concepts in Technical Trading Systems for some excellent ideas on developing a trading method that adapts to the changing markets.

One of my favorite ideas of his is what he calls the reaction-trend system. It influences how I interpret divergences in strong trends.

Basically, the system assumes the market is in a sideways trend, and sells strength and buys weakness, within certain bounds. Typically, these are very short term trades that are closed out within a few days.

When those bounds are exceeded, the system assumes a trend is in place, reverses and follows that trend, using a trailing stop loss to exit. It then assumes the market is in a sideways trend.

When I see a divergence, I be inclined to use a short term counter trend strategy, betting on a return to the moving average I'm using to define the trend.

If volatility is large enough, that might justify a short position for a swing. Often, it simply means waiting before entering long positions.

As for GS, the value of a divergence depends upon where the price is when it occurs. If I get a divergence, but volatility is low while the price is not too far from support, I might wait for a better entry point, or wait for a breakout for a clearer signal.

Those October divergences came when price had not to recently broken out of a rectangle bottom to a new 52 week high.

A technician would have to assume at that point the trend is up for the foreseeable future. But does that mean the stock is in a good position to buy? Usually not. Breakouts often fail, and it is usually better to wait for a pullback of some sort, unless there are other factors (ie. possible bankrupcy for short, or possible buyout for longs) that suggest chasing the breakout.

Volume expanded on the breakouts, but tended to drift down, and price made a new high in October.

I'm using a percent true indicator to pick the specific time when the divergence could be seen on a MACD. I get an initial divergence signal on 10/11/06, that lasted until 10/20/2006. The net price move during that time was 0.8%, with a max rise of 3.49% (using closes) from divergence to highest close, and a decline of -2.60% from that peak to the end of that divergence signal.

A few days later, a new divergence signal came, and lasted from 10/24 through 11/1.

The max price rise during the time from the initial signal was 2.95%, while the max decline from the highest price in that time frame was -4.33%

Now, how does this compare to when there is no divergence?

My indicator where there was no negative divergence starts on 9/14--just a few days after the breakout, and remains in "trend mode" until 10/11/06.

Taking a long trade there would have netted 10.12%, with drawdowns of less than 1%.

The next part of the trend where there is no divergence:
11/1 thru 11/16. That period saw a rise of 6.34, with minimal drawdowns (ie. a loss to longs about -0.12%).

Suffice it to say, taking signals only based on divergences, without giving consideration to other factors--ie duration of the trend, volume, and broad market action, is likely asking for trouble.

With divergences, you can get to keep very close stops. That is one of the reasons I like them so much.

If you have the courage to completely reverse yourself when a divergence fails, I think you could make lots of money. Failed divergence signals tell you the trend is strong and worth betting on.
dcostello
Posted : Sunday, March 11, 2007 8:58:22 AM
Registered User
Joined: 11/3/2004
Posts: 39
I agree with rmr in that you can't just trade divergences, like you can't just trade any other single signal. Confirmation is needed, entries and exits planned, etc. I tend to use MACD divergences as warnings/confirmations of trend reversals or continuations and then use other indicators to enter/exit a trade.

Continuing with GS. Pull up a weekly chart with MACD histo 12/26/9. I have mine set to zoom 5 at the moment. Starting on the left, AMJ 06, you can watch price work its way down long with MACD until we print some very nice doji's as price flattens. Notice, however, the rise in MACD. Do we trade now? All depends on your trading style, agressiveness, risk aversion, etc. If you follow along, price makes higher highs and MACD rises accordingly. The divergence in AMJ JAS 07 is one that would attract my attention. Price runs up nicely in A07, capping a year long run from A06. The we drop into a flat range but notice the MACD slowly creeping up while price flounders. 9/14 we print a very nice candle and MACD crosses positive. We're also getting close to a 52 week high and are looking to complete at least a Pan/Saucer pattern or possibly a Cup and Handle pattern. Do we trade now? That's up to your style.

diceman
Posted : Sunday, March 11, 2007 9:37:19 AM
Registered User
Joined: 1/28/2005
Posts: 6,049
One thing that I would add. GS is a relatively high priced
large cap stock. (it happened to be the first stock on my
screen when I opened telechart)

I'm a big believer in the accumulation of success/failure.

If we take the simple concept that a stock is in buy
mode when TSV28 is above its 21 day mav and compare
it with OBV above its 21 day mav.

We get buy signals on GS in the 9/13/06 zone at about
160.86. (I'm being loose with the numbers)
We get a TSV sell at about 10/30/06 at about 189.75
We get an OBV sell at about 12/12/06 at 200.

If you took a $10,000 position in GS this would work
out to:

TSV profit approx. 1796
OBV profit approx. 2433
difference approx. 637
-----------------------------------------------------------------------------
If we now look at SGG (a stock with a similar pattern)
we get buys at about 9/29/06 at 6.35. We get TSV
sells at about 11/10/06 at 7.54 and OBV sells at
1/25/07 at about 8.31.

If you took a $10,000 position in SGG this would work
out to:

TSV profit approx. 1874
OBV profit approx. 3086
difference approx. 1212
------------------------------------------------------------------------

As can be seen this "error" is more costly
(most likely) with lower priced stocks and
increased volatility.

My guess is the same type of "error" would be
created using MACD or STOC as an exit tool.
----------------------------------------------------------------------
Realize that this should in no way diminish the
usefulness of TSV, MACD,STOC. It just highlights
the concept that you should understand what you
are looking at and what it implies.

(It can also be seen that switching to TSV>0 after
the position was established. Would have allowed
you more profit from the position)

Thanks
diceman

rmr1976
Posted : Sunday, March 11, 2007 11:35:57 AM
Registered User
Joined: 12/19/2004
Posts: 457
Diceman,

You write:
"If we take the simple concept that a stock is in buy
mode when TSV28 is above its 21 day mav and compare
it with OBV above its 21 day mav."

TSV is a much more volatile indicator than OBV. Using OBV provides practically identical signals as using price vs. 21 day moving average (simple).

It is a bit extreme to call these early sell signals "errors." The trend just happened to continue in these cases. I could easily pick out loads of stocks where the divergence got you out just a few days before a price collapse.

You need to compare the price action during a trend where there is no divergence, to the price action in a trend where there is a divergence.

There is also the issue of volatility. The more volatile the stock, the more likely short term trading is going to render profits.

Considering the fact that gains and losses are not symmetrical, the less time in the market means less risk.

The whole point of divergences is that they LEAD price. They are SUPPOSED to be early.

People are treating divergences as COINCIDENT indicators, expecting price to turn immediately after a divergence. This is not correct.

There needs to be an analysis of the price momentum in the context of where the price is in relation to the trend. Is the price very far above or below the moving average? What has volume done at these high levels? How long has the price spent there?

I discussed this in a Worden Report in April 6, 2005.

First, I discussed the need to monitor momentum indicators on a number of time frames. It helpful to compare the 5, 20, and 50 day rates of change, or whatever number you find useful. It is very important to understand where the stock is on the weekly charts.

Expecting 1 momentum indicator to pick up the turning point where multiple price cycles converge is asking too much.

I then point out that momentum leads price, and that it is often more profitable to wait before exiting. You can tighten stops and put above market sell orders when you see a divergence.

There is no need to act on them immediately. Pick your spots.
bustermu
Posted : Sunday, March 11, 2007 11:52:28 AM
Registered User
Joined: 1/1/2005
Posts: 2,645
QUOTE (diceman)
If we take the simple concept that a stock is in buy mode when TSV28 is above its 21 day mav and compare it with OBV above its 21 day mav.


diceman,

Your comparison is more a comparison of different filters than of TSV and OBV. The raw data of each is:

TSV1

OBV1.0-OBV1.1

respectively. For TSV, you have placed an SMA28 on the TSV raw data and then placed an SMA21 on the SMA28. For OBV, you have plotted the OBV raw data as a cumulative indicator and then placed an SMA21 on the cumulative indicator. Those are two very different filters. If you wish to compare TSV and OBV it seems that you should place the same filters on the raw data of each.

Either:

1) Place an SMA28 on the raw data of each and then place an SMA21 on the SMA28, or

2) Plot the raw data of each as a cumulative indicator and then place an SMA21 on the cumulative indicator,

or both.

Please notice you would have obtained essentially the same results by replacing TSV28 Simple by PROC28 and replacing OBV by C. This is because, in your example, you are not using TSV or OBV for divergences from Price and that, as far as I know, is the only value claimed for either.

Thanks,
Jim Murphy
diceman
Posted : Sunday, March 11, 2007 1:04:33 PM
Registered User
Joined: 1/28/2005
Posts: 6,049
rmr, bustermu

You are pointing out the obvious.

The whole purpose of my post is to illustrate the
"flawed" thinking that goes on in traders minds.

Indicators are compared as if anything can be used
anyway at any time.

Realize that I have shown nothing new. I see traders
and magazines use indicators in this style all the
time.

The whole point is to illustrate that you must know what
you are doing and what you are looking at.

Thanks
diceman
rmr1976
Posted : Sunday, March 11, 2007 1:20:10 PM
Registered User
Joined: 12/19/2004
Posts: 457
Diceman,

I agree with what you were trying to point out in your initial post on the structure of indicators and divergences, although I would prefer to phrase it differently.

In my way of thinking, momentum divergences reflect pauses in the direction of the trend. If the trend is up, then the future price trend is more likely change from sideways to down.

The probabilities will depend upon other information available to the trader.

bustermu
Posted : Sunday, March 11, 2007 3:02:50 PM
Registered User
Joined: 1/1/2005
Posts: 2,645
QUOTE (rmr1976)
TSV is a much more volatile indicator than OBV.


rmr1976,

That is an interesting statement. What caused you to draw that conclusion?

Thanks,
Jim Murphy
sharkattak
Posted : Sunday, March 11, 2007 6:36:39 PM
Registered User
Joined: 4/17/2006
Posts: 271
This is a great thread! Lots of info and subtleties in deciphering indicators.
hohandy
Posted : Tuesday, March 13, 2007 4:24:16 PM
Registered User
Joined: 12/21/2004
Posts: 902
I've been trading about 6 years now. Basically learned my stuff from the Bill O'Neil model - setting my screens to look for breakout stocks in hot industry groups moving on higher than average volume and looking for cup-and-handle patterns. Basic stuff. My inital TC2000 screen would generate maybe 50-300 charts a night and I'd put them on a slide show and make my picks or experiment with other indicators to whittle them down to a watch/buy list. And I did pretty well, considering.

I was given Dr Alexander Elder's latest book, Entries and Exits, for Christmas. He basically describes 16 trading rooms and trading styles and follows and comments upon 2 trades in each - fascinating book. But Elder is HUGE on MACD-H - an indicator I'd never really payed much attention to in the past - and especially big on MACD-H divergences, which he considers to be one of most powerful trading signals out there. So, beginning in January, I started taking my O'Neil flavored daily watchlist and further filtered it down to those stocks whose MACD-H performances were matching their breakout performance and rejecting those in divergence. And, surprisingly, maybe only about half the stocks on my daily list actually met that category. Big surprise for me - I thought there would be more. And then as we got into February, the number started really dwindling (but of those that remained - like the Agricultural Chemical stocks such as TRA, were *choice* and I made really nice money). But most breakout stocks were hitting new price highs, but the MACD-H wasn't doing the same. Since I was only looking to pare my list down to a few long candidates, I didn't really mind - and not really being all that experienced with following MACD and MACD-H, I didn't really recognize the significance - those stocks that weren't matching their MACD-H to their price rise were in a divergence, and, overall, the share of breakout stocks in a MACD-H bearish divergence was becoming significant.

It was only after February 27 that that significance became apparant to me.

The lesson I've taken from this is that MACD-H divergences are quite important - not just as measuring the "health" of a single uptrending (or downtrending) stock, but also, when observed daily in a universe of uptrending (or downtrending) stocks, as an indicator of the "health" of the market as a whole - the higher the percentage of stocks in the breakout group with MACD-H behaving, the stronger the overall prevailing trend is - the higher the percentage of stocks in the breakout group with MACD-H divergences, the more tenuous the prevailing trend. If I had only figured that out before February 27!
scottnlena
Posted : Tuesday, March 13, 2007 11:54:14 PM

Registered User
Joined: 4/18/2005
Posts: 4,090
So let me see if I got this right. You are looking for stocks to be in agreement with macd. but why is MacD-H (histogram?) more improtant than a mac d... dont they display the exact informatino but in a diferent visual way?

Stocs moving up while MacD in general is going down (or averaging down) would be a negative divergence... and would be avoided... possibly poasted to a short sell watch list.

Stocks falling on a rising MacD would be in a positive divergence and represent stocks with underlying hidden strength.... those presumeably would be ideal long entries? Or all divergences are rejected on the grounds that Macd is misbehaving and there for can't be trusted.

Is there any concern that widely published strategies will become less effective as more people use them? If for no other reason than the predators know the ways to beat them? I have been playing with MacD my self and find it ... historically anyway a good timing indicator.... but settings may vary for each stock.

Have you played with the marekt indicators in BLOCKS yet? some realy neat stuff there... % stocks above their moving average is pretty nifty. You can set the moving average period....and the watchlist sampled. Seems like a great market timing tool.... works similar to MacD.
scottnlena
Posted : Tuesday, March 13, 2007 11:59:33 PM

Registered User
Joined: 4/18/2005
Posts: 4,090
So in general it might be a good strategy to favor divergences over nondivergent entries... but not to exclude non divergent entries. however grooming down to only divergences might focus a watchlist into more potential power. I'm speaking in terms of preparing a list for trading int he future... as stocks begin to move then indicators confirming movement would be preferable.
scottnlena
Posted : Wednesday, March 14, 2007 12:34:42 AM

Registered User
Joined: 4/18/2005
Posts: 4,090
hohandy,

how would you view AVT? TSV 24 is divergent as I see it. MacD (with the same settings you mentioned under the ask atrainer thread)would apear to be divergent set as reg macD or histogram..to my thinking this is a prime short example.
diceman
Posted : Wednesday, March 14, 2007 12:51:56 AM
Registered User
Joined: 1/28/2005
Posts: 6,049
I done see it.

This looks like the to 2 incorrect examples
I gave in my post.

(of course it could always go down due to
market conditions)

Thanks
diceman


hohandy
Posted : Wednesday, March 14, 2007 12:54:11 AM
Registered User
Joined: 12/21/2004
Posts: 902
Aspll - I approach things only from the long side (looks like I'm going to have to learn to change my stripes though) - so when I was looking for divergences, I was solely looking for confirmation of long candidates indicated by the absence of a divergence. Look at a daily chart of TRA beginning around November - beautiful chart from a price standpoint. But look at the MACD and especially the MACD-H. Price was steadily rising in a nice orderly manner - nice upward legs with sideways consolidation - nice regular support at the 13ema. The MACD itself for most of the period was slowly rising, but nothing really special except that it stayed about zero. But look at the MACD-H - beginning in November, with each price leg rise, the MACD-H "bump" would rise even higher than the previous leg, indicating that the trend was growing stronger as it went higher. And from a "geometric" standpoint, the area in the MACD-H when it was above zero was consistently much larger than when it was below. These are indications of a strong stock with an intact and growing stronger trend that wouldn't have been really apparant from just looking at the MACD lines themselves. From a standpoint of identifying and filtering long candidates, I've found this extremely attractive. And if one was looking foe short candidates I'm sure this would work too to confirm downtrends.

But onto divergences - what I talked about in my earlier post - I was sort of surprised that of my list of daily "breakout" stocks" when I started looking in January, that fully about 1/2 of what should have been attractive stocks indicated a bearish MACD-H divergence. I didn't do anything with them, but just noticed as time went along that fewer and fewer of the daily breakout list had "quality" MACD-H to match the price rises. The point of my post was that at a MACRO level, I thnk if one follows this over time, one could possibly predict or anticipate strength levels of general market trends (as opposed to using the divergences on individual stocks at the MICRO level). But on an individual stock level, theoretically bearish divergences should indicate a price drop. Dr Elder talks about divergences on longer time-series charts tend to be strongest indicators.

As to the MACD vs MACD-H - I'm not good at theory so I'll go broadbrush and maybe some of the others on here can fill it in better - but I know that the MACD lines themselves are 2d derivative of 2 price series - a fast one compared to a slower one. When the fast one is higher than the slower one, there is an uptrend, when it's lower, there is a downtrend. The MACD-H measures something a little different - the difference between the fast and slow MACD lines - this measures the strength of the movement - if the immediate movement is sharply up (say 4-5 good up days in a row), the shorter term MACD line will be much higher than the slower one and the MACD-H level will be higher. As the price movement slows, the slow line approaches the fast line and the MACD-H level drops. If they become equal (i.e., no shorterm price movement strength) it drops to zero and then into the negative. The MACD lines themselves can be well above zero, but if the price movement itself has stalled, the MACD-H can be below zero. So they do show more or less different things.
hohandy
Posted : Wednesday, March 14, 2007 1:11:02 AM
Registered User
Joined: 12/21/2004
Posts: 902
re: AVT - on a daily chart, AVT is in a MACD-H (and MACD) divegence - however on a weekly chart, the divergence isn't as pronounced and the MACD itself is still rising (red flag for shorts according to Elder).

Elders' rules basically set up conditions that must be met - weekly MACD and 13EMA must be in the direction that you want to go, a volume indicator such as TSV should be in the direction you want it to go, but preferably be below zero. You use the weekly charts to pick your stock candidates, the daily for timing of entry. I don't have experience with shorts, so I'm not a good person to ask as to what other criteria should be met in picking a good short. If I was considering this for a long stock, I would pass as the daily divergence definitely indicates weakness, if not an actual short candidate - sorry for being wimpy here


OTOH, I personally think the market is going to tank after failing the retest of the March low tomorrow or Thursday, so if I was going short the most important indicator that I would look at right now is the Beta.
Apsll
Posted : Wednesday, March 14, 2007 7:47:34 AM

Registered User
Joined: 3/21/2006
Posts: 4,308
Hohandy:

You have shared a very powerfull tool with us here today, and I for one am very glad that you have poped onto the scene.

It sounds however that Diceman is not fully convinced of this indicators ability to gage the strength of a given trend or to uncover its weakness, since I highly value his opinion in these matters I would feel better to have his full endoresment.

Either way I hope that you will decide to stay on and become a regular contributor to this forum. Your obvious experience and indicator knowledge are much needed and appreciated...

Apsll...
Craig_S
Posted : Wednesday, March 14, 2007 7:55:04 AM


Worden Trainer

Joined: 10/1/2004
Posts: 18,819
To see how MACD is constructed and what a MACD histogram really shows, check out this video: Understanding MACD

- Craig
Here to Help!
hohandy
Posted : Wednesday, March 14, 2007 8:51:16 AM
Registered User
Joined: 12/21/2004
Posts: 902
Thanks Apsll - very nice of you. I've never really paid much attention to MACD until this past winter when I read Elder's books and I've just really started using it and finding uses for it but I'm not necessarily wed to it - and I share your respect for the Diceman's experience (I periodically lurk so I'm familiar) so it's not something I'm living and dieing by - and I would certainly use it in conjunction with other indicators in my aresenal (until I become disenchanted with it LOL).

But in the limited context of the question posted by scottnlena in the beginning of this thread - I have found the MACD-H divergence (or lack thereof) very useful in the few months that I've been using it on both a micro and macro level. I would recommend Elder's books - especially his last two - Come Into My Trading Room and Entries and Exits - as very interesting and informative reads.
rwstic
Posted : Wednesday, March 14, 2007 9:33:02 AM
Registered User
Joined: 12/25/2004
Posts: 51
I've found divergences useful for endorsing or warning midst a trend, but not as a trade trigger. In other words like overbought or oversold, a divergence may persist a very long time and become greater and greater without a reversal. In contrast a security with divergences warns face of the move may be unreliable and masking greater risk.

I say divergences clue to risk...but trading on divergences is so ever risky as to perhaps not be better than random walk. That's due to undependable timing of any reversal based on divergence! It's my experience and what I find if I walk a historical chart forward from past attempting to paper trade a security without first seeing how the divergence plays out. They don't look as "good" that way as trading trigger as they do analysed by a larger chart view seen in retrospect. Try it!
dcostello
Posted : Thursday, March 15, 2007 5:19:19 PM
Registered User
Joined: 11/3/2004
Posts: 39
hohandy: You might want to look for Elder's first book, Trading For A Living. It gives some additional detail not found in Come Into My Trading Room.

scottnlena
Posted : Thursday, March 15, 2007 6:59:13 PM

Registered User
Joined: 4/18/2005
Posts: 4,090
I remember getting that from the library and not realy being floored by anything in there. A big section on the psychology of trading some TA. and a few strategies as I recall ... the only one that i was really after was the range bound stock strategy...... but it has been some time.. perhaps I should revisit them.
hohandy
Posted : Thursday, March 15, 2007 7:20:11 PM
Registered User
Joined: 12/21/2004
Posts: 902
QUOTE (dcostello)
hohandy: You might want to look for Elder's first book, Trading For A Living. It gives some additional detail not found in Come Into My Trading Room.



Thanks for the suggestion - but yes I've read that one twice - but I liked his other 2 better - especially the last one!
scottnlena
Posted : Friday, March 16, 2007 12:01:47 AM

Registered User
Joined: 4/18/2005
Posts: 4,090
Hohandy

which Elder title is your favorite. I'm reading an options strategies book right now... and my brain is fighting it... I think I prefer the stocks.... cant seem to get enough of that. Recently finished "technical analysis of stock trends" by McGee then read one of Don Wordens books. Actually liked the Don book better even though it dosent get into deep technical analysis, and devalidates allot of McGees work..... probably legitimately so.
Users browsing this topic
Guest-1

Forum Jump
You cannot post new topics in this forum.
You cannot reply to topics in this forum.
You cannot delete your posts in this forum.
You cannot edit your posts in this forum.
You cannot create polls in this forum.
You cannot vote in polls in this forum.