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 |

Negative divergence Rate this Topic:
Previous Topic · Next Topic Watch this topic · Print this topic ·
bittwiddler
Posted : Wednesday, January 12, 2005 3:51:48 AM
Registered User
Joined: 12/29/2004
Posts: 7
Hi,

I thought that "negative divergence" meant that the indicator was moving down, while price was moving up.

If that definition is correct, I don't understand many of the comments that Mr. Worden makes in his daily reports. He'll remark on the negative divergence of tsv, yet when I look at the chart it seems to me that tsv and price are moving in the same direction.

Any clarification would be greatly appreciated!

BT
Craig_S
Posted : Wednesday, January 12, 2005 6:34:00 AM


Worden Trainer

Joined: 10/1/2004
Posts: 18,819
Negative Divergence only means that the indicator is more bearish or less bullish than price. You give one example in you post above - price is moving up (bullish) and the indicator is moving down (bearish, for some indicators).

Here are some other possibilities:

Price moving sideways while the indicator drops (Price is not going up but is more bullish than the indicator).

Price is dropping and the indicator is experiencing a worse drop. This is not measured by visually looking at the slope but by comparing current levels to previous levels. For example, if the indicator is making new lows and price is not... that is a negative divergence.

(all of the above assumes that the indicator going up is bullish and going down is bearish)

Divergence can be more subtle than just one up and one down. Divergences can occur when one makes a new high, the other fails to. One makes a new low while the other fails to. I would HIGHLY recommend watching Peter Worden's videos in the strategy forum.

CLICK HERE for Peter Worden's posts in the Strategy forum.

You might also enjoy his videos and chart reading

CLICK HERE for the Peter Worden videos on CD-ROM.

The examples I gave don't even scratch the surface on divergence. Sometimes it can be very obvious (Price makes new highs while MoneyStream makes new lows) or very subtle. Go back and look at the reports and the charts he is referring to. If you look carefully can you see where the price is less bullish than the indicator?

- Craig
Here to Help!
Tanstaafl
Posted : Wednesday, January 12, 2005 5:58:37 PM
Registered User
Joined: 10/7/2004
Posts: 799
Location: Duluth, GA
Hi, Craig:

I hope you won't mind if I differ with one of your comments, that:
"Price is dropping and the indicator is experiencing a worse drop" is a legit example of a negative divergence. For reasons having nothing to do with Technical Analysis or interpretive opinion, that is not a correct statement. However it *is* an impression that many people have, so please bear with my explanation below, which I hope will clear up the matter.

The Price and the Indicator are plotted on entirely separate scales, and the visible slope of the divergence lines, however they are drawn, is (in TC) a function of the Zoom and other factors. Change the zoom, and the slope steepness changes - this is ALWAYS true of the Price slope, and is ALSO true of the slopes of unbounded indicators such as Moneystream or MACD.

Also, the Price and the Indicator have entirely different sets of units. That, independent of the prior issue, also disallows direct comparisons of their values, or derivatives (like LinReg) of their values.

Therefore it is *never* legitimately useful to refer to the downslope of a price based line being "steeper" or whatever than the downslope of an indicator-based line, in the process of trying to find a divergence. To do so would be the same kind of mistake as if one were to plot the Indicator itself (such as Stoc or MACD or MS or whatever) on the Price pane, and make up rules about "crossovers" of Price and that indicator. Pretty, on the screen, but meaningless.

HOWEVER ... it *is* legit to compare the SIGNS (i.e the directions) of the slope of price vs indic, regardless of the scaling or the units. That is, if the Price is UPsloping (at any steepness), and if the Indicator is DOWNsloping (at any angle), then you have a legitimate negative divergence. Or vice-versa for a positive divergence.

You can, if you wish, consider an "exactly horizontal" line to be a simple case of either up or down.

There ARE ways of evaluating strength of divergence in relation to steepness, but they require that you compare apples to apples. I have found that two methods work quite nicely:

1) compare the steepness of, say, a 10-bar LinReg of an Indicator today to the steepness of the SAME 10-bar LinReg of that Indicator, plotted on the same chart with the same scale, as of 5 days ago. Its direction of rotation can be used as a "vote" for strengthening or weakening divergence.

2) compare the steepness of a 10-bar LinReg to that of a 20-bar LinReg, both anchored to today, on the same chart and the same scale. Again, their rotational direction can help evaluate strength of divergence.

Both #1 and #2 can be applied to price as well. Just be sure to keep it apples and apples re the scaling and the system of units.

This analysis method can be applied nicely both using pure charting and SortBy, or more directly with (somewhat complex) PCF's.

I hope this helps.

Jim Dean

StockGuy
Posted : Wednesday, January 12, 2005 6:50:20 PM

Administration

Joined: 9/30/2004
Posts: 9,187
Jim,

I don't think Craig could have been any more clear when he said:

"Price is dropping and the indicator is experiencing a worse drop. This is not measured by visually looking at the slope but by comparing current levels to previous levels. For example, if the indicator is making new lows and price is not... that is a negative divergence."



Tanstaafl
Posted : Thursday, January 13, 2005 10:07:23 AM
Registered User
Joined: 10/7/2004
Posts: 799
Location: Duluth, GA
I'm sorry if I misunderstood what Craig was saying. I believe that there is a bit more to it than that, when the "classic" divergence approach is used (which I *think* is what was being referred to, maybe), rather than the "Linear Regression" approach, which is what TC has implemented internally for MoneyStream, and offers as a tool for general use. My comments were with regard to that LinReg approach, for which there are no "lows" or "highs" applicable. So, in retrospect, I think that we are both correct, but just approaching it from different perspectives, using different tools.

The classic method for divergence analysis relies completely on eyeball interpretation, notably in the choice of pivot highs and lows to be used to draw the lines with. Since that is very subjective, I personally prefer the more automated, consistent approach that TC's nifty LinReg tools offer.

The classic approach, as I understand it from a number of publications, depends on the simultenaity of price and indicator pivots ... that is, the lows in question should be occuring on about the same bars. However, again, that is a subjective aspect of it that some may not agree with.

Again, I apologize for the misunderstanding.

Jim Dean

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.