Does Divergence Signals Really Work?

[UrmaBlume 10]

Divergence on price based indicators such as RSI, MACD and Stochastic on time constant charts do not carry the weight or are not nearly as effective/useful as divergences demonstrated by volume/order flow based indicators run on volume bars.

 

There have been significant divergences from such indicators on volume bar charts on at least one of the session extremes in each of the last five trading day sessions in ES as shown below:

Just click to enlarge any of these graphs to see txt, time & dates

 

 

 

 

 

 

 

 

Price makes a poor input to predict price and the passage of time does not motivate price. It is the occurence of trade that motivates the movement of price and more specifically it is an imbalance between buying and selling in that trade that spurs price.

 

 

Cheers

 

UrmaBlume

[Do Or Die 10]

Do or Die, in your days as a floor trader were you able to witness this phenomenon of price rejection during the consolidation phase of a trend move?

Hi Phantom,

 

Definitely yes, it's a very useful observation. I once met an ES trader who would draw next 4-5 candlestick bars on a notebook every time he felt a reversal is about to happen. He used to trade on 5 minute time frame. There is so much to properly reading the market, now people can see how tough it can be to automate trading.

 

I have a theory why the MACD will show divergence during corrective moves in a trend.

 

....

 

One can oftentimes see indication of price rejection (dependent upon the time frame one is looking at, of course) in the form of hammer bars while no such indication of price rejection exists around the "a" point.

 

However, you may like to rethink about your explanation on what causes divergences. Indicators such as RSI or MACD take only closing prices into account and candlestick shapes will not effect their outcome. I think we agree other than on candlestick shapes. I explained it in the terms of prices closing towards the limits of their past n days trading range Construction of RSI If fact, Relative Strength in context of RSI means simply the measure where a stock is trading in reference to its past range.

[phantom 16]

However, you may like to rethink about your explanation on what causes divergences. Indicators such as RSI or MACD take only closing prices into account and candlestick shapes will not effect their outcome.

 

Quite the other way around. Closing prices will affect the outcome of the candlestick shape. The hammer is merely a representation of the fact that the closing price of a bar is near its extreme.

 

Its looks like we are actually in complete agreement here...

 

 

Phantom

[hunter1 10]

I have not read the whole thread since I just discovered it so maybe it has already been said.

 

Divergence does work. Reverse dvg which signals trend continuation gives longer moves on average than regular dvg which signals potential trend shift.

 

Trend is chart specific. So trend on a 1 min chart maybe counter to trend on 5 min or 15.

 

I use volume charts. Dvg based on oscillators alone is not that profitable . If you wait for the osc to actually turn over or cross etc to give a trade signal you may get into a trade to late. This is because oscillators lag.

 

The best dvg is that which does not depend on oscillating indicators. Oscillators can provide you with a accurate read of where you are in the current cycle however the dvg you are looking for is the dvg between price and order flow. This does not lag.

[phantom 16]

RE: "Price makes a poor input to predict price and the passage of time does not motivate price."

 

W.D. Gann himself disagreed with your statement, UrmaBlume.

 

He said that "when time is up, the market MUST change!"

 

I think I'll go with Gann on this one.

 

 

Phantom

[phantom 16]

It is the occurence of trade that motivates the movement of price and more specifically it is an imbalance between buying and selling in that trade that spurs price.

 

So, if Florida orange crop experiences a severe freeze, the limit move on the next opening was due to the trade imbalance, and not the weather???

 

Many would argue that fundamentals drive price.

 

 

Phantom

[UrmaBlume 10]

RE: "Price makes a poor input to predict price and the passage of time does not motivate price."W.D. Gann himself disagreed with your statement, UrmaBlume. He said that "when time is up, the market MUST change!"I think I'll go with Gann on this one.Phantom

 

Not me. I don't believe the passage of time influences price except in the case of premium decay. I believe that changes in price are not motivated by the passage of time but rather are motivated by more buying that selling or vice versa..

 

It doesn't make sense to me that becasue a certain amount of time has passed that prices must reverse themselves.

 

Every single trading day I can demonstrate instances of where as buying volume increased prices rose and vice versa. I chanllenge you to show me a unit of time that changes price in the sub-session time frame on any consistent basis.

[UrmaBlume 10]

So, if Florida orange crop experiences a severe freeze, the limit move on the next opening was due to the trade imbalance, and not the weather???Many would argue that fundamentals drive price.Phantom

 

DOH - Obtuse to the max.

 

The freeze itself didn't do anything more than produce the buyers when created an imbalance in trade and it was that buying that drove prices not the fact that it was cold in Florida.

 

If an orange juice producer from anywhere in the world received a huge order and he decided to secure his profits by buying in the futures market - if he bought enough certainly it was supply and demand and in this case buying produced by demand that drove prices.

 

To say that buying and selling imbalances don't push prices to me is a huge non sequitur.

 

 

UrmaBlume

[phantom 16]

I figured you'd respond with something similar to that...

 

Keep trading. I'll be glad to take your money.

 

 

Phantom

[zdo 20]

mathematically...

Divergences using RSI

 

They can be quantified to a good extent (the analysis can be automated). but the problem is that coding them requires more than just familiarity with ninja/tradestation. I personally know people who have automated their discretionary trading systems and their code runs into thousand of lines.

 

Wasn't really talking about the issue of "quantifying" / coding with the use of the word "mathematically"

I brought this up because in my experience a large percentage of ‘normals’ reading about and looking for ‘divergences’ literally see price as strong and the indicator as weaker… when actually it was price action (of closes in rsi, macd, etc) in the interim btwn the price peaks that created the ‘inferior’ indicator reading.

'Mathematically', the interim action could have produced an equivalent, non diverging, extreme in the indicator and the trading signal still be valid.

ie In the debate about the value or efficacy of ‘divergence’ , absolute indicator readings are just as valuable as divergence readings…context…

It's not as 'mathematically' correct and it looks like he's backing use of indicators out of the mix altogether , but choicecap1 was getting at the same concept early in the thread with "Divergence is nothing more that price rejection faster at certain level,so if you can identify other indications of where to expect price to have this type of action(sop/resist) then you dont even need the indicator to find it."

 

Do or Die, I don’t think we’re really arguing here. For example you said,

“In fact, Relative Strength in context of RSI means simply the measure where a stock is trading in reference to its past range.”

which to me is ~= to

“‘mathematically’, … divergences are, in large part, created / made possible by the form, extent and duration of the most recent correction before the current thrust which is exhibiting indicator divergence …”

 

“form, extent and duration” would also establish limits on movement away from moving average that create MACD divergences, etc etc.

Edited August 2, 2011 by zdo

[Do Or Die 10]

RE: "Price makes a poor input to predict price and the passage of time does not motivate price."

 

W.D. Gann himself disagreed with your statement, UrmaBlume.

 

He said that "when time is up, the market MUST change!"

 

I think I'll go with Gann on this one.

 

 

Phantom

 

I never bothered to read/study about Gann techniques, but would agree with you to disagree on that statement written with hubris.

[phantom 16]

I never bothered to read/study Gann techniques, but would agree with you to disagree on that statement written with hubris.

 

It is said that Gann took $50M out of the markets...mostly in grains.

 

Very esoteric techniques but certainly effective, wouldn't you agree?

 

 

Phantom

[Do Or Die 10]

I need to study those techniques in the first place to agree or disagree

[clmacdougall 11]

It is said that Gann took $50M out of the markets...mostly in grains.

 

Very esoteric techniques but certainly effective, wouldn't you agree?

 

 

Phantom[/quote

 

It's impossible to compare trade done today during the information age with trade done during an age of relative ignorance. UB's insights into the market are in keeping with the increased flow of information available to us now. Your argument can't stand.

Gann used time and price because that was all there was available to everyone at that time. Now with more information available people don't need to use proxy's anymore, when they now have actual trading information available.

I'm sure if Gann were here today he'd be using the greatest depth of information available and not the "square of nine," technique he used back then.

[phantom 16]

Your argument can't stand.

Gann used time and price because that was all there was available to everyone at that time.

 

I don't care if he was rolling dice...

 

If he did pull 50 mil from the markets, I'd say that his methods were pretty darn effective.

 

If you don't agree with that statement per se, there's not much else for us to discuss...

 

 

 

Phantom

[clmacdougall 11]

I don't care if he was rolling dice...

 

If he did pull 50 mil from the markets, I'd say that his methods were pretty darn effective.

 

If you don't agree with that statement per se, there's not much else for us to discuss...

 

 

 

Phantom

 

You should care if he were "rolling dice," chance and luck are hard to repeat. As to whether or not his methods were effective in his day, what difference does it make, as it was back "in his day."

In the now his methods are rubbish and will leave you "hoping" your next trade will work out..

[phantom 16]

In the now his methods are rubbish and will leave you "hoping" your next trade will work out..

 

Whatever you say, MAC...

[clmacdougall 11]

Whatever you say, MAC...

 

Scroll up to the top of the page, you'll notice it says, "Traders Labratory." Everything posted on this forum is open to an objective style analytical debate.

If you don't want your thoughts or ideas debated or refuted simply don't post. Nothing personal, it's just the purpose of the forum.

[calsprdr 10]

You should care if he were "rolling dice," chance and luck are hard to repeat. As to whether or not his methods were effective in his day, what difference does it make, as it was back "in his day."

In the now his methods are rubbish and will leave you "hoping" your next trade will work out..

 

 

I'm not trying to dis you here, but what about Gann's methods are "rubbish"? Time analysis is very big in the Fibonacci trading methodologies, which are hardly considered out of date.

If you want to reject a person's ideas please explain *where* you think he is wrong.

[phantom 16]

Everything posted on this forum is open to an objective style analytical debate.

If you don't want your thoughts or ideas debated or refuted simply don't post.

 

I don't mind debate as long as the line of reasoning is somewhat logical. But when one makes blanket statements without any evidence to back them up, then one is simply arguing for argument's sake.

 

I happen to know several successful pros who use Gann's methods in their trading and it involves much more than just the square of nine.

 

I don't use the stuff because, and if you've followed my thread you'd know this already, I'm into simplicity.

 

Anyway, no hard feelings.

 

 

Luv,

Phantom

[clmacdougall 11]

I'm not trying to dis you here, but what about Gann's methods are "rubbish"? Time analysis is very big in the Fibonacci trading methodologies, which are hardly considered out of date.

If you want to reject a person's ideas please explain *where* you think he is wrong.

 

I believe that he is wrong in disregarding the fact that trade motivates predictive behaviour and not time. Gann focused on the motivation of time and that's no longer a leading factor in predictive technologies. The length of time a car stops at a fork in the road does not lead you to know where it might go next.

There's just too much more information available than price and time which can give you insight into where price might go. With hardly any movement in price, the internal structure and balance can change tremendously, this is a good way to predict where price might go next, not simply by looking at where price has been.