[UrmaBlume]

We do not trade longer than the session and usually consider reference to longer time frames more of a distraction than a help but the magnitude of late in the divergences between longer term price movemtments and buying and selling volumes is historic.

On the second of May we made an intermediate term high in the S&P. At the same time, as reflected by the chart below, there was a huge negative divergence between price and our indicator of Net New Trade. For the next 2 months price clung to the areas of this high while NNT continued to reflect an increasing divergence between trade volumes/order flow and price action.

Price broke and made a bottom on the 9th of August. The S&P then rallied for about 40% of the total drop into a zone of consolidation.

During that rally of 109 S&P points NNT rose by a total of 87,082 contracts. Of note is that in mid April there was a smaller rally of 83.25 points and the attendant rally in NNT was 350,829 contracts. This divergence could prove to be as significant to pice as the divergence that spawned this most recent break in price.

While this sort of notation is way beyond the time frames of consideration of the intra-session trader it is useful background for traders of any timeframe. Of note also is the old TA axiom that as a rule a horizontal consolidation is a formation of continuance, not reversal, around 2/3 of the time, and that further the move out of the consolidation is usually in the same direction as the move into the consolidation and of equal or greater length.

For the trading done by me and mine this is an observation of a time frame that is of the LEAST importance but the nature of the formations and the size of the divergences are hard to ignore.

Please click to enlarge image

Cheers

UrmaBlume

[UrmaBlume]

While we seldom play in these higher time frames, a new low in Net New Trade within the session usually foretells a new low in price.

Today's day session trade in ES summed to -87,575 contracts which produced a new low in the longer term NNT as shown below:

Please click to enlarge image

cheers

UrmaBlume

[getwore]

UB,

Thanks for posting. This is interesting and different perspective that i used to.

[clmacdougall]

Pat, when it comes to changes in net new trade are you basing it on differences in average cumulative volume for a time period or are you actually basing it on buyer seller imbalances? Thanks, Cory

[UrmaBlume]

Originally Posted by clmacdougall »

Pat, when it comes to changes in net new trade are you basing it on differences in average cumulative volume for a time period or are you actually basing it on buyer seller imbalances? Thanks, Cory

Cory,

We think that of the time/price/volume continuum, time is by far the least important and is given practically no weight in any of our calculations. The only time based chart we ever use is the daily shown here.

The only place where time motivates price, that we know of, is related to premium decay in certain derivatives. The passage of time does not motivate change in price, it is the occurence of trade that motivates price. More specifically it is an imbalance in that trade that is the prime mover of price.


cheers

pat

[clmacdougall]

Thanks Pat

When I check out your website it seems to break down imbalances in normalized trade and normalized volatility in your scanset 2.0. Are both of these metrics based upon imbalance in the now or are they based upon a comparative behaviour of x number of days worth of average volatility / volume. The reason I ask is wouldn't the chosen number of days in your average begin to skew your results/analysis if the nature of the markets change in the short term.

[UrmaBlume]

Originally Posted by clmacdougall »

Thanks Pat When I check out your website it seems to break down imbalances in normalized trade and normalized volatility in your scanset 2.0. Are both of these metrics based upon imbalance in the now or are they based upon a comparative behaviour of x number of days worth of average volatility / volume. The reason I ask is wouldn't the chosen number of days in your average begin to skew your results/analysis if the nature of the markets change in the short term.

Cory,

To some degree, here, we are starting to talk about both apples and oranges. Scanset 2.0 is more about target/instrument selection than it is about trading protocol.

The operative terms when talking acout this application are Relative and Normalized.

It takes an unusual commitment of capital to produce unusual relative relative volume and it takes unusually imbalanced trade to produce unusual relative price volatilities.

To make these calculations relative they must first undergo a process we call "serial normalization," that is normalized for instrument, time segment and time of day.

Since all of the target instruments are normalized under the same process then changes in the longer term market becomes irrelevant as the results are relative rather than absolute. 200% of relative volume still compares the same way to an instrument with 80% of relative volume regardless of the time/days considered.

This kind of processing is an essential part of preprocessing inputs to the development of intelligent agents and data mining operations.

cheers

pat

[clmacdougall]

Thanks Pat, my learning curve is still steep. Forgive my ignorance!! I'm learning to identify imbalance and entrance of strong anti trend participation. Thanks for the reply!