UrmaBlume

BlowFish, If "when" refers to time of day, the answer is anytime that the price and momentum models these guys use trigger - and that can be anytime. They also occur with some regularity during the night session but the thresholds and other settings are entirely different and are adjusted according to time of day normalized volume measurements. These measurements are seen in Blue on our HUD. I couldn't be more delighted that the brightest among you, like Blowfish and the others that are actually doing some work in this area, are finding value in our work. cheers

Here are some shots from yesterday's trade that show an upgrade to the original - both of which run in Trade Station. http://www.traderslaboratory.com/forums/f34/commercials-very-active-today-filtered-intensity-6064.html#post66410

BlowFish, Thanks for the kind words. Yesterday certrainly was a great day for those that use this type of indicator. Ours finds 6 -12 such trades per day session in the ES. As to the color you will notice blue bars at the tops and red at the bottoms. This demonstrates rare instances when what is shown as selling volume is really buying by the commercial and vice versa. These bottoms (reverse for tops) are not made by traders suddenly taking the asked they are made when downward momentum runs into commercial buying that is bigger than the market at that moment in time and price. This size is executed by automated order placement and the size is almost never shown on market depth. We also read this on our HUD as CPM (contracts per minute) which is measured as of every transaction and with time granulated to milliseconds. During a normal day session in the emini if CPM is < 3k price is not going to go very far and begins to have more drive above 6k. cheers

We have been working on our indicators of the intensity of trade and have developed a couple of new filters. Today the commercials were more active than normal. We found 13 trades based on intense commercial activity - 11 were profitable here are charts of the 3 best of the day (Times are PST). Hint for filters - size and threshold.

Right on BrownsFan. Price is not motivated by the passage of time, but rather, is motivated by the occurence of trade and more specifically an imbalance to the buy or sell side within that trade. Tick charts record transactions and treat each transaction equally regardless of size. When measuring trade or the imbalance in trade how useful can a chart that treats a 1,000 lot the same as 1 lot be? Below is a Volume Chart (4k) of the last half hour of yesterday's trade in the emini S&P showing flat prices and an indicator which is our calculation of net trade by commercial traders. While prices are flat the indicator, based on an imbalance of trade, shows net distribution - THE PRO's WERE SELLING THE CLOSE. The result was an immediate drop that continues as I write. Price does not predict price - it is the balance or imbalance in order flow/trade that motivates price and in most cases it is apparent before price responds. That imbalance is best read on Volume Charts not price or tick charts.

Programmer, You couldn't be more spot on. Thanks for the chuckle. I do a bit of training and have heard all of that buschwa again and again before the wannabees begin to learn. If it was that easy, you couldn't get a cab. cheers

I am not implynig, I am making a statement born of experience that most trade during the nite session is done by commercial traders, many outside the US. Also much of this trade is to either hedge or speculate on equity indexes other than the S&P. As to the 15k, this number is not an arbiter but rather a zone of interest. The precise over and under here is a matter for each trader to reasearch- I also didn't mention what method I used to differ UpVol from DnVol - I gave the hint - you should do at least some of the work yourself. Simple research will allow you to either verify my number or come up with a better one. The point here is that there is valuable information available from night session trade and that of that information, most important is the size and balance of trade during that period. cheers

"How to spot the beginning of a Trending Day?" While this is written for the emini S&P the same principles apply to most other markets. Before the open: Night Session trade is mostly done by professional traders which gives us a very good idea of professional sentiment in world equity markets. Night session trade in this contract currently averages around 250k contracts. The first clue of a trending day is above average trade - 300k or more. If price trended during the night session on an above averge trade this is your second pre-open clue of a trending day. If the difference between UpVolume and DnVolume is more than 15k contracts during the night session this is your third pre-open clue of a trending day. Of note is that over the long term over half the time one extreme of the 24 hour session is made during the night session. The open: Again over a fairly high percentage of the days the open is near (20% of the day's range) one extreme or the other. On days where price passes through the open more than a few times during the first hour trend days are unlikely. On the other hand on days where the three pre-opening conditions mentioned above are in effect and price departs decisively from the open during the 15-30 minutes of trade you have a faily good indication that the open has established one extreme and that trend trading away from the open will establish the other.

The examples posted were all taken from TradeStation. The problem with TradeStation having a time stamp that is only granular down to the minute is overcome by accessing the windows kernel. Ticks are useless as they don't count volume only transactions and without volume how can you determine intensity? cheers

Michal, Thanks for the kind words. We are a small private group of traders, programmers and poker players from the US and the EU. I am the old man of the group, 65, the rest are mostly in their late 20's. We each bring different points of value to the group - technical skills, market experience, trading skills and/or poker skills that we find to be most useful in today's hi-frequency enviornment. In spite of a busy schedule running our company, trading and writing, I do plan on making more posts to this board and I hope that you will continue to find them to be of value. cheers UrmaBlume

We believe that each instrument is a market unto itself and from there we try and discover the interests and activity of the major participants. In the case of the stock index instruments we believe that the commercial speculator is the force that drives price and we make efforts to track his trade via such indicators as Trade Intensity, Buy/Sell Volume Harmonic, Net New Commercial Trade and others, most of which have been disscussed in previous posts to this board. cheers

As both a poker player and a trader - game theory and predictive reasoning plays a big part in what I do. In the clip bleow a very smart man provides insight into basic game theory and predictive reasoning - insights that, for those that can handle it, provide a glimpse into a new dimension in trade decision support information processing. http://digg.com/d1o9iD

Thanks Michal, From my post you can see that I used that indicator almost 30 years ago. It is dated and with today's volume numbers would need to be re-engineered. However, if you are still interested in the MBI it is, as explained in my post, fully disclosed in Humphrey Lloyd's book - "The Moving Balance System." Today we use the indicators and others that I have demonstrated in other posts to this board and we trade in micro trime frames which was not possible for off-the-floor traders 30 years ago. In those days there was no real-time volume information, execution was by phone and very slow, there was no real-time online accounting and even discounted commissions were several times what they are today. Today we have complete online data, electornic execution, member rates for commissions and a plethora of state of the art data mining technologies for analysis. With all of that some of my more aggressive young traders get up to 150 round trips per session in the S&P. That picture was taken over 20 years ago in a marina in Hawaii after a 17 day crossing from San Diego. Good Luck with your trading. cheers UrmaBlume

In the ultra short-term we find a very high correlation between price and the short term surges in buying and selling as reflected in the pie charts on the application below. Below is a shot of our Market Heads Up Display - HUD which we belive is indicative of the non candle, non bar, non profile future of trade decision support processing and presentation. This application is written in C++ and is driven by data exported from Trade Station via easy language and our own dll. Below are some screenshots of indicators that often lead price. They are all calculated from different methods of volume analysis and price is no part of their calculation. While the charts below demonstrate their efficacy as trading indicators we believe their optimal application will be as inputs to high-frequency automated trading system driven by logic developed by a cascade of technologies such as neural networks, polynomial generators, genetic optimization routines and trade rule generators. While my younger traders use a more robust setup w/8 screens this is the setup I use for my personal trading: This first chart is a calculation that when run on constant volume structures produces a "Harmonic" of buying and selling power. The chart demonstrates that oftentimes turns in the volume graph precede the turn in price. This charts shows how we calculated net new trade this session by commercial traders. The object of the indicator is project accumulation or distribution, which we consider to be the same as money flow. Here you can see the accumulation under flat price movement prior to a fed announcement. You may also take interest in the indicators of trade intensity which sparked some discussion on this board. Keep at it Rusty - you can't be ahead of the pack if you are following it and the pack never makes it - only those ahead of it. cheers

I couldn't agree more. It is shocking that with data processing technologies smart enough to track a missile in three dimensions at Mach 2, that there is the slightest problem, for all but a select few, in predicting the price of the S&P a few minutes from now. Almost all of our work is starts with your premise that there is a huge gap between the state of mainstream technical analysis today and what it could be today if its practicioners availed themselves of the benefits of modern, intelligent, predictive data processing technologies. My first forray in to this area was almost thirty years ago and based on an indicator derrived from the advances and declines and the up and down volume on the NYSE. The "Moving Balance Indicator" is fully disclosed in a book by Humphrey Lloyd called "The Moving Balance System." At that time the equity options on the CBOE were relatively new and the S&Ps were yet to come. I traded a crossover of the MBI and its five day simple moving average by buying puts and calls on high beta stocks whose options were being traded on the CBOE. I was young, living in the mountains and learning to trade. The work gave me the jump on the huge bull market in the early 80's and bought me my first yacht - pic below. Good luck in your quest. You can see other such non-traditional indicators in some of my previous posts to this board.

Zen-Fire is very fast real time data. You can get it for free with Ninja from any of several brokers. My internet connection is a couple of steps higher than t1 and I use Trade Station and one other service I would rather not name. It's getting late in Germany - time to take your meds and go night-night.

I want what he smokes.

For me there is a difference between the definition and the calculation. I have introduced a concept. I have provded my definition. I have indicated how others make a similar calculation, I have provided hints on how I make the calculation and have even provided clues as where to look for answers. The rest is up to you. Don't forget - We all have access to the same base unit of data, the transaction/tick. What makes all the difference for all of us is how we process those ticks.

There are several algorithms and methods deployed to spot, ever more elusive, commercial trade. Some do it by transaction size alone, trade intensity is another while still others deploy variations of the motivated buyer/seller themes. As to our definition of commercial trade, we define it as the speculative, not arb, trade done by professional, size traders. Our calculations use some of the methods mentioned above and others designed to pick out or more heavily weight commercial trade that is desinged to span the current session. Knowing where these guys do their spec trade is all you need because it is this trade that sets many local extremes. As I have mentioned before these traders go to great lengths to disguise their entries and exits. The use automated executions sometimes separating orders by only a few miliseconds and sometimes even putting up opposing trades to do as the old timers used to say, "sell a little to buy a lot." While they can hide their methods and whether the trade was an opening or closing transaction they CAN NOT hide the fact of their transactions and that's where the clues begin.

The CTI2 data is from days of old when Peter Steidlmayer was still very influential at the board of trade and reflects trade by known commercial traders. Our calculation is broader than that and is designed to catch the trade of traders such as hedge fund operators which certainly fit the description of commercial traders but at the same time would not met the criteria of CTI2 trade. Almost 3 decades ago I was an avid follower of Peter's work and spent weeks at his ranch in Butte Meadows forming what has become the base of most everything I do. Along these same lines you might find value in a post I made on "Order Flow - Then and Now" at http://www.traderslaboratory.com/forums/34/predicting-breakouts-accumulation-distribution-5289-2.html#post60085 Thank you for the kind words, I am glad you found some value in my work cheers

Accross the top are Session High and Low, Volume and Net New Trade by commercials which I have discussed before. The red, yellow, and green trade signals are taken from indications of price and net new trade on 200, 400, 1,000, 2,000, 4,000, 8,000, 12,000, 18,000, 25,000, 60,000 and 120,00 contract volume charts. The blue is time of day normalized average volume over previous 62 days presented for 6 intra-session time frames. The pie charts show the amount of up volume and down volume in in the most recent 2,000, 4,000, 8,000 and 12,000 contracts traded. The red and green bars to the right of the pice charts show our measure percentage of time of day normalized trader commitment over 6 intra session time frames over the past 62 days. Bottom left is a graph of the net new trade number shown at the top and the bottom right is a Jurik MA smooth of the mid point of prices taken from a 500 contract volume bar in the ES.

In our trade we pay close attention to the night session. As markets, especially equity markets, become more global, money managers have a wider choice of derrivatives with which to both hedge and speculate. Because the S&P is the most liquid of them all, night session trade is is done with vigorous foreign commercial participation. What makes this trade of note to the day session trader is that on a majority of days one or the other extremes of the 24 hour session is made in the night session and most all of the trade done in the night session is done by commercial/professional traders. If the night session reflects strong conviction in one direction or the other there is every likelyhood that the same sentiment will be operative during the day session. The screenshots below were taken 2 hours before Monday's 3/2 open. The first graphic below is of our Market Heads-Up Display which shows night session volume is significantly above average in all time frames. It shows sell signals in both the very long and very short intra-session time frames. Also you can see the long term trader commitment is negative by a factor of 2-300% in the longer term intra-session time frames after a sharp rally during the previous 12 minutes. Net New Trade shown at the top is near its session lows and is of a level to predict more selling to come during the day session. The second chart is an overlay of both the day and the 24 hour session in the S&P. It is 2 hours before Monday's 3/2 open and the open of the session is at the high and the current price is near the low after having taken out the November lows on Friday.

We use Trade Station. I am the oldest contiuous user of products developed by Bill Cruz. I even bought his Temp S&P system almost 30 years ago before there was TradeStation, SuperCharts or Omega Research. TradeStations does have some issues. For us many are centered around the lack of granularity with the TS time stamp. For applications like our Market HUD we use dll's to export the data to applications we have written in C++. For finer granularity in the usage of time we access the windows kernel. A couple of months ago we investigated the possiblity of changing to Ninja and ZenFire data. We were able to speak to management at both Ninja and ZenFire and found the ZenFire data stream to be great but that the way Ninja handled the data stream made the data not suitable for our purposes plus Ninja is very young software and needs a couple of more versions before it will be ready for real power users. The best we got out of the discussion was meeting with Mirus futures. Great attitude, great rates and super day margins. They even support member rates and the CME volume incentive plan that gets total round trips in the ES down to $1.80 round trip. While Ninja is not yet ready for our level of trade decision support processing we may have to get it just for executing trades at Mirus and keep Trade Station for the analysis part. And thanks again for the kind words.

This is an entirely different discussion that happens to use the same chart. Because it is different it is not appropriate to post it on the other thread.

Before computerized exchanges pit traders held one huge advantage over the public and other off the floor traders - THEY COULD SEE THE ORDER FLOW. In the pits to signal a buy the trader showed both hands palms inward - one hand indicated price and the other the size. Palms out to sell. Orders come from either outside the pit via the phone banks or from traders inside the pit. Inside the pit are several different kinds of traders - individual speculators, brokers, locals and most important of all - the commercial traders. These commercial traders execute large size speculative and hedge trades for the big firms and funds. Everybody in the pit makes sure he knows exactly who these guys are and knows that when the commercial trader puts his palms out to sell that he probably has enough selling to do to put in at least a short term top. With this understanding, should a trader look toward the phone bank and see a sea of palms he knows that a wave of selling is about to hit the market. At the same time should he see a known commercial with palms out he knows size selling is coming. In our trading we use an indicator of session net new commercial trade as shown below. This chart shows several hours before an announcement from the Fed - While price was completely flat - accumulation by commercials was obvious. While it is easy to miss the excitment and energy of the floors of old at the Merc and CBOT - we online traders have it better than they ever did.