There are three possibilities for the way open interest changes with each trade in a futures market, depending on whether the long and short sides of the trade are closing out already existing positions, or opening up new positions. If both sides are opening new positions, open interest goes up by the number of contracts traded. If one side is opening a new position and the other is offsetting an existing position then open interest is unchanged. If both are offsetting existing positions then open interest is reduced by the number of contracts traded. The way I think of it is: every `trade' of a contract in a futures market actually creates a completely new contract between the long and the short side, whose initial present value is exactly zero for both parties. So in all trades, the number of existing contracts goes up. It's not really that the seller of a contract transfers that contract to the buyer. Instead the short and the long make a new contract, and the seller thereby obtains a new short position, which may or may not offset an already existing long position. So, when determining how much money will flow at the end of the day from the long to the short side of the market (or vice versa) it's not the total number of contracts long and short that matters, but only the net number of contracts on the long and short sides, those which have not so far been offset by contracts in the opposite direction. That number is the open interest. In the ES example, the open interest was 1541K in the March contract on the prior day, and the volume was 1032K. But we're missing an important piece of information before we can say what happened: how much the open interest changed from the day before. It's very unlikely of course, but in principle, if the open interest on the day before were 509K, all 1032K of that volume would correspond to new long and short positions that were opened up. If the open interest only changed by a number much smaller than 1032K, then the situation is different.
What is real about markets is how big they are and how money is made. In the ES, who is doing what is more important when looked at from the opportunity it represents. "E mini and the daily volume" is the topic. Some people look at the day's volume a a single number others are viewing it as a sum of many transactions. Those who held contracts previously purchased and still hold them do not show in the daily total volume nor the sum of the transactions during the day. Seeing the volume of 1 million accumulate as a consequence of trading has a consideration in it that is helpful in seeing the open interest level value's meaning. It is 1.5 miliion. I see it as a fairly steady value and most all of it outside of the consideration of what causes price movement from day to day and what causes price movement within the day. I think that what is important is to have a handle on daily trading and especicially the trading during the day. You view different things as important, perhaps and I was commenting on the logic of the considerations. This is pseudo bullshit to you and that is important to express. I feel that I must consider volume to be able to make money. I try to make money without affecting the flow of price. The million transactions come from the bars of the day within the day. Their accumulation does define, day aftr day, who is holding long term, but this is known only indirectly by coding the market volume. Posting that kind of stuff is more BS if it comes from me. I find that any matrix derived from using volume is powerful and important for making money. By combining volatility and volume, it is even more powerful for seeing money being made in ES. The market operating point as a function of volatility and volume per unit time, is an incisive real and real time market evaluation technique. You can make up an EOD version of the attachemnt for a whole 3 month run of the ES contract. Do it seasonally as well so you can contrast the seasons of the year. The attachment of the ES for a month of ES during the ESM6 contract is for 5 minute durations instead of the EOD level that you desire for a whole period's variation. You will see from what you make, the relationships you desire. It will be just as elegant as the 5-minute period seen over a month (1606 data elements) I use it to deal with propensities during parts of days. Your version will reveal propensities during parts of months over the three month run. By having the four portions of the year (3 months) each you can see how the current Q is unfolding a day at a time and keep a tab through the current Q. Others have told you about how the days interrelate from the beginning of the Q. By tabbing, you get to see the probable volume and volatility values for the next day because the market operating point migrates and is, at most, subject to periodically scheduled econ data announcements. just use five ranges for volume on the EOD instead of the fineness represented by the deci gradation of the attached. If you tab on a transparency that overlays the last Q and the past yearlycorresponding Q, then you get to have a context as well. For me, as a daytrader, I use the intraday attached version is several ways. Three blanks a day for the am, midday and pm. People usually do not work to learn about markets in depth. So not many people in ET will do the EOD version of the 5 min that is attached. Who cares. If a person wants to find out how things work, it takes an effort and that can lead to understanding by being logical in the thought processes. to get to logical thinking is a task in itself. The attachment is most powerful in being able to see how market pace affects the potential for making money.
I think I overshot the target (participation on a topic) with the above post. Here is an ES EOD distribution (RTH) for 250 days ending 11Jan07. (rounded to nearest point.) This is a year's worth in other words. NB: there is a volume price thread elsewhere and "efficiency" , a person, posts the opposite of what this tab shows is true.
Jack Hershey, What relation do you see between the volatility in size of the different market depth levels and the accumulated volume on an intraday basis for the E-mini S&P ? Could this relation = correlation coefficient have anything to do with the efficiency of this market on an interday time frame ?
Looking at accumulated volume on an intraday basis is a good idea. At any time during the day you have a status report by regarding he DOM at that time. One stastisitic is building through the day (Acumulated volume) and you can see this on an MP for greater detail. Appropriately, the DOM may be considered in relation to these. So lets look at the DOM. It may be familiar territory or a fringe situation. That can be sized up in jig time. That under our wing, we look further to see how its going in a deeper context. (I have my three charts out on money velocity and money velocities relationship to the time of day as well). You can use the 5 min volatility/ Pace chart here to see, on a go/nogo basis, at first, to understand whether the DOM range is hitting up against the impossibility of the volatility (for the given pace) to exceed the DOM range or not. If it may be exceeded, you are looking at games being played on the extremes of the DOM rather that a "protection" wall being there as caused by the more "fearful" players "in" the market. This tells you your strategy for holding through the near present. Both are positive: you out game the BS'ers and you get the "cascading" of the loser types having their stops hit as they practice "discipline" in trading to learn more about failure, R/R and having low sharpe ratios. So look further. That is, past the next profit taking. You take those profits or you have to wait (conditions of go /no go not applicable) and see where the "wall" forms on the DOM. At that time repeat the above and play accordingly. What is generally going on? You are making money and not doing any of that premature exit stuff anymore. Lets look at the trade after this trade. Only on lower volatility can you see the full range of the distribution. This puts you into the rythmic type trading where few games and few protection is at hand. This is a time in the market when sudden changes are anomolies BUT they can happen. We look back at accumulation and the MP to see where prior failure of the losers is dominating for the day. This is "the edge of the earth" and "don't get burned again" arena. The price only people live here a lot. So MP gives you accumulation places where losers have lost and then beyond this region we see the future of price movement. Yuo have to do the rythmic stuff for a while and then you get to go to "trends" out side of this place. To learn this well, do brackets after sidelining on low volume on the accumulation. If you pick up the three charts whose abbscissa is time, you can best see when this is coming to pass. It is usually in concert with the "settlement" of the mutuals and HF's right after "lunch". You will have to reread this. Sorry. as a person slips on up the skill path, then some interelationships become pertinent for making sure you are getting all that is there out of the market. The DOM and volatility are important here and you have to continually regard the pace/ volatility map to understand that the next forming bar is going to be adjacent to the current bar and then you see vertically and horizontally the Gaussian distribution on the matrix cells. We plan to take all of this up this year as the SCT trading goes to intermediate and then to expert. We are going to do a fast four day transition on this using recent full trading days that were vieo recorded without annotation. from the 30 to 40 hours of class time. we will cull a story from these annotations and produce about 400 pages of illustrated narration to back up the editied videos. Good Q's. thanks for your inquiry.