camron.systems.australia section 23 of 30
Targeting is concerned with the range for the day resulting from a high and a low, not the final price.
The range can be achieved from any number of locations (ie permutations) of high and low.
Once the probable high is in place the probable low can be calculated. OR
Once the probable low is in place the probable high can be calculated.
Consider a marathon. The distance or range is predetermined at 42kms. Once the start location is established the finish point can be determined. Suppose the chosen location is flat terrain - a circle with a radius of 42 kms or a diameter of 84kms. The race can be run in a straight line. Straight out from centre to perimeter. Thats 360 possibilities with a fixed start location and fixed distance. If the course of the race is instead 22kms out and 20kms back in ("only 22 as the crow flies') that becomes 720 possibilites. Shift the start location 1 km and you have 1440 possible finish points. and so on. It is not possible to determine the maximum distance out ("as the crow flies') until the start point and course shape have been established. As the SPI travels vertically, there is only 1 permutation, not 360. If the SPI was a marathon where 1 point = 1 kilometre, the number of possible locations is limited to 42.
The absolute range of the SPI is a straight line. up-down. ("as the crow flies") The distance is a semi-predetermined length which defines a number of high and low permutations. Not many. Broker projections and preopen range give plenty of warning what the distance will be. All that's left to work out is where the course will be. Much of the uncertainty can be removed. The pre-open (warm-up) session range, and course traversed, is a template for the day. The preopen range (or an iteration) is best used to establish where the spi won't go rather than where it will go.
There are two separate races being run at the same time. The SPI and the CASH. Different courses. Different distances. Each has its own target distance. As the SPI covers its distance, it will stop the minute the cash reaches its finish line or target range. Never the other way round. If the SPI has not achieved its distance when the CASH stops after reaching or passing its target, it stops. If the cash covers extra distance extending its target range, the SPI will cover the extra ground, and extend its range by the same amount. Once "happy hour" arrives all connections with CASH are terminated.
The SFE system is closed between sessions, during which orders cannot be entered.
The spi pre-open auction session begins at 09:40:00am
The spi auction balance match-off occurs at 09:49:30am
Trading for the day session in the outright market begins at 09:50:00am
At the commencement 09:40, unfilled GTC orders in the system are exposed to view.
An aop cannot occur until the bid and ask cross over. As sellers offer prices lower than bid prices, or buyers bid prices higher than ask prices. As new orders are entered, the SFE preopen auction algorithm automatically re-calculates the aop for those cross overs. During the next 9 minutes as new orders are entered, the aop is automatically recalculated, fluctuating through a range. Usually sedate with a small to medium range the norm.
The pre-open session is usually indicative of the behaviour for the day. Periodically the sedation wears off and the "extreme sports" department takes over. A larger than normal range occurs. Care should be exercised. The range can explode on one rogue order which is immediately withdrawn. Withdrawal of a large order which causes a preopen high, does not negate the preopen high, which stands as a high. Such is identified by 'aop' changing 10 points in a single move then immediately reversing 10 points in the following move. A single change of 10 points or more indicates the entry of a large order. A delayed reversal of the move indicates a change of heart or the entry of a compensating order.
Large orders front running the queue. Infrequent. Shows extreme confidence or fear. If the range is large. Not withdrawn. It can be considered valid. A large SELL order front running the queue, will abnormally depress the open. A single large order in the middle of the queue will change the 'aop' 10 points in one move. Open will be low, with a momentary reaction to the upside as price seeks to return to a normal equilibrium. (vice versa for BUY)
Technicians would expect the price to fill the gap left behind. True if the open is depressed by a single large front running order. Doubtful if a low open is due to a number of medium sized orders rather than one single large order.
The behaviour of the pre-open process and its size, or an iteration of its size, usually dictates whether the control zone described below will operate.
pre-open live is available from BourseData Ltd
see it in the "action-replay" download at camron.com.au/download
Targets are based on a control zone. It marches to an invisible drum.
Evidence of the operation of a control zone is absolute.
The SPI is driven by a root number Rn. It is a fixed number. It is not a fibonacci number.
The default control number Cn is an iteration of its root Rn.
The cash is driven by a root number and its control number is also an iteration of its root.
The cash root and its iterations, are different to the spi. Neither are fibonacci numbers.
There are 3 possible explanations consistent with the behaviour.
Explanations 2 and 3 require the concept of two types of day.
(a) natural day where the required range occurs naturally as a result of action in the physical cash market.
(b) artificial day where action is light, and required range is not achieved naturally.
(•) most large range days are natural days.
(•) ½ small range days are natural days.
(•) ½ small range days are artificial days.
(•) most days are small range days.
(•) most broker estimates reflect small range days.
The controlzone operates independent of any other process.
It is applicable only to intra-day operations.
It is a mathematical function of controlled behaviour. It is not technical.
Complete intel details of the controlzone can be purchased from camron.com.au
The three probable causes listed below are now obsolete. It's retained for historical purposes.
Written in 2002 before the broking industry consolidated and before the transition of the majors from execution-broking to prime-broking, and the disappearance of "locals". Read in conjunction with sections on "market-making" and "algorithms". A market-maker now achieves the same effect. see market-maker
Possibly the control number is a Benoit Mandelbrot chaos theory fractal.
The iteration characteristic of the control number is consistent with fractal theory.
Fractal theory does not claim that there is a constant pattern in nature.
Possibly the control number is consistent with the fibonacci sequence.
While there are some mandelbrot characteristics ..... the incidence of the control zone is too consistent to be random. (Like, exactly the default control zone achieved 4 days in a row. Not random. And the way it gets there). This explanation is inconsistent with the behaviour displayed in the samples below.
probability assessment - low.
The control number may be a fixed pre-determined number defining the area within which a single institution accepts a fixed tolerance level in its daily trading range. eg. the institution wishes to buy 3000 contracts within plus or minus x points of y. The repetition is consistent with the controlled behaviour of one institution. Its fixed because their actuaries have established the limits of risk within it. That institution, acting independantly would be overwhelmingly dominant to avhieve the level required. It would also require the passive willingness of other institutions to take the other side. An alternative explanation would have the one institution buying through one broker and selling through another broker. Does not make sense. Nothing is achieved. Alternatively 2 institutions acting in collusion. Again nothing is achieved. This explanation is inconsistent with the behaviour displayed in the samples below.
probability assessment - low.
The rabbit is the mechanical rabbit used at greyhound tracks.
If you are familiar with the insurance industry you will know it operates what is called a knock-for-knock system. A client of companyA has an accident with a client of companyB, both companies just repair the damage, debit the cost to the knock-for-knock account, and get on with business. If companyA client was at fault, companyB knows that at some time in the future one of their clients will be at fault. It all balances out and keeps the legal costs down. This is just a general principle, and disregards the fact that where possible insurance companies will try and recover the cost from the insurer of the party at fault. More so with un-insured parties.Exchanges make money from exchange fees. Brokers make money from brokerage. Both parties need action and turnover in order to make money. To facilitate action, the NYSE has market makers in most stocks, to ensure there is always a buyer or seller willing to take the opposite side of a trade for a private individual. Official Market makers exist in the Australian Options Market and The ASX warrants market. In the days of the floor, locals acted as market makers, and still do, although not obligated to do so. Deutsche Bank have undertaken to make a market in $AUD futures.
Large buyers and sellers will be found at the base Cn boundary because it is known it just doesn't go past it, mainly because the brokers are too familiar with the Cn control zone. (behaviour, behaviour, behaviour. familiarity breeds contempt). Why change it. Just watch the market depth at the Cn point boundaries for yourself. Doesnt mean it cant go past it. It can. To the next iteration. Just not often.
The control zone has validity as an activity target at cn points. It's consistent repetitive behaviour would indicate a process where brokers undertake to ensure the SPI trades in a reasonable range each day, using a roster system. The rabbit run occurs late. By 15:30 if the SPI has moved X points, and X is less than Cn, the rostered broker has the honour of achieving the targeted range. Often the target is hit right at the closing bell. Nearly always after 1600. Sure it costs him. But he will be the seller next time when it is the next guys turn to perform the task. It all washes out in the end. (knock for knock). They look after one another. Why Cn points?. Dont know. It may change one day. The period after 1600 is known as "happy hour" for obvious reasons. An example of "happy hour" is shown in the data extract below for the time period 15:50 through 16:30
probability assessment - high.
On artificial days a rostered broker steps in.