Developing a System for Maximum Efficiency

Discussion in 'Strategy Building' started by eeleusnosaj, Jan 30, 2004.

  1. When I first heard of systems trading, I thought that it wouldn't be viable because of the dynamics of the market. When a system's optimal parameters change to a large enough extent, the system will yield different results, equal to or less than the maximum profitability of the system, which is why all systems have periods of sustained profits and sustained losses.

    The sustainability of profits and losses tells me that the changes in a system's parameters are directional (changes are in the same direction for a given set of bars), and therefore can be patternized or correlated.

    But then something occured to me:
    "In order to develop a system that will be effective all the time, adapt it's parameters to the current market conditions"

    Thus, designing a system to change with the market conditions will always present you with the most profitable trades to take.

    How is this done? First, you must realize that new data is being presented every business day, whether it is in minute, hourly... etc.

    This in itself gives the neural net a chance to 'learn' from its mistakes.

    Note: this 'learning' should be restricted to a rolling range, such as a rolling 2 month database. Forward step through this range to ensure that the system only uses parameter values that are relavent in today's market. But, some events will happen less frequently, so for those, a longer learning range should be used. The optimal values for these ranges can be defined by the neural net.

    Decide on what time frame to use, and gather some previous data to build the basis of your neural network (eventually, all of the main time frames 1,2,5,10,15,30... will be correlated to each other.

    Develop a database and algorithm that will mathematically define how any 3 consecutive bars relate to each other. I chose '3' for this value because it was suggested previously on this thread, but I will eventually have the neural net decide what is optimal.

    Develop a second database that will define what will happen in the next 'X' amount of bars after a given set of 3 consecutive bars are encountered.

    'X' will be defined by the neural network as this parameter will adapt itself to find the optimum amount of data to use.

    Note: Both databases can be further broken down in a tree fashion so data can be retrieved faster.

    If the program encounters a new variety of a 3 bar combination, then add new entries to the first and second databases.

    You also want to add another indicator in the neural net that determines the correlation of price movements and patterns with which side of resistance you are on at many time frames.

    Now, with all patterns and results mathematically defined, set up the neural net to define what the best current price target and stop are. These parameters also change over time.

    Finally, the program can calculate the absolute probabilities that a certian event will happen given a known set of 3 bars. You've hit GOLD!

    Not all combinations will predict certian events with satisfactory probability. In fact, most will be split among hundreds, if not thousands of different probabilities, each taking it's piece of the 100% pie. But every once in a while, you will come across a particular pattern that will produce a given result with high confidence and probability.

    I believe that the degree of accuracy is related to the method one chooses to define the 3-bar set. There is a trade off when designing your method because the more loosely you define your sets, the more garbage will get sucked up into the overall probability.

    Anyways, this is something i've been researching for the past 4 months, and I plan on building a prototype of this system within the next 3-8 months.

    The very thought of making a system that adapts to current market conditions and makes optimal trades makes my mouth water. On top of that, once the neural net is designed, there is no need to tinker with it because it is designed to find the best parameters that work in current market conditions. This, to me, is the way to live life - using your intellect to make you money.
  2. I think your concept piece is a classic.....for how do design a lagging indicator system.

    At least by using 3 bars you have come up with a finite # of possibilites that will fit into a finite # of subsets.

    The most important thing you will learn is that the market migrates rather than jumps around.

    As each 3 bar group transforms as a delete/add sequence you will find three basic transformation classes. The most frequently occuring one (Class A for convenience) is the class of no significant change in the sub set collection of transformations.

    The other two classes are the next most powerful for making money and are on equal footing. Obviously the most common one cited above is the best for making money.

    One of them I would call the "blocker class". This is the sub set of variations on the transformations which bring class A to a close (this is the signal for trading action). What makes this subset so important is that it is a definitve subset in the Venn sense of what is "outside" this subset.

    Lastly, the third class is what allows for the beginning of making money and accelerating money velocity.

    The factor that makes for debate is the fineness of measure of a 3 bar set. In limiting cases the aurgumentation is that infinity is a possibility. That is fine but then you approach lesser possibilities of any repetition at all. The goal for making money is obviously high rates of repetition while loosely coupled. All this, since change is required for making money and thus the more extended the opportunity is with an extreme 3 bar unchanging set.

    A Venn universe matching a Ballentine overlapping triad is probably an optimum. The seven componet subsets show that risk is minimized within the exclusive extremes (you definitely know what stasis exists). You recognize that you cannot get to another exclusive state without a transition (this is a second level comment upon the fact that between a trend starting, a trend continuing and a trend ending, that there is a mixed interlude between each condition). The central subset where all classes overlap provides a major setting. Here the market is "on" but not functional. It is a location that is defined as "noise". It is the only place where random effects are valid and is a place where money cannot be made. It represents the focus of"random walk" and "chaos" machinations.

    Once you have all the 3 bar sets located in these 7 categories by overlapping the three major classes. You can avail yourself to a magic place, not often seen by most. Take yourself through the seasons of the year. Enjoy what happens as quarterly reports and annual reportsimpinge. Then work your way out and then in. You need to do seven trips.

    After that you will be able to see where efficiency lies, where risk is prevalent and what is the potential of each market in which you work.

    If you are orderly, you can use the Venn as a map. It will be possible with overlays to define trends, S and R definitions, all formations and best of all the effect of volume 3 bar pairs (pairs with price 3 bar elements) for use as anticipatory indicators of price.

    Did you ever notice how smart money bar sequences and cash bar sequences in future indexes corrolate?

    What I think is best learned from this is that prediction is not necessary. And that anticipation is best acomplished by seeing all that is not possible at any particular time. It is all related to going with the flow since the flow is so orderly as a migration at all times.

    The best single example is characterizing the differences between a BO and a FBO.
  3. nitro


    nitro :confused:
  4. Grob

    Please explain what you mean by "smart money bars".
  5. Here is my view, briefly.

    I have gone through a reasoning process that it will be easy for most people to accept or throw away.

    I think that the minority control the market at all times.

    So I do look at the minority as much as possible.

    Secondly, I believe that the market is orderly and moves by migrating.

    Lastly, I believe that a trader is more efficient if he trades a lagging price scenario compared to one that runs ahead of where he trades.

    I often build triads as stable types of contructs, especially for making money.

    Most people will not drop out at this point. People will shut out what I say here depending upon a few things. I am skipping that list here.

    To synthesize the above three items, I get to smart money bars.

    Where are the few really rich people whose opinions are always there to see? My answer is the Dow Jones Industrial Average and all the technology focussed upon it for really eaking out what is potentially available. When young, I experienced this right after I left my only professional job I ever had. From Greenwich, I commuted to Wallstreet to advise an institutional investment firm. There is smart money in NYC and it has a pervasive problem, then and now. It is a dilemma in nature.

    So I look at the DJ04H and it's poor people's buddy YM04H. Before emini's I traded the DJXX. Since 1957, I have tracked the INDU daily whether here or living in Switzerland or working at the WH.

    I chart the INDU (I am using qchart and IB symbols here for convenience) on a 2 min fractal. I superimpose a MA of one unit of YM04H on the chart. This gives me bars and a continuous line. always easily differentiable by the senses.

    In the context of the 3 bar stuff, I crudely monitor the migration of the markets on the chart (there are two markets, one dependant upon the other). I look at one comparitive possibility. That is whether smart money accepts the moves of the INDU. The situation is this precisely. Fewer trades occur on DJ04H compared to YM04H but those occuring come before the INDU data moves.

    The small minority that deals with change is best seen as the change initiators on YM04H. This is the small smart money component of the YM04H. None of them are showing ahead of time on the DOM. They control what is the BBid/BAsk by their use of market tools.

    Small note. The DOM thread currently running in ET hypothesizes several misconceptions. As you read this it will settle out what is really what.

    How "Smart money accepts the moves of INDU" is the key single facet that I regard. It turns out to be a leading indicator. In maths, think of it as a first derivative of the potential to make money under change conditions. Change is a requirement to make money.

    As the cash changes in INDU occur, it is my belief that smart money "judges" them. I abide by this judgment that is continually made.

    Simply stated, during trends, smart money leads the trend.

    When change is in control, smart money judges the efficacy of the change.

    You can see people dropping out like flies by now. The key subjunctive force they feel is related to timing. There is something "fuzzy" or they feel that the message has not continued to be pristine.

    What they see as a paradox is "How can a person "judge" and still be "leading". I will skip the causal factors for these difficulties they aren't dealing with.

    Go to a consideration of whether or not you prefer continuous or descrete data elements. Continuous (smoothed) is very nice and not possible nor available. All that is possible in monitoring are descrete snapshots always delayed as is pointed out by people in ET who deal with this stuff. As an engineer/designer I do not dwell on limits like this.

    I use the "flapper" of the YM04H MA waving at me to focus on change. As a Go/No go there is flap or no flap. I increased line weight as a relaxor vis a vis sensing. This machination is "beating" against INDU data. They concur or do not concur.
    That is understood and processed in a glance, all ahead of the market I am trading to make money.

    Relative to what I do, it is all slow motion from two respects* and acts as a confirmation of what is at hand for making money.

    There are several caveats** as well that need to be regarded. All reinforcing and relaxing.

    But at this point, I have covered your Q.

    * I can chat about this but not just now in this post. I want to keep your focus.

    ** These are indirectly related to having a properly bounded view of things.
  6. Thank you. I got qcharts to check this out but have noticed it lagging IB a number of times. Working off a 2 minute chart do you not have problems with this lag?
  7. Jack,

    So you are saying you look for the MA of the YM to confirm or fail to confirm the cash? Or is it vice versa? Isn't that the same as watching the PREM for the S&P? Certainly a quick rise can signal program activity, but you have to be awfully quick to catch it.

    Also, you previously mentioned the concept of relative volume in individual issues and pointed out that QCharts reports this as "unusual vol". Do you know what the number they report actually represents? If it is a ratio to 65 day average volume, do they adjust it for time of day? What do you consider an unusual enough number to warrant action?
  8. Lol. Its funny how things work. Like Scientist said in the other thread, keep an open mind. Whats the old Zen saying.........I cant remember but Im sure theyve got one to cover the subject.
  9. LOL

    Qcharts can be a real drag on making money effectively.

    I can only suggest at this point in time, that I am a little sanguine about the unecessary impediments we all have to enjoy while piling up the profits using qcharts.

    In the DAX thread a while back is where this stuff first came up. Thee I noted that the qcharts display is one set of data away from NOW. NOW means the present instant in time that moves through time. I use DOM of ES as my vernier on NOW. It comes from IB and I watch the qcharts T&S stream to see how bad what you point out is relative to using my market tools on IB.

    Because qcharts is still quite quaint in it's ability to deliver adequately, I have ust been giving in to not extracting as much as it possible as I trade.

    I have these quick note error items I note. Qcharts insufferably accounts for a lot of the excuses I come up with for E1's (acting on anticipations that are still not set in stone)and E2's (hesitating after the facts are clear). This slot is like the visible light spectrum compared to he energy transmission spectrum.

    Why qcharts isn't technically sized up to serve it's clients, I keep thinking is temporary. But it is an improvement from my starting point of penciling daily charts in 1957.

    So in this context, the smart money input is really valuable. See the upcoming post response to AA in the B.
  10. You got the right idea, Pal!:cool:

    #10     Jan 31, 2004