Market Forecast with Artificial Intelligence

Discussion in 'Trading' started by k3m, May 14, 2012.


  1. 1. You are speaking about a commonly held myth. Markets have no flaws, no noise and no anomalies.

    Also, you need to know the proofs of the above. Step ahead for a moment by taking the above as an assumption. Then, your discussions on AI would be different; to whit:

    What it is in the markets that requires intelligence is several things; They are:

    a. Taking all of the market's offer using pool extraction.

    b. Using the pattern of the maorket.

    c. Processing information using the mathematics of the market.

    2. Wikipedia (the great scholastic sourse) falls short. Information is processed as it appears. Skip the clicking stuff. Use the leading information of price to simply trade price segment by segment of the full offer of the market.

    What makes pool extraction of the full offer a certainty, is the foundation that all markets have a fine granularity. From the small inconsequnetial seeming thing, everything else needed to do the proofs of no flaws, no noise and no anomaties, follows deductively.

    Summary

    Unfortunately, you are taking a course in Business and Economics that is built from the Conventional Wisdom (CW) of the financial industry. This is a iron bound tradition that winds all the way down to small colleges from the lofty heights of the Ivy League. This is how you go to school to meet friends who families own the businesses of the financial industry. These C+ history major families hire people to do what Morgan Stanley just did in London (had a very majority control of an instrument where the minority (which is always correct) took their pants off)

    Get high grades by knowing the criteria for high grades. Meet a family member. Get a job and have a typical life style. Be crew on a sailboat owned by an exec at a trading outfit; ski at a western destination resort or Stowe or Mad River.

    Trading extracts 3 to 6 times the ATR daily. Events follow a specific order. Carving the Sentiment turn is telegraphed by leading indicators of price.
     
    #11     May 16, 2012

  2. do we owe you anything?

    If it's for Stanford or MIT,...I "might" consider completing the form


    if it's for some community college, or train wreck school like Wesleyan University then forget about it.....

    Before all the Wesleyan grads provide rebuttal here's my source:

    http://www.collegeguide.org/itemdetail.aspx?item=486fb85a-5d15-4d1f-a8f5-5ce2804c3129&page=14
     
    #12     May 16, 2012
  3. To achieve perfect results comes down to always "knowing that you know". Read this as light humor, but do notice when my words take on a more and more curious mysterious high density (Hum the themes of Phantom of the Opera) that your mind cannnot keep pace. Its okay. Try to read it again and again. I make student's set the pace as you will see below.

    The mathematics of the market lets you do that. This is Boolean algebra.

    By deducing the order of events of a trend just once by thinking, you apply that known order to the markets for the rest of life's opera.

    Speaking, in the simplest terms, all trends to come to an end. If you are able to reason this through, you now have two points for every trend: a beginning and an ending.

    you will notice that during a trend it has a Sentiment.

    If you can figure out that a trend has to end with a failure, then you can figure out that the sentiment change is during the interval when the trends overlap.

    Add these detailed events to your first brief list of all the venets in a trend. briefly it minght look like: begin, fail, end, overlap.

    Now you notice you started the overlap but you didn't add the word for the end of the overlap. Lets do that: old fail, new begin, old end, both overlap new BO of old RTL, new hold, new fail and end.

    So many categories so suddenly in their information content.

    Keynes said define them by their variables (independent and depedent) and express this a a complete in kind Hypothesis set. along with that note the Parmetric Measures used in the HS.

    Can you imagine yourself at Wharton and I'm the Adjunct Prof. Its the same when you are in a third form class at a private school like Choate or Rosemary Hall. Children can learn to trade in a suburb of New York minute. Their parents are astounded.

    Look at all the parts of any trend. They all follow the same order of events.

    The HS gives you the independent and dependent variables. the math is true/false of Boolean algebra.

    80% is absolute BS. 100% is the answer.

    So where did the flaws and noise and anomalies come from in the myth you quoted from some Prof or inaccurate book? the Ivy League or the Oyster Bar after dark. Tea in the main lounge or in the basement at Trader Vic's (since moved).

    the point 1 is the beginning, you get to point 2 by going through the BO of the RTL, At point 3 you draw the container new RTL from point 1 to point 3. Add a parallel line through point 2 called the LTL.

    After point 3 is where trends fail in their third move.

    Myths abound and you see them as the trend goes through its pattern while you use ONLY Boolean algebra on the questions that have true of false answers.

    So you made up a 10 question quiz to find out about buiness and economics.

    There are only 5 questions. And only one correct path in the 2^5 tree of answers.

    Here is how teaching goes:

    I walk into the class and ask for questions.

    I answer questions.

    questions begat questions.

    the course always finishes early the text book is finished.

    Now we get on with the real questions.

    Do you want to be right or rich?

    Pick rich.

    Learn how to ask about how to "read" the market.

    It does one thing at a time and in an order of events.

    If you make a log of three interlocking fractals where the trading fractal is in the middle and you write one event on a row for 5 market cycles, you have a fully built mind for always "knowing that you know".

    Go wreck your class. LOL....

    Take 100% of the market's offer every day. Just trade segments of sentiment profits on the trading fractal you choose. Don't work at Morgan Stanley; they are all dummies who believe the Ivy League myths.
     
    #13     May 16, 2012
  4. #14     May 16, 2012
  5. says the man with an $80,000 IRS tax lien and mooching off his partner...

    did you clear that lien up yet?
     
    #15     May 16, 2012
  6. river

    river

    This is at least the second time in the past week you have written about a binary path of 5 or 6 questions that takes a beginner to expert.

    Maybe I'm making this much more difficult than you intended but I'm drawing a blank on this set of questions.

    Any hints for us?

    -river
     
    #16     May 16, 2012
  7. If you develop an accurate prediction model you will not have to sell services because you can just trade your own money using your accurate prediction model.
     
    #17     May 16, 2012
  8. I think everyone goes through getting up to speed trading by seeing several forks in the road.

    The # of Q's could be figured out by looking at the ratio of winners to total traders. Probably 1 out of 10 are still in trading. those really making the money are probably 1 in 100.

    The questions that I believe are important are:

    1. Pro or Amateur. I chose amateur to keep my reporting to a minimum.

    2. Boolean or probability? I felt it was very important to always "know you know" rather than betting. Not having to deal with risk is an impotant question to deal with. I do not have to deal with risk by my choice.

    3. Getting the routine straight. I choose MADA and most choose OODA. Here in this choice is where you choose up/down thinking or right/left thinking. Markets make money going to the left. If your mind is working on only one process for the dominant sentiment all the time, you are not foooled as in doing dog fighting s a pilot would. Win win is different than win lose.

    4. Making money. You choose between compounding profits or taking profits. You have to make the choice W. J O'Niell emphasized. Compound.

    5. Thinking regime. The choice is deduction of induction. You can see that the financial industry always chooses induction. It doesn't work, of course. Deducing the the whole set of building blocks was very important. You only have to do it once.

    6. The Biggy is deciding how to learn. Loosers usually are learningto make money as the learning goal. They get to learn fear, anxiety and anger instead. I lucked out by focussing on building my mind by doing what I was taught in everything else I learned... doing drills. In modern times neuro science is proving this is how to go about it. My emotions center on being supported (See hurricane post ...lol), feeling comfortable by always knowing that I know.

    There are some neat consequences of the above.

    And there is the history of how trading successfully came to be.

    By missing out on any of the forks, the rest of the way is futile. Any once you go through a fork, the mind is not able to back track at all.

    The whole of the financial industry is based on sales and marketing to get OPM which enables fees and commissions. Computer Science people who work in that domain learn to do stats and filtering mostly. Yet it is true they sit in a computer atmosphere based on Boolean algebra and never figure that out.

    All those algorithms are based on reasoning and the reasoning usually goes to doing stats to prove what data says.

    A log only works if the columns are set up properly. Computer Science has a way of making that work: RDBMS. The huge trick to see it that it doesn't have time in the data for knowing that you know. No wrist watches are needed. Events are what are ordered in markets.

    These various categories are the juice. they each have unique definitions.

    It is very hard (and priceless as they say) to simply understand that at any event, that is the unique place where that event fits.

    My views are just what I figured out and no else really has to depart from where they are going. But, on the other hand, it is really neat to get there the way I went. I built my mind so there is a spectrum and everything fits along that spectrum. I call it "having a fully differentiated mind".

    When all the pieces are available at any moment, I say I have facility. RDBMS says it all: Relative Data Base Management System.

    Picture a market and its streaming data.

    Think up how money is made in a market.

    AHA!!!

    The sentence you said was: Money is made in markets through .... price change.

    At Goldman Saks they end with sales and marketing to get fees and comissions. BUT... they, today, still have to think up way to do that in markets. they charge for trading money not caring about anything else. One department does one thing another epartment does the opposite.

    the corollary to money is made through price change is..... "you have to keep on the correct side of the market".

    Covel keeps getting it wrong so he makes tapes of people who get it right. He follows them and he thinks they trend follow. THEY DO NOT.

    They monitor and analyze instead.

    In markets, we all know that price is the dependent variable. (Except for those in business and ecomomics.)

    Price change in the dependent variable is lead by the independent variable.

    So we watch the independent variable to know what will follow for the ependent variable, price.

    So here we are in another post. This is where Covel would turn off his video or tape recorder. So far he can't make transscripts for either..... lol... I have dragon so I don't have his problems in that arena. Everything a person needs to know is already written down.

    But it cannot be put in a person's mind to make future use of. None of it can be thought through thoroughly enough. It is just "gibberish" as Covel says. For him it IS gibbrish. His mind is not differentiated and cannot be for the rest of his life.

    Volume tells you what price will be doing .

    By repitition, you log the variables and get to see their inter-connections in each and every possible and singular finite way.

    when leibnitz and Newton used to call and talk o0n the phone (in Latin), they argued over the pieces. One wanted to put them together and the other was into chopping the whole thing up into bits. Both agreed the smallest piece was so cool functions they eal with were considered continuous functions.

    The market does no have continuous functions and it is Greek to most people instead.

    We have these pieces....they always fall into the same orderr of events.

    Everyone knows exactly what "good" piece is. You know by repitition. Two pieces move the market (the good pieces) The rest let you wait for a good piece. BUT this only describes the dependent variable.

    If you see a good piece, you better look at the volume, to see whats next.
     
    #18     May 16, 2012
  9. And prediction is NOT even necessary.

    Needing to predict is a MYTH.

    Do some time therapy and find out when and where you got hooked on this mis-information.
     
    #19     May 16, 2012
  10. Perhaps the distinction will seem a bit esoteric, but much more valuable than a "reliable market forecast" is a "reliable trading algorithm".

    See, the thing about forecasts is that they contain all kinds of variability as regards to how they come to pass. Let's say I made a forecast for the ES at the end of today's trading that said, "The market will go to 1370 by next Wednesday". Seems so straightforward, doesn't it?

    But, what if the market goes to 1290 tomorrow? Does that invalidate the forecast? If a trader bought at today's close based on my forecast, he'd already be down 30 points tomorrow. OK, so what does he do now? More traders who followed my forecast would be asking themselves that question tomorrow than worried about whether 1370 would get hit next Wednesday. They'd be worried about 1270 getting hit and soon.

    Then, let's say that the market recovered from 1290 and went up to 1340 by the end of next Tuesday. My original forecast was basically for a 50 point move (today's close ~1320 and I was saying it would go to 1370), so was that really what the forecast was about, the size of the move? Or, is 1370 still in play, in which case the hypothetical trader would continue to hold?

    Then, let's day the market closes at 1340 because it does nothing next Wednesday. Is that it? Should the trader close the position because the time window of the forecast has "finished"? Or, should he hold for 1370 because that was the price forecast?

    There are a few (at least) other things that make forecasts pretty useless without specific step-by-step instructions on how to actually trade them. Truly unexpected news, for example. What if a forecasting model relies on some piece of data that comes in way off? Is the forecast invalidated or not? If the forecaster makes certain assumptions about how markets react to news and how they discount the future, then no information could be truly "unexpected" and throw a forecast off, but that's hardly comforting when your position is 20 points underwater because of a bad unemployment claims number, for example.

    Anyway, just some thoughts from a trader who, as much as humanly possible, avoids forecasting at all costs. While there is inevitably an element of "I forecast that this trade I am entering will work", the only time I am "forecasting" is when I enter a trade and that only happens when the market does 1 specific thing. I don't really care about whether the market goes up or down, I just watch for my set-up and take it when it comes. A "general market forecast" would never be "reliable" for me unless it encompassed all of the facets I touched on above. At which point it ceases to be a "forecast" and is instead (not that the line is completely clear here) a trading algorithm.
     
    #20     May 16, 2012