Do you see patterns in Random Walks?

Discussion in 'Technical Analysis' started by atlTrader666, Aug 10, 2011.

  1. Sure but this is a classic logical fallacy. Random generated patterns are not market related so you cannot conclude that markets are random.

    I have studied patterns for years and I can assure you they exist but they are like light bulbs in that they have a mean life. I think anyone who objects to patterns should read this blog by a guy who has done a lot of work in this area. especially with short term related formations. The book mentioned in the blog is out of print but scanned copies shoudl be available around.
     
    #21     Aug 11, 2011
  2. the1

    the1

    So you mean to say the degree of randomness in the markets is zero? Your position that people with "habits" and "trading styles" interacting with each other suggests the markets aren't random is heavily flawed. Trader 1 uses a MACD crossover, T2 uses a histogram divergence, T3 uses breakouts, T4....you get the idea. Every trader on the planet has his or her unique set of criteria for entering and exiting a trade. Think about several million strategies (with some being the same, of course) interacting with each other at time = unknown. Considering that, how could there possibly be NO randomness in the markets?

    The markets are not absolutely 100% random for sure. I've seen big players push a stock below a certain low to intentionally trigger stops and remove supply for the subsequent move up. If you can get the sellers out NOW the stock will travel further and with greater ease if a certain amount of supply has been removed. This type of event is a pure non-random event. The market is less random than academia believes and more random than traders believe.

     
    #22     Aug 11, 2011
  3. That would be mostly correct. MACD crossovers and all that is basically simplified TA ( the part we gave away to you to distract you). The definition of true TA is actaully usuing price to look for patterns. this means a constant learning process.

    As for there not being any randomness in the markets you are mostly correct

    However, on january 1st of this year I saw this randomness you speak of. it was a tuff night for the spy, but thanks to that i saw a couple of points that could trigger our current down fall.

    But seriously, where price will react is gaugebale what it will do is a bit more tricky.

    Our only limitation is that as we approach 0 there is an unjust resistance. and that time changes, causeing a strong tendency for price to move to the right. It skews us over.


     
    #23     Aug 11, 2011
  4. the1

    the1

    Attached is a very, very basic Monte Carlo Simulation. This simulation is based on the assumption that there are three possible outcomes on the next tick: +1, -1, or 0, meaning, the market will trade 1 tick above or below the previous tick or it will be a flat tick. Naturally, this doesn't take into account unexpected news, which makes the market more random because you never know when unexpected news will be announced, nor do we know whether it will be good or bad news. At any rate, on the attached output we can clearly see trends, patterns, false breakouts, true breakouts, consolidation. Pretty much everything you'd see in the stock market. Any simulation beyond this one would be more sophisticated but I'm not about to spend the next few days writing C++ code. There absolutely is an element of randomness in the stock market. Anyone who says the stock market has zero randomness is truly lost.
     
    #24     Aug 11, 2011
  5. the1

    the1

    Another one with an S&P downgrade :D

    Edit: Gurus would be talking about 980 support around data point 631
     
    #25     Aug 11, 2011
  6. Yes... this is the best simple answer.

    To add, I believe there are environments when the market behaves less randomly (usually during panic selling... it's tough to reverse momentum when there is obvious liquidation). But for the most part the market does act random enough to prevent traders from extracting consistent profits.

    This is why trading every day is so difficult. You have to be selective because the market doesn't provide good opportunities on a daily basis. You just need to have the conviction to put a lot of size during those times that you interpret as special moments.

    If you believe there's a perfect opportunity then go for the jugular. No need to churn your account on countless noise trades... risk more than 2% of your account on these opportunities. Don't overdiversify and settle for mediocrity.
     
    #26     Aug 11, 2011
  7. I think you are falling behind here and may need to study up. What`s your background?

    A lot has happened in this field since Eugene Fama launched his efficient market hypothesis in the early 70s.

    It seems like you are not really interested in the truth, but rather interested in confirming your viewpoints and thus discarding anything that opposes it. If you are however interested and I`m wrong, you might want to read up on the works of Andrew Lo and other material in the field of behavioral economics.

    Randomly generated data may contain patterns and runs, but that does not mean that price data is random although they at times take on similar characteristics. A random set of data is normally distributed. Price data is not normally distributed, it has fat tails.

    I`m sorry if you lost a lot of money in the markets :(
     
    #27     Aug 11, 2011
  8. Wow thanks for the chart! its intresting how these news events are all in the middle of large swings as opposed to the beginning.


    Lets assume that this mid point is important.

    In between the 2 events there are 2 high pivots. that stopped half way (they failed to reach higher then price was before initial drop)

    there is no news there.

    what does this tell us ?
    why is it that when good or bad news is presented we see a 100% increase in direction? Is the mid point a key place?

    where are we right now ? are we at a bottom or are we at a mid point ? what will happen if good or bad news comes out or no news?

    If anticipation to news is equal to reaction to news, does that not mean that reaction to news can be guaged by past price movement? Is that not a pattern in the simplest form?

    If we had bad news at our last pivote low what would of happened?

    and what would happen if we would have bad news tomorrow

    TA is not just about drawing simple trend lines its about developing ideas and getting in the bare bones of data. using indicators, will never let you see this. maybe one day some one can make a simple indicator that takes news and past pivots in to account. but that is where the indicator will stop. It will not build on that.

    Did we predict the news? Or did people set themselves up in a manner they let the news fall in the middle ?

    This is not something i have ever looked at but just form your chart, i can explain why some have said that i can predict news at times. Its because uncertenty and anticipation is written in the chart

    We are taking data of a set group of people making a decision. all these people reacted similarly. there reaction was charted on a sloping line. and if we put a line through these points we get something like a moving average. So we are now hopefully in agreement that in this case peoples anticipation and reaction to news has a pattern... though this is the simplest of things in TA. It should make one reconsider that each pivot up was a fluke. it was actually an anticipation for good news, and when the news hit we had a reaction.


    But in general its a really nice chart and reconfirms that the news is something i can integrate in to my TA. as TA is an analysis of data and news can be considered data.




     
    #28     Aug 11, 2011
  9. i hope your lines were accurate. and not pointing just to the down fall.. was the line precisely pointing to the center?
     
    #29     Aug 11, 2011
  10. There is a universe of difference between saying that the market is random versus saying the market is a non-stationary stochastic chaotic process. The first is completely unpredictable, the second is predictable a good portion of the time.

    If you see "random", put the magnifying glass on and you will see trends, and fat tails again. The same effect will happen if you see "random" and look from a longer term perspective.

    When you see "random" in your selected timeframe, it is a symptom that what you are really looking at an "strange attractor in price". There are time attractors, price attractors, and volume attractors.

    When you look at your hand and see a hand, what you are really looking at (mostly) is a bunch of electrons vibrating "randomly" in a pattern. The market is almost the same thing.:eek:
     
    #30     Aug 11, 2011