Predicting randomness

Discussion in 'Trading' started by oddiduro, Nov 3, 2005.

  1. #511     Nov 27, 2005
  2. tommo

    tommo

    I cant believe the random walk theory has lasted this long its ridiculous. If it were random do you think investment would even exist?

    To say a trend doesnt exist is madness, look at a stock announcing profit warning after profit warning and you think its declining price is a fluke and the odds of it going up were just as even?
    Random walkers like to use the coin test experiment to prove trends emerge in random events, but only one of two things can happen. This is not the case in trading, price can go up one tick then nothing for five seconds, or can race up 5/6 ticks in a second or two. Why does this happen? Because at that time there was huge buying/selling pressure, why was there such big trade volume? Because at that moment in time there was a massive consensus that the price should move in that direction.

    Markets are ruled by psychology, if a price goes up to x and reverses who are we to say why it reversed there? it may be because news came out that made people change their valuation of the stock or maybe a major player is defending this level (as happens very often in forex where banks keep a currency pegged at a certain level)? We couldnt predict in advance what the news would be, or what key levels are being defended, that is random, however we can see price is not going up anymore (for whatever "random" reason) and that the price bounces off a certain figure consistently so a trader will close their longs at this point, or establish new shorts, this reinforces this level. There is not a "rule" that says you have short here, but as a human you are going to say to yourself there are lots of people dumping their stock/currency/commodity at this level so you will also because this is human nature because you know it irrational to go against what everyone else is doing and think you are bigger than the market.

    I could pull a gun out and point it at your head, that is random, you couldnt see that coming, but I am certain you will get worried, you may jump around screaming, you may stand still, you may faint, I cant predict exactly how you will react but I know you will be scared. Human nature is very predictable, just like when a bad figure is released everyone panics and sells their positions.
    I could then turn around and say the gun i pointed at you was just a water pistol, again I cant predict exactly how you will react, you may hit me for being so thoughtless, you may take a deep sigh of relief, you may laugh and see it as a joke.
    Just like when the market digests this bad news and realise it isnt so bad after all the price quickly stabalises and shoots up.
    If you plotted the price movement of the news release it would spike down then up throwing all indicators out the window, market "theorists" say you cant predict that.
    Just like your heart rate during the gun incident would look random shooting up then down.
    But at the core of both is psychology. The name of the game isnt predicting when a gun is going to be pulled/ news released/ support broken but bet on the high probability you will react in a certain way when this happens.

    This is why the majority of indicators dont work and why you cant backtest a system effectively, because the majority of the systems tested look at what has happened then assume it will continue (MA corssover systems ect). This isnt what a good trader does. A good trader puts themselfs in the shoes of what everyone else is thinking and reacts with them as it starts to happen. Without sounding too poetic and up my own arse about it, you become part of the market and prediction doesnt really come into it
     
    #512     Nov 27, 2005
  3. I'd like to know where the "mutual" part comes in.

    I notice (and I am sure others did) that you stopped posting for quite a while there. Presumably you were checking out the concept I was talking about. Thats good. I offered it in the spririt of helping other traders. Apparently I missed the part where you said something aimed at helping me (or anyone else for that matter).

    Maybe it was the post where you reply with the witty one word repartee "please..."

    Thanks, I felt "self-improved" right away.

    What a guy!

    Steve
     
    #513     Nov 27, 2005
  4. www.maa.org
     
    #514     Nov 28, 2005
  5. #515     Nov 28, 2005
  6. sabotage

    sabotage

    But look - consider there is some real bad news in the market. It doesn't quite price it in immediately. Usually it starts going down real quick and, if you are fast, you can take the boat and ride it. In short: bad news -> bad fundamentals + panic -> market valuations are too high over a short period -> free ride down the hill. I think it is possible to extrapolate this in the longer term and in the context of lesser changes to fundamentals.
     
    #516     Nov 28, 2005
  7. Hey guys,
    Still going?
    Instead of "Predicting randomness", why don't you give "Predicting market-price" a serious try. With some real work (and patience) it might turn out to be much simpler and less wasteful than the proposition of this thread.

    nononsense
     
    #517     Nov 28, 2005
  8. I submit that you cannot consistently predict market price either, nononsense.....

    Now, I know that you SEEM to be the type to make unqualified statements, so saying that you can and that the rest of us are lost in the woods is a method to escape the argument.

    There is also a method that you SEEM to use called sophism.

    Please avoid these if you please....

    Now, do you have a methodology that SUGGESTS that predicting price on a consistent basis is possible?
     
    #518     Nov 28, 2005
  9. Uh, well, at the risk of seeming to use "sophistry" I will say that I can "predict" the market pretty well on an intraday basis.

    Let me qualify my remark in the following way.

    I don't have the ability to say "OK, we are at 1234, and tomorrow we are going to 1244". Thats not what I am talking about.

    What I mean to say is that I can look at the open (first hour) and then "predict" that the market is going to move up or down from there by at least x points.

    If my prediction is correct (and it often is) then I can go a step further and "predict" that it will move to another price point. I do this regularly. By "regularly" I mean that I am able to do it accurately about 3 times a week. By "accurately" I mean I am able to have enough confidence in my "prediction" that I am willing to bet on it, to overcome expenses, and to make a nice living.

    When I am wrong (the other 2 days of the week) often it is that my "prediction" is correct in principle, but is inaccurate. For instance, I may predict that the market will move up 5 points, then instead of moving up 5 points it moves 2 and then retraces.
    That is what I am talking about.

    Now I realize that some may say, that isn't predicting, but for my purposes (in order to make a living), that is all I need to do. I have been able to do it for about 10 years.

    I have already commented on this before, there are many good professionals out there who can do this (or better) on a daily basis. Just to provide an example, AMT4SWA (Chris) was even better than I at this. I recall seeing him demonstrate his skill on several ocassions, once here on ET during a "live trading contest". Interested parties can use the search function to find that "event" in the archives. I believe the contest was the brain child of an ET member whose handle was "T-Rex"...

    Good luck
    Steve
     
    #519     Nov 28, 2005
  10. Understood.

    As I have said, most traders CAN make money once they learn what not to do in the trading arena.

    However, being right 60% of the time is not exactly predicting. That is more like taking advantage of random market anamolies( you do not know when they will occur), which most of us can do.
     
    #520     Nov 28, 2005