Predicting randomness
- Thread starter oddiduro
- Start date
I think the point you made for dynamic positions of overnight trading should be reasonable and insightful in order to reduce the risk impact of price gaps. Potentially that would also enhance profitability in some cases, I guess.
I'm currently working on over-weekend trading of forex gaps.
http://www.elitetrader.com/vb/showthread.php?s=&threadid=56791
Here, you definitely are in the right thread. Predicting randomness is an excellent technique applicable to this.
Can you please explain a little more about what you were talking about with all that you need is 3 bars of price to stay on the right side of the market. Maybe even show a chart explaining what you were talking about in your post.
It seems that you were stressing the importance of picking the correct time frame to trade in that you can keep up with while monitoring the market.
Actually I'll not try to predict the upcoming gap, if any. I'll just adjust my directional positions more frequently than before in response to any price movements during weekends. One scenario would be similar to the dramatic movements on Friday's EUR/USD jumping up and down 200 pips within 2.5 hours.
It occurs to me that a lot of the discussion on this thread is coming from a definition of randomness.
I f I say the chance of the next event occuring is 90% and 10% of not happening that to me is random.
On the other hand, if by random you mean the chance of any particular event is equal. That is, a uniform distribution. So since, pretty much, anything can happen the probability of any event approaches zero.
If that is what is meant by random, then I don't belive you can make money in the market.
It seems pretty obvious to me that the mechanism operating in the markets is not of this form.
As such there are regularities over certain time scales.
For example over billions and billions of years the solar system is random but over the time scale of our pitiful lifetimes its pretty regular!
Okay that's my epiphany for today
I f I say the chance of the next event occuring is 90% and 10% of not happening that to me is random.
On the other hand, if by random you mean the chance of any particular event is equal. That is, a uniform distribution. So since, pretty much, anything can happen the probability of any event approaches zero.
If that is what is meant by random, then I don't belive you can make money in the market.
It seems pretty obvious to me that the mechanism operating in the markets is not of this form.
As such there are regularities over certain time scales.
For example over billions and billions of years the solar system is random but over the time scale of our pitiful lifetimes its pretty regular!
Okay that's my epiphany for today
Simply put: When is a signal (most) random?
When its autocorrelation function is the dirac function or, which comes to the same, when it has a constant spectral density function. This is called 'white noise'.
Now, if the above doesn't hold, a signal can be 'colored noise' and still be random.
Generally, in order to define a random signal you can show how it is build up or transformed from a relation involving a classical well defined probabilistic process like the increments of the Wiener process, Poisson process, etc.
In a more general way, a random process can be defined as a solution to a stochastic differential equation.
What is random in this thread? The definition of random.
BINGO Quote from traderNik:
Right.
To me, the logical extension of the random walk theory is a grander, 'everything is random' theory, which suggests that companies 'slowly taking market share' from their competitors are doing so randomly, that there is no cause and effect operating, and that the efforts of the entrepreneurs whose ideas are successfully realized are actually 'random' events. All achievement then becomes random, and not the result of directed efforts by active agents. I don't have the theoretical chops to describe what I am feeling, but I just keep coming back to this question - how is the movement of the company illiquid described, or the cotton market during a period of blight, random??
The weather may be random but human responses to it are not.
And it doesn't do any good to say that there are 'periods' of non-random behaviour but that the overall market is random. Once you admit that there is causality in the markets, every move must be seen in those terms.
I can't tell you how bizarre this whole idea of non-causality in the markets seems to me.
This argument seems to me to be akin to the one about religion or pro choice/pro life. People, including me of course, hold their beliefs and there isn't much that can be said that would change anyone's mind. This is so bizarre because an objective analysis disproves the random walk theory so easily
The mere act of discussion is mind expanding to the trader. Some newbie is reading this thread, and your words have influenced their behavior.
You may have created the next Market Wizard via the butterfly effect.