Predicting randomness

[oddiduro]

Quote from traderNik:

oops... that seems dangerously close to thinking that you know better than the markets.
Don't get me wrong, you may well know better than the markets on this or that occasion, but to say that there is no rational basis for any market movement based on one presumably irrational movement might be dangerous, no?

Sure, there have been plenty of examples of irrational exuberance on both the bull and bear sides.

You asked if there's a rational basis to supply and demand. Suffice it to say that Ayn Rand is a hero of mine. I don't agree with everything she says but she's got it right re: supply and demand. To me, it's one of the most rational things there is.

Anyway oddi, I wish you luck in your trading approaches, even if they are predicated on beliefs which are irrational to me As you said, your trade management is (just as?) (way more?) important than any of this.

My trading is just as random as yours

But how do I, or you, really know? Maybe it just works right now. Maybe I will blow up today, especially if I bet the farm on that test of support on lower volume that I know is coming today...or is it higher volume

Anyway, as another poster posted, there are so many variables that we cannot possibly know them all. Even if we did, would we know what the outcome of the interaction of said variables would be??

 

[oddiduro]

Quote from Equalizer:

Correct, like someone who makes a trade that moves the market because he is drunk like a skunk. But, tell me, even though the reason is "the pratt is drunk - or - he wants to get this trade in quickly, in order to get a cheeky shag from the bimbette he just picked-up", does this appear as a random event or as a deterministic phenomenon to the outside observer unaware of the inebriated(/shag) state of some Joe_NQ_drunk_Trader?

Read that again...
...outside observer...

If you think that some spike like the one described above is not random, then you must be smoking some exquisite stuff. If you were a HF trader and that spike took out your stops, then you got shagged. Random, or deterministic? Hmm... Perhaps deterministic to take out your stops... or wait, was the pratt drunk?

Of course, in the scheme of things, such a trade in isolation (regardless of its size) would probably (though not always) have little effect on the longer term price trajectory.

Read that again...
...longer term price trajectory...

So I guess it all comes down to this:
- Are prices deterministic or random - to the outside observer - accross ALL TIMEFRAMES OF OBSERVATION?
- What is is random?
- What is deterministic?
- Are some timeframes better than others?

Hmm...

A. There is a reason for every single trade ever made - regardless of how silly the reason might be - i.e. not all trades are ...ahem... "rational" events.
A2. You have no hope of knowing the reason behind every trade.
A3. There is no way you can know this a-priori, unless of course you are afflicted with an acute case of Crystal Balliitis.

Q. Given A, A2, and A3, can you still profit from the market?
A. Yes.
A2. Shhh, you are giving too much away!
A3. OK.

Chill-out.

PS. ...and my glass is empty, so I need a refill of Jim Beam...
PPS. I am amazed I can even produce such a Deterministically-Random-Rant given my inebriated state

Here a quick study on pattern formations

http://www.agribiz.com/merchdiz/cointoss/cointoss.html

 

[science_trader]

Quote from autocorr:

The question about trends ist if the markets have any kind of positive autoregression. The question about every kind of technical analyses is if there is any kind of autotocorrelatation positive or negative . The first poster is right ,you can not find positive autocorrelation since mid of 60s at least in the Dow.

I made a lot of these analyses but "trends" are really phantasy. They only exist in your eyes watching a chart

First clever post I find on this thread. Question : you are talking about linear, two points, correlations. What about the rest of all possible correlations ?

 

[oddiduro]

Quote from traderNik:

oops... that seems dangerously close to thinking that you know better than the markets.
Don't get me wrong, you may well know better than the markets on this or that occasion, but to say that there is no rational basis for any market movement based on one presumably irrational movement might be dangerous, no?

Sure, there have been plenty of examples of irrational exuberance on both the bull and bear sides.

You asked if there's a rational basis to supply and demand. Suffice it to say that Ayn Rand is a hero of mine. I don't agree with everything she says but she's got it right re: supply and demand. To me, it's one of the most rational things there is.

Anyway oddi, I wish you luck in your trading approaches, even if they are predicated on beliefs which are irrational to me As you said, your trade management is (just as?) (way more?) important than any of this.

We agree on supply and demand, but what is the engine that drives supply and demand? If that engine is random, then the results from that engine will be random as well.

If all of Americas loans were called today, the US dollar is worthless.

Yet many will still say that the Treasury note is a low risk venture.

LTCM thought the same thing about Russia, if you remember.

They were fooled by randomness.

http://www.agribiz.com/merchdiz/cointoss/cointoss.html

 

[science_trader]

Quote from oddiduro:

We agree on supply and demand, but what is the engine that drives supply and demand?

I don't agree with supply and demand theory. Actually I make most of my money in trading disagreeing with this economical principle.

 

[Epiphany]

So I take it you will be buying the QQQQs and forgetting about the market for the next 25 years?

 

[nononsense]

Predicting randomness?
You got to define things so that the question makes sense though.
Then you can start with Wiener filtering. A classical exercise is the Wiener Optimal predictor, which indeed predicts "randomness" of a suitable kind. This is indeed possible.
That's roughly what you might want to dream about in markets. It ain't that simple though if you want to make money with it.
Prediction, mathematically, is possible. You have to understand what is meant by it.

 

[oddiduro]

To those who played the coin toss simulator there were some great trends in there weren't they?

All we need to do now is backtest the data and out comes a system!

Backtesters......Fooled by Randomness.

 

[oddiduro]

Quote from nononsense:

Predicting randomness?
You got to define things so that the question makes sense though.
Then you can start with Wiener filtering. A classical exercise is the Wiener Optimal predictor, which indeed predicts "randomness" of a suitable kind. This is indeed possible.
That's roughly what you might want to dream about in markets. It ain't that simple though if you want to make money with it.
Prediction, mathematically, is possible. You have to understand what is meant by it.

Mathematics, to my very limited understanding of it, becomes probability beyond a certain point.

I read somewhere that there are some axioms that still have no proofs.

 

[science_trader]

Quote from oddiduro:


I read somewhere that there are some axioms that still have no proofs.

That's why they are 'axioms'...