Print

Print


To add to your point Nick the whole premise of B&S was wrong. It led many that should have known better to say ' the world is a normal distribution and so this won't probably happen in your lifetime' but people have a very shaky understand of probability and they applied the data to Hungarian Equities as well as the root study, US Bond Prices from 1934 - 1977 (Milken). They took the whole thing out of context.

And to take Doug's point about avoiding them: You don't want to avoid them, just manage them. If you put in all the controls to avoid then you've lost the economy because we need volatility to make a profit at one level of the system: It's how you view the whole that matters.

Stefan

On 13 Nov 2009, at 03:10, Nick Green wrote:

my point is there were probably theoretical mistakes in using normal rather than power law assumptions. There were definitely practical mistakes because there was no feedback on performance: sacrificing rationality for commercial confidentiality. There was no public market in these secondary instruments. That doesn't mean we abandon the method of quantitative description of processes. Anyhow Black and Scholes company went bust in 1998 a year after the Nobel so even emotional traders might have guessed something was wrong. Like it or not you look at the trading history and guess the future market price on the basis of current conditions- if you know them. If you don't it's just luck as Taleb has it. Martingales work for the house not the punter. These simple facts are getting lost in all the vacuous posturing. Trivial to fix: set liquidity on the risk performance and add a few percent for luck- just like always.
 
 
From: Doug McDavid
Sent: Friday, November 13, 2009 1:23 AM
Subject: Re: The world is deterministic

Interesting erudition, but i think we're forgetting how this originated.  This guy was claiming he could predict (from past performance) how financial markets will move.  The myth of the rational market has been debunked.  Animal spirits (Keynes) are loose, as always.  The quants and their bogus so-called rationality and over mathematization of human emotions have pretty much been discredited yet again in late 2008 and early 2009. 

Just be ready for the next bubble/crash.  They can't be be banned or avoided.

On Thu, Nov 12, 2009 at 4:24 PM, Nick Green <[log in to unmask]> wrote:
Chaos is deterministic- but sensitive to initial conditions. When at Cardiff we found that chaotic oscillators could be forecast i.e. to find within 5% for the next 3 or 4 points- but this was interpolation off a hyper dimensional surface constructed from say 10 nearest neighbours.
 
Indeed the difference in nearest neighbour variance for white noise was 20 times bigger than a chaotic signal. ( A sine wave would be four orders of magnitude variance smaller than white noise on say 1000 data points and ten nearest neighbours.)
 
Extrapolation is where the problems really start i.e. futures have to include outlier forecasting. Mandelbrot and Taleb seem to be saying the variance around the mean is usually not standard normal it's skewed and not stationary i.e. data is generally of power law type not symmetric about a stationary mean. That means Wiener has to redo his treatment of the random walk, gambler's ruin etc. and trouble for the Law of Large Numbers when applied to physical systems. Black and Schools assumed the Wiener process to define their prices and to have a better chance they needed a Mandelbrot/per Bak (self-organized criticality - see wiki) power law. When you do the random walk it seems plausible that the frequency of steps longer and shorter than the mean is equal. But the step length can never be shorter than zero and may be infinitely long (the walker/actor evaporates or at least crosses the containing coherence boundary) or more simply can be long (and unbounded). The upper bound of the longest step length of a Brownian walk increases with time.
 
At constant temperature in a long experiment the length of the longest step will increase with observation time- this is the simplest way to put it. Given these are abstractions of very large restricted n-body problems we are certainly in difficult territory. Experiments show power laws in the frequency of word use, size of cities, length of rivers, blood vessels, sizes of meteorites etc http://en.wikipedia.org/wiki/Power_law#Universality 
 
But recall Stafford applied Harrison and Stevens for testing the significance of a new data point (transient, trend or step change) so very short term forecasting only...Science and mathematics is certainly regards forecasting as its major concern. Pask would probably speak of the production of applicable descriptions. e.g. forecasting the length of wire to put a fence around an x by y rectangular field- even that gets tricky if the ground is not flat or x and y have astronomical values and thus need corrections from General relativity
 
Was von F talking about the halting problem perhaps?

Sent: Thursday, November 12, 2009 1:03 PM
Subject: Re: The world is deterministic


I believe that it isan open system as it needs energy from the sun

It is certainly a non-linear system and any non-linear system has the potential to be chaotic

As far as I understand there are very narrow limits for many parameters and if they transgress the results are large i.e. small imput - large outcome which is a sign of Chaos

I dont know what von Forster said about non-trivial machines but to me deterministic imples predictable

Alfredo

----- Original Message -----


From: Dirk Vriens <[log in to unmask]>

Date: Thursday, November 12, 2009 12:14 pm

Subject: Re: The world is deterministic


> Dear Alfredo,
>
> Sorry for our intrusion into this discussion, but we would like to
> know: if
> the world is a chaotic open system, to which input is it open?
> Input coming
> from outside the world? Hmm, weird. Or input from within the world
> - but
> then one would rather use the term closed system, don't you agree?
> And,besides, even though a system is deteministic it can still be
> utterlyunpredictable (or transcomputational, as von Foerster
> claims as he disusses
> non-trivial machines).
>
> Kind regards,
>
> Jan Achterbergh,
> Dirk Vriens
>
>
> _____
>
> From: Forum dedicated to the work of Stafford Beer
> [mailto:[log in to unmask]] On Behalf Of Alfredo
> MoscardiniSent: donderdag 12 november 2009 11:32
> To: [log in to unmask]
> Subject: Re: The wo ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ For more information go to: www.metaphorum.org For the Metaphorum Collaborative Working Environment (MCWE) go to: www.platformforchange.org METAPHORUM eList Archive available at - https://listserv.heanet.ie/ucd-staffordbeer.html Archive of CYBCOM eList available at - http://hermes.circ.gwu.edu/archives/cybcom.html ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ For more information go to: www.metaphorum.org For the Metaphorum Collaborative Working Environment (MCWE) go to: www.platformforchange.org METAPHORUM eList Archive available at - https://listserv.heanet.ie/ucd-staffordbeer.html Archive of CYBCOM eList available at - http://hermes.circ.gwu.edu/archives/cybcom.html ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~



--

Doug McDavid
[log in to unmask]
916-549-4600
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ For more information go to: www.metaphorum.org For the Metaphorum Collaborative Working Environment (MCWE) go to: www.platformforchange.org METAPHORUM eList Archive available at - https://listserv.heanet.ie/ucd-staffordbeer.html Archive of CYBCOM eList available at - http://hermes.circ.gwu.edu/archives/cybcom.html ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ For more information go to: www.metaphorum.org For the Metaphorum Collaborative Working Environment (MCWE) go to: www.platformforchange.org METAPHORUM eList Archive available at - https://listserv.heanet.ie/ucd-staffordbeer.html Archive of CYBCOM eList available at - http://hermes.circ.gwu.edu/archives/cybcom.html ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ For more information go to: www.metaphorum.org For the Metaphorum Collaborative Working Environment (MCWE) go to: www.platformforchange.org METAPHORUM eList Archive available at - https://listserv.heanet.ie/ucd-staffordbeer.html Archive of CYBCOM eList available at - http://hermes.circ.gwu.edu/archives/cybcom.html ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~