|
PureBytes Links
Trading Reference Links
|
:Amazing article
:"A Multifractal Walk down Wall Street" by Manelbrot - Scientific American
:Feb 1999. It really hammers on the validity of MPT. The assumptions behind
:the normal measures of risk (i.e. based on Std Deviation) are flawed, or at
:least limited.
:The simulations are indistinguishable from real markets. it's uncanny.
Well, it's been my belief that the basic math used by academics and
corporate finance people have been flawed from the beginning ("A Random Walk
down Wall St."). What these people forget or don't realized is that we have
gaps in our data (gap opening up or down). This results in discontinuous
data, which mean we cannot use the standard Gaussian statistics. So throw
out your linear regression models. ;-) You're leaving out an unknown factor.
Check out the Navier-Stokes equation.
|