2014年3月11日
Briefly Speaking
日期:2014/03/10
Sir Harold Jeffreys recommends that the simplicity of a model be counted as the number of degree, the order, and the sum of the absolute values of the coefficients of the differential equation that models. He believes it is an immutable law of science that all great discoveries fall into a sum less than 7 or so . (See Ackermann [8 page pdf] for review and critique).
I wonder if the moves of markets can be modeled usefully in simple laws like this. The simplest solutions of an exact equation are
y dx + x dy = 0 which derives from xy = c and dx/x + dy/y = 0 which derives from ln ( xy) = c
Which markets move like that during a day or week and can useful predictions about the continuation of this relation for further parts of the period be made? Are there other simple models which work like
x (y)(y) = c or (x)(x)(y) = c
that are just a tad less simple that work as well.
On another note, a visit to the Drexel Museum of Natural History reveals the interesting fact that even though the lion is classified as the king of the jungle, old lions are often eaten by hyenas and leopards. One can see that playing out in the corn belt and those businesses that rely too heavily on yoga.
The beaver on the other hand, one learns often sends a seasoned emissary to help his colleagues build a new dam before returning home.
訂閱:
張貼留言 (Atom)
沒有留言:
張貼留言