![]()
Visiting guest. Why not sign in?
|
|
Big Brains Rule Trading Floor Agents in Finance | |
Big Brains Rule Trading Floor
There's a Reuters wire discussing the world of finance, and how things have changed since the advent of computers and smart agents capable of trading -- Big Brains Rule Trading Floor. It's all about predicting trends, and having enough confidence in them to act straight away. Experts are still often required to confirm these hunches, since a wrong move in such contexts can be dramatic. I remember that financial model that was defined last decade, whereby small losses could be offset by the purchase of other stock... that all fell to bits due to earthquakes Asia, if my memory serves me correctly. Though, still traders believe you'll need a maths or AI Ph.D. soon to get into Wall Street! |
|
Black-scholes equation
The model used in the late 80s was the Black-Scholes equation. The equation models the price of options based upon stock price, volatility of the stock, and past performance. It is related to the 4D (3 spatial, 1 time) wave equation by a simple change of variables. A group of investors used it to make approx $1.5 trillion (US dollars). This worked phenomenally well, until the Asian economy started to break down. Unfortunately, they trusted their model to much, and while most of the experts were getting out, they remained in. The organization went bankrupt, shortly thereafter. |
|
Crunching for Dollars with NN/GA
There's another similar article on Tech Review, that describes how big computers are required to perform complex AI based calculations -- Crunching for Dollars Neural Networks as pattern predictors, genetic algorithms as classifier systems, and fuzzy logic for rule based systems impose serious demands on the computer... it just makes me wonder how they modeled the problem for it to be so complex!! |
|