Main | September 2006 »

August 31, 2006

Probability Theory

We discussed the rules of probability theory with an application to finance.

P1010036a.jpg

P1010039a.jpg

P1010041a.jpg

P1010042a.jpg

Posted by bparke at 09:29 PM | Comments (0)

August 29, 2006

Probability Theory vs. Statistics

The first point to understand is the relationship between probability theory and statistics.

Most economics students are familiar with calculus. Differentiation and integration go from f to f' and back again. Integrating (going from f' to f) is much more difficult than differentiating because the former cannot be done by applying a simple set of rules.

P1010018a.jpg

Probability theory tells us how the data would look given the true model. Statistics attempts to infer something about the true model given how the data looks.

P1010020a.jpg

A short course in probability theory follows.

P1010020b.jpg

P1010020c.jpg

P1010022a.jpg

P1010022b.jpg

The upper-left diagram is representative of the truth in most economics classes. The upper-right diagram gives a realistic picture of the truth in a world where there stochastic forces affecting the realized data. The lower-left diagram is a typical picture of realized data. The lower-right diagram is a competing potential true model that would also account for the data.

P1010023a.jpg

The point here is that we will not learn much about the truth (be it #1 or #2) without first assuming that we know truth #1 to be the right model. That is, to estimate beta we must first assume that the true model is in the upper-right quadrant.

Posted by bparke at 09:28 PM | Comments (0)

August 24, 2006

Welcome

Syllabus, Wall Street Journal.

Posted by bparke at 09:24 PM | Comments (0)