by Patrick Burns.
Abstract: The quality of stock market predictions based on the winner of the Super Bowl is examined using permutation tests. These tests are very easy to perform in modern computing environments like the R language. One key point that comes to light is that the success rate of a prediction is not a good measure of its usefulness. Statistically significant success in prediction does not automatically lead to economically profitable strategies.
This version: 2004 January 02 (pdf)
Associated software is in the BurStMisc R package and updated data is available via ”The US market will absolutely positively definitely go up in 2012”.
jun 25, 26
The function in question is scriptSearch. I’m not much for superlatives — ”most” and ”best” imply one dimension, but we live in a multi-dimensional world. I’m making an exception. The [...]
jun 25, 26
I recently gave a talk at the R in Finance conference in which I introduced the marketAgent package for R. Here is the source for the package if you'd like [...]
jun 25, 26
Executive summary Surprisingly good. And it’s not like my expectations were especially low. Structure There are 20 chapters. I mostly like the chapters and their order. Hadley breaks the 20 chapters [...]