S&P 500 daily movements
Daily percentage changes in the S&P 500 from 2001 to 2005 - a deliberately hard classification problem where the honest answer is "barely better than chance".
Daily movements in the Standard & Poor’s 500 index over the five years from 2001 to 2005, with the percentage change on each of the five preceding days, trading volume, and whether the market went up or down. An Introduction to Statistical Learning uses it as a classification example whose value is negative: the predictors carry almost no signal, and a model that appears to work on the training data collapses on held-out years. It is the best available antidote to reading a good training score as success.
Key columns
Representative fields, refer to the source for the full, authoritative schema.
| Column | Type | Description |
|---|---|---|
| Year | integer | Calendar year, 2001 through 2005. |
| Lag1 … Lag5 | number | Percentage change in the index on each of the five previous trading days. |
| Volume | number | Shares traded that day, in billions. |
| Today | number | Percentage change on the day itself. |
| Direction | string | Whether the market moved "Up" or "Down" - the classification target. |
License: Distributed with the book’s companion R package; see the source page for terms.