One type of trading strategy is commonly known as regression trading or correlation trading. Essentially the idea is to compute a “typical” price relationship between two securities, indexes, or curves over a period of time using regression and correlation statistics. When the relationship moves out of its typical range and your chosen investment vehicles become more or less correlated, you can take a position for profit.
To get started, you will need two basic things:
1) a period of historical data observations on which you will calculate regression and correlation statistics, and
2) a tool to calculate the statistical measures.
Assuming you work with daily or weekly period data, the first requirement can be met by just downloading free data from Yahoo! Finance or another free source. The second requirement can be met by purchasing a statistical template package for Excel. This should conveniently and quickly calculates multiple statistical measures for use in your trading, backtesting, and performance analysis models.