Overview — Implementation and capstone

A reproducible forecast must state which observations were available when it was made. This topic tests that discipline on seeded synthetic series. The lab estimates an AR(1) mean and checks residuals. The capstone estimates a hedge ratio on a formation sample, freezes it, and evaluates spread forecasts on later observations.

By the end of this note you should be able to:

• simulate AR(1) or GARCH(1,1) paths with a fixed seed and estimate a conditional mean;
• compute a residual ACF check and a one-step forecast with output in comments;
• construct a cointegrated pair, estimate a hedge ratio, and evaluate a one-step forecast;
• predict how changing a DGP parameter moves residuals or forecast errors.

1. Contents
2. The big picture