Time Series Forecasts and Intervals

Free SOA Exam SRM (Statistics for Risk Modeling) lesson in Time Series Models. 28 min read, ~4,150 words.

Point forecast = conditional expectation given history. Use the fitted recursion: plug in known values, set future shocks to zero, iterate one step at a time. Forecast error variance grows with horizon h, governed by the model's shock-propagation structure. Random-walk forecast variance is, growing linearly and unbounded. Stationary AR(1) variance...