Out of the 120 sample observations, if the first 110 samples were taken to regress the AR (1) model. Then, using the estimated model to predict the next 10 observations, the root mean squared error is 0.90.

If, however, the first 100 in sample observations were used to estimate the AR(1) model. Then, with the estimation to predict the next 20 observations, the root mean squared error is 0.75.

1. Which part of the data has better forecasting capabilities?

Answer rating: 100% (QA)

To determine which part of the data has better forecasting capabilities we can compare the root me... View the full answer