Questions 6 through 9 below utilize the Excel data file titled "STAT 510 Examination 3 Data Set 2". This file presents quarterly data from the first quarter of 1936 through the second quarter of 2021 on U.S. treasury bill rates (TBILL). For these questions, I do not need to see the Megastat files you used for your underlying work. I only need to see the work you did on calculating the within-sample RMSE's, an indication of which model you chose, and the actual forecast values. For each question, you will be asked to forecast the next four quarters (2021 Quarter 3 through 2022 Quarter 2) and to use the within sample forecasting technique in determining your forecasting model. It is possible to have negative forecasted values even though treasury bill rates are not going to be negative (we hope).

Data set: https://docs.google.com/spreadsheets/d/19N8Wsl99d1FLrs2FTXlsKLhITFa3DnxhOq4exirN1kQ/edit?usp=sharing

6. Use the entire data set in developing your forecast for 2021 Quarter 3 through 2022 Quarter 2. To save time, consider only the following possible forecasting models:

First Order Trend Polynomial

Second Order Trend Polynomial

Third Order Trend Polynomial

Fourth Order Trend Polynomial

Again, provide your work on calculating the within RMSE values and the actual forecast. (15 points)

7. Repeat Question 6 except, in this case, only use the data from 1987 Quarter 1 through 2021 Quarter 2. Provide the same information. (10 points)

8. Compare your results for the two questions above (6 and 7). When you look at the data (or a graph of the data), why do your comparative results make sense? (5 points)

9. Use the entire data set in developing your forecast for 2021 Quarter 3 through 2022 Quarter 2. To save time, consider only the following possible forecasting models:

Simple Exponential Smoothing Model with = 0.90

Holt Exponential Smoothing Model with = 0.90 and = 0.10

Again, provide your work on calculating the within RMSE values and the actual forecast. (10 points)