EC422/522 Economic Forecasting Winter 2022 Skills Project #4 Objectives You will practice estimating and forecasting ARMA models. You will create tables and charts to report your results. You will practice using the following new RATS commands/procedures: @BJAUTOFIT, BOXJENK, DO Introduction The objective of this project is practice estimating and forecasting ARMA models. Econometrics 1. Forecast the quarterly series Nonfarm Business Sector: Labor Productivity (Output per Hour) for All Employed Persons (PRS85006091) from FRED. Start date = 1984:1 End date of regression = 2019:4 Start date of forecast = 2020:1 End date of forecast = 2021:4 (a) Test the series for time dependency. Report your conclusions.

(b) Estimate an ARMA model over the sample 1984:1 to 2019:4. Briefly describe the process by which you chose your model, being sure to include analysis of the autocorrelations and partial autocorrelations of the series, AIC and SIC criteria, the covariance-stationary and invertibility properties of the model, and an analysis of the residuals.

(c) Forecast the series. Chart your forecast with a 95% confidence interval. 2. Forecast inflation rates using the monthly series Consumer Price Index (CPIAUCSL) and the Consumer Price Index Less Food and Energy (CPILFESL) from FRED. Compute the monthly inflation rates as 100 * the log difference of each series. The former is known as headline-CPI and the latter core-CPI. Start date = 1995:1 End date of regression = 2019:12 Start date of forecast = 2020:1 End date of forecast = 2021:12 (a) Briefly summarize the information content of this series, what report it comes from, and the other data available in the report. Hint: Find information about this report on the website BLS.gov.

(b) Test each series for time dependency. Report your conclusions.

(c) Estimate an ARMA model for each over the sample 1995:1 to 2018:12. Briefly describe the process by which you chose your model, being sure to include analysis of the autocorrelations and partial autocorrelations of the series, AIC and SIC criteria, the covariance-stationary and invertibility properties of the model, and an analysis of the residuals.

(d) Forecast each series. Chart your forecast with a 95% confidence interval.

(e) Which series is more difficult to predict? Comment with an economic interpretation. Hint: What is the difference between core and headline CPI?

3. Download the quarterly series Real Gross Domestic Product (GDPC1) from FRED. Convert to quarterly annualized growth rates: