Economics 427.  Economic Forecasting

**Subject to Change**
Last Revised April 15, 2010

Prerequisite

Economics 321 (Introduction to Statistics) or AREC 310 or SOCS 225 or equivalent.

Student Learning Outcomes

1. To be familiar with time series econometric methods for forecasting in business and economics, and to understand the statistical basis for these approaches. Topics include statistical methods for modeling and forecasting trend, seasonal and cyclical components of economic time series, multivariate regression models, forecast evaluation, and forecasting in the presence of unit roots. 
2. To be able to apply time series econometric forecasting methods to practical forecasting settings, particularly those relevant to the U.S., Asian and Hawaii economies.  To be able to use standard econometric modeling software for forecasting.
3. To be able to work in a forecasting team to successfully complete a model-driven forecasting project.

Readings and Software

The required text and software bundle for this course is:

            Diebold, Francis X., Elements of Forecasting, 4th ed., (with EViews Software 6.0) Thomson South-Western, 2007. ISBN 0324817606.

This bundle is available from University bookstore.  I will also hand out a number of supplementary readings during lectures and in conjunction with problem sets.  You are responsible for all assigned textbook readings, assigned supplementary readings, and the content of my lectures.

Course Requirements

Grades for the course will be based on one midterm exam (20%), one comprehensive final exam (25%), a term group project (20%), problem sets (25%), and participation (10%).

Due dates are firm!  Any problem set, project component, or other work submitted after the start of class on the day it is due will be marked down half a grade per day until I have received it.

Exams

The midterm and final exams must be taken at the dates and times given in the schedule below.  Except for medical emergency, I will not schedule makeup exams.

Problem Sets

There will be 6 problem sets, covering analytical lecture material and applying the forecasting methods that we are working with to practical forecasting problems.   Examples of possible application problems include:  trend modeling of U.S. retail sales, analysis of seasonality in Hawaii visitor arrivals, cyclical modeling of Japanese employment, and a regression model of U.S. automobile demand. These will take a substantial amount of time, so plan ahead!

Group Forecasting Project

You will work with partners on a forecasting model for an aspect of the U.S, Hawaii or global economy.  You will review existing literature on the behavior and determinants of the indicator, perform preliminary statistical assessment, develop a forecasting model, evaluate forecast performance, and write up your results.  Each groups will make brief forecast presentations in addition to the written report.  I will provide you with detailed guidelines in a few weeks.

Class Participation

Active participation helps to deepen understanding of course material. To facilitate this, I organize class in an informal lecture and discussion format, and I expect you to participate. You will be evaluated on your class attendance, your preparation, and your contribution to class discussions.

Please be prepared for class. Read handouts and textbook chapters before I lecture on the material. Do not miss class.

Plus-Minus Grading

I will award plus and minus grades for course work and the overall course grade, according to this plus-minus grade schedule, applied to adjusted (curved) scores.

Academic Integrity

The University has strict standards on academic honesty and severe penalties for dishonesty. Please review carefully this page on honesty and the language in the University Catalogue.

Students with Disabilities

If you feel you need reasonable accommodations because of the impact of a disability, please (1) contact the KOKUA Program (V/T) at 956-7511 or 956-7612 in room 013 of the QLCSS, and (2) speak with me privately to discuss your specific needs. I will be happy to work with you and the KOKUA Program to meet your access needs related to your documented disability.

Econometric Software

We will use EViews, a leading econometric modeling and forecasting program.  EViews is a graphical object-oriented statistical package that combines powerful modeling capabilities with a simple-to-use interface.  The Diebold text provides Eviews  examples throughout.  Note: EViews runs under Windows (not Mac).

For more information on eviews see http://www.eviews.com/.   A couple quick-start guides to basic eviews functionality that I have found are here and here (See links in online version of this doc.) 

Selected Additional Resources on Forecasting

Books

Makridakis S., S. Wheelwright, & R. Hyndman (1997), Forecasting: Methods and Applications, J. Wiley & Sons.  A classic forecasting text with a wealth of practical advice.

Robert Pindyck and Daniel Rubinfeld (1998): Econometric Models and Econometric Forecasts, Fourth Edition, McGraw-Hill.  A comprehensive introduction to econometrics with a notable chapter on the use of multi-equation econometric models.

Evans, Michael K. (2003): Practical Business Forecasting, Blackwell.  Emphasis on business forecasting.

Web Sites

International Institute of Forecasters, Forecasting Principles web site, http://www.forecastingprinciples.com.

Frank Diebold’s home page, http://www.ssc.upenn.edu/~diebold/.

The Fairmodel, http://fairmodel.econ.yale.edu/main2.htm. Ray Fair’s site houses complete macroeconometric forecasting models of the US and global economies that can be simulated online.

Project LINK, http://www.chass.utoronto.ca/link/.  The United Nations collaborative project for global forecasting and policy analysis.

Online Data Sources

Datasets to accompany Diebold, Elements of Forecasting. (Download .zip file and extract files.)

Economagic, http://www.economagic.com, comprehensive online economic data source

Econdata.net, http://www.econdata.net, a guide to regional data sources

FRED database, St. Louis Federal Reserve Board, http://research.stlouisfed.org/fred2/

U.S. Bureau of Economic Analysis, www.bea.gov.  Source for U.S. GDP, personal income, balance of payments data, state personal income data.

U.S. Bureau of Labor Statistics, www.bls.gov.  Source for U.S., state and local data on employment, unemployment, consumer prices, and other data.

State of Hawaii Department of Business, Economic Development and Tourism, http://www.hawaii.gov/dbedt. Primary source for Hawaii data.

University of Hawaii Economic Research Organization, http://uhero.prognoz.com.  Lots of Hawaii data and lnks to recent forecast reports.

Course Schedule

The following is a plan for the course, showing exam and assignment dates (firm) and a lecture schedule (tentative).

Date	Begin Topic:	Assigments:	Text Chapters
Jan. 12        14	The forecasting challenge.   Forecasting problems and forecast objectives.		1, 3
19        21	Statistical review:  linear regression and diagnostics; graphing		2, 4
26	Modeling and forecasting trends		5
28		Problem set 1 due:  forecast objectives, statistics
Feb.  2	Modeling and forecasting seasonality		6
4
9	Characterizing economic cycles		7
11
16	Modeling cycles: MA, AR and ARMA models	Problem set 2 due:  trends & seasonality	8
18
23	Forecasting cycles		9
25
Mar. 2		Problem set 3 due:  cycles
4	Midterm Exam
9	Putting it all together: trends, cycles and seasonality		10
11
16	Forecasting with multivariate regression models		11
18		Problem set 4 due: time series forecasting
23	Spring Break --No class
25	--No class
30
Apr.  1		Project: Preliminary results due
6	Introduction to unit roots, stochastic trends		13
8	[no lecture--project work]
13		Problem set 5 due: regression models
15
20	Forecast evaluation	Project: Final report due	12
22
27	Project: Presentations	Problem set 6 due:  unit roots and forecast evaluation
29	Project: Presentations
May   4	Review

Final Exam:  Tuesday, May 11, noon-2:00 PM.