Economics 451

ECON 451: Forecasting for Business and Economics
Prerequisites: STAT 200 and one of the following: MATH 126, MATH 169, MATH 171 or ECON 207.

Credit hours : (3)
Instructional Method : Three hours lecture.

In this hands-on course, students will learn how to make forecasts using business and economic data. Emphasis will be placed on identifying patterns in different types of data (trends, seasons, cycles, etc.), designing forecasting models, and evaluating model performance. Students will learn how to use SAS (a statistical software package favored by firms) for predictive analytics.  

2. Detailed Description of Course

Forecasting is the art of predicting future conditions based on past data. As such, forecasts give us an added edge when trying to make better (more strategic) business and economic decisions. Creating and evaluating forecasts is a skill very much in-demand by firms and other organizations who want to put the large amounts of empirical data they have accumulated on past business and economic outcomes to good use. However, a firm may end up making choices with disastrous consequences if an employee has constructed a poor forecasting model, not fully tested a forecasting model, or failed to inform an organization as to the true limitations of forecasting. This course introduces students to analytical techniques that are especially useful in forecasting economic and business data (such as sales, expenditures, inventory, GDP, interest rates, stock market returns, etc.). Students will learn how to use a statistical software package (SAS) to design and test forecasting models using real data. All findings will be reported in a way that provides an accurate interpretation of potential future conditions. Topics covered in this course include:
    1) An introduction to the value and limits of forecasting.
    2) A review of essential statistics.
        a. Useful math (natural logs, growth rates).
        b. Summary statistics (mean, standard deviation, correlation).
        c. Linear regression (OLS).
    3) Visual analytics: how to identify patterns in graphs and tables.
    4) Decomposing data: trends, seasons and cycles.
    5) Identifying trends, evaluating trend models and creating trend forecasts.
    6) Identifying seasons, evaluating seasonal models and creating seasonal forecasts.
    7) Identifying cycles.
        a. Cycle properties.
        b. Cycle modeling and model evaluation.
        c. Cycle forecasting.
    8) Combining forecasts.
    9) Alternative forecasting techniques.
        a. Prediction markets.
        b. Pattern recognition and artificial intelligence.

Detailed Description of Conduct of Course

Many organizations (businesses, government agencies, unions, non-profits, etc.) rely on forecasts to aid in the development of effective strategies. This course will expose students to basic forecasting methods and help students develop analytical skills that potential employers value. As such, the course will be conducted in a hands-on manner. Students will complete weekly theoretical and applied assignments utilizing real and fabricated (conceptual) data. Applied assignments will be completed using SAS, a statistical software program favored by firms. Students will submit applied assignments in an “on-the-job” medium (i.e. business reports) to expose students to data-oriented business writing. Students will conduct their own forecasting project (with the aid of the instructor) in which they will analyze a data set, formulate a forecasting model and explain its predictions, assess the model’s accuracy and summarize its value. Students will produce a detailed forecasting report highlighting their findings. To ensure students have complete knowledge of the material, students will take two exams (a midterm and a final) which will cover the fundamental techniques. Students will form research groups to aid their performance.

Student Learning Outcomes

Students who complete this course will be able to:
    1) Describe the uses and limitations of different forecasting techniques.
    2) Clean and prepare time series data.
    3) Identify the components of a time series (trend, seasonal, cyclical and stochastic components).
    4) Construct and assess trend and seasonal forecasting models.
    5) Construct and assess mixed moving average and auto-regressive (cycle)
       forecasting models.
    6) Combine multiple forecasts to improve forecasting accuracy.
    7) Justify model selection and compare different models using measures of fit and common sense.
    8) Demonstrate proficiency using SAS to create and evaluate forecasts:
        a. Create a SAS data file.
        b. Use basic SAS commands and functions.
        c. Use SAS forecasting commands.
        d. Read and interpret SAS output.
    9) Write a technical, data-laden report in a business style to summarize and interpret model forecasts.

Assessment Measures

    1) Weekly forecasting exercises
    2) Midterm exam
    3) Forecasting project
    4) Final exam

6. Other Course Information

Students will rely on SAS to perform analyses. SAS resources (tips, instructional videos, help forums) are abundantly available online. Students will use as much real-world data as possible in their work to enhance their experience. Students will gain experience working with a variety of business and economic data, including data on sales, costs, profits, yield, prices, production levels, input levels, spending, inflation, interest rates, unemployment, GDP and exchange rates. Data from the World Bank, US Federal Reserve, OECD, Bureau of Labor Statistics, Bureau of Economic Analysis, Keiser Foundation, NZADAQ, UN, and other local and international sources is often freely accessible and will be used during the class.

Review and Approval