MANAGERIAL ECONOMICS 12 th Edition By Mark Hirschey

Forecasting Chapter 6 © 2009, 2006 South-Western, a part of Cengage Learning

Chapter 6 OVERVIEW l l l l Forecasting Applications Qualitative Analysis Trend Analysis and Projection Business Cycle Exponential Smoothing Econometric Forecasting Judging Forecast Reliability Choosing the Best Forecast Technique © 2009, 2006 South-Western, a part of Cengage Learning

Chapter 6 KEY CONCEPTS Ø Ø Ø Ø macroeconomic forecasting microeconomic forecasting qualitative analysis personal insight panel consensus delphi method survey techniques trend analysis secular trend cyclical fluctuation seasonality irregular or random influences linear trend analysis growth trend analysis business cycle economic indicators Ø Ø Ø Ø composite index economic recession economic expansion exponential smoothing one-parameter (simple) exponential smoothing two-parameter (Holt) exponential smoothing three-parameter (Winters) exponential smoothing econometric methods identities behavioral equations forecast reliability test group forecast group sample mean forecast error © 2009, 2006 South-Western, a part of Cengage Learning

Forecasting Applications Ø Macroeconomic Applications l Ø Microeconomic Applications l Ø Predictions of economic activity at the national or international level, e. g. , inflation or employment. Predictions of company and industry performance, e. g. , business profits. Forecast Techniques l l Qualitative analysis. Trend analysis and projection. Exponential smoothing. Econometric methods. © 2009, 2006 South-Western, a part of Cengage Learning

Qualitative Analysis Ø Expert Opinion l l l Informed personal insight is always useful. Panel consensus reconciles different views. Delphi method seeks informed consensus. Ø Survey Techniques l l Random samples give population profile. Stratified samples give detailed profiles of population segments. © 2009, 2006 South-Western, a part of Cengage Learning

Trend Analysis and Projection Ø Secular trends show fundamental patterns of growth or decline. l l Ø Constant unit growth is linear. Constant percentage growth is exponential. Cyclical fluctuations show variation according to macroeconomic conditions. l Cyclical normal goods have εI > 1, e. g. , housing. Seasonal variation due to weather or custom is often important, e. g. , summer demand for soda. Ø Random variation can be notable. Ø © 2009, 2006 South-Western, a part of Cengage Learning

© 2009, 2006 South-Western, a part of Cengage Learning

Business Cycle Ø The Business Cycle is a rhythmic pattern of economic expansion and contraction. Ø Economic indicators help forecast the economy. l l l Leading indicators, e. g. , stock prices. Coincident indicators, e. g. , production. Lagging indicators, e. g. , unemployment. Ø Economic recessions are periods of declining economic activity. © 2009, 2006 South-Western, a part of Cengage Learning

Exponential Smoothing Ø One-parameter Exponential Smoothing l Used to forecast relatively stable activity. Ø Two-parameter Exponential Smoothing l Used to forecast relatively stable growth. Ø Three-parameter Exponential Smoothing l Used to forecast irregular growth. Ø Practical Use of Exponential Smoothing Techniques © 2009, 2006 South-Western, a part of Cengage Learning

Econometric Forecasting Ø Advantages of Econometric Methods l l Models can benefit from economic insight. Forecast error analysis can improve models. Ø Single Equation Models l Show Y depends on X variables. Ø Multiple-equation Systems l Show many Y variables depend on several X variables. © 2009, 2006 South-Western, a part of Cengage Learning

Judging Forecast Reliability Ø Tests of Predictive Capability l Consistency between test and forecast sample suggests predictive accuracy. Ø Correlation Analysis l High correlation indicates predictive accuracy. Ø Sample Mean Forecast Error Analysis l Low average forecast error points to predictive accuracy. © 2009, 2006 South-Western, a part of Cengage Learning

Choosing the Best Forecast Technique Ø Data Requirements l l Scarce data mandates use of simple forecast methods. Complex methods require extensive data. Ø Time Horizon Problems l Short-run versus long-run. Ø Role of Judgment l l Everybody forecasts. Better forecasts are useful. © 2009, 2006 South-Western, a part of Cengage Learning