80 likes | 184 Views
Why forecasting?. Understanding future possibilities. Understanding the possible changes in business conditions. Learning how to cope with the possible changes. Learning how to influence the changes. Knowing what is expected to happen to the future demand of goods and services.
E N D
Why forecasting? • Understanding future possibilities. • Understanding the possible changes in business conditions. • Learning how to cope with the possible changes. • Learning how to influence the changes. • Knowing what is expected to happen to the future demand of goods and services. • For decision making purposes, one must analyze what has taken place in the past, what is happening at present, and predict the future based on these findings.
How to Forecast • Consider as many factors as possible which are deemed to be relevant for a good forecast. • Analyze the historical record, i.e. past trends, cycles, or seasonal patterns. When carefully modified, the past serves as an important guide to the future. • Study exogenous (outside) factors which effect specific businesses, industries, national, and global economy. • Identify the expounding factors.
Meaning and Validity • A valid and meaningful forecast must reflect the interactions of social-economic relationships. A forecast can have validity only if it incorporates the relevant strategic factors and variables. • A forecast can have a meaning if its results are useful, i.e. if it can be applied for the formulation of right policies and making correct decisions. • A valid and meaningful forecast has to reflect reality.
Scientific Forecasting • Systematic strategy • Establishes a cause-and-effect relationship among economic variables. • The method makes intensive use of mathematical and statistical techniques applied to economic theories and models. • Research using a systematic strategy is replicable, i.e. once a model is defined, same results should be obtained by others using the same data. • The researcher must follow very clear, precise, logical, and explicitly laid out steps that lead from assumptions to the generation of the predictive model.
Scientific Forecasting • Advantages of the systematic approach • Forces the forecaster to be as objective as possible. • Keeps the forecaster from pretending that he/she has worked out an objective analysis when in fact he/she has not. • Helps discarding obsolete and erroneous theories. • Can be applied to many intricate problems. • Yields the first approximation of a forecast, on which the researcher must continue to build. • Mathematical Model • A mathematical model brings into focus the relationships among different variables in order to test and clarify the results which they render.
Scientific Forecasting • Variables and data • Select the dependent variable for which a forecast has to be made. • Make a list of independent variables which cause the behavior pattern of the dependent variable (i.e. general economic performance, GNP, consumer and producer price indices, growth rate of a particular industry, etc.) • Carefully evaluate which of the variables are quantitative and which are qualitative. • Investigate whether relevant data is readily available for all quantitative variables, both dependent and independent, for a sufficiently long period of time.
Demand Model • Mathematical function • Q = f(P) • Functional relationship • Q = a - b(P) • where • a = intercept (value of Q when P = 0) • b = slope or coefficient (change in P/change in Q)
Words of Caution • There is no absolute guarantee that a brilliant forecast will always produce good results. • No model can yield correct forecasts if the researcher makes false assumptions. • The specification of the right exogenous variables is very important, but difficult. • All models assume that certain relations between economic variables remain sufficiently stable in order to be used for forecasting purposes---never completely stable. • Models are only to be relied upon as a first approximation. Forecasts have to be refined with the intuition and judgment of the researcher.