Page 347 - Encyclopedia of Business and Finance

eobf_F 7/5/06 3:02 PM Page 324

Forecasting in Business

method is difficult because it is hard to identify an appro- the identified relationship to predict the future. Econo-
priate sample that is representative of the larger audience metric models are also called regression models.
for whom the product is intended. There are two types of data used in regression analy-
sis. Economic forecasting models predominantly use
QUANTITATIVE FORECASTING time-series data, where the values of the variables change
MODELS over time. Additionally, cross-section data, which capture
Three quantitative methods are in common use. the relationship between variables at a single point in
time, are used. A lending institution, for example, might
want to determine what influences the sale of homes. It
Time-Series Methods. This forecasting model uses histor-
ical data to try to predict future events. For example, might gather data on home prices, interest rates, and sta-
assume that an investor is interested in knowing how long tistics on the homes being sold, such as size and location.
a recession will last. The investor might look at all past This is the cross-section data that might be used with
recessions and the events leading up to and surrounding time-series data to try to determine such things as what
them and then, from that data, try to predict how long the size home will sell best in which location.
current recession will last. An econometric model is a way of determining the
A specific variable in the time series is identified by the strength and statistical significance of a hypothesized rela-
series name and date. If gross domestic product (GDP) is tionship. These models are used extensively in economics
the variable, it might be identified as GDP2000.1 for the to prove, disprove, or validate the existence of a casual
first-quarter statistics for the year 2000. This is just one relationship between two or more variables. It is obvious
example, and different groups might use different methods that this model is highly mathematical, using different sta-
to identify variables in a time period. tistical equations.
Many government agencies prepare and release time- For the sake of simplicity, mathematical analysis is
series data. The Federal Reserve, for example, collects data not addressed here. Just as there are these qualitative and
on monetary policy and financial institutions and pub- quantitative forecasting models, there are others equally as
lishes that data in the Federal Reserve Bulletin. These data sophisticated; however, the discussion here should provide
become the foundation for making decisions about regu- a general sense of the nature of forecasting models.
lating the growth of the economy.
Time-series models provide accurate forecasts when THE FORECASTING PROCESS
the changes that occur in the variable’s environment are When beginning the forecasting process, there are typical
slow and consistent. When large-degree changes occur, steps that must be followed. These steps follow an accept-
the forecasts are not reliable for the long term. Since time-
able decision-making process that includes the following
series forecasts are relatively easy and inexpensive to con- elements:
struct, they are used quite extensively.
1. Identification of the problem. Forecasters must iden-
The Indicator Approach. The U.S. government is a pri- tify what is going to be forecasted, or what is of pri-
mary user of the indicator approach of forecasting. The mary concern. There must be a timeline attached to
government uses such indicators as the Composite Index the forecasting period. This will help the forecasters
of Leading, Lagging, and Coincident Indicators, often to determine the methods to be used later.
referred to as Composite Indexes. The indexes predict by 2. Theoretical considerations. It is necessary to deter-
assuming that past trends and relationships will continue
mine what forecasting has been done in the past
into the future. The government indexes are made by using the same variables and how relevant these data
averaging the behavior of the different indicator series that
are to the problem that is currently under considera-
make up each composite series.
tion. It must also be determined what economic
The timing and strength of each indicator series rela- theory has to say about the variables that might
tionship with general business activity, reflected in the
influence the forecast.
business cycle, change over time. This relationship makes
forecasting changes in the business cycle difficult. 3. Data concerns. How easy will it be to collect the data
needed to be able to make the forecasts is a signifi-
cant issue.
Econometric Models. Econometric models are causal
models that statistically identify the relationships between 4. Determination of the assumption set. The forecaster
variables and how changes in one or more variables cause must identify the assumptions that will be made
changes in another variable. Econometric models then use about the data and the process.

324 ENCYCLOPEDIA OF BUSINESS AND FINANCE, SECOND EDITION