Forecasting Consumer Installment Credit: An Application of Parametric Time Series Modeling

Abstract

One of the more interesting developments in the United States financial markets the last half-century has been the remarkable growth in consumer installment credit. Whereas use of consumer installment credit was relatively limited even up to World War II, it has today reached a significant dollar volume. For example, total consumer installment credit outstanding at year-end 1971 amounted to approximately $109 billion. Another significant dimension of consumer installment credit in addition to its large dollar volume is its cyclical and seasonal variability. These cycles are evident not only in the aggregate data but for each component of consumer credit as well as source of credit.' Because of these considerations, variations in installment credit have sometimes been singled out as either a. key factor in generating short-run movements in national income or as a significant offset to short-run monetary and fiscal stabilization policies.2 In view of this it is surprising that there have been only a relatively few studies analyzing short-run movements in consumer installment credit or studies concerned with forecasting future movements in this series. The objective of this study is to derive a forecasting model for consumer installment credit using the parametric time series modeling and forecasting approach developed by Box and Jenkins.3 Using this procedure, forecasts can be obtained using only the information contained in the past history of the series. The following section briefly discusses the Box-Jenkins methodology and utilizes this technique in deriving a forecasting model for consumer