Nested Rotterdam model: Applications to marketing research with special reference to telecommunications demand

Nested Rotterdam model: Applications to marketing research with special reference to telecommunications demand

0.00 Avg rating0 Votes
Article ID: iaor1989191
Country: Netherlands
Volume: 4
Start Page Number: 193
End Page Number: 206
Publication Date: Jul 1988
Journal: International Journal of Forecasting
Authors:
Keywords: marketing, forecasting: applications
Abstract:

The Rotterdam model is a highly structured econometric model that can be used to model and forecast market demand in a system-wide manner. Since marketing researchers are primarily interested in a subset of consumer goods, the application of the Rotterdam model to marketing research requires the assumption of some separability conditions such as preference independence or block independence on the utility function. If the consumer’s preferences for consumer goods can be characterized by a utility function which is the sum of individual utility functions for all goods, it is called preference independence. Similarly, block independence refers to the situation where the consumer’s preferences for consumer goods can be expressed by a utility function which is additive in the groups of goods. Under these separability assumptions, it is possible to restrict attention to a subset of goods. Such a model is called a conditional form of the Rotterdam model. This is in contrast to the unconditional form of the Rotterdam model which deals with the consumption of all goods and services as a whole. This paper discusses an embedding procedure that has been developed for linking the conditional and unconditional forms of the Rotterdam model and several statistical issues associated with use of the model. The resulting unconditional model can be used for forecasting the market demand at a future time. The data required for estimating the Rotterdam model are usually a time series of goods and services being studied.

Reviews

Required fields are marked *. Your email address will not be published.