Primary market and aftermarket competition in the bicycle component industry

Abstract

This study examines the bicycle component industry. This industry is characterized by one dominant firm, Shimano Inc., and four or five smaller players. Firms in the component industry produce components for sale in two related markets, the market for original equipment manufacturers (OEM) and the component aftermarket. A unique data set containing information on aftermarket prices and OEM market shares is used to determine whether or not market power in the aftermarket is a function of OEM market power. The results indicate that concentration in the OEM market is positively related to aftermarket price, while an individual firm's OEM market share is inversely related to aftermarket price.

Acknowledgements

We would like to thank Janet Netz for valuable feedback and thoughtful insights. Participants at the 1996 Midwest Economic Association Annual Meeting and the 1998 Western Economic Association Annual Meeting contributed useful comments. Sarah Jackson (Western Washington University) provided able research assistance.

Notes

Firms accused of such behaviour include Kodak, Prime Computer, Unisys, Xerox, Hewlett-Packard and Siemens. See Voortman (1993) and Borenstein et al. (1995) for details.

The enthusiast market is defined as bicycles sold through independent bicycle shops, sporting goods stores, bicycle mail order companies, and small custom bicycle builders. This market excludes the lower priced bicycles sold through mass merchants.

One firm, SRAM Inc., began by producing only shifters and is currently expanding into other components.

The main source for OEM market information was Bicycling Magazine's Annual Buyer's Guide issue. Only data from 1990–1996 was used because these are the only years for which a full set of data was available. The aftermarket price data came from two different mail-order catalogues, Bike Nashbar and Performance. Non-sale prices were used from catalogues issued in the Spring of each year. In cases where the two catalogues listed different prices for the same component the average price was used. Sales data come from the 2000 NBDA STATPAK produced by the National Bicycle Dealers Association.

Theory does not suggest the optimal lag. Several different lags were tried and the results did not change qualitatively.

Again, with no guidance from theory, several different lags were tried with no qualitative difference.