A Triple Hazard Model for Price and Sales Crashes of New High-Technology Products

48 Pages Posted: 17 Jul 2015

Date Written: July 6, 2015

Abstract

The author propose a triple hazard model to analyze a phenomenon that occurs for information and high-technology new products that face short life cycles: the sales crash, the price crash, and the sales recovery. The model can untangle three important sources of variation in these interdependent events: i) lagged-event causality, ii) heterogeneity due to observed factors, and iii) heterogeneity due to unobserved (and possibly) correlated factors.

Results suggest that the price crash involves a 23% drop in the introductory price. The sales crash amounts to 60% drop in peak introductory unit sales. The occurrence of a price crash significantly decreases the hazard of a sales crash whereas the occurrence of a sales crash significantly increases the hazard of a price crash. The latter effect is significantly stronger than the former. The price crash significantly increases the hazard rate of a sales recovery whereas the sales recovery has a positive but insignificant effect on the occurrence of a price crash. The findings and model have important implications for managers of information and technology new products.

Keywords: Duration models, Hazard models, New products, Sales saddles, High-tech products, Video-games, Bayesian models

JEL Classification: M31, D12, C01, C02, C11, C15

Suggested Citation

H. Mireles, Carlos and Effraimidis, Georgios, A Triple Hazard Model for Price and Sales Crashes of New High-Technology Products (July 6, 2015). Available at SSRN: https://ssrn.com/abstract=2631705 or http://dx.doi.org/10.2139/ssrn.2631705

Carlos H. Mireles (Contact Author)

University of Texas at Austin ( email )

2317 Speedway
Austin, TX 78712
United States

Georgios Effraimidis

Qualcomm, Inc. ( email )

5775 Morehouse Dr
San Diego, CA 92121
United States

Register to save articles to
your library

Register

Paper statistics

Downloads
91
Abstract Views
225
rank
284,628
PlumX Metrics