Repository logo
 

The dynamic spillovers of entry: an application to the generic drug industry

Accepted version
Peer-reviewed

Type

Article

Change log

Authors

Ronald Gallant, A 
Hong, H 
Khwaja, A 

Abstract

This paper examines if experience from entry in one market can potentially enhance profitability at a future market opportunity for a related product. We formulate and estimate a dynamic game of entry in which forward looking firms make decisions not just based on present benefits of past entry but also anticipating potential future benefits of current entry. Dynamic spillovers of entry are incorporated through a firm specific unobservable (to the researcher) cost that depends on past entry decisions. The unobserved costs may also be serially persistent. Thus, the model allows for firm specific unobserved heterogeneity that evolves based on firm actions. The challenge of estimating a dynamic game with serially correlated unobserved state variables subject to endogenous feedback is overcome by embedding a particle filter based technique in a Nested Fixed Point Algorithm. Using an application to a stylized model of entry in the generic pharmaceutical industry we underscore the motivation for the model specification and the methodology developed. Our estimates imply positive spillover effects of entry. Moreover, these spillovers suggest heterogeneity not just across firms but also within firms over time based on their history of entry decisions. Our results illustrate that entry may potentially provide firms with additional strategic advantage in later markets, and that entry spillovers may be an important factor to consider in the equilibrium evolution of the generic drug industry.

Description

Keywords

dynamic discrete games, particle filter, unobserved state variables, endogenous feedback, serial correlation, dynamic entry, entry spillovers, generic pharmaceuticals

Journal Title

Management Science

Conference Name

Journal ISSN

0025-1909
1526-5501

Volume Title

64

Publisher

Informs
Sponsorship
H. Hong acknowledges financial support from the National Science Foundation [Grant SES1459975] and the Stanford Institute for Economic Policy Research