Better Together? Retail Chain Performance Dynamics in Store Expansion Before and After Mergers∗

We study firm performance dynamics in retail growth using a dynamic model of expansion that allow these dynamics to operate through an unobserved serially correlated process that evolves both endogenously and stochastically. The model is estimated with data on conveniencestore chain diffusion across Japanese prefectures from 1982 to 2012, whereby an actual merger between two chains takes place in 2001. Given the presence of serial correlation and selection biases in observed revenue, we combine recently developed particle filtering methods for dynamic games with control functions in revenue regressions. The estimated structural model provides us insights about how performance dynamics evolve before and after the merger. Our findings suggest that despite major increases to revenue and profit growth, the underlying unobserved performance dynamics for the merged entity does not improve following the merger.