COMPETITION AND ITS VIRTUE TO NEW STARTUPS UNDER THE POTENTIAL THREAT OF ENTRY

Abstract

An innovation racing model simulates the strategic investment of rapidly growing but recently established U.S. startups, defined as rapid-growth startups, with differentiated original markets from pre-existing incumbent markets. For this, an entry signaling game is considered. Large incumbents send a signal either 'enter' or not enter' to the startups' original markets, which will affect the relative innovation qualities of the startups. If a 'not enter' signal reaches to the startups' original markets, the market perceived innovation of the startups would increase and vice versa. Combined with a high or a low value of market shock, I construct four business scenarios: (high market perceived innovation, high market luck), (high market perceived innovation, low market luck), (low market perceived innovation, high market luck), and (low market perceived innovation, low market luck). The simulation is designed to test a necessary and a sufficient condition. A necessary condition is that rapid-growth startups must be able to grow to be market leaders and a sufficient condition is that rapid-growth startups must overcome hostile M&A threats by the large incumbents. Differently from a general presumption that startup companies lacking initial endowments are less likely to grow rapidly under the potential threat of entry, the model predicts two conflicting outcomes. First, rapid-growth startups are more likely to maintain their market leaderships when their original markets are expected to become more competitive. Therefore, crisis would visit, surprisingly, when they enjoy strong market position rather than when they are exposed to tense competition. Second, when their original markets become competitive, rapid-growth startups are able to acquire their early competitors. These simulation results suggest an important message to policy makers: competition is desirable to raise rapidly growing startups. A panel fixed effect model supports the simulation predictions.