Learning and collusion in new markets with uncertain entry costs

Abstract

When new opportunities appear – the development of new products, to enter new geographic markets – firms begin to acquire information. They gradually assess these new events through studying the markets and testing products. Only once they have become sure that the new market is profitable do they decide to enter it. While such a strategy is clear for a firm that has a monopoly, it is much less clear in an oligopoly where two competing firms acquire information independently and secretly about the fixed costs of entry. In this theoretical article, Bloch, Fabrizi and Lippert study a competitive learning model. A firm sees that its competitor has not gone into a new market, and does not know whether it took that decision after observing a negative signal, or simply on the basis of not having any information about the cost involved. As time passes, the probablility that no information has been received diminishes and the firm becomes more and more optimistic about its future profits. Thus, at a set time, it chooses to enter the market even if it has still not received any signal about the entry cost. This is also how the model leads, like the 1985 Fudenberg and Tirole model, to a pre-emption equilibrium; however, this equilibrium is due not to a exogenous change in costs as in the latter model, but to an endogenous change in beliefs. The authors then examine the collusion. How can firms avoid full-frontal competition on the new market and come to an understanding about the sharing of profits? This question is especially thorny when the firms do not know which of the two has the lowest entry costs, and when information about the costs is acquired gradually and in hidden fashion. A netting agreement allows a firm that has entered the market to pay its competitor to remain outside that market. But this compensation is necessarily inflated – because the firm does not know whether its competitor has received a negative signal or not. The more time passes, the more a firm appears reticent to compensate a competitor for which it thinks there is a strong possibility that the cost of entry would be too high for it to threaten the market. These netting agreements can only be efficient if a leading firm receives a positive signal about the cost of entry early enough. To put it another way, a compensation agreement between two firms is only possible if the difference in information between them is not too great; in addition, the agreement must give to the second firm a proportion of the profits sufficient to equalise them (the “leader” and the “follower”) and thus reduce excessive pre-emption on the market.