Product market competition: The impact of observability on the demand for cost system accuracy
This study examines markets of Cournot and Bertrand competition with cost uncertainty to better understand how sources of information about a firm’s and its rivals’ costs affect the firm’s competitive behavior and quantity or pricing decisions. Bertrand oligopolies are an example of a game with strategic complements while Cournot oligopolies are an example of a game with strategic substitutes (Eichberger and Kelsey 2002). Consequently, firms in Cournot markets will react differently than firms in Bertrand markets to cost information about their own costs as well as their rivals’ cost. There is an extensive theoretical literature studying the role of cost information in imperfect markets under uncertainty. An underlying theme in this literature is that the competitive dimensions of a firm’s product market environment affect the impact of cost uncertainty, and consequently, the need for increased accuracy in cost signals, in order to enhance profits. Cournot markets have the been the focus of much of the literature. This paper contributes to this steam of literature by examining the need for increased accuracy in cost signals in order to enhance profits in Bertrand competition with cost uncertainty. It also provides a basis for comparison of the need for increased product cost accuracy between Cournot and Bertrand markets. When the precision of the cost signal is observable, firms in Cournot markets will chose as precise cost signals as possible. When the variance of the cost signal can be reduced costlessly, firms can maximize profit by choosing a cost signal with no variance. Firms competing in Bertrand markets will always choose infinite variance. This is consistent with the opposite results usually associated with Cournot/Betrand comparisons. When the precision of the cost signal is unobservable, firms facing Cournot competition as well as firms facing Bertrand competition will choose the most precise cost signal affordable. Further, when the precision can be increased costlessly, all firms will chose to know their true costs. We also find that for all market conditions, firms facing Bertrand competition will chose more precise cost signals than firms in Cournot markets. When considering individual market parameters, the opposite results usually associated with Cournot/Betrand comparisons reappear with the exception of the cost of investment.