1: \begin{abstract}

2: We consider a price competition between two sellers of perfect-complement goods.

3: Each seller posts a price for the good it sells, but the demand is determined according to the sum of prices. This is a classic model by Cournot (1838), who showed that in this setting a monopoly that sells both goods is better for the society than two competing sellers.

4: %\mbcomment{Some people that only knows Cournot's work on competition by quntities might thing we are crazy. Maybe say something like ''This is a classic, yet less known, model by Cournot (1838),... ''}

5: 

6: We show that non-trivial pure Nash equilibria always exist in this game.

7: We also % prove bounds on the ``Price of Stability'' that

8: quantify Cournot's observation

9: %\mbedit{that splitting a monopoly to two complementary sellers is worse to sociaty and to the sellers, studying equilibirum welfare and revenue comapred} %

10: with respect

11: to  both the optimal welfare and the monopoly revenue.

12: We then prove a series of mostly negative results regarding the convergence of best response dynamics to equilibria in such games.

13: \end{abstract}

14: