Week 7: Practice

Pricing Electric Turbines

In the 1960s and 1970s, General Electric (GE) and Westinghouse made 90% to 100% of the large electric turbines sold in the United States. Here is a simplified version of their pricing problem.

Production cost is zero, or equivalently, prices are quoted net of cost. Each firm can quote one of three prices: $2 million, $6 million or $8 million. There are 100 identical customers, each willing to pay up to $10 million for one turbine, and they see the two firms' turbines as identical. Each firm can supply all 100. Customers buy from the firm with the lower price; if the prices are equal, 50 buy from each. The firms quote their prices secretly and at the same time.

Question 1

Write the payoff matrix. Does either firm have a dominated strategy? Find the Nash equilibria.

Reveal the solution
Profits in millions of dollars.

No price is worse than another against every rival price, so neither firm has a dominated strategy. $8m comes close: it never does better than $6m and does worse against $6m and $8m, but against $2m both earn nothing.

Best responses: against $2m, quote $2m (100 against 0); against $6m, quote $6m (300 against 200 or 0); against $8m, quote $6m (600 against 400 or 200). There are two Nash equilibria, ($2m, $2m) and ($6m, $6m). Both firms prefer the second, so they might well coordinate on it.

Question 2

Is ($6m, $6m) a good prediction of what the two firms would charge if they could quote any price? Explain using the Bertrand result.

Reveal the solution

No. ($6m, $6m) is an equilibrium only because the price grid has nothing between $2m and $6m. If GE could quote $5.9m, it would take all 100 customers and earn 590 instead of 300. Westinghouse would then undercut in turn, and so on. With any price allowed, identical products, equal costs and enough capacity, the only equilibrium is the Bertrand result: both quote marginal cost, here zero. Read the matrix as a warning: a coarse set of strategies can hide the undercutting that drives price competition.

The real turbine market didn't end at cost, which tells us that some Bertrand assumption failed. One was capacity: building turbines takes years, and a firm with a full order book can't take every customer. The full version of this exercise, with capacity limits, needs mixed strategies, which this course doesn't cover. Repeated play was another: GE and Westinghouse met in the same market year after year, and in the 1960s their executives were convicted of price fixing.

Concept atlas