Using Opportunity Costs to Measure Margins and Market Power

In a recent article for the ABA’s The Antitrust Source, Partner Jeremy Sandford, alongside NERA’s Daniel Greenfield, demonstrate that an export price is the opportunity cost of a domestic sale and can thus be used to identify a firm’s economic margin and market power as a domestic seller. The paper identifies settings in which prices and market shares alone determine economic margins, as an alternative to relying on accounting data or demand estimation which can be difficult to use in antitrust settings. 

The views and opinions expressed in this paper are those of the authors and do not necessarily reflect the views of Econic Partners, NERA or either firm’s clients.

Introduction

Measuring incremental profit margins—i.e., the percentage of the price of the last unit of the good supplied that accrues to its supplier as profit—is a fundamental task in most antitrust inquiries, relevant to both market definition and analysis of competitive effects. Despite the importance of margins, they are notoriously difficult to measure in practice with the information that is typically available in antitrust settings, and mismeasured margins can lead to incorrect inferences.

This paper describes a novel method for measuring a firm’s incremental margin using observed prices and market shares when a firm sells the same product from the same production facilities into multiple geographies. The idea is that a price in a geographic market represents the value of selling in that location, and thus the opportunity cost—the economic term for a foregone benefit— of selling the same unit in a different location. For instance, if a firm sells in its domestic market at a given price and sells at a lower price (net of transportation costs) in an export market, additional domestic sales crowd out additional export sales, in that the firm forgoes a unit sale abroad when it sells a unit domestically. Thus, a domestic sale “costs” the firm the incremental revenue it would have received from a sale of an exported unit. This opportunity cost is equal to the firm’s marginal revenue in each market and its (economically relevant) marginal cost of supplying an additional unit, since the firm maximizes profit by choosing levels of output in each market such that its marginal revenues from each market equal its marginal production cost. In short, the two markets are distinct sources of demand but nonetheless connected from the firm’s standpoint because the firm sells the same product in both markets and thus faces the same marginal production cost irrespective of where it sells the incremental unit.

We demonstrate our methodology for estimating margins in two common settings. First, we consider a firm that is a price-taker in an export market (meaning that the firm is too small as an exporter to materially influence the price in the export market) and has increasing marginal costs. In this setting, the opportunity cost of a domestic sale is simply the price in the export market, because price equals marginal revenue when a firm has no market power. The firm’s marginal cost can be directly measured using data on the price the firm receives as an exporter, net of any additional transport costs the firm incurs as an exporter [1]. It follows that the firm’s domestic margin is simply the difference between the domestic price and the export price, net of transport costs.

Second, we consider a setting where the firm has market power as an exporter, meaning that a greater quantity exported results in a lower price received for exported units [2]. While this situation is more complicated because the firm’s marginal revenue as an exporter is less than the export price and not directly observable, we show that prices and market shares alone determine the firm’s margin as a domestic seller under the common setting of Cournot competition, meaning that the firm is one of several suppliers of a commodity product [3]. Our result depends only on an assumption that market demand elasticity is equal (or close to equal) in both markets; this assumption is reasonable because geographic markets are distinguished on the basis of customer arbitrage, i.e., the transportation costs of moving goods across geographies, and not on the basis of different customer preferences or price sensitivities across geographies. If elasticities are not equal, we develop bounds on the firm’s domestic margin—for instance, if the firm has minimal market power abroad, the export price (net of transportation costs) will be a close (if inexact) proxy for the opportunity cost of a domestic sale. Finally, we show that our methodology extends to firms selling into multiple product markets within the same geography if the firm is able to reallocate (or swing) its supply between the two products [4].

We have found the linkage between a firm’s margin as a domestic seller and the price the firm receives for exports to be useful in commodity industries where firms sell into multiple markets; for instance, industries involving chemicals, oil and gas, cement, metals and materials, agricultural products, electricity, or other bulk purchased industrial inputs. We have worked on several antitrust cases in commodity industries where a domestic seller exports residual units, typically with a transportation cost disadvantage and a small market share in its export market [5]. In contrast, for firms selling differentiated products, the link between sales in multiple markets is more complicated, and thus less useful for making inferences about a firm’s domestic margin.

Read the full article here.

[1] For instance, if the firm sells to an exporter who transports the product to a foreign market, the relevant export price would simply be the price paid by the exporter. If the firm were to ship the product to a foreign market itself the transport cost would be any additional marginal cost realized by the firm above and beyond that of a domestic sale.

[2] When a firm has at least some market power, its marginal revenue is below its price, reflecting the fact that to sell an additional unit the firm must lower the price not just on the incremental unit, but also on units that would have been sold at a higher price.

[3] Under Cournot competition, firms sell a homogeneous good (e.g., a commodity like oil and gas, cement, coal, or chemicals) with price in each market set so that total supply equals total demand. Such products are frequently at issue in antitrust inquiries. For instance, see Fed. Trade Comm’n v. Tronox Ltd., No. 18-cv-1622, slip op. (D.D.C. Sept. 12, 2018) (at 31, describing the Cournot model employed by the Federal Trade Commission’s expert witness). See also Fed. Trade Comm’n v. Peabody Energy Corp., No. 4:20-cv-00317-SEP, slip op. (E.D. Mo. Oct. 5, 2020) (at 62, describing the Cournot model used by the Federal Trade Commission’s expert witness).

[4] Similarly, our methodology assumes that the exporting firm is producing at a single plant (or a set of plants across which it equates marginal costs) with a single transportation cost to the export market. The analysis would be more complicated if firms have multiple production facilities each serving multiple markets at differing transportation costs.

[5] See, for example, Initial Decision at 29-35, In re Tronox Ltd., FTC Docket No. 9377 (Dec. 7, 2018) (finding that “chloride TiO2 is a commodity product” (at 35) and that domestic firms exported some quantities of TiO2 (at 29-30).