Why Price Cuts Escalate—and How Firms Can Compete Without Colluding

A payoff matrix makes assumptions visible

A payoff matrix lists each firm’s available actions and the outcome associated with each combination. In a simplified price-war model, each firm can hold price or discount. If both hold, each earns a healthy margin. If one discounts while the other holds, the discounter may gain volume and the other loses customers. If both discount, volume shares may resemble the starting point while both margins fall. The matrix clarifies the temptation: discounting can be individually attractive regardless of the rival’s action, creating a jointly poor result.

The numbers are not facts until grounded in demand and cost. A firm needs contribution margin, expected switching, capacity, fulfillment cost, competitor response and duration. A matrix that ignores stockouts may exaggerate the gain from discounting. One that treats all customers as identical may miss a premium segment. Managers should build ranges and ask which assumptions reverse the preferred action. Game theory disciplines the question “what will the rival do?” by connecting possible responses to incentives, but it does not grant access to the rival’s plan.

Worked example one: an illustrative one-period price game

Two local delivery firms independently choose whether to hold a $20 fee or offer $16 for one month. Illustrative contribution payoffs, in thousands, are: if both hold, each earns 80; if A discounts and B holds, A earns 95 and B earns 45; if A holds and B discounts, the outcomes reverse; if both discount, each earns 55. For A, discounting yields 95 rather than 80 when B holds, and 55 rather than 45 when B discounts. Discounting is A’s dominant action in this constructed matrix. The same reasoning applies to B, so both arrive at 55.

The outcome resembles a prisoner’s dilemma, but the illustration omits brand, service quality and future periods. It also assumes the $16 fee remains above variable cost and that both companies can serve the gained volume. If A’s capacity is full, discounting may produce delay and churn rather than 95. If B has much lower cost, A may not have a symmetric choice. The matrix is useful because it exposes what must be estimated. It should be labeled illustrative and rebuilt with company data, not copied as evidence that price cuts always destroy profit.

Repeated games change incentives but do not legalize coordination

Firms usually meet more than once. In a repeated game, today’s action affects expectations about future responses. A company may avoid a broad discount because it expects rapid matching and years of lower reference prices. Repetition can support stable outcomes when firms value future profit and can observe conduct, but real markets contain entry, innovation and noisy demand. A rival may cut price because its inventory expires, not to start a war. Misreading a temporary promotion as aggression can create escalation neither side intended.

Legal boundaries are essential. Competitors cannot use meetings, private messages, trade associations or signaling arrangements to agree on prices, output, customers or bids. An unlawful agreement need not be a formal signed contract. Firms should set prices independently using their own costs, customers and public market information, with legal review where conduct creates risk. Game-theory language about cooperation or punishment is dangerous if translated into communications with rivals. Repeated interaction explains incentives; it is not a playbook for reaching a common price.

Capacity and commitment can decide whether a threat is credible

Low pricing is not automatically unlawful. Antitrust law generally protects vigorous price competition, and a firm can lawfully charge less because it is efficient or wants to win customers. Specific doctrines such as predatory pricing have legal tests that cannot be inferred from a rival’s complaint or a simple below-average-price observation. Managers should avoid using legal accusations as a substitute for cost analysis. Counsel can assess facts; commercial teams should concentrate on their own sustainable response and preserve records showing independent decision-making.

Capacity determines the value of extra demand. An airline seat departing tonight has near-zero salvage value, while accepting another passenger may carry modest incremental cost if capacity remains. A factory already running overtime faces a different curve: a discount can push volume into expensive production or missed service. Rivals infer from observable capacity, inventory and cost whether a low price can persist. A temporary offer from a constrained firm is less threatening than a price supported by a new low-cost facility. Strategy should model marginal cost across the relevant volume range.

Commitments alter future options. A long-term low-price guarantee may attract customers but make later increases costly. Building capacity signals an ability to serve volume, yet creates fixed cost that can intensify discounting when demand disappoints. Loyalty benefits, contracts and product architecture can also commit a firm to a value proposition. A credible commitment is costly to reverse; an empty threat is not. Before matching a rival, managers should ask whether the move reflects temporary inventory, a durable cost advantage, a new segment or a commitment that changes the market’s structure.

Worked example two: differentiate instead of matching every cut

An illustrative business-software provider charges $120 per user annually. A rival cuts to $90. Immediate matching would reduce annual revenue by $3 million across 100,000 existing seats before considering retention. Customer research shows that one segment is price sensitive, while regulated customers value audit logs and response time. The provider independently introduces a $92 limited tier with fewer integrations for new price-sensitive accounts, keeps the full product at $120, and invests $600,000 in faster support and compliance evidence. The action changes the game by reducing direct comparability rather than asking the rival to restore price.

The provider measures migration from full to limited tier, win rate, support cost, renewal and contribution margin. If existing customers downgrade heavily, segmentation may destroy value; if the limited tier attracts a distinct segment, it can defend volume without repricing every account. This example is not a recommendation or a forecast. It shows how differentiation, service and scope create more actions than hold-or-cut. The provider must describe tier differences honestly and make its decisions unilaterally. Publicly inviting the rival to follow a price or divide segments would raise a different legal issue.

Compete through value, cost and reversible experiments

Non-price competition includes reliability, speed, convenience, design, warranty, financing, service, interoperability and trusted expertise. These investments are not automatically positive-sum; they cost money and may be easy to copy. Their advantage is that they can expand willingness to pay or lower total customer cost without teaching the entire market to wait for discounts. Process improvements can support a genuinely lower sustainable price. A firm with a cost advantage may rationally pass some savings to customers, which is competition working rather than a destructive war.

A practical response process begins with diagnosis. Estimate whether the rival’s move is broad, funded, capacity-backed and durable. Segment customers by switching motive. Model contribution after likely matching and operational strain. Test a bounded offer where legal and feasible, with a prewritten stop rule. Review public statements and competitor contacts with antitrust counsel. The objective is not to avoid competition or preserve an industry profit pool. It is to choose an independent action that the firm can sustain while giving customers a lawful reason to choose it.

1. Populate payoff matrices with contribution, capacity, switching and duration rather than invented profit labels.
2. Test whether a rival’s low price is temporary, cost-supported or backed by a durable commitment.
3. Use repeated-game analysis to anticipate responses, never to coordinate conduct with competitors.
4. Consider segmentation, service, quality and cost redesign before repricing the entire customer base.
5. Keep pricing unilateral and obtain legal guidance for competitor communications, signaling and trade-association activity.

Sources

• US Department of Justice, Antitrust Laws and You
• Federal Trade Commission, Guide to Antitrust Laws
• Federal Trade Commission, Price Fixing