Week 08 · Complete notes
Market entry and barriers to entry
How can entrants overcome, and incumbents create, strategic barriers to entry?
This print edition combines all five readings. Interactive definitions become ordinary text when printed.
01
Getting in, keeping out
When does a threat to fight an entrant actually keep it out?
Monopoly and oligopoly profits attract attention. In Week 3, entry competed profit away in a competitive market. In concentrated markets, the incumbents have reasons and means to resist.
This reading separates barriers that come with an industry from those an incumbent chooses, writes entry as a game and shows why the most common threat, a price war, often fails to deter.
1.1
Profits attract entrants
In Week 3, positive profit in a competitive market drew in new firms until price fell to minimum average cost. In Week 4 we saw that a monopoly can earn profit for years, and in Week 7 that a few oligopolists can share it. Those profits are a standing invitation. An entrant asks a simple question: if I come in, what will I earn after the incumbents respond?
The answer depends on two kinds of obstacle.
- come with the industry: large economies of scale, high capital requirements, scarce inputs, licenses and patents. Week 7 listed them as reasons why some markets have few firms.
- are chosen by the incumbent to change the entrant's answer: price cuts, extra capacity, loyalty programs, exclusive contracts with suppliers or retailers, bundles that tie a new product to an established one, and large patent portfolios.
Structural barriers can be analyzed with the tools of Weeks 2 to 4. Strategic barriers need game theory, because what matters is how the entrant expects the incumbent to behave.
Structural or strategic?
A new steel plant must produce at least 10% of national demand to reach the industry's lowest average cost.
Economies of scale come with the technology. No incumbent chose them to keep anyone out.
An airline gives its frequent flyers status that they lose if they switch carriers.
The incumbent designed the program, and part of its value is that it makes customers costly for an entrant to win.
A beverage maker pays bars to stock only its drinks.
Exclusive dealing is chosen, and it can deny an entrant access to customers. Competition law looks closely at such contracts when the incumbent is dominant.
1.2
The entry game
In the 1940s Alcoa produced nearly all the primary aluminum in the United States. Suppose Reynolds is considering entry. Reynolds moves first: it enters or stays out. If it enters, Alcoa responds: it fights, flooding the market and cutting the price, or it accommodates, sharing the market at a reasonable price. The tree labels accommodation "Share".
- If Reynolds stays out, Alcoa keeps its monopoly profit of 10 and Reynolds earns 0.
- If Reynolds enters and Alcoa accommodates, each earns 5.
- If Reynolds enters and Alcoa fights, Alcoa earns 4 and Reynolds 1.
Solve it backward. After entry, Alcoa earns 5 by accommodating and 4 by fighting, so it . Reynolds anticipates this: entering earns 5, staying out earns 0. Reynolds enters, and Alcoa accommodates.
Alcoa could announce that it will fight any entrant. If Reynolds believed it, Reynolds would compare 1 with 0 and still enter, so the threat wouldn't work even if believed. In other games a believed threat would deter, and then the question from Week 6 decides it: would Alcoa carry it out? Here it wouldn't. Once Reynolds is in, fighting costs Alcoa 1.
1.3
Why price wars rarely deter
The entry game captures a general point. A price war hurts both sides. Once the entrant has sunk its investment, the incumbent's question is no longer "how do I keep the entrant out?" but "what's my best response, given that it's in?" Usually that's to share the market at a decent price, not to fight.
An entrant who understands this ignores the threat. That's why incumbents who want to deter entry do something before entry that changes their own payoffs after it. Reading 2 shows how.
Summary. Structural barriers come with the industry; strategic barriers are chosen by the incumbent to change the entrant's payoff. In the basic entry game, Alcoa earns 5 by accommodating and 4 by fighting, so its threat to fight isn't credible, and Reynolds enters. Once an entrant is in, the incumbent picks its best response given entry, which is usually to share the market rather than wage a costly price war.
02
Commitment and capacity
How can an incumbent make a threat credible?
A threat is credible only if carrying it out pays when the time comes. An incumbent can make that true by acting first, in a way that changes its payoffs later.
This reading adds a capacity decision to the entry game, shows where such payoffs come from in a simple market and explains why a low price before entry isn't enough.
2.1
Alcoa builds capacity first
Now Alcoa chooses its capacity before Reynolds decides. Normal capacity leaves the game as it was. High capacity costs more to build and run, so it lowers Alcoa's payoff in every outcome: 7 instead of 10 if Reynolds stays out, 2 instead of 5 if Alcoa accommodates, 3 instead of 4 if it fights. But it also makes fighting cheaper relative to accommodating. With the extra capacity already paid for, flooding the market costs Alcoa little more, while sharing it leaves expensive plant idle. A fight now leaves Reynolds with −1.
Solve backward.
- After high capacity and entry, Alcoa earns 3 by fighting and 2 by accommodating. It fights. Reynolds would earn −1 from entering, so it stays out, and Alcoa earns 7.
- After normal capacity, the game is the one from Reading 1: Reynolds enters, Alcoa accommodates, and Alcoa earns 5.
Alcoa compares 7 with 5 and builds high capacity. In the credible Nash equilibrium, Reynolds stays out. The threat to fight is now credible, because after high capacity, fighting really is Alcoa's best response to entry.
2.2
What makes a commitment work
High capacity lowers every one of Alcoa's payoffs. Taken by itself, it's a bad investment. It pays only through its effect on Reynolds. That's the defining feature of a strategic : it makes you worse off in a way that changes what your rival expects you to do.
Three properties make it work.
- Visible. Reynolds must see the capacity before it decides. Secret capacity deters no one.
- Costly and hard to reverse. If Alcoa could sell the extra capacity cheaply after Reynolds stayed out, or before it fought, Reynolds would know that, and the commitment would unravel.
- It changes the later choice. The point isn't to look strong. It's to make fighting Alcoa's best response if entry happens.
A commitment works by removing options. Alcoa's cheap way to accommodate is gone once the capacity is built. We saw the same logic in Week 6, when Akubra's promise to match prices kept Barmah from undercutting, and in the Advertisement Game, where an in-house department committed a firm to advertise more.
2.3
Where the payoffs come from
The Alcoa numbers are illustrative. Here is a market in which deters entry for reasons we can compute. Demand is \(p=900-q_1-q_2\), and production costs nothing up to a plant's capacity. A small plant costs 50,000 a year and can produce 100 units; a large one costs 175,000 and can produce 700.
Firm 1 is the incumbent. With no threat of entry, it compares a small plant, producing 100 units at a price of 800 for a profit of \(80{,}000-50{,}000=30{,}000\), with a large plant, producing the monopoly output of 450 at 450 for \(202{,}500-175{,}000=27{,}500\). Alone, it would build small.
Now Firm 2 can enter after seeing Firm 1's plant, and then the two compete on quantity, as in Week 7.
- Firm 1 small, Firm 2 small: both produce at capacity, 100 each, at a price of 700. Each earns \(70{,}000-50{,}000=20{,}000\).
- Firm 1 large, Firm 2 small: Firm 2 produces its 100, and Firm 1's best response is \(450-\frac{100}{2}=400\). The price is 400. Firm 2 earns \(40{,}000-50{,}000=-10{,}000\).
- Firm 1 large, Firm 2 large: the Cournot outputs, 300 each at a price of 300. Each earns \(90{,}000-175{,}000=-85{,}000\).
If Firm 1 builds small, Firm 2 enters with a small plant and both earn 20,000. If Firm 1 builds large, any entry loses money, so Firm 2 stays out and Firm 1 earns 27,500. Firm 1 builds large. The large plant is worth building only because of the entrant it keeps out. The practice set works through this game in full.
2.4
Limit pricing
Could the incumbent deter entry just by keeping its price low? This is : charging less than the monopoly price, but above cost, before entry, to make the market look unattractive.
The trouble is the same as in Reading 1. The entrant cares about the price after it enters, and after entry the incumbent will choose its best price given entry, whatever it charged before. A low price today is not, by itself, a commitment to anything tomorrow.
Limit pricing can still work when the price carries information. If the entrant doesn't know the incumbent's costs, a low price may signal that the incumbent's costs are low and that competing with it would be unprofitable. It can also work when today's price shapes tomorrow's market, for example by locking in customers on long contracts. Without information or lock-in, a low price before entry gives up profit and deters no one.
Summary. A commitment is a visible, costly, hard-to-reverse action that changes what you'll want to do later. Alcoa's high capacity lowers all its payoffs but makes fighting its best response to entry, so Reynolds stays out and Alcoa earns 7 instead of 5. In a capacity model with demand \(p=900-q_1-q_2\), the incumbent builds a large plant that it wouldn't build as an unthreatened monopolist, because the large plant makes any entry unprofitable. A low price before entry deters only if it carries information or locks in customers.
03
Price wars and predation
Can a dominant firm drive rivals out by pricing below cost, and how does the law tell?
Predatory pricing is the most aggressive form of strategic deterrence: deliberately losing money to drive out a rival or scare off entrants, then raising prices once the market is clear.
This reading asks when such a strategy makes sense, why game theory is skeptical of it, and how competition law tries to separate it from the low prices that competition is supposed to produce.
3.1
How predation is supposed to work
has three phases.
- The price war. The predator cuts its price below its own cost, so that a rival or entrant loses money.
- Exit. The rival leaves the market, or potential entrants decide not to come in.
- . The predator raises its price above the competitive level and recovers its losses, and more.
Every phase is costly or uncertain for the predator. During the war, the predator usually loses more than the prey, because it sells more units at the loss-making price. Exit isn't guaranteed. And recoupment requires that the high prices afterward don't simply attract new entrants.
3.2
Why backward induction is skeptical
Suppose the prey understands the plan. It knows that the predator will stop the war as soon as fighting no longer pays, and that a war is costly for the predator. If the prey can borrow to survive the war, it can wait the predator out. Knowing this, the predator shouldn't start. It's the entry game again: once the rival is in, sharing the market usually beats fighting.
The argument extends to many markets. Picture an incumbent operating in 20 towns, facing an entrant in each in turn. In the last town, there are no future entrants to impress, so the incumbent accommodates. In the second-to-last town, the incumbent knows it will accommodate in the last, so fighting now can't scare anyone, and it accommodates again. Backward induction unravels the whole chain: the incumbent accommodates everywhere. This is the chain-store paradox. Like the centipede game in Week 6, it's logically sound and not quite how people behave.
3.3
When predation can work
Predation becomes possible when an assumption of the skeptical argument fails.
- Deep pockets. If the prey can't borrow, because lenders can't tell a temporary price war from a doomed business, a richer predator can outlast it.
- Reputation. If entrants aren't sure the incumbent is a rational accommodator, perhaps because some incumbents genuinely fight, then fighting early on builds a reputation that deters later entrants. A firm in many markets has more to gain from that reputation. This is the repeated-game logic of Week 6 at work.
- Signaling. A low price may convince the prey that the market or the predator's cost is worse for it than it really is.
Two cases show how hard predation is to pin down.
- Qualcomm and Icera. In July 2019 the European Commission fined Qualcomm €242 million for selling 3G baseband chipsets below cost to two large customers, Huawei and ZTE, between 2009 and 2011, with the aim of eliminating Icera, a small rival. The General Court upheld the finding in 2024 and trimmed the fine to about €239 million; Qualcomm has appealed again.
- Amazon and Diapers.com. Amazon cut its prices on diapers sharply in 2009 and 2010, according to Brad Stone's account of the episode by as much as 30%, while Quidsi, the owner of Diapers.com, was growing fast. Quidsi agreed to sell to Amazon in November 2010 for about $545 million. Internal Amazon emails released by a US congressional committee in 2020 described a plan to weaken Diapers.com, but no court or regulator has found that Amazon priced predatorily.
A third case is a good discussion question. For years, ride-hailing companies funded by venture capital charged fares below their costs in many cities. Was that predation, an investment in building a network, or both? What would recoupment look like?
3.4
Predation and the law
Low prices are what competition is supposed to deliver, so competition law has to separate predation from vigorous price-cutting, and the cost of a mistake runs both ways. Courts and agencies look for some combination of three things.
- Market power. A firm without it can't hope to recoup.
- Prices below an appropriate measure of cost. In the EU, prices below average variable cost are presumed abusive for a dominant firm. US courts require below-cost prices and a dangerous probability of recoupment.
- Harm to competition, not just to a competitor.
In Australia, the misuse-of-market-power provision, section 46 of the Competition and Consumer Act 2010, was rewritten in 2017. A firm with a substantial degree of market power must not engage in conduct that has the purpose, effect or likely effect of substantially lessening competition. Below-cost pricing is judged under that general test; the earlier provision aimed specifically at predatory pricing was repealed. The cases show both outcomes. The ACCC lost its predatory-pricing case against Boral in the High Court in 2003, because the court found Boral lacked substantial market power in the Melbourne concrete-block market. In 2010 Cabcharge paid $14 million in penalties after admitting contraventions that included supplying taxi meters below cost.
For a manager, the lesson is double. Don't count on a rival's threat of predation, which is costly and rarely credible. And if your firm is dominant, deep price cuts aimed at a rival carry legal risk as well as cost.
Summary. Predatory pricing means pricing below cost to drive out a rival or deter entry, then recouping the losses with higher prices. Backward induction is skeptical: a rival that expects the war to end can wait it out, and the chain-store argument says an incumbent should accommodate in every market. Predation can work with deep pockets against a rival that can't borrow, with a reputation for fighting, or with signaling. Because low prices are what competition is supposed to deliver, the law looks for pricing below cost and harm to competition, not just low prices.
04
Strategies for entrants
How can a newcomer get in without provoking a fight it would lose, or persuade customers to rely on it?
Entrants have strategic moves too. One is to make customers trust that the entrant's position won't be used against them. Another is to come in small enough that the incumbent prefers to leave the entrant alone.
This reading looks at both: Intel's second-source licenses, and the judo strategy of Kiwi Airlines and Red Bull.
4.1
The trust game
A firm with a breakthrough product can face a different entry problem: its customers fear its power. In 1978 Intel introduced the 8086 microprocessor. Computer makers (OEMs) that designed their machines around it would be locked in. Once they had committed, Intel could raise the price.
Write it as a game. The OEMs move first: adopt the 8086 or don't. If they don't adopt, both sides earn 2. If they adopt, Intel prices high (OEMs 0, Intel 10) or low (5 each).
After adoption, Intel prices high, 10 against 5. The OEMs anticipate it and don't adopt. Both earn 2, though both would earn 5 if Intel could promise to price low. A promise alone isn't credible.
4.2
Giving up power to gain trust
Intel's answer was to license other chipmakers to produce compatible chips, so that customers had second sources. With competitors making the same chip, a high price would win Intel only 3, against 5 from a low price. Second sourcing makes the low price Intel's best response.
With licenses, the OEMs adopt and Intel prices low, earning 5. Without licenses, Intel earns 2. Giving up market power is the profitable move, because it's the only way to make the low price credible to customers who must commit first.
The history is close to this. When IBM chose Intel's 8088, a version of the 8086, for its first personal computer in 1981, IBM's purchasing rules required a second source. Intel signed a technology exchange agreement with AMD in 1982, and other manufacturers also produced the chip: around ten in all. Intel accepted a smaller share of a much larger market. Once the x86 had become the industry standard, customers no longer needed reassurance, and Intel kept later generations, starting with the 386, largely to itself. AMD's challenge to that decision went to arbitration in 1987 and through the California courts before a settlement in 1995. A commitment that is worth making early, when buyers must be persuaded to commit, may not be worth renewing later.
4.3
The judo strategy
A small entrant can't win a head-on fight with a large incumbent. A avoids the fight: move quickly into a position the incumbent doesn't value much, stay flexible, and use the incumbent's size against it. Matching a small entrant's low price often means cutting prices to all of the incumbent's own customers, which costs the incumbent far more than it saves.
Kiwi Airlines, a low-fare carrier, considered entering Delta's Atlanta to New Orleans route. Delta's main customers were business travelers. Kiwi could fly 200 or 400 seats a day and charge $100 or $180. Delta could keep its fares or fight with aggressive discounts.
- With 400 seats, Kiwi would take enough passengers that Delta would fight. Kiwi's $100 fare is dominant, Delta's best response is to fight, and Kiwi earns $6,000 a day.
- With 200 seats, Kiwi can't take many of Delta's passengers. Fighting would cut Delta's fares on its whole flight for the sake of 200 seats, so Delta keeps its fares, and Kiwi earns $7,000 a day.
Kiwi does better by entering small. Limited capacity is a commitment, like Alcoa's large capacity, but pointing the other way: it promises not to threaten the incumbent, so the incumbent's best response is to leave the entrant alone. This is the Week 8 assignment, which works through the full game.
4.4
Red Bull
Red Bull launched in Austria in 1987 and entered the United States in 1997, starting with a few states in the West and Southwest. The cola giants dominated soft drinks; energy drinks were a tiny niche. Red Bull didn't fight them on their ground. It sold a small, expensive can through bars, clubs and convenience stores chosen account by account, sponsored extreme sports instead of buying mass advertising, and cultivated an image the big brands couldn't copy without damaging their own. Coca-Cola's first energy drink, KMX, reached US stores only around the end of 2000.
Why didn't the incumbents respond sooner? The niche was too small to matter to them at first. A response risked cannibalizing their own brands and lending legitimacy to the category. And the product's edgy image didn't fit their family-friendly marketing. Each reason turns the incumbent's size against it, which is the core of the judo strategy. By the time the giants entered, Red Bull had defined the category and held most of it.
Summary. Entrants have strategic moves too. When customers must commit first, a firm may need to give up power to be trusted: Intel's second-source licenses made the low price its best response, and the OEMs adopted, raising Intel's profit from 2 to 5. A judo strategy avoids a fight the entrant would lose. Kiwi earns more with 200 seats than with 400, because a small entry makes accommodation Delta's best response. Red Bull entered a niche the cola giants didn't value, and by the time they answered, it had defined the category.
05
Hold-up, and putting it together
Why do firms underinvest when an investment ties them to one partner, and how can one decision use the whole course?
Entry often requires investments that are worth much more with one partner than without it: a parts plant next to one carmaker, a spur line to one mine. Once the investment is sunk, the partner can renegotiate.
This reading shows how that threat distorts investment, what firms do about it, and then works through one entry decision with the tools of all eight weeks.
5.1
Specific investments
A is worth more inside a particular relationship than outside it. Examples:
- A coal mine that depends on one railway to reach the port.
- A distributor that spends heavily on marketing one manufacturer's product.
- A supplier that does early design work for one buyer's new model.
- A firm that builds a pipeline whose returns can later be changed by regulation.
In each case, once the investment is made, the investor depends on the other side. If the terms of the deal can be reopened after the investment, the other side can take a large part of the return. Anticipating that, the investor invests less, or not at all. This is the .
5.2
NiceBody and CoolCar
NiceBody makes car bodies and is negotiating with CoolCar, which is planning a new model. NiceBody must invest in tooling. An investment of size \(x\) costs NiceBody \(x^2\) million dollars. The tooling is worth \(3x\) million to CoolCar, but only \(2x\) million to other buyers, because it's partly built around CoolCar's design.
Suppose the price is set after NiceBody invests, and CoolCar names it. CoolCar offers just enough to beat NiceBody's other buyers: \(2x\). NiceBody keeps \(2x-x^2\) and chooses \(x\) to maximize it:
NiceBody invests 1 and keeps \(2-1=1\). CoolCar keeps \(3-2=1\). Total surplus is 2.
The efficient investment maximizes total surplus, \(3x-x^2\):
At \(x=1.5\), total surplus is \(4.5-2.25=2.25\). Hold-up costs 0.25 of surplus. NiceBody bears the whole cost of each extra unit of investment but receives only the part of the return it could get elsewhere, \(2\) per unit instead of \(3\). The specific part, 1 per unit, goes to CoolCar, so NiceBody stops early.
5.3
What firms do about it
The problem comes from bargaining after the investment is sunk. Every remedy changes that.
- Agree the price before investing. A long-term contract that pays NiceBody \(3x\), or a fixed price for a specified tooling, lets NiceBody keep the return on its investment. The limit is that contracts are : if quality can't be verified in court, or circumstances change in ways nobody foresaw, the contract gets renegotiated, and hold-up returns.
- Change who has the bargaining power. If NiceBody names the price after investing, it asks for \(3x\), keeps \(3x-x^2\) and invests the efficient 1.5. Bargaining power should sit with the party making the specific investment.
- Integrate. If CoolCar owns NiceBody, there's no price to renegotiate, and the joint owner invests efficiently. has costs, such as weaker incentives inside a large firm, but it's a common answer when specific investments are large.
- Keep alternatives open. Making the investment less specific, or using several suppliers, raises the investor's outside option. Intel's second sources did this for the OEMs.
- Rely on reputation. A buyer that deals with many suppliers over many years loses future business if it holds one up. That's the repeated-game logic of Week 6 again.
OMBA's EnerTech scenario is a good case to discuss. A solar developer signed a 20-year contract with a grid operator, invested billions in a remote site and then faced a demand to renegotiate when market prices fell. Ask which of these remedies could have protected it, and at what cost.
5.4
Putting it together: one entry decision
Suppose you're advising a mid-sized Australian company considering entry into a regional market served by one large incumbent. Here's how the course's tools line up.
- Value and cost (Weeks 1 and 2). What will customers pay, and what does entry cost? Count opportunity costs, and separate the sunk cost of entering from the avoidable costs of operating.
- The market (Weeks 3 and 4). Is the incumbent's profit a scarcity rent that entry will compete away, or a monopoly protected by structural barriers? What does demand elasticity say about how much the incumbent marks up?
- Pricing (Week 5). Can the entrant serve a segment the incumbent prices out, with a different version or a different price?
- Strategy (Week 6). Write the game. After entry, what is the incumbent's best response? Is any threat it makes credible?
- Competition after entry (Week 7). Will the firms compete on price or on capacity? Are the products identical or differentiated? Will they learn to coexist, or tacitly collude, if they meet every day?
- Commitments (Week 8). Has the incumbent committed to fight, for example with excess capacity? Can the entrant commit not to threaten it, by entering small or in a niche? Does entry require specific investments that a partner could hold up?
A recommendation that answers these questions in order is a managerial economist's recommendation: it says what the entrant should do, and why the other side will respond the way you expect.
Summary. A relationship-specific investment is worth more inside one relationship than outside it, so once it's sunk the partner can renegotiate and take the specific part of the return. Anticipating hold-up, the investor underinvests: NiceBody invests 1 instead of the efficient 1.5 when CoolCar sets the price afterward. Remedies change the bargaining after investment: price agreed in advance, bargaining power with the investor, vertical integration, outside options or reputation. Contracts are incomplete, so none is free. The course's tools combine in one entry decision: value and cost, market structure, pricing, the game, competition after entry, and commitments.