CLASS 5: ENTRY DETERRENCE (chapter 12-14)
Chapter 12: Entry Deterrence and predation
Why do large dominant firms persist in the economy?
- Why have smaller companies had difficulty in gaining market share in such markets?
- Why are no entrants attracted by the profits?
- Natural monopolies are not all that common
One explanation could simply that the emergence of large dominant firms is a natural process
12.1 Market structure over time: random process & stylized facts
Before we begin to model incumbent firm strategies to deter rival’s entry, we need to consider more
fully what theory and evidence can tell us about how an industry’s structure might evolve over time
Example:
• Industry of 256 firms à each firm has sales of $10 million
• In each period sales may increase, decline or stay the same:
§ 25% probability sales will decrease by 30%
§ 25% probability sales will increase by 30%
§ 50% probability sales will remain the same as last period
Sales ($M)
Period 2.40 3.43 4.90 7.00 10.00 13.00 17.90 21.97 28.56
0 256
1 64 128 64
2 16 64 96 64 16
3 4 24 60 80 60 24 4
4 1 8 28 56 70 56 28 8 1
Table shows the evolution of firm size over just 4 periods. We can see that even over this short
amount of time, the distribution of firm sizes is becoming very unequal
à largest firm is now nine times as large as the smallest
à the top nine firms account for roughly the same amount of output as the bottom forty
à this inequality will grow even larger as we let the process run over more periods of time
= Gibrat’s Law
- Let Xt be a firm’s size at time t and vt a random shock
- Let this be related to its size at t – 1 by the following process: Xt = (1 + vt )Xt-1
If we take the natural log of both sides, we end up with:
The size of the firm x at time t is the sum of the prior random shocks plus its original size
The distribution of the natural log of firm sizes follows a normal distribution
But the distribution of actual firm sizes follows a skewed distribution
In real world, we see companies that have been able to preserve their
competitive positions for a long time..
(SLIDES!!!)
,This ‘mechanistic’ process does not take into account:
- Innovation
- Mergers
More importantly, Gibrat’s law does not take into account strategic action that may affect entry in
the first place
Let’s look at an entry situation (textbook example 11.3)
11.3 Credibility of threats and nash equilibria for dynamic games
2 firms:
- Microhard (incumbent)
- Newvel (potential entrant)
Sequential game:
- Newvel decides to enter or not
- Microhard chooses whether to “fight” of “accommodate”
Microhard
Fight Accommodate
Newvel Enter 0,0 2,2
Stay out 1,5 1,5
What is the outcome of this game? à two Nash equilibria
Let’s take a look at the 2nd equilibrium (stay out, fight) = (1,5)
Why would Microhard “fight” (keep price low) if Newvel stays out?
à to make Newvel think it would fight so it doesn’t come into the market
à by adopting the Fight strategy, Microhard says to Newvel: “I am going to price high as long as you
stay out, but if you enter the market, I will cut my price and smash you”
à the problem is that this threat suffers a serious credibility problem
If Newvel enters the market, it is better for Microhard to accommodate instead of fight. So, why
should Newvel believe that threat in the first place?
à any Nash equilibrium strategy combination based on non-credible threats is not very satisfactory
à we need to strengthen our definition of Nash Equilibrium to rule out such strategy combinations
à subgame perfection of a subgame perfect Nash equilibrium
A Nash equilibrium is subgame perfect only if at that point in the game when a player is called upon
to make good on a promise or a threat, doing exactly that and fulfilling the promise of threat is what
would be the player’s best response
If any promises or threats are made in one period, carrying them out is still part of a Nash
equilibrium in a later period should the occasion arise to do so.
à therefore we prefer to use an extensive or tree representation of the game
Fig
h t (0,0)
r
E nte Microhard
Ac
Newvel c om (2,2)
Sta mo
yo
ut
(1,5) da
t e
,à a subgame is defined as a single node and all the actions that flow from that node
In the extensive game illustrated, there are 2 subgames:
- There is a full game starting from node Newvel (the full game is always a subgame)
- There is a subgame starting at node Microhard, and including all subsequent actions that
flow from this node
A strategy combination is subperfect if the strategy for each player is a best response against the
strategies of the other players for every subgame of the entire game.
In this case, it is apparent that for the subgame beginning at node Microhard, the best response
strategy is Accommodate and not fight.
à Hence, the strategy combination (stay out, fight) can not correspond to be subgame perfect
equilibrium
à the only such equilibrium in this case is that of (enter, accommodate)
Tip: the simplest way to identify the subgame perfect equilibria is to work backwards, eliminating
branches that will not be taken until we have reduced the game tree to having a single branch from
each node.
11.4 The chain paradox
What is Microhard faced entry in multiple markets?
à It may want to establish a reputation as a ruthless competitor in one market to prevent entry in
other markets
What is the outcome of this game?
à entrant will enter each market and Microhard will accommodate in all markets
The fact that extension of the above game to many markets and to other rivals may not lead to a
different outcome is a famous result dubbed by Selten as “The Chain-Store Paradox”
So far, incumbent’s threats to fight entry are not very credible
à are there tactics that incumbents can use to effectively deter entry by potential entrants?
• Limit pricing
• Predatory pricing
• Capacity expansion
• Bundling
, 12.2 Predatory conduct and Limit pricing
Predatory actions come in two broad forms:
- Limit pricing: prices so low that entry is deterred
- Predatory pricing: prices so low that existing firms are driven out of the market
The outcome in each case is the same – the monopolist retains control of the market
Anti-trust legal actions tends to focus on predatory pricing.
à Why? Because there is an identifiable victim: a firm that was in the market but left
When is conduct “predatory”
Predatory actions are those that are profitable only if they deter competing firms from entering or
staying in a market
à This is not the same as normal responses to competitive pressure
à Predatory conduct often appears to lower profit and is thus “irrational” (except for the additional
profit an incumbent would earn by driving other firms out of the market)
Let’s look at the Limit pricing first…
Consider the following model:
• A Stackelberg leader chooses output first
• A potential entrant believes that the leader is committed to this output choice
• Entrant has decreasing costs over some initial level of output
• The potential entrant’s cost curves are ACe and MCe
• Demand is D
In principle, the entrant can enter the market
Suppose that the incumbent commits to an output Qd
à the residual demand for the entrant shifts to the left
(to Dd)
à price pe is not sufficient to cover ACe, thus entry is
unprofitable
By committing to Qd the incumbent deters entry
Similarly, the incumbent can commit to a price pd to
deter entry
Committing to output Qd may be aimed either at eliminating a existing rival or blocking a potential
entrant