Brainscape's Knowledge GenomeTM

Browse over 1 million classes created by top students, professors, publishers, and experts.


See full index

FHS > Industry > Flashcards

Industry Flashcards

1
Q

Define vertical restraint

A

One firm in the vertical chain influences the behaviour of another firm

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
2
Q

Double marginalisation diagram

A
  • MCd = w
  • vertical separation induces a higher price of the final good
  • Vertical externality: each firm doesn’t take into account that an increase in its final price reduces the profit of the other firm
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
3
Q

Other vertical externalities

A

Downstream moral hazard

  • demand function for U includes promotional services ‘S’
  • service forcing (write into contract)
  • FF: set at optimal profit given both choices of p and S

Input substitution
- distortion to how producing good as 1 U monopoly and other U is competitive. Will only buy from competitive

Pre-sale problem with free-riding

  • undermines incentive to do pre-sale
  • RPM with FF can eliminate externality but led to too much effort -> inc profit but dec CS
  • e.g. CMA investigated perfume market as they refused to sell to Superdrug as well as Debenhams
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
4
Q

Uncertainty over cost and demand with FF and RPM

A

Rey and Tirol (1986)

Demand uncertainty:

  • With FF: A < E(profit) which is costly for U
  • RPM: U chooses w until D makes no margin so change in q has no impact on D so no compensation needed

Cost uncertainty
- With RPM: D has cost uncertainty needs to be compensated for

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
5
Q

Exclusionary contracts model

A

1) U offer D exclusive contract with payment t for signing
2) D decides whether to accept
3) E decides whether to enter
4) Active upstream sellers set prices

Assume if E enters, sets p = cu
Assume E enters in absence of contract e.g. (cu - ce)Q(cu) > F

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
6
Q

Segal and Whinston on exclusionary contracts

A

2000:
- now 3 D’s
- Buyers are not in competition: profit from each buyer is 9
- if p = cu then buyer surplus is 12
- assume unable to coordinate to defeat tactic
- firms must operate only at or above minimum efficiency scale

Simultaneous:

  • optimally offer t = 12 + e to 2 and t = 0 to 1
  • dominant strategy to accept

Sequentially:

  • publicly observable and distinct offers
  • unique SPE: U sign up all 3 D or almost free and E doesn’t enter
  • as no. of D inc, buyers become relatively small and U almost certainly exclude for free
  • contracts not expiring at same time and other discriminations in contract: foreclosing potential stronger
  • Each D doesn’t internalise the fact chaat when it signs it dec no. D E can sell to so dec likelihood of entry
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
7
Q

Partial exclusion through stipulated damages

A

Aghion and Bolton (1987)

  • contract: (p,d)
  • Assume F = 0

Certainty: (p,d) = (cu, cu-ce)
- signs and enters, conspire by penalty. Redistribution

Uncertainty: cu = 1/2

  • if E enters: p = pm - d
  • d set before ce known
  • joint surplus: v - (p-d) max when p = 1/4
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
8
Q

Example of exclusionary contract

A

Intel and Dell 2002-2007

  • Intel share > 70%, make integrated circuits for computers
  • AMD outperform Intel: threat
  • arrangement with Dell to buy exclusively from them via patent reward
  • Intel fined > €1bn by EC

1991 Mars complaint as Schiller and LI exclusive agreements with retailers

  • EC decided infringed Article 81
  • make available freezer that only store their products
  • no after or pre-sale services with ice-cream
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
9
Q

Secret deals issue

A

Hart and Tirol (1990):

  • public offers contract vs secret deals
  • have 2 Ds and assume can’t contract multilaterally with both Ds to maintain Qc/2
  • Bilateral contracts: (q,T) T payment from D to U
  • Public: (Qc/4, pmQc/4)
  • If U and D1 agree to the above, then incentive to max residual demand with D2 and set Q > Qc/4
  • Anticipating this, D1 wouldn’t accept -> equilibrium in unique duopoly cournot
  • presence of externalities between the 2 contracts. As generalise to n, profit U -> 0
  • here nondiscrimination clauses can restore monopoly power: exclusive dealing, RPM public promise, VI
  • from competitive point of view its good to have secret deals
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
10
Q

Profitability of horizontal merger without synergies

A

Cournot:

  • insiders internalise negative pecuniary externality and dec their output
  • outsiders inc output which moderates the price increase
  • lower q rom insiders not compensated by inc price in industry
  • Motta: R shifts left as internalise negative externality
  • Salant, Switzer Reynolds (1983): with linear demand, profitable for insiders only if m/n > 0.8
  • Stigler (1950): even if profitable, may be more profitable to let others merge. Violated if n large, no synergies, competition not too tough

Bertrand:
- Motta: R shift right as internalise previously setting price too low relative to price that would max joint profits

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
11
Q

Welfare effects of merger

A

Williamson (1968)
- trade-off graph with higher price, lower cost

Farrell and Shapiro (1990):

  • case where output-reducing merger is welfare increasing
  • output increasing is good for everyone
  • 1) CS inc as p dec if P(Q) - Cm > sum p(Q) - ci). Post merger profit margin at pre merger price > sum of merger partners; pre-merger profit margins. Corollary: Cm < C2 (lower one)
  • 2) Aggregate surplus inc if s < 0.5
  • intuition: low share, smaller impact on price and higher c pre-merger so reallocate output to more efficient firms
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
12
Q

Limits of Farrell and Shapiro (1990)

A

Cournot
Abstracts from collusion
Abstracts from entry
Looks at merger in isolation

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
13
Q

Assessing merger with price competition

A

Farrel and Shapiro (2010):

  • price setting and differentiated
  • Diversion ratio: measure of substitutability between products
  • D = no. customers switch to 2 when 1 raises price / no customer stop buying from 1
  • measured via surveys
  • T=inter-firm externality= (p2-c2)D12 = dprofit2/dq1
  • UPP high if close substitutes and high profit margin of 2
  • Impact T - E
  • don’t have to define the market, market shares not v informative about closeness of substitutes
  • quantities the risk that prices rise based on rate at which firms cannibalise sales from each other
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
14
Q

Lerner index

A

(p-c)/p = s/e

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
15
Q

Empirics of assessing mergers (equations)

A

Single-industry:

  • pi = a0 + B1Ni + yXi + u
  • X: exogenous data
  • i is a different market within the same industry
  • have omitted variable bias: Xi not included that impacts pi and Ni. N is endogenous.
  • e.g.congestion charge impacting costs

Panel data:

  • pit = ai + B1Nit + yXit + u
  • ai: absorbs impacts like demand so not in u
  • picks up the time-constant, location-specific uobserved differences across locations
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
16
Q

Example proposed merger analysis

A

Ashenfelter 2004:

  • Stables and Office Depot, FTC challenged in 1997
  • used panel data: locationally separate individuals Staples outlets at different times
  • data on N (due to entry) and p for each i across time
  • Q: how much would Staples price increase in markets where Staples and Office Depot compete if all Office Depot stores were converted to Staples stores?
  • price inc of 4%
  • Difficulties: defining the market (superstore Walmart), methods hold competition constant (strategic effects), measurement error bias
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
17
Q

Example actual merger analysis

A

Hastings (2004):

  • ACRO acquiring Thrifty petrol stations
  • didn’t need to specify structural specification of demand and competition (quasi)
  • cit: (1 if Thrift within 1 mile, 0 otherwise)
  • ‘treatment’ is the loss of competition from an independent retailer
  • Panel data to compare changes in p between treated and control
  • Results: treated 2-3 cents below control pre-merger and 2-3 cents higher after. Effect = 5c. Consistent across locations
  • pit = ai + dct + o cit + uit
  • dct: city-time effect (common trends per city)
  • cit: treatment
  • panel so ai means cit uncorrelated with sit
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
18
Q

Other things to consider with merger analysis

A

Not actually profitable for insiders:

  • Roll (186): overbidding
  • Morck et al (1990): empire building

Assumes private gains for merging partners are social gains -> e.g. not tax savings

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
19
Q

Examples of blocked mergers

A

Sainsburys and Adsa:
- used surveys -> D and UPP

Siemens and Alstrom:

  • future competition from CHina’s CRRC not impacted by competition authority
  • EC applied competition law by looking at the market as it currently stands

UPS and TNT:

  • EC blocked in 2013: price increase in 15 states
  • FedEx bought in 2014
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
20
Q

Points from Rotemberg and Saloner

A

(1986) :
- perfect information, uncertainty re demand
- monopoly price is independent of demand state.
- ω ∈ {H,L}: Dω(p) = θωD(p)
- Since θH > θL, π HM > π LM .
- If δ < δ ≡ (n − 1)/n price above marginal cost cannot be sustained 􏰆
- if δ ∈ [δ, δ), the pM in low demand state can be sustained while, in the high demand state, a price pH ∈ (c,pM) can be sustained. This price pH is such that the associated industry profit πH ≡ [pH − c]D(pH) satisfies the no-cheating constraint in the high demand state, (1) with equality.
- collusion harder to sustain than stationary
- incentive to cheat in high demand state is bigger
- Move along equilibrium collusive path, as collusive price higher in low demand states, no firm deviates.
- if δ ∈ [δ, δ), there are “price wars” during boom times

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
21
Q

Points from Green and Porter

A

(1984) :
- signal extraction problem
- firms never observe demand state
- if sell nothing dk if under-cut or slump
- - optimal collusive scheme: if q=0 then revert to punishment phase (p=mc) for T periods
- T minimised so that no incentive to cheat
- predicts finite price wars following unobserved demand slump
- no cheat but collusion break down

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
22
Q

Evidence for Rotemberg and Saloner

A

Scherer (1980): pure discipline broke down when large order from Armed Service Procurement Agency

Ashenfelter and Grady (2005): Sotheby’s and Christies. Fixing in 1995. In 1997 market recovered: S waive c, C make donation.

Abrahamson (2011): spreads (fees for IPOs) increased when low volumes. Collusion when low volumes.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
23
Q

Evidence for and against Green and Porter

A

Ellison (1994): JEC - 1880s rail cartel in US.

  • models what expect demand to be and found price 66% higher in cooperative periods.
  • Less likely place to apply RandS - RandS have transparency over prices

OECD (2017): algorithms - able to distinguish intentional deviations and natural reactions to market changes (prevent unnecessary retaliation)

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
24
Q

Multimarket contact authors

A

Multi-market contact

  • reneging on a collusive agreement in one market triggers price wars in all other markets
  • oligopolistic to competitive -> more environments can sustain collusion.
  • punishment is worse than zero
  • Bernheim and Winston (1990): identical, CRS, perfect monitoring then don’t affect opportunities for tacit collusion
  • Porter (1983): destabilising spillover
  • Evans and Kessides (1994): golden rule
  • time series and cross sectional data on 1000 largest domestic city-pair routes
  • regional cost adv so heterogenous
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
25
Q

Transparency impact on collusion

A

Albaek et al (1997):
- Danish concrete industry. 15-20% ands convergence.

Algorithms: like price-guarantees without the clunkiness - immediate responses

Frequency:
- bus tickets vs buying aircraft

26
Q

Common ownership in collusion

A

Azar et al (2018):

  • via tracker funds
  • looked at financial institutions e.g. Citigroup and JPMorgan
  • 3-7% price increase
  • Intuition: ease of sharing info and incentive to cheat reduced
  • Blackrock acquire Barclay’s: 10-12% - difference-in-difference analysis
27
Q

Types of price discrimination (list)

A 
1st degree
2nd 
3rd 
Oligopoly 
Bundling
Durable
28
Q

Talk about first degree price discrimination

A
  • firm has complete info and makes take-it-or-leave-it-offer
  • charge WTP
  • MR becomes AR curve so get full triangle and social welfare maximised
  • e.g. De Beer’s sale of rough diamonds: buy or never buy again
  • too much info to be feasible in most cases
29
Q

Talk about 3rd degree price discrimination

A

Exogenous signal that firm uses to classify customers into groups

  • know overall demand conditions (condition of each sub-market but not individual)
  • p-c/p = 1/e
  • E.g. Varian sold £50 UK but £94 in US

Welfare:

1) open market: gain some monopoly profit and CS in previously unserved market
- e.g. anti-retroviral drugs to Africa - sold to pharmaceuticals who sold back to US not straight to customers

2) misallocation effect: total output inefficiently distributed as MU not equal - consumers optimising against different budget lines (diff price ratio) so different MRS. Profits inc but CS dec by more
Output effect: increase output good for welfare. Pig (1920): liner demand then output unchanged. Welfare reduced.
- Agire et al (2010): curvature L < c H then raise welfare

Schmalensee (1981): necessary condition for discrimination to inc welfare is output inc

30
Q

Talk about 2nd degree price discrimination

A
  • customer self-selection
  • know proportions of H and L usage but not which is which
  • menu of tariffs: T = A + pq
  • 1st is incentive incompatible: Profit = 2B
  • Improve: Profit 2B + E
  • Further improve: design a L package sufficiently unattractive for H
    Profit = 2b + x + y + E
  • damaging goods to artificially provide a range of qualities so self-select
  • L get all surplus extracted, H keep some.
  • E.g. Dupuit (1849): make 3rd class uncomfy to frighten the rich
  • E.g. IBM printers, 8-12 page per min
    E.g. Intel maths coprocessor switched off
31
Q

Talk about oligopoly price discrimination

A
  • Use hotelling model
  • No discrimination: profit = t / 2
  • Good deal equation pa + tx = pb + t(1-x)
  • Compete until less favoured profit = 0
  • pa = c + t(1-2x)
  • apart from 0 and 1 everyone pays less
  • profits = t/ 4 (triangles from squares)
  • game theory: both discriminate as dominant strategy
  • Welfare: no change in t. With monopoly not all served.
  • Robinson-Patan Act (1936): prohibited discrimination in US to stop chain stores getting lower prices than smaller retailers - can actually be beneficial for consumers.
32
Q

Talk about bundling

A

No bundling: qa = 1 - pa and qb = 1-pb
-Profit = 2 x 0.5 x 0.5

Bundling:

  • at p< 1, don’t buy is left triangle = 0.5P^2
  • demand = 1- 0.5P^2
  • p = (2/3)^0.5 < 1
  • profits and CS higher
  • assumptions: consumer demand -ve correlated and heterogeneity in tastes
  • Adam and Yellen (1976): prevalent market strategy: restaurants complete dinners, banks multiple services for a fee.
  • Profitable even if complementaries in consumption and cost saving absent
  • Designed to retain those with lower variance in reservation prices between products
  • just need joint distribution of v not individual
33
Q

Talk about durable goods issue and overcoming it

A
  • Intertemporal price discrimination
  • Maths: q2, q1, sub to get p2, profit
  • coase conjecture: incentive to reduce price in second period for residual demand, so not willing to accept higher price in p1
  • pacman economics: finite number of consumers, only drop price if some buy, some buy immediately as highest surplus - buyers follow ‘get-it-while-you-can strategy’
  • Burlow (1982): less profit as no way of guaranteeing won’t reduce price: will want to keep selling until p = mc so c wait and loose monopoly power.
  • Overcoming: reputation (Disney available limited time), contractual commitment (best-price clause), limited capacity, planned obsolescence (Varian: new every year and lecturers use new one)
34
Q

Empirics for price discrimination

A

Shiller and Walgvogel (2011): digital music

  • iTunes sold for 0.99 all songs
  • Other: two-part tariff, bundling
  • Survey data on students to try get reservation prices
  • Alternatives could raise producer consumer surplus by about 17% in one sample and 30% in another
35
Q

Entry deterrence steps for model

A

1) I chooses K at cost rK
2) E chooses whether to enter at cost F
3) Active firms engage in Cournot competition

36
Q

Credibility in entry deterrence

A

Credibility depends on sunk cost nature of investment and whether profitable.

37
Q

3 cases in Dixit Spence model

A

1) Blockaded profit (A) < F
2) Ineffectively impeded profit(B) > F
3) profit (A) > F > profit(B): point(d1,d2) where profitE = 0
- if d1 < m1 which is monopoly output then blockaded
- if d1 > m1 then either deter by producing d1 or accommodate and choose best point on A-B: S else B

38
Q

Equations:

  • tough vs soft
  • substitutes vs complements
  • direct vs strategic
A
  • T/S: dprofit2/dk
  • C/S: dR2/dx1
  • dprofit (x1,x2,k)/dk = dprofit/dx1dx1/dk (0) + dprofit/dx2dx2/dk + dprofit/dk
    = 0 + strategic + direct
  • x2: firm 2’s ex-post behaviour
39
Q

What is means by investment is soft

A
  • over investment can be a strategic handicap as can reduce the incentive to respond aggressively to competitors
40
Q

Advertising model of entry deterrence

A

Underinvest in advertising as then lower stock of ‘goodwill’ so credible threat to cut prices in the event of entry. Strategic can dominate direct effect (which works in opposite direction)

41
Q

Reasons for preventing pre-emption

A

1) Blockaded
2) Uncertainty
3) resource constraints
4) perfect information (price scenario)
5) Public good issue (free-ride - Lieberman)

42
Q

Empirics for entry deterrence

A

Ghemawat (1984):

  • DuPont in titanium dioxide 1972-77- costs 21-23% lower
  • demand had grown steadily since 1955
  • eco downturn and high price sensitivity: demand-supply imbalance reverse in 1974
  • preemption is a hazardous process in which miscalculations can depress profitability in entire industry for years to come

Lieberman (1987): 38 chemical product industries of 20 years - high fixed cost, n small, economies of scale

  • only built excess capacity in 3 products and all cases proved ineffective
  • market growth (size) and depreciating (time) capital reduces the potency of excess capacity
43
Q

Price predation

A

Milgorm and Robert (1981): signalling model

  • c is private info and p is credible signal of MC
  • asymmetric info crucial role
  • less rational than merging as preserves profits but banned
  • re-entry possible but ignores low recovery rate of fixed costs from exiting
  • if credible, threat rarely carried out in practice - reputation effect of Kreps and Wilson (1981)

Long price theory: can drive out weak financially constrained

44
Q

Issue with RandD

A
  • Its a public good - non-r and non-e
  • Non-rival: once completed it is ex post (static) efficient to price the new info at the marginal cost of replication. Appropriability problem: innovator doesn’t receive rewards.
  • ## Need to make a private good w patents - make excludable. Causes monopoly: “second best policy”
45
Q

Assumptions of RandD model

A
  • patent has indefinitely long-life
  • Memoryless poisson: independent of time, probability of obtaining patent doesn’t depend on past RandD expenditure
  • no uncertainty over type of innovation (whether drastic or not)
46
Q

Talk about replacement effect

A

Reinganum (1983):

  • monopoly replaces himself -> ‘rest on his laurels’
  • E innovates more
  • by inc x, monopoly removes discovery date forward and therefore hastens own replacement
  • marginal productivity of RandD £ decreases with initial profit
  • entrant doesn’t forgo a flow profit when discovering
  • GandN: this assumes away potential entrant. Monopoly with free-entry should have incentive B as well as A
47
Q

Talk about efficiency effect

A

Gilbert and Newbury (1982):
As competition reduces profits, the monopolist’s incentive to remain a monopolist is greater than the entrant’s incentive to become a duopolist
- with drastic innovation, no dissipation of monopoly rent so no efficiency effect

48
Q

Policy instruments for RandD

A
  • govt provision
  • patents
  • copyright
  • subsidies e.g. RandD tax credits
  • prizes
  • research joint ventures (avoid wasteful duplication)
49
Q

Aggregate expenditure on RandD relative to the cooperative optimum and social optimum

A

Reinganum (1989):

  • Too many firms: only considers MB from investment, not the reduction it imposes on expected value of each other’s investment
  • Each invests too much: entry continues until all profits are dissipated. Forgo inter temporal efficiencies which would be realised if invested at lower rate over a longer planning horizon

Comparison with social-maximising surplus is harder since innovator can’t appropriate all surplus -> depress investment so ambiguous.

50
Q

Oligopoly, innovation and entry

A

Dasgupta and Stiglitz (1980):

  • Cournot, RandD to reduce cost, free entry
  • a = elasticity of marginal cost w.r.t RandD
  • RandD/Revenue = a/1+a =
  • with cournot = 1/ne
  • 1/n = ae/ 1+ a
  • higher a: more valuable to do RandD = do more RandD = higher fixed cost = more entry deterred
51
Q

Patent life model

A

Nordhaus model:

  • endogenising size of reduction of cost
  • Higher T raises the gain to the firm (thus x) -> more long term benefits to society but delays the date at which the maximum benefits accrue.
52
Q

“Pay for delay” in RandD

A
  • Monopoly pay entrant to not enter and to retain monopoly profits since > duopoly profits
  • e.g. killer acquistions -> shut-down
53
Q

RandD empirics

A

Cockburn and Henderson (1994):

  • 10 pharmaceuticals - 17 years
  • not accurate: (output across firms negative)
  • spillovers and races with multiple prizes

Lerner (1997):

  • disk-drive industry. Follower greater propensity
  • tech opp and finance constraints not cause

Igami (2017):

  • Evidence of leap-frogging
  • Cannablisation explain 57% of gap , only half incumbents ever innovated to 3.5 inch from 5.25 inch
54
Q

Bundling assumptions

A
  • 2 periods
  • no discounting
  • v uniform (0,1)
  • mc = 0
55
Q

Bertrand with capacity constraints

A

Kreps and Scheinkman (1983)

- the unique SPE coincides with the equilibrium outcome associated with Cournot

56
Q

Definition of price discrimination and when feasible and why done

A

Sold at prices that are in different ratios to marginal costs (Varian 1989)

  • sort customers, no arbitrage, market power
  • to convert CS to PS
57
Q

Intuition for FandS condition (1)

A
  • when inc output, a merged firm has a lower MR at the pre-merger outputs as it internalises then effect of a fall in prices on both firms output therefore c must be correspondingly lower for output to be maintained
58
Q

Define incentive incompatibility with 2nd degree

A

If a consumer whom a given budge is directed, may choose a bundle directed to another consumer

59
Q

Example of predatory pricing

A

Darlington buses

  • 3 operators
  • over bussing and offering free services
  • bankruptcy of DTC
  • scale of Busway meant could absorb losses
60
Q

What impacts collusion

A
  • n
  • fluctuating demand
  • growing market
  • multimarket contact
  • asymmetric costs
  • transparency
  • common ownership

FHS (8 decks)

Brainscape helps you reach your goals faster, through stronger study habits.
© 2024 Bold Learning Solutions. Terms and Conditions