Transcript The GE/Honeywell Case: To merge or not to merge?

The GE/Honeywell merger
and the issue of bundling
The situation
To Merge or not to Merge?
The European Commission
prohibited the US-merger
 2 arguments:

Market foreclosure through bundling of GE
and Honeywell products

Market foreclosure through the vertical
integration of Honeywell with GECAS, GE’s
leasing arm
1. The bundling issue
 What is bundling?


When complementary products are offered in
a package
Pbundle<  Pindividual items
 Why valuable for GE/Honeywell?

Integration of an external price effect:
Complements A and B:
When Pengine , QHoneywell
Qa
0
Pb
Could be beneficiary for the enlarged GE-group!
Commission’s Reasoning
 Post-merger, GE may offer package
discounts to customers who buy both GE
engines and Honeywell products
Prices would
Competitors’ profits would
Rivals will be unable to finance R&D
Forced exit of competitors
 Result:
Strengthening of GE’s dominant position
Prices would , consumer welfare would
Is this reasoning economically sound?
 Cournot model: internalising a pricing
externality

Two monopolists

Complementary goods: jet engines (GE) and
avionics (Honeywell)

q = 1 - (p1+ p2)

Coordinated pricing leads to higher profits and
to lower prices
Welfare increasing
Mathematics of the Cournot model
Uncoordinated
Coordinated
q = 1 - (p1+p2)
pb = p1+p2
П i = pi*q = pi – pi * (p1+p2)
so: q = 1 - pb
Maximize profit: ∂ П i/∂ pi = 0
Profit function of combined firms:
For i = 1: 1 - 2p1 - p2 = 0
П1+2 = pb*q = pb - pb²
For i = 2: 1 - 2p2 – p1 = 0
Maximize profit: ∂ П 1+2/∂ pb = 0
Substitution gives:
1 - 2pb = 0
p1 = p2 = 1/3
pb = ½
q = 1/3
q=½
П 1 = П 2 = 1/9
П1+2 = pb*q = ½ * ½ = ¼
Expansion of Cournot necessary
 Need for expansion of basic Cournot model:


GE and Honeywell are not monopolists
Dynamics: what are the possible reactions of
competitors?
 Solution:



Model of differentiated oligopoly
Fixed location of firms
Pure bundling and mixed bundling
Differentiated oligopoly
-fixed location of firms Four firms:
 Two differentiated versions produced of each good
 Engines by firms A1 and B1, avionics by A2 and B2
 A firms located at 0, B firms at 1
 Customers distributed uniformly with preference
(x1,x2), so distance to A firm is xi and to B firm (1- xi)
Differentiated oligopoly
 Cost for consumers:
 Linear transportation cost: C(distance) = t*d(x) and t=1
 Cost of component i is [xi + pAi] or [(1-xi) + pBi]
 Find the marginal consumer for each good:
 V- pAi- xi = V- pBi-(1-xi)
 xi(pAi, pBi) = ½ + ( pAi- pBi)/2
 Profit functions:
 П Ai = pAi * xi(pAi, pBi)
 П Bi = pBi * (1- xi(pAi, pBi))
 Solution is Nash equilibrium:
 pAi = pBi = t = 1
 П Ai = П Bi = t/2 = 1/2
Differentiated oligopoly
-pure bundling Case 1: each firm acts independently
 pA1 = pA2 = pB1 = pB2 = 1
 П A1 = П A2 = П B1 = П B2 = ½
 Market is evenly split, consumers match their
bundle and pay a price of 2
 Case 2: bundle vs bundle (price coordination)
 pA = pB = 1
 ПA = ПB = ½
 Market evenly split, price and profits cut in half
No incentive to bundle
Differentiated oligopoly
-pure bundling- (2)
 Case 3: bundle against components (A firms
coordinate their price)





pA = 1.45; pB1 = pB2 = 0.86; pB = 1.72
П A = 0.91; П Bi = 0.32
The A bundle is sold at a discount
The A firms gain market share but lose profit
Dynamics and reaction of competitors result in
lower profits for the merged firm
No incentive to bundle
Differentiated oligopoly
-mixed bundling Approximate equilibrium prices and profits:
 pA = 1.63; pAi = 1.21; pBi = 0.89; pB = 1.78
 П A = 0.97; П Bi = 0.40
 Profits of the A firm are not an incentive to
bundle
 Gain in market share (up to 55.4%) and fall of
the competitor’s profits (-21%) give an
advantage
 Potential to expand market: even if small, it
can make bundling profitable
There may be an incentive to bundle, but
this
will not automatically lead to
foreclosure
Bundling in the aircraft industry?
 Dynamics and possible reactions of
competitors make that firms do not always
have the incentive to bundle
 If firms decide to bundle, this will result in
lower prices and a raise in consumer welfare
 The economic theory does not support the
prediction that the GE/Honeywell merger will
be anticompetitive, because bundling will not
necessarily lead to foreclosure
2. The vertical integration issue
 Market foreclosure through the vertical integration of Honeywell
with GECAS
 Commission’s Concern:




GECAS has a “GE-only policy”
Fear that GECAS would extend this policy to Honeywell
products
GECAS would only buy GE/Honeywell-based airplanes for
their leasing activities.
Result:


Airframers (Boeing, Airbus,…) would face the irresistible
incentive to use GE/Hon-components for the aircraft they build
so it can be potentially purchased by GECAS.
Competing component suppliers squeezed out of the market
Critics on the Commission’s analysis
 Not based on equilibrium-analysis. What
about the reactions of other market players?


Other leasing companies would shift away
from GE/Hon products in order to differentiate
Competitors can respond by also affiliating
with a leasing company
 Honeywell-only policy is not credible
 Large quantity of avionics per airplane
 GECAS would lose flexibility
However, vertical integration of products and
financial services can lead to market foreclosure
Conclusion
 Bundling
 Will lead to lower prices and higher consumer welfare
on the short term
 But will it necessarily lead to market foreclosure?
 Vertical integration through GECAS
 Although Honeywell-only policy not thrustworthy, the
combination of financial services with products can be
a real threat for competitors
 Commission based its decision on theoretical
potential and not on waterproof economic models
 European competition law applied in a preventive
way