Price pressure analysis and merger simulation: a UK retrospectiveChris WaltersChief Economist, OFT

LEAR Conference: The Economics of Merger Control, 27 June 2013

OverviewUK merger

control and UPP, GUPPI, IPR, PPI

Market definition is

dead!

Price pressure analysis is too

hard!

No wait, it’s too easy!

Either way, it’s too

interventionist!

What’s happened to

merger simulation?

BackgroundUK merger control● Is voluntary, not mandatory● Is administrative, not judicial● Is bicameral, not unitary (for

the time being!)● Is well resourced (relatively

speaking)● Lacks compulsory information

gathering powers at first phase

Alphabet soup: ingredients

Market definition is dead!

08/09 09/10 10/11 11/12 12/135 2 4 4 7

24 20 17 26 2551 50 52 70 68

Price pressure analysis

Other CRM de-cisions

Non CRM de-cisions

The great majority ofOFT first-phase mergerdecisions still rely onthe ‘traditional’ approach

Clearances Decision

Lovefilm/Amazon 15/04/08Tesco/Brian Ford 22/12/08Morrison’s/CGL 10/07/09Sainsbury’s/CGL 09/11/09Carpetright/Allied Carpets

13/09/10

Asda/CGL 11/01/11Amazon/Book Depository

26/10/11

Sainsbury’s/Rontec 07/06/12

Conditional clearances Decision

CGL/Somerfield 22/10/08CGL/LBA 06/03/09Asda/Netto 23/09/10Unilever/Alberto Culver 18/03/11Prince’s/Premier Foods 22/06/11Shell/Rontec 03/02/12Jewson/Build Centre 08/02/12Edmundson/Electric Centre 11/05/12Mid Counties Co-op/Harry Tuffin’s

18/10/12

Rexel/Wilts 26/10/12References Decisio

nHolland & Barrett/Julian Graves

20/03/09

Booker/Makro 08/11/12Barr/Britvic 13/02/13Cineworld/City Screen 30/04/13

Lots of retail mergers

It’s all too hard!Myth● The analysis is too new● It’s too complicated● It always uses consumer

surveys and they’re too expensive

Reality● Used by OFT regularly since

2005● The ingredients for alphabet

soup are the same as for market definition. But when products are differentiated, markets can look too narrow and price pressure analysis is more sensible. So the focus on margins and diversion is not a consequence of using price pressure analysis, it is the cause of it

● It uses evidence from internal documents, supply disruptions, market research and econometrics as well as consumer surveys

No wait, it’s all too easy!Myth● The

ingredients may be poorly measured

● The analysis requires lots of hidden and unrealistic assumptions

In defence of assumptions

● This is no more true of alphabet soup than of other evidence

It’s too interventionist!Myth● The analysis is determinative● By comparison to the

traditional market definition and market share approach, it leads to more intervention

Reality● Analysis is seen in the light of

other evidence: “...the evidence on diversion and the GUPPI calculations corroborate this view.”

● In the UK, it is the parties’ commercial decision to offer divestments at first phase

● Of the 22 ‘price pressure’ cases above, only 4 were referred—same as OFT’s overall reference rate, 18%

● In the 10 conditional clearances, the analysis has been exculpatory to the traditional approach in 8 cases and corroborative of it in 2

Why has merger simulation not delivered on its promise?

Aggregation aggravation• No-one paid 235.4126706p for this steak &

kidney pie!• Are retailers’ and consumers’ preferences the

same?

If retailers and shoppers think that past prices matter (‘was/now’ reference pricing), then why aren’t they in our econometric models?

Consumer behaviour in the real world• 86% of shoppers don’t use a shopping list• 40% of purchases made are on promotion• Shoppers prefer bulk buy deals (BOGOF, 3-for-2)

to explicit price discounts

Thank you. Any questions?

Chris WaltersChief Economist, OFT

LEAR Conference: The Economics of Merger Control, 27 June 2013