Econometric Analysis of The Whole Foods/Wild Oats Transaction by d9n1aQO


									     Unilateral Effects Case Study:
The Whole Foods/Wild Oats Merger

             Competition Commission of India
                          October 25-26, 2010
I. Introduction
Background to the Merger
• Whole Foods Market (WFM) and Wild Oats (WO) were national
  supermarket chains selling organic and natural foods

          Store Locations

           Wild Oats

           Whole Foods

• WO was in financial difficulty
• WFM sought to acquire WO     NUGGET

• FTC alleged that WFM and WO competed “head to head” in 21 geographic
  markets and claimed that the merger would create a monopoly in
  “premium natural and organic supermarkets” in nearly all of those

I. Introduction
Key Antitrust Issues
• Did WFM and WO constrain each others’ prices to such a degree that the
  merged firm would be able to profitably raise prices without fear of
  significant customer diversion to “conventional” supermarket chains?

• Central issues
    – Product market definition / direct evidence of effects
       - FTC alleged market was “premium natural and organic supermarkets”
       - In a broader product market, WFM and WO had very small shares (usually <5%)

    – Consumers’ willingness to switch to “conventional” supermarkets
       - FTC contended there were many “core customers” who would not switch
       - Parties contended that most consumers cross-shop and many would switch, and that price
         discrimination against remaining “core customers” was infeasible

    – Ease of repositioning, expansion and entry
       - FTC argued supply responses would be unlikely, untimely and insufficient
       - Parties pointed to many examples of repositioning and entry, and event study analysis

II. Product Market Definition
FTC’s Argument and Analysis
• Product market is “premium natural and organic supermarkets” (PNOS)
    – WFM, WO and a very few local niche competitors

• Qualitative evidence
    –   WFM and WO stores looked most alike among all competitive alternatives
    –   WFM tended to located its stores in close proximity to WO
    –   Internal documents reflecting historical rivalry
    –   Testimony from WFM CEO downplayed competition from “conventional” chains
    –   Testimony that merger would avoid “nasty price wars”

• Empirical evidence
    –   When a WFM store opened near a WO, it took substantial sales away from WO
    –   Generally supported by WFM’s internal analyses
    –   WFM margins generally higher than “conventional” supermarkets’ margins
    –   WFM margins tended to be lower in markets with a WO store present

• FTC did not attempt formal SSNIP test, favoring Staples approach
  primarily focused on margins analysis (and qualitative evidence)

II. Product Market Definition
Parties’ Argument and Analysis
• Product market includes multiple other competitors — conventional
  supermarkets, specialty food retailers, club stores, mass merchants

• Qualitative evidence
    –   Conventional chains carried many of the same products
    –   Internal business documents recognize competition from non-PNOS supermarkets
    –   WFM regularly price-checked conventional supermarkets
    –   WFM tracked entry and store remodeling by conventional supermarkets

• Empirical evidence
    – Price analyses showed uniform regional pricing by WFM and WO: prices did not
      seem to vary based on WO’s presence or absence
    – Event studies showed no evidence of significant price responses following WFM store
      openings and WO store closings

• Parties contrasted absence of direct price effects evidence with Staples
  case to argue market was much broader than PNOS

II. Product Market – Direct Effects Evidence
Event Study Analysis
• Natural experiments
    – Entries of WFM stores near existing WO stores and exits of WO stores near
      existing WFM stores
    – Opening a WFM store should have little effect on WO if WO’s pricing was
      already constrained by conventional supermarkets and great effect if it was
    – Closing a WO store should have had little effect on WFM if conventional
      supermarkets constrained WFM’s pricing and great effect if they did not

• Compare average weekly prices paid by shoppers for thousands
  of grocery items sold by WFM and WO before and after the entry
  and exit events
    – Compare item pricing at “treatment stores” (nearby to entries and exits)
      versus “control stores” (in same region, but not close to entries and exits)

• Compare before vs. after prices on an item-by-item basis

II. Product Market – Direct Effects Evidence
Price Change Comparison, Wild Oats Exits

             Source: Whole Foods Market

II. Product Market – Direct Effects Evidence
Price Change Comparison, Whole Foods Entry

             Source: Whole Foods Market

II. Product Market Definition
Discussion Issues
• How useful / informative is market definition in a highly
  differentiated product merger?

• Should direct evidence of competitive effects be viewed as a
  substitute or complement to formal market definition?

• What are the risks to defining the market unduly narrowly – e.g.,
  PNOS rather than “all supermarkets”?

• How should quantitative evidence (data) vs. qualitative evidence
  (documents, testimony) be weighed and reconciled when they
  appear to be at tension?

II. Product Market – Direct Effects Evidence
Further Readings

• Carlton Varner and Heather Cooper, “Product Markets in
  Merger Cases: The Whole Foods Decision,” Antitrust Source
  October 2007

• Jonathan B. Baker and David Reitman, “Research Topics in
  Unilateral Effects Analysis,”

• FTC Unilateral Effects Analysis and Litigation Workshop,
  February 12, 2008 transcript

III. Consumers’ Willingness to Switch
     FTC’s Argument and Analysis

• WFM and WO marketed themselves to a core clientele

• “Core customers” would not switch to non-PNOS
      – Products and services were highly differentiated

• Even if many consumers would switch, core customers
  would be subject to price discrimination
      – Higher prices?
      – Reduced shopping convenience?

III. Consumers’ Willingness to Switch
     Parties’ Argument and Analysis
• Parties contended that consumers cross–shop widely
      – WFM and WO documents
      – Consumer survey findings
      – Observations from the real world

• No identifiable mechanism to price discriminate against
  “core” customers
      – Conditions to support price discrimination were absent

• “Critical loss analysis”
      – Calculate minimum loss of sales needed to make a post-merger
        price increase unprofitable
      – Qualitative evidence on whether “enough” consumers would switch
        or divert their purchases to make the price increase unprofitable

III. Consumers’ Willingness to Switch
     Issues for Discussion
• Who defines the relevant market – the core customer or the
  marginal customer?
      – The marginal customer determines price, absent price
      – The marginal customer defines the boundaries of the market,
        absent price discrimination

• What needs to be proven to support price discrimination
      –   Identify targeted buyers with inelastic demand
      –   Charge higher prices to targeted customers
      –   Prevent re-sale (arbitrage) from low-price to high-price buyers
      –   Prevent customer diversion to substitute products

III. Consumers’ Willingness to Switch
     Further Reading

• Sumanth Addanki and Alan J. Daskin, “Who Defines the
  Relevant Market – The Core Customer or the Marginal One?,”
  NERA Antitrust Insights, Summer 2008

• Charles J. Biggio, “Whole Foods’ Impact on Unilateral Effects,”
  Global Competition Policy, September 2008

IV. Repositioning, Expansion and Entry
     FTC’s Argument and Analysis
• Barriers to supply-side responses
      – Reputational barriers
      – Conventionals would need to “abandon” their own “core” customers
      – WFM/WO “core” customers would not switch

• Likelihood, timeliness, sufficiency
      – Unlikely – see above
      – Not timely – reputational barriers?
      – Insufficient – past examples were limited and did not elicit significant
        responses from WFM or WO

• Elasticities should already reflect the effects of
  repositioning on the likelihood of a unilateral post-merger
  price increase

IV. Repositioning, Expansion and Entry
     Parties’ Argument and Analysis

• No barriers to supply-side responses
      – Competitors had strong motive and means to enter/expand
      – Uncommitted entry was feasible

• Multiple significant entries had taken place and were
      – Conventional stores’ private labels
      – Stores within a store
      – New store concepts and banners

IV. Repositioning, Expansion and Entry
Empirical Analysis

• Direct effects evidence also can come from analyzing other
  events – entries, expansions, repositioning by competitors
  alleged to be outside of the product

• Supply-side responses by non-market participants should
  not affect prices by firms within the alleged market

• Event studies
     – Conventional entries and remodelings affected WFM and WO prices
       ―oftentimes by more than WFM entries and WO exits

IV. Repositioning, Expansion and Entry
     Issues for Discussion

• How high are the evidentiary hurdles to assess likely, timely
  and sufficient supply responses?

• How probative are case examples of past entry?

• How should the possibility of post-merger repositioning by
  the merging firms be factored into the unilateral effects

IV. Repositioning, Expansion and Entry
     Further Reading

• Peter Boberg and John Woodbury, “Repositioning and the
  Revision of the Horizontal Merger Guidelines,” Antitrust Source
  December 2009

• Amit Gandhi, Luke Froeb, Steven Tschantz and Gregory J.
  Werden, “Post-Merger Product Repositioning,” Journal of
  industrial Economics 2008

V. Relevant Case Materials
Further Reading
• Complaint

• FTC’s economic expert reports

• District Court opinion

• Court of Appeals opinion

To top