Economic analysis of retail mergers at the Competition Bureau

September 15, 2014

Renée M. Duplantis, PhDFootnote *

On this page

  1. Purpose
  2. Downstream retail analyses
  3. Upstream analyses
  4. Conclusion

1. Purpose

The Competition Bureau ("Bureau") has recently reviewed several retail mergers, including Canadian Tire/Forzani, Leon’s/The Brick, Agrium/Viterra, Safeway/Sobeys, La Coop Fédérée/Groupe BMR, Canadian Tire/Pro Hockey Life and Loblaw/Shoppers Drug MartFootnote 1. Factors relevant to an economic analysis of retail mergers can differ from case to case. These include the number of products offered by the merging parties (e.g., which can range from thousands of products offered for sale at a grocery store versus a more limited number of products offered for sale at a mattress store); whether the products offered are homogeneous or differentiated; and the extent of the parties’ geographic reach. Retail mergers are also unique in that the merging parties may have tens of thousands of direct customers and the merging parties may be competing on bundles rather than individual products. Like other mergers, retail mergers also can involve both downstream and upstream issues.

While every acquisition is unique in some respects, and while any Bureau investigation is tailored to the specific facts and evidence relating to that particular merger, the Bureau usually relies on a common core of economic tools when analyzing retail mergers. This paper discusses some of the economics tools and techniques the Bureau has used in its past investigations of retail mergers to provide additional transparency to stakeholders and the general public about these investigations. It is not a step‑by‑step guide to how the Bureau conducts merger reviews. It does not replace any other Bureau publications and should be read in conjunction with the Merger Enforcement Guidelines (the "MEGs")Footnote 2. The paper also does not provide an exhaustive list of the analyses the Bureau currently conducts or may conduct in the future as the field of economics continues to change: instead it outlines those tools the Bureau has employed or considered in recent retail merger reviews.

The first section of this paper discusses the Bureau’s approach to analyzing the downstream (retailer to consumer) portion of a retail merger, while the second section focuses on the analysis of the upstream (supplier to retailer) portion of a retail merger.

2. Downstream retail analyses

The Bureau analyzes the downstream and upstream markets separately. This section discusses the downstream market.

2.1. Market definition

There are two dimensions to market definition: product market and geographic market. Both are important. As noted in the MEGs, the Bureau will typically define relevant product and geographic markets, but that is not a required step, nor is it necessarily the first step in the analysisFootnote 3.

In determining the relevant product market, the Bureau assesses the degree to which the products produced by the merging parties are substitutable in order to determine whether a hypothetical monopolist over those products could profitably raise prices by at least a small but significant and non‑transitory amount over some minimum time period. The relevant product market could comprise only one product or, particularly in retail mergers, those stores which sell a basket of products (e.g., groceries) purchased in a "one‑stop" shopping experienceFootnote 4. In determining the relevant product market in retail mergers, the Bureau will consider qualitative evidence of substitutability across stores, including any consumer switching studies that the parties may have commissioned during the normal course of business.

In retail mergers, the relevant geographic markets are typically local in natureFootnote 5. As with the product market, the relevant question is whether a hypothetical monopolist in that local area is able to profitably raise prices by a small but significant non‑transitory amount over some minimum time period. In recent retail merger investigations by the Bureau, the relevant geographic markets have been defined to be as small as a one‑kilometre‑radius circle around a retail location and as large as a metropolitan area.

To properly define the relevant geographic market, the Bureau will consider the documents provided by the merging parties and third parties, the views of market contacts, and empirical evidence. With respect to the latter, there are several techniques the Bureau can employ when defining the relevant geographic market.

First, the Bureau can use maps and trade or draw areas to delineate the boundaries of the geographic market. This typically entails determining how far customers are willing to travel (either from home or work) to shop at the retail location. This information may come from maps of trade‑draw areas (or catchment areas) maintained by the retailer, address information for loyalty‑card customers, postal‑code information captured at the point‑of‑sale, customer surveys conducted by the retailer, or other normal‑course documents. The Bureau can use this information to determine the customer distribution around a given store, normalized for population, as the basis for identifying the relevant geographic market.

Other techniques that can be used for both product and geographic market determinations include diversion ratio analyses, critical loss analyses, price correlation/cointegration analyses and regression analysesFootnote 6. Diversion ratio analysis of retail mergers often examines customer switching patterns to determine to which stores customers would switch when faced with an increase in price. In the absence of switching studies, the Bureau can calculate diversion ratios from market shares when logit demand is a reasonable assumptionFootnote 7. Critical loss analyses involve measuring the minimum amount of sales a hypothetical monopolist would need to lose to make a 5‑10% price increase unprofitableFootnote 8. Price correlation and cointegration analyses measure the extent to which prices of different products move together as an indication that the two products are substitutesFootnote 9. Finally, regression analyses (which are discussed in more detail below) can also be used to quantitatively determine the relevant geographic or product markets.

2.2. Market shares and concentration

Once the relevant product and geographic markets are defined, the Bureau can undertake a concentration analysis of the merger, which entails looking at the market shares of the merging parties in the relevant product and geographic markets. As stated in the MEGs, "[t]he Commissioner generally will not challenge a merger on the basis of a concern related to the unilateral exercise of market power when the post‑merger market share of the merged firm would be less than 35 percentFootnote 10." Mergers with post‑merger market shares (usually based on revenues) of greater than 35% are not presumed to be anticompetitive, but they will usually require a more detailed analysis of the market, including, but not limited to, assessing barriers to entry and/or expansion, as well as determining the effectiveness of remaining competition.

2.3. Competitive effects analyses

2.3.1. Unilateral effects

The Bureau may undertake a unilateral competitive effects analysis when reviewing retail mergers. Unilateral competitive effects refer to the incentive that a merged firm has to raise prices irrespective of the reactions of competing firms. Competitive effects analysis will generally focus on assessing the price effects expected from a mergerFootnote 11, but the Bureau will also consider non‑price effects such as impacts on innovation, service, quality, product variety, store formats or hours of operation.

Unilateral competitive effects analyses include empirical examination of natural experiments (which may involve cross‑sectional or time‑series regression analyses), upward pricing pressure ("UPP") analyses, and more complex analyses, such as merger simulation models.

UPP models rely on estimates of diversion ratios and gross margins to derive an indication of significant potential price effects. To date, the Bureau has used general UPP models in order to provide an initial screen for potential unilateral competitive effects rather than as a quantification toolFootnote 12.

Natural experiments, such as differences in market structures across geographic markets or changes in market structure over time, can provide an important predictor of the effect of a merger. For example, some local areas may have competition occurring between the merging parties, while other local areas may not. Using regression analysis when the appropriate data are available, the Bureau can exploit these natural differences by comparing prices, discounts or margins in those areas where the parties compete to those areas where the parties do not compete to quantify the potential impact the parties have on each other. The results of this type of cross‑sectional regression analysis can then be used to predict the potential price effect from this mergerFootnote 13.

In some cases, however, a cross‑sectional analysis may not be possible, as, for example, when all firms in the market set prices on a nation‑wide basis. Of course, even with national pricing, there could still be price effects due to local competition, especially if local or regional managers have some degree of discretion over discounts and promotions, or are allowed to match competitors’ prices. In those cases, the Bureau could empirically assess how competition impacts local promotions or manager discounts through the use of cross‑sectional or time‑series regression analysis.

The Bureau also reviews natural experiments that may have occurred over time, such as mergers or prior entry/exit from this or similar marketsFootnote 14. When data are available, the Bureau uses time‑series regression analysis to estimate the impact these events may have had on prices, discounts or margins and then uses those estimates to predict the potential price effects from this merger.

Finally, in recent investigations, the Bureau has used merger simulation models to predict the price effect of the merger. These types of analyses can require the estimation of both own and cross‑price elasticities, as well as assumptions on the functional forms of both the demand and supply functions. The results of merger simulation models can vary depending on how these functional forms are specified and thus, correct specification can be criticalFootnote 15. The Bureau may also implement simplified merger simulation models that rely on estimates of diversion ratios and margins, along with assumptions to the form of the demand curve, to generate estimated price effectsFootnote 16.

2.3.2. Coordinated effects

In addition to unilateral effects, the Bureau will also consider whether the merger makes coordination at the downstream retail level more likely or more effectiveFootnote 17. As discussed in more detail in the MEGs, the Bureau will consider certain indicia that coordination (either explicit or tacit) would become more likely post transaction. These indicia include

  1. recognition of mutual benefits to coordination;
  2. the ability to monitor conduct and detect deviations;
  3. the ability to respond to deviations by using deterrent mechanisms; and
  4. external factors that may limit coordinationFootnote 18. The analysis of coordinated effects is typically conducted qualitatively, by reference to the firm’s documents as well as to information received from market contacts.

As with any merger review, the particular fact situation at hand, as well as the evidence and data available to the Bureau, will dictate which of the analyses discussed above will be employed by the Bureau when reviewing the downstream aspects of retail mergers.

3. Upstream analyses

The analysis of the upstream market involves how a merger will affect the sale of the product by the supplier to the retailer. Given all of the possible market structure combinations and distribution arrangements, it is difficult to provide concrete examples of the analyses undertaken to assess whether a transaction could lead to a substantial lessening or prevention of competition upstream. It is possible, however, to highlight some of the factors the Bureau will take into consideration when assessing these transactions.

3.1. Horizontal unilateral effects

The Bureau will examine any horizontal overlap between the parties in the upstream market and assess whether the merged entity will have the incentive and ability to raise prices unilaterally after closing the transaction. This assessment follows the standard unilateral effects theory in the MEGs and focuses on the extent to which products supplied by the merging parties are close substitutesFootnote 19.

3.2. Vertical effects

The Bureau will assess the likelihood that the merger will result in anticompetitive effects stemming from any vertical aspects of the transaction. Potential concerns include the effect of the merger on the incentive and ability of the merging parties to foreclose a downstream competitor from access to an input, either through full input foreclosure (refusing to supply a downstream competitor) or partial input foreclosure (raising the input costs of a downstream competitor). The Bureau may also be concerned about potential upstream foreclosure through full customer foreclosure (refusing to purchase inputs from a rival supplier) or partial customer foreclosure (disadvantaging upstream rivals by restricting their distribution options), as well as the potential for increased coordinationFootnote 20. In particular, the Bureau will assess contracts between upstream suppliers and retailers to determine whether increased concentration could result in potential anticompetitive effectsFootnote 21.

3.3. Monopsony power

The Bureau will assess whether the transaction could be expected to create, maintain or enhance monopsony power. This assessment follows the standard analytical approach the Bureau takes to merger review, albeit applied to the buying, as opposed to the selling, side of the marketFootnote 22.

The Bureau will also assess whether the transaction could have other impacts on input prices or otherwise distort downstream competition. For example, while buyers that gain bargaining power may benefit from lower supply prices, other buyers may face higher supply prices as a result. This is sometimes referred to as the waterbed effectFootnote 23.

4. Conclusion

The facts and particular circumstances surrounding a merger investigation will dictate the types of economic analyses the Bureau will undertake to assess the likely competitive effects of retail mergers. To provide increased transparency to our stakeholders, this paper highlights several of the economics techniques that the Bureau has used in recent investigations of retail mergers.

Date modified: