Ex-post assessment of price squeezes: approaches and key issues

Rodrigo Barreto, Principal Consultant

Price (margin) squeeze tests are used to limit the practice by dominant operators of predatory pricing and excessive pricing. In this article we look at the approaches and key issues found in ex-post assessment of price squeezes.

What is a price squeeze?

A price or margin squeeze occurs when a vertically integrated operator leverages its market power in the wholesale market (upstream production) to drive out competitors in the retail market (downstream transformation).

Figure 1 Typical setup for a price or margin squeeze

Source: Ovum 

Competitive issues due to price or margin squeezes are best known for the high-profile cases, often involving substantial fines, that have been widely reported in the media. Examples of these cases include:

• July 2007 – the EC fined Telefonica for unfair prices in the Spanish market for broadband access (€151 million).
  
  
• February 2005 – the ACCC revoked a competition notice issued to Telstra for anti-competitive conduct relating to its wholesale pricing of high-speed Internet after Telstra lowered its wholesale prices (rebate of A$6.5 million to affected wholesale customers).
  
  
• November 2003 – Oftel (now Ofcom) decided that BT had not infringed the Competition Act and that the margin between BT’s residential broadband wholesale and retail prices was sufficient to allow a reasonably efficient operator to compete, by reference to BT’s own costs.
  
  
• July 2003 – the EC fined France Telecom for predatory pricing for ADSL-based Internet access services (€10 million).
  
  
• May 2003 – the EC fined Deutsch Telekom for a margin squeeze in the German narrowband and broadband markets (€12 million).
  
  

Predatory pricing at retail level occurs when the dominant operator sets its retail prices at such a low level that it is not possible for other alternative operators (OAOs) to continue competing in a profitable way.

Excessive pricing at wholesale level occurs when the dominant operator sets its wholesale prices at such a high level that OAOs, which use these services as essential inputs to provide services at retail level, are not able to gain enough margin to sustain their business. In addition, the dominant operator may charge different wholesale prices to its downstream operation and to OAOs, in this case acting in a discriminatory way.

Figure 2 Types of price squeeze

Source: Ovum 

The assessment of price squeezes has different particularities and is used with different objectives depending on whether it is done ex-ante or ex-post.

Common challenges for ex-post assessment

Ex-post analysis of price squeeze allegations is usually undertaken by national competition authorities. The main goal of the analysis is to identify whether the alleged price squeeze has exclusionary effects (i.e. if competitors exit the market as a result).

From an ex-post perspective, the types of price squeeze are classified as:

• discriminatory price squeeze: where a dominant operator charges its retail rivals a higher wholesale price for essential inputs than it charges its own downstream operation
  
  
• non-discriminatory price squeeze: where a dominant operator raises the price of essential inputs across the board, i.e. both to OAOs and to its own downstream operation
  
  
• predatory price squeeze: when a dominant operator prices its retail services at a level that leaves insufficient (or even negative) margins for OAOs to continue in the market.
  
  

The analysis of an alleged price squeeze under competition law involves the following:

1. Assessment of dominance in the wholesale market. The dominant operator has significant market power (SMP) in the wholesale markets for services required to produce the retail service where the alleged price squeeze occurs.

2. Evaluation of the level of competition at the retail level. There are high barriers to entry or exit in the retail market and the dominant operator leverages these barriers to maximise its profits in the long term.

3. Characterisation of vertical integration. Dominance in the retail market is not a necessary condition, but it should be demonstrated that the operator also has an operation in the retail market that will benefit from the exit of OAOs in this market.

4. Substitutability analysis. There are no substitutes for the inputs provided by the dominant operator. There are no close substitutes to the retail service using a different set of wholesale inputs.

5. Imputation test. Margins for OAOs are insufficient. This should be demonstrated for two cases: when the OAOs are as efficient as the retail division of the vertically integrated operator (equally efficient operator – EEO) and when the OAOs are more efficient (reasonably efficient operator – REO).

6. Analysis of duration. The alleged price squeeze has had sufficiently long duration to have an exclusionary effect.

Competition authorities are particularly interested in understanding the long-term effects on competition and consumer welfare, and will be very careful to characterise a case as a price squeeze before being sure that the above conditions are all met.

The imputation test

The imputation test is a key stage in the evaluation of a price squeeze, and due attention should be given to assumptions and sources of input for these tests.

Data inconsistencies may lead to error and ultimately to loss of efficiency because of negative impacts on the ability of a vertically integrated operator to exploit economies of scope and scale (for type 1 error, i.e. finding a price squeeze where there is none).

The logic behind imputation tests is simple; the net margin at retail level should be at least equal to an appropriate rate of return.

Net margin

Net margin (NM) is calculated as follows:

NM = price at retail level – (price of wholesale inputs + retail costs)

‘Price at retail level’ and ‘price of wholesale inputs’ are those charged by the dominant operator. The appropriate measure of costs is that of avoidable or incremental costs – whichever is lower. Avoidable costs refer to those costs that the vertically integrated operator could avoid if it decided to close its downstream operation but continue to provide the related wholesale inputs to OAOs. Incremental costs refer to the costs that an OAO would have to incur if it were to efficiently operate in or enter the retail market.

Some of the complexities associated with the net margin calculation are:

• Retail and wholesale tariffs have several components, including setup fees, fixed charges, usage-dependent charges (with variances depending on time and/or distance) and event-dependent charges. A price index must be identified.
  
  
• The service may use several essential wholesale inputs (some shared with other retail services). A clear methodology for apportioning of costs must be adopted.
  
  
• The service may be provided as part of a bundle. The methodology used should consider replicability tests at both individual and bundle levels.
  
  
• It is hard to determine historic retail costs of the vertically integrated operator.
  
  
• Identification of a REO may be contentious.
  
  
• To calculate the retail costs of a hypothetical efficient OAO it is necessary to develop a bottom-up model.
  
  
• A decision must be made on whether to use static (period by period) or dynamic (based on discounted cash flow valuation) assessment of price squeeze.
  
  

Appropriate rate of return

Finding an appropriate rate of return may also be difficult. The initial challenge is to define which rate of return is preferable:

• Return on turnover has been used in a number of asset-light retail markets (e.g. directory services).
  
  
• Weighted average cost of capital (WACC) is widely used by regulators and has, for instance, been used by DG COMP in the Telefonica case.
  
  

The appropriate rate of return might differ for the EEO and REO test cases. For example, the retail division of a vertically integrated firm might have a lower cost of capital because integration reduces some types of business/financial risks. Therefore, the appropriate rate of return depends on the retail costs used:

• In the EEO case, the rate of return should be that of the retail division of the vertically integrated operator.
  
  
• For the REO test case, the rate of return should either be an average of the rates of return of similar OAOs or the rate of return that a hypothetical OAO would require to enter in that specific retail market.
  
  

In either case, there are difficulties to identify the right rate of return:

• The WACC for the vertically integrated operator may only be available for the company as a whole and it may be difficult to identify a standalone WACC for the retail division only. The DG COMP, for instance, although acknowledging it to be an approach, has considered the WACC for the dominant operator (and not for the individual retail division) when evaluating margin squeeze cases.
  
  
• Trying to benchmark rates of return of OAOs may prove difficult as many of these providers are privately held and do not disclose disaggregated financial information. Alternatively, the rate of return may be obtained by adjusting upwards the WACC for the retail division of the dominant operator, as OAOs usually bear more financial and operational risks.
  
  

Rodrigo is a Principal Consultant in Ovum’s Regulatory and Policy Practice.