Competition Law Risk From Pricing Algorithms

Understanding Competition Law and Pricing Algorithms

Competition law in the UK governs how businesses set prices and compete in the marketplace. The Competition and Markets Authority (CMA) enforces these rules to ensure fair trading practices. When retail organisations use pricing algorithms, they must understand that these automated systems can inadvertently create anti-competitive outcomes.

The fundamental principle is that businesses must not engage in practices that restrict competition or harm consumers.

Algorithms that set prices automatically can raise concerns when they lead to coordinated pricing behaviour between competitors. This happens when businesses use similar algorithms that result in similar pricing outcomes without direct communication. The CMA has investigated cases where companies used algorithmic pricing models that produced identical or very similar prices across different retailers.

Such outcomes can suggest anti-competitive coordination even when no explicit agreement exists.

Common Risk Scenarios in Algorithmic Pricing

• Price Fixing through Similar Algorithms – When multiple retailers use similar machine learning models to set prices, they may unintentionally align their pricing strategies. For example, if several clothing retailers use algorithms that respond similarly to stock levels or competitor pricing, their final prices might converge without any direct communication between businesses.
• Market Sharing through Price Discrimination – Algorithms that segment customers or markets may unintentionally create arrangements that prevent competition. A grocery chain using an algorithm that sets different prices for different customer groups might inadvertently create geographic or demographic price barriers that reduce competition.
• Collusion through Information Sharing – When algorithms share data or intelligence about pricing decisions, they might create opportunities for competitors to coordinate pricing strategies. Retailers using similar data sources or sharing pricing intelligence through industry platforms might develop similar pricing models that reduce competitive pressure.

Specific examples from retail demonstrate these risks. A supermarket chain using an algorithm that automatically matches competitor prices might unintentionally create a situation where multiple retailers set identical prices. This scenario occurred when several high street retailers adopted similar pricing models that responded to competitor data in the same way.

The result was widespread price alignment that appeared to reduce competitive pressure in the market.

Managing Compliance and Risk Controls

Organisations must implement specific controls to prevent algorithmic pricing from creating competition law issues. The first step involves conducting regular algorithmic impact assessments. These assessments should examine whether pricing algorithms might produce outcomes that reduce competition or harm consumers.

Retail businesses should review their pricing models at least quarterly to identify potential problems.

Specific measures include establishing clear governance frameworks for algorithm development and deployment. Companies must ensure that pricing algorithms do not simply copy competitor strategies or respond to similar market data in identical ways. The approach should focus on creating algorithms that respond to genuine market conditions rather than mimicking other businesses.

This involves testing algorithms against various scenarios to ensure they don’t produce similar outcomes across different organisations.

• Independent Testing – Regular testing of algorithms against competitor data to ensure they don’t produce identical or very similar pricing outcomes
• Monitoring Controls – Continuous surveillance of pricing decisions to identify unusual patterns or convergence
• Documentation Requirements – Maintaining detailed records of algorithm design, testing, and deployment decisions

Implementation of these controls requires investment in monitoring systems and staff training. Retail organisations must develop processes that allow them to detect when pricing algorithms might be creating anti-competitive outcomes. This includes establishing clear reporting procedures for unusual pricing patterns or when algorithms respond similarly to market conditions across different businesses.

The goal is to maintain competitive pressure in the marketplace while using automated pricing systems.

Training staff on these requirements is essential. Managers must understand that algorithmic pricing decisions have wider regulatory implications. The focus should be on creating pricing strategies that reflect genuine market conditions rather than simply matching competitors.

Regular updates to these controls ensure they remain effective against evolving algorithmic capabilities and changing market conditions.