Competition Law and Strategic Forecasting Advantages and Antitrust

1. Introduction

Strategic forecasting refers to the use of data, algorithms, artificial intelligence, market intelligence, predictive analytics, demand forecasts, pricing models, capacity forecasts, customer-behaviour predictions, and other forward-looking information to anticipate market conditions and competitor behaviour.

Forecasting itself is generally pro-competitive. A business may legitimately forecast:

• consumer demand;
• inventory requirements;
• production capacity;
• commodity prices;
• logistics requirements;
• future costs;
• technological developments;
• customer churn;
• investment needs; and
• market growth.

The competition-law concern arises when a firm obtains or uses strategically sensitive forward-looking information about competitors, or when forecasting systems facilitate coordination, exclusion, information exchange, discriminatory access, or algorithmic collusion.

The central antitrust question is therefore not simply:

"Is forecasting being used?"

but rather:

"Does the forecasting advantage improve independent competition, or does it reduce uncertainty between competitors and thereby facilitate coordination or exclusion?"

2. Meaning of Strategic Forecasting Advantage

A strategic forecasting advantage exists where an undertaking possesses substantially better information or predictive capability concerning future market conditions than its competitors.

For example, a dominant digital platform may possess:

• millions of real-time transactions;
• historical customer behaviour;
• competitor-price information;
• search queries;
• inventory information;
• supplier data;
• advertising conversion data;
• logistics data; and
• machine-learning models predicting future demand.

This can create several layers of competitive advantage.

A. Information advantage

The undertaking knows more about the market than competitors.

B. Predictive advantage

The undertaking can predict future demand, prices or consumer behaviour more accurately.

C. Timing advantage

The undertaking can react to market changes before competitors.

D. Strategic-response advantage

The undertaking can anticipate how competitors are likely to respond.

E. Algorithmic advantage

An AI or machine-learning system may continuously transform large quantities of data into commercially useful forecasts.

F. Ecosystem advantage

A platform controlling several complementary markets can combine information from one market to forecast developments in another.

3. When Does Forecasting Become an Antitrust Problem?

Strategic forecasting can raise competition concerns through several mechanisms.

4. Strategic Forecasting and Article 101 / Cartel Law

Under EU-style competition law, the principal concern is Article 101 TFEU, particularly agreements or concerted practices involving the exchange of competitively sensitive information.

Future-oriented information is particularly problematic.

There is an important difference between:

Historical information

For example:

"Our sales last year were €50 million."

and

Strategic forward-looking information

For example:

"We intend to increase prices by 15% next quarter."

The second category can materially reduce uncertainty about future competitive behaviour.

Information concerning:

• future prices;
• planned output;
• future capacity;
• investment;
• discounts;
• strategic commercial plans;
• product launches; and
• market expansion

may therefore present significantly greater competition risks.

5. Strategic Forecasting and Article 102 / Abuse of Dominance

Forecasting advantages can also become relevant to abuse of dominance.

A dominant undertaking might possess an exceptionally valuable predictive-data infrastructure and then:

1. deny competitors access to essential data;
2. provide inferior data to rivals;
3. use competitors' commercially sensitive data against them;
4. favour its own downstream operations;
5. tie access to forecasting services to other products;
6. use predictive information to discriminate between customers; or
7. leverage the forecasting advantage into adjacent markets.

The legal analysis depends heavily upon market structure, dominance, foreclosure effects and the particular conduct involved.

6. Six Important Case Laws

1. United States v. Airline Tariff Publishing Co. — U.S. Airline Fare Information

Background

The U.S. Department of Justice challenged practices involving the airline industry's computerized fare-information system.

Airlines could communicate fare information through systems that allowed competitors to observe proposed or future fare changes.

Competition issue

The concern was that sophisticated publication of future pricing information could allow competitors to:

• observe intended price changes;
• communicate indirectly;
• coordinate responses;
• test competitor reactions; and
• reduce uncertainty concerning future pricing.

Legal significance

The case is one of the classic examples of the competition problem created by public or semi-public disclosure of future strategic information.

The important lesson for modern forecasting systems is that a technological system does not become competitively harmless merely because competitors communicate through an intermediary or information platform.

Relevance to strategic forecasting

Modern AI systems can perform a function similar to an extremely sophisticated information exchange:

Competitor data → forecasting system → prediction of competitor conduct → strategic response.

Thus, competition authorities may examine whether the forecasting infrastructure facilitates coordinated conduct.

7. T-Mobile Netherlands BV v Raad van bestuur van de Nederlandse Mededingingsautoriteit

Facts

The case concerned contacts and information exchanges among mobile telecommunications operators.

The European Court of Justice considered whether information exchanges between competitors could constitute a concerted practice.

Principle

The Court emphasised that competition law protects the independent determination of market conduct.

Where competitors exchange information capable of reducing uncertainty concerning their future competitive behaviour, the exchange may create a concerted practice.

Strategic forecasting relevance

Forecasting systems can amplify precisely this problem.

Suppose competing telecommunications companies provide information concerning:

• expected price increases;
• anticipated capacity;
• future investment;
• subscriber forecasts; and
• network deployment.

An AI system could transform this information into extremely accurate predictions of competitor behaviour.

Key lesson

Forecasting does not eliminate the information-exchange problem; it can intensify it.

8. Eturas UAB and Others v Lietuvos Respublikos konkurencijos taryba

Court

Court of Justice of the European Union.

Facts

The case involved an online travel-booking platform used by travel agencies. A technical message communicated through the platform concerned a limitation on discounts.

The technological system therefore became relevant to the coordination of commercial behaviour.

Legal significance

The Court examined the circumstances in which businesses using a common electronic platform could be regarded as participating in a concerted practice.

Strategic forecasting relevance

The case is particularly valuable for modern digital-market analysis because it demonstrates that:

A software platform can become the mechanism through which competitively significant conduct is communicated or coordinated.

A modern forecasting platform could similarly communicate:

• recommended prices;
• expected competitor responses;
• discount levels;
• capacity forecasts; or
• market predictions.

The competition-law issue would depend upon the undertaking's knowledge, participation, communications and surrounding circumstances.

Principle

Technology is not legally neutral merely because the coordination is implemented through software rather than a traditional human meeting.

9. AC-Treuhand AG v European Commission

Facts

AC-Treuhand concerned a consultancy/advisory undertaking that assisted cartel participants.

Legal principle

EU competition law can extend beyond the companies directly producing or selling the relevant products where an undertaking intentionally contributes to an anticompetitive arrangement.

Strategic forecasting relevance

This principle becomes important for:

• AI providers;
• data intermediaries;
• forecasting vendors;
• industry-information platforms;
• pricing-software providers; and
• market-intelligence companies.

A company supplying forecasting infrastructure is not automatically liable merely because its software is used by competitors.

However, the case illustrates the broader proposition that active assistance to an anticompetitive arrangement can itself attract competition-law scrutiny.

Important distinction

There is a substantial difference between:

selling neutral forecasting software,

and:

knowingly designing or operating a system intended to facilitate coordination among competitors.

10. Google Shopping — European Commission / General Court

Background

The European Commission found Google dominant in general search and concluded that Google had treated its own comparison-shopping service more favourably than competing comparison-shopping services.

The EU courts subsequently examined important aspects of the Commission's reasoning.

Strategic forecasting relevance

The case is important for understanding the relationship between:

• data;
• digital platforms;
• vertical integration;
• information advantages; and
• self-preferencing.

A vertically integrated platform may possess enormous quantities of information about:

• consumer searches;
• clicks;
• purchasing patterns;
• merchant performance;
• demand trends; and
• competitor behaviour.

It may then use those insights to improve its own downstream service.

Competition concern

The relevant issue is not simply that the dominant firm has better forecasting capabilities.

The concern can arise where the dominant platform uses its control over an important infrastructure or information channel to favour its own downstream business and disadvantage competing services.

Principle

Superior predictive capability may be legitimate competition, but leveraging control over an important platform to disadvantage competing businesses can raise Article 102 concerns.

11. United States v. RealPage, Inc. — Algorithmic Rental Pricing

Background

The RealPage litigation concerns the use of algorithmic pricing systems in the rental-housing sector.

The controversy involves the use of software and data to recommend rental prices to landlords.

Competition significance

The case illustrates a modern antitrust concern:

Can a common algorithm or pricing-information system facilitate coordination among otherwise independent competitors?

Traditional cartel theory often imagines competitors communicating directly.

Algorithmic markets create another possibility:

Competitors → common data/software → algorithmic recommendation → similar pricing behaviour.

Strategic forecasting relevance

A sophisticated pricing algorithm can forecast:

• demand;
• occupancy;
• competitor pricing;
• expected vacancies;
• consumer willingness to pay; and
• market responses.

If competing businesses rely on the same system and the system incorporates information about competitors, the resulting market dynamics can raise significant antitrust questions.

Important qualification

The mere use of an algorithm by several competitors does not automatically establish an antitrust violation. The precise facts, communications, data inputs, contractual arrangements, algorithmic design and effects matter.

12. United States v. Apple Inc. — Digital Ecosystem Information Advantages

The U.S. government's antitrust litigation against Apple concerns alleged exclusionary conduct involving Apple's control over its ecosystem.

Although the litigation is not simply a "forecasting case," it illustrates a broader modern competition problem concerning strategic information and ecosystem control.

A major digital ecosystem can obtain information regarding:

• user behaviour;
• app activity;
• transactions;
• payment behaviour;
• device usage;
• developer activity; and
• emerging competitive threats.

The resulting information advantage can improve the platform's ability to anticipate competitive developments.

Competition-law relevance

Where a dominant ecosystem controls an important technological layer, competition authorities may examine whether that control is used to:

• restrict rivals;
• disadvantage competing services;
• limit interoperability;
• raise rivals' costs; or
• protect adjacent markets.

The case therefore provides a useful framework for analysing forecasting advantages as part of broader ecosystem power.

13. Forecasting Advantage and Algorithmic Collusion

One of the most important emerging issues is algorithmic collusion.

Traditional cartel:

Firm A communicates with Firm B → agreement → prices increase.

Algorithmic environment:

Firm A + Firm B → common data/algorithm → continuous prediction → automated responses → reduced competitive uncertainty.

The second model can be substantially harder to detect.

Example

Suppose four airlines independently use a common forecasting platform.

The platform predicts:

• Airline A's likely fare;
• Airline B's capacity;
• Airline C's expected discount;
• Airline D's expected response.

Each airline then receives recommendations based partly upon the others' predicted behaviour.

Even without a conventional meeting, the system could potentially reduce strategic uncertainty.

The competition-law analysis would depend on whether there is:

• communication;
• knowledge;
• agreement or concerted practice;
• facilitation;
• algorithm design;
• exchange of competitively sensitive information; or
• exclusionary conduct.

14. Forecasting and Tacit Coordination

Forecasting systems can change the economics of tacit coordination.

Normally, competitors face uncertainty:

"If I raise my price, what will my rival do?"

A predictive system attempts to answer precisely that question.

If forecasting becomes sufficiently accurate, competitors may be able to anticipate one another's behaviour.

This can potentially make coordinated outcomes more stable.

However

Tacit coordination is not automatically unlawful merely because firms independently observe and respond to market conditions.

Competition law generally requires a legally cognisable form of coordination or exclusionary conduct, depending upon the jurisdiction and applicable legal provision.

15. Forecasting and Information Exchange

A useful classification is:

Low-risk information

• old historical sales;
• publicly available macroeconomic data;
• general market statistics;
• industry-wide government statistics.

Intermediate-risk information

• recent sales data;
• aggregated market forecasts;
• anonymised industry information.

Higher-risk information

• future prices;
• planned discounts;
• planned output;
• future capacity;
• investment plans;
• customer-specific strategies;
• product-launch plans.

Particularly sensitive information

Information that allows a competitor to determine:

what another competitor intends to do before the market discovers it independently.

That information can have substantial competition significance.

16. Forecasting and Data Advantage

A dominant undertaking can develop a forecasting advantage through a data feedback loop:

More users

↓

More data

↓

Better forecasting

↓

Better product/service

↓

More users

↓

Even more data

This can create a self-reinforcing competitive advantage.

The advantage itself is not necessarily unlawful.

Competition concerns arise where the undertaking uses exclusionary conduct to prevent competitors from developing competing capabilities.

17. Forecasting and Essential Facilities

Suppose a market depends upon a particular forecasting dataset.

For example:

• a national transport-data system;
• electricity-demand forecasting data;
• airport capacity information;
• financial-market data;
• weather data;
• navigation data;
• healthcare demand data.

If a dominant undertaking controls the indispensable dataset and refuses access, competition authorities may examine the matter under theories involving:

• refusal to deal;
• essential facilities;
• discriminatory access;
• leveraging;
• interoperability;
• data foreclosure.

The precise legal test differs considerably between jurisdictions.

18. Forecasting as a Barrier to Entry

A new entrant may possess technology comparable to incumbents but lack comparable historical data.

For example:

The entrant may therefore face a data-driven forecasting disadvantage.

This becomes a competition-law concern particularly where the incumbent has acquired or protected the data through exclusionary practices.

19. Forecasting and Predatory Conduct

Forecasting can also be relevant to predatory pricing.

A dominant undertaking might possess superior information concerning:

• competitor cash reserves;
• demand elasticity;
• competitor costs;
• likely market exit;
• future demand;
• competitor financing.

It might then use this information to determine how long it can sustain losses.

The relevant antitrust question would be whether the conduct satisfies the jurisdiction's legal requirements for predatory pricing or another exclusionary theory.

20. Forecasting in Mergers

Strategic forecasting is also important in merger control.

A merger may combine:

• two major datasets;
• two demand-prediction systems;
• competing AI models;
• logistics forecasts;
• consumer behavioural data;
• financial-market predictions.

The authority may consider whether the combined undertaking would acquire a strategically important information advantage.

Relevant theories can include:

Horizontal effects

Two competing forecasting providers merge.

Vertical effects

A forecasting-data provider merges with a downstream competitor.

Conglomerate effects

A platform combines forecasting capabilities across several markets.

Innovation effects

The merger may eliminate an emerging competing forecasting technology.

21. Competition Between Forecasting Systems

Forecasting itself can become a market.

Examples include:

• AI demand forecasting;
• financial forecasting;
• energy forecasting;
• traffic prediction;
• logistics optimisation;
• advertising prediction;
• healthcare demand forecasting;
• agricultural forecasting.

Competition authorities may therefore analyse:

1. market definition;
2. data access;
3. interoperability;
4. switching costs;
5. network effects;
6. algorithmic advantages;
7. economies of scale;
8. exclusive contracts; and
9. vertical integration.

22. Strategic Forecasting and Consumer Welfare

Forecasting can produce substantial consumer benefits.

For example, better forecasting can result in:

• lower inventory costs;
• fewer shortages;
• lower transportation costs;
• improved electricity-grid management;
• more accurate demand planning;
• lower prices;
• better product availability;
• reduced waste.

Therefore, antitrust law should not treat predictive analytics itself as suspicious.

The legal concern arises from the manner in which the forecasting advantage is obtained, shared or exploited.

23. Compliance Framework for Businesses

Companies using strategic forecasting should establish safeguards.

1. Information classification

Classify information as:

• public;
• historical;
• aggregated;
• confidential;
• competitively sensitive; or
• highly sensitive.

2. Restrict competitor information

Do not unnecessarily collect or distribute competitors' future strategic plans.

3. Algorithm governance

Document:

• data inputs;
• model objectives;
• pricing variables;
• competitor information;
• human intervention; and
• automated decision rules.

4. Independent decision-making

Businesses should independently determine:

• prices;
• output;
• capacity;
• investment;
• discounts.

5. Audit third-party software

Contracts with forecasting and pricing providers should be reviewed for antitrust risks.

6. Data-access controls

Sensitive competitor information should not be unnecessarily available across business units.

7. Human oversight

High-impact competitive decisions should have appropriate compliance review.

24. Key Doctrinal Distinction

The most important distinction can be represented as follows:

Legitimate competitive forecasting

Internal data → independent analysis → better prediction → independent competitive decision

versus

Potentially problematic forecasting

Competitor-sensitive information → common platform/algorithm → reduced uncertainty → coordinated behaviour

and:

Potentially exclusionary forecasting

Dominant data infrastructure → exclusive control → superior predictions → foreclosure of rivals → reduced competition.

These are legally and economically different situations.

25. Case-Law Synthesis

26. Emerging Competition-Law Issues

The next generation of antitrust disputes is likely to involve increasingly sophisticated questions concerning:

A. AI forecasting

Can an AI system independently predict competitors' strategies without creating unlawful coordination?

B. Predictive pricing

When does prediction become a mechanism for coordination?

C. Data asymmetry

When does a superior data stock become a durable barrier to entry?

D. Forecasting-as-a-Service

Can the same provider supply competitively sensitive prediction tools to competing businesses?

E. Autonomous pricing

Who bears responsibility when an algorithm independently changes prices in response to predicted competitor behaviour?

F. Strategic information markets

Should certain highly sensitive forecasting information be subject to special access or confidentiality rules?

G. Cross-market forecasting

Can data gathered in one market be used to disadvantage competitors in another?

27. Conclusion

Strategic forecasting is not inherently anti-competitive. It is often an important source of innovation, efficiency and better consumer outcomes.

The competition-law problem emerges when forecasting:

1. facilitates coordination between competitors;
2. reduces strategic uncertainty through sensitive information exchange;
3. enables algorithmic collusion;
4. creates exclusionary data advantages;
5. allows a dominant platform to leverage information across markets;
6. forecloses competitors from essential predictive data; or
7. is used by a common intermediary to facilitate anti-competitive conduct.

The fundamental antitrust principle is therefore:

Competition law protects the competitive process, not equality of forecasting ability.

A company may legitimately win because it has better technology, data analysis and forecasting. The legal concern arises when that advantage is obtained or deployed through collusion, unlawful information exchange, exclusionary conduct, discriminatory access, or other practices that weaken the competitive process.

For modern digital markets, strategic forecasting should consequently be analysed at the intersection of information exchange, algorithmic coordination, data concentration, platform dominance, vertical integration and barriers to entry.