Competition Law and Digital Twin Modelling of Markets

1. Introduction

Digital twin modelling of markets refers to the creation of a dynamic, data-driven virtual representation of a real market in which the behaviour of firms, consumers, prices, supply, demand, entry, exit, innovation, and regulatory interventions can be simulated.

A conventional digital twin models a physical object or system—for example, a factory, aircraft, supply chain, or infrastructure network. A market digital twin extends the concept to an economic system. It can continuously ingest market data and simulate how competitive conditions may change when a firm changes prices, acquires a rival, restricts access to data, changes an algorithm, enters a new market, or adopts a particular business strategy.

Digital twins are increasingly relevant to competition policy because they can model complex systems involving real-time data, algorithms, platforms and interconnected markets. The European Commission has specifically identified digital twins as an important enabling technology and has noted the importance of access to live data, computing power, interoperability and competitive conditions.

The important legal point is that digital twin modelling does not create a new category of competition law by itself. Rather, it gives competition authorities and courts a more sophisticated tool for applying existing doctrines concerning:

• market definition;
• market power;
• dominance;
• abuse of dominance;
• anti-competitive agreements;
• algorithmic coordination;
• mergers and acquisitions;
• exclusionary conduct;
• innovation competition;
• data access;
• interoperability;
• network effects; and
• competitive effects of technological conduct.

2. Meaning of a Market Digital Twin

A market digital twin can be represented conceptually as:

Real Market → Data → Digital Model → Simulation → Competitive Prediction → Regulatory Decision → Real Market Feedback

The model may contain:

A. Firms

The digital twin can represent:

• market shares;
• pricing strategies;
• production capacity;
• advertising expenditure;
• investment;
• innovation;
• entry and exit;
• switching costs;
• platform policies.

B. Consumers

It can model:

• consumer preferences;
• willingness to pay;
• switching behaviour;
• price sensitivity;
• multi-homing;
• network effects;
• privacy preferences;
• quality preferences.

C. Market structure

The model can incorporate:

• number of competitors;
• concentration;
• barriers to entry;
• vertical integration;
• platform ecosystems;
• access to data;
• interoperability;
• distribution channels.

D. Dynamic variables

Unlike a static market study, a digital twin can simulate changes over time.

For example:

Firm A acquires Firm B → data concentration increases → entry becomes more difficult → rivals lose access to customers → prices or quality change → innovation declines.

The model can test this chain before and after the transaction.

3. Why Digital Twins Matter to Competition Law

Traditional competition analysis frequently relies upon historical evidence.

A digital twin can potentially move enforcement towards predictive competition analysis.

For example, suppose a dominant digital platform proposes acquiring a rapidly growing start-up.

The conventional analysis may ask:

What is the present market share of the parties?

A digital twin can ask:

What would competition look like five years after the acquisition if the start-up remained independent?

That distinction is extremely important in digital markets because a small company may be a future competitive constraint even when its present market share is modest.

Digital twin modelling can therefore be particularly useful in:

1. merger control;
2. digital platform investigations;
3. algorithmic pricing investigations;
4. abuse of dominance cases;
5. market definition;
6. data-access disputes;
7. interoperability disputes;
8. innovation competition;
9. essential-facility questions; and
10. regulatory impact assessment.

4. Digital Twin and Relevant Market Definition

Market definition traditionally identifies the products, services and geographic areas in which competition occurs.

The digital twin can simulate hypothetical substitution.

Suppose there are:

• Search Engine A;
• Search Engine B;
• Social Media Platform C; and
• AI Assistant D.

A traditional analysis might consider whether consumers regard them as substitutes.

A market digital twin could simulate:

What percentage of consumers would move from A to B if A increased price or reduced quality?

In digital markets, the "price" may be zero.

Consequently, the model may instead simulate:

• reduction in privacy;
• deterioration in search quality;
• increased advertising;
• reduction in functionality;
• increased subscription cost;
• slower service;
• reduced interoperability.

This is important because competition in digital markets often occurs through quality rather than monetary price.

5. SSNIP and SSNDQ Analysis

Traditional competition economics often uses the SSNIP test—Small but Significant and Non-transitory Increase in Price.

Digital markets require adaptation because many services are supplied at zero monetary price.

A digital twin could therefore model an:

SSNDQ test

Small but Significant and Non-transitory Decrease in Quality.

For example:

1. Platform A reduces privacy protection.
2. The digital twin estimates consumer switching.
3. Consumers move to Platform B.
4. Platform B becomes a competitive constraint.
5. The relevant market may therefore include B.

The model could simultaneously test:

• price;
• quality;
• privacy;
• advertising intensity;
• data collection;
• switching costs.

Thus, digital twin modelling can make market definition more multi-dimensional.

6. Digital Twins and Market Power

Market power is not determined solely by market share.

In digital markets, a digital twin could incorporate:

Direct network effects

More users make the platform more valuable.

Indirect network effects

More users attract advertisers or developers, which attract still more users.

Data advantages

More users generate more data.

More data may improve algorithms.

Better algorithms attract more users.

This creates a feedback loop:

Users → Data → Better Algorithm → Better Service → More Users

The digital twin can model whether this feedback creates a durable competitive advantage.

7. Digital Twin and Algorithmic Pricing

One of the most significant applications concerns algorithmic pricing.

Suppose competing firms use pricing algorithms.

The algorithms may independently observe:

• competitor prices;
• market demand;
• inventory;
• consumer behaviour.

Even without an explicit agreement, algorithms might repeatedly produce similar prices.

The competition-law question becomes: