Our Expert in Bulgaria

GOLD

Ivelina Cherneva

GOLD

Algorithmic Pricing in Bulgaria: Comply with the “excessively High Price” Rules

By Global Law Experts

– posted 1 hour ago

Algorithmic pricing bulgaria is now a live compliance concern under the Protection of Competition Act (Zakon za zashtita na konkurentsiyata), which enables the Commission for Protection of Competition (CPC) to scrutinise excessive and unfair pricing by dominant undertakings and to exercise strong investigatory powers. For e‑commerce sellers, marketplace operators and the product teams who build repricers and dynamic pricing engines, this is a shift from abstract antitrust theory to concrete operational risk. This guide sets out how machine‑driven pricing systems attract regulatory scrutiny, and, most importantly, the practical steps you should take before an inquiry lands.

You will find a side‑by‑side risk comparison, a step‑by‑step compliance playbook, a CPC response plan, and a decision framework designed for in‑house counsel and pricing managers who need to act now, not later.

Who should act: in‑house counsel, pricing and product managers, marketplace compliance teams, and e‑commerce sellers operating in Bulgaria. What you get: a risk map, a prescriptive compliance programme, sample contractual language, and an enforcement‑response checklist. For the broader legal context, see the Competition Law Bulgaria country overview.

The legal framework for excessive pricing in Bulgaria

Bulgarian competition law is set out primarily in the Protection of Competition Act, enforced by the CPC, and applied alongside Articles 101 and 102 of the Treaty on the Functioning of the European Union (TFEU). The Act and its amendments are published in the State Gazette, with the underlying legislative process available through the National Assembly. Together these instruments give the CPC both substantive grounds to challenge unfair or excessive pricing by dominant undertakings and a broad procedural toolkit. This makes algorithmic pricing bulgaria a topic that boards, not just compliance officers, need to understand.

The “excessively high price” concept, legal elements

Under the Protection of Competition Act, the direct or indirect imposition of unfair purchase or selling prices can constitute an abuse of a dominant position, mirroring Article 102(a) TFEU. In substance the analysis asks whether a price bears a reasonable relationship to the economic value of the product or service, or whether it is objectively excessive by reference to cost, comparable markets or benchmark prices.

This concept is not invented from nothing: it draws on established EU jurisprudence, most notably the two‑limb test set out by the Court of Justice in United Brands (Case 27/76), which asks first whether the difference between cost and price is excessive and second whether the price is unfair either in itself or by comparison with competing products.

An important threshold point is that excessive‑pricing analysis generally requires the undertaking to hold a dominant position on a relevant market; it is not a free‑standing rule that applies to every seller. Where dominance exists, however, the practical consequence for pricing teams is that the business must be able to articulate an objective justification for a high price. Where a price is set or escalated by an automated system that cannot articulate a cost‑based or value‑based rationale, that evidential gap becomes an enforcement risk. The excessive pricing bulgaria standard therefore rewards transparency and documentation and penalises opaque, unexplained price levels.

CPC powers (inspections, interim measures, fines, remedies)

The CPC has extensive investigatory and remedial powers, aligned with EU enforcement standards including the ECN+ Directive (Directive (EU) 2019/1), which requires national competition authorities to have effective tools for detection and deterrence. In practice this means the CPC can conduct unannounced on‑site inspections (dawn raids), request extensive documentary and electronic evidence, impose interim measures to halt conduct pending a decision, and levy significant fines. Under the Protection of Competition Act, fines for abuse of dominance and restrictive agreements can reach up to 10% of the undertaking’s total turnover for the preceding financial year. Guidance on the CPC’s procedures and powers is published on the Commission for Protection of Competition site.

Who is liable? Sellers, marketplaces, platform operators, individuals?

Liability under the regime is broad. A dominant seller that sets excessive prices through its own repricer is an obvious target, but so is a marketplace operator whose buy‑box algorithm, default pricing tools or vendor rules shape the prices consumers ultimately pay. Platform operators cannot treat pricing as purely the vendor’s problem where the platform designs or controls the algorithmic infrastructure. Corporate liability is the norm. Where personalised pricing raises consumer‑fairness questions, there is also overlap with the Commission for Consumer Protection (CCP), which enforces consumer‑protection and unfair‑commercial‑practices rules.

Why algorithmic and dynamic pricing increase CPC risk

Modern pricing systems are built to react, to demand, to competitor moves, to inventory, to time of day. Those same reactive properties are precisely what make algorithmic pricing bulgaria a magnet for regulatory attention. An automated engine can produce, within seconds, a price outcome that a human pricing committee would have flagged, debated and rejected. Where a firm is dominant and that outcome is objectively excessive, the fact that “the algorithm did it” is not a defence.

Three distinct legal concerns map onto typical algorithmic behaviour. First, excessive pricing: surge or demand‑based logic can push prices to levels the CPC may treat as unfair where the seller is dominant. Second, price discrimination: personalised pricing by user attributes can generate exploitative or discriminatory prices, engaging both competition and consumer law. Third, tacit collusion: when competing sellers all deploy repricers that read the same signals and mirror one another, prices can stabilise at supra‑competitive levels without any express agreement, a risk the OECD’s work on algorithms and collusion examines in detail.

Common algorithmic features that trigger scrutiny

No human override. A repricer that can raise prices without any ceiling or human sign‑off removes the last line of defence against an excessive outcome.
Automated escalation loops. Two or more algorithms reacting to each other can ratchet prices upward in a feedback spiral that no one intended and no one is monitoring.
Third‑party data feeds. Live competitor price feeds and shared pricing‑intelligence services can facilitate coordination, or at least make it far harder to show that price levels were set independently.
Opaque optimisation objectives. A model told simply to “maximise revenue” with no fairness or legality constraints will find excessive‑price outcomes if the market permits them.

Distinguishing excessive pricing from collusion and price‑fixing

It is essential to keep these categories apart because they engage different legal tests and different mitigations. Excessive pricing is a unilateral abuse concept, it requires a dominant position and turns on the United Brands‑style question of whether the price is unfair relative to economic value. Collusion, by contrast, concerns coordination between competitors, whether by explicit agreement or through algorithmic mechanisms that produce coordinated outcomes, and does not require dominance. Dynamic pricing compliance therefore has to address both: guardrails against setting an objectively excessive price where you may be dominant, and design choices that prevent your repricer from mirroring rivals or facilitating tacit alignment.

Side‑by‑side comparison: algorithmic pricing bulgaria practices vs CPC enforcement concerns

The table below is the analytical centrepiece of this guide. It maps common algorithmic and dynamic pricing practices to the reasons the CPC may view them as risky, the legal trigger, the practical control that reduces exposure, and a priority rating. Use it directly to triage your own systems.

Dimension	Typical practice	Why the CPC may view it as risky	Legal trigger	Practical control	Priority
Automated surge pricing	Prices auto‑increase on demand spikes without human review	Can produce excessive‑price outcomes where the seller is dominant	Abuse of dominance (unfair prices) under the Protection of Competition Act / Art. 102 TFEU	Human price caps, business rules for maximum allowed increases, logging of demand signals and decision rationale	High
Competitive data ingestion	Live competitor price feeds and auto‑matching	May facilitate coordinated outcomes or aggravate abusive price setting	Restrictive agreements / concerted practices; aggravating excessive pricing	Limit competitor data, anonymise inputs, avoid exact mirroring, document design choices	High
Black‑box ML repricer	ML model optimises revenue without transparency	Hard to show objective justification for the price level	Burden on firm to justify price; absence of justification raises risk	Model explainability, retained training data, ex ante pricing impact assessments	High
Personalised pricing	Price discrimination by user attributes	Can create exploitative or discriminatory prices	Consumer protection + competition overlap; need objective justification	Risk assessments, segmentation limits, documented lawful basis and non‑discrimination rules	Medium
Competitor‑based undercutting	Continuous undercutting to win the buy‑box	Can trigger collusion analysis or drive high prices elsewhere	Depends on market power; CPC assesses market structure	Monitor market‑power metrics; guardrails against escalation loops	Medium
Supplier‑side contracts	Algorithms impose minimum resale prices or disincentives	May amount to resale price maintenance or price rigidity	Article 101 TFEU / restrictive vertical agreements under the Act	Revise supplier contracts; remove RPM clauses; add compliance and audit rights	High

The comparison makes the pattern clear: the highest‑priority exposures share a common failing, the inability to explain and constrain the price the machine produced. Where a system can escalate prices without a cap, ingests data that ties it to rivals, or optimises in a black box, the firm cannot readily demonstrate the objective justification the excessive pricing bulgaria standard demands.

Three urgent fixes emerge above all others. First, implement logging and a human override so that every material price movement has a recorded rationale and a ceiling. Second, run a tailored pricing impact assessment that stress‑tests your algorithm against demand shocks and competitor‑feed scenarios. Third, update supplier and platform contracts and vendor‑onboarding rules to strip out resale price maintenance and to secure audit rights over third‑party repricers. These three actions address the bulk of the High‑priority rows in a single coordinated programme.

Practical compliance playbook for e‑commerce sellers and marketplaces

This section turns the risk map into a stepwise programme. It is deliberately prescriptive so that legal, product and data teams can divide the work. Treat every template and sample clause here as a starting point to be reviewed by counsel, not a finished document. Effective dynamic pricing compliance is a cross‑functional exercise, and the fastest path to a defensible position is to run these steps in sequence.

Step 0, Scope: identify the systems, markets and sellers affected

Begin by mapping every algorithmic component that touches price: repricers, surge modules, promotional engines, personalisation layers and any marketplace buy‑box logic. Catalogue the data feeds each system consumes, internal cost data, inventory, and any external competitor feeds. Establish the coverage periods for which you hold historical pricing data. This inventory is the foundation for everything that follows; you cannot govern what you have not identified.

Step 1, Legal and economic risk assessment

Screen each product line for market power, because excessive‑pricing risk arises only where the business is dominant in a relevant market. Run simple price‑cost tests to identify products where the margin could look objectively excessive. Add consumer‑vulnerability checks for any personalised pricing. The output is a risk register (sample template) that ranks each pricing system by likelihood and severity of enforcement exposure and links directly to the mitigations in Step 2.

Step 2, Algorithm design and governance controls

This is where most enforcement risk is either created or contained. Build the following into your pricing architecture:

Human‑in‑the‑loop. Require human review before any price movement exceeds a defined threshold, and before any surge event above a set percentage takes effect.
Default caps. Hard‑code maximum permissible increases so the algorithm physically cannot produce an excessive price without escalation.
Rollback mechanisms. Ensure any automated price change can be reversed instantly and that the reversal is logged.
Model explainability notes. Maintain documentation of what the model optimises, which inputs it uses, and why each input is legitimate.
ML documentation. Record data provenance and training sets so you can reconstruct why the model behaved as it did at any point.

A workable human override escalation flow looks like this: the system detects a proposed price above the cap; it freezes the change and raises a ticket to the pricing owner; the owner either approves with a recorded justification or rejects; the decision and its rationale are written to an immutable log. This single flow converts an opaque automated event into a documented, defensible business decision, the difference between an easily explained price and an unexplained one.

Step 3, Monitoring and evidence

Enforcement outcomes often turn on evidence quality. Maintain, at minimum: complete price series with timestamps; the decision inputs behind each material change; snapshots of any competitor feed used; and logs of every human intervention. These records are your primary means of showing objective justification if the CPC asks. On retention, adopt a policy proportionate to the limitation and review periods relevant to competition matters; retaining logs and model documentation for contested pricing events for several years is a prudent baseline, subject to confirmation with counsel and to data‑protection obligations.

Step 4, Contractual and marketplace controls

Marketplaces and sellers should embed compliance into their contracts. Key clauses include a compliance covenant requiring vendors to operate lawful pricing algorithms, audit rights over vendor repricers, an indemnity for infringements caused by a vendor’s system, data‑sharing constraints to prevent feeds that facilitate coordination, and an express prohibition on defined repricer behaviours such as exact mirroring of rival prices. A vendor‑onboarding checklist should confirm each of these before a seller goes live.

Step 5, Compliance training and sign‑offs

Train the people who build and operate pricing systems, not only lawyers. Product managers, data scientists and category managers should complete competition‑awareness training at onboarding and at least annually. Capture compliance attestations, signed confirmations that each person understands the caps, the override flow and the prohibited behaviours, so that responsibility is documented and cultural.

Step 6, Internal escalation and remediation

Define in advance when to pause an algorithm: for example, when monitoring detects a sustained price above cap, an escalation loop between repricers, or a consumer complaint pattern. Set out preservation steps to be triggered the moment an investigation is anticipated, and a legal notification matrix identifying who is informed, in what order, and within what timeframe. Rehearsing this before an incident dramatically improves your response when one occurs.

If the CPC opens an inquiry or inspection: timeline and response plan

Even a well‑run programme can attract an inquiry. A calm, prepared response protects both your legal position and your ability to keep trading. The overarching rules are simple: preserve everything, engage counsel immediately, and protect legal privilege where it applies.

Day 0–3, immediate preservation and notification checklist

Preserve all pricing logs, model documentation and communications; suspend any routine deletion.
Notify in‑house counsel and, where the matter is material, external competition counsel without delay.
Identify and ring‑fence potentially privileged material.
Brief a small, controlled internal team and instruct staff not to alter or discuss the relevant systems.

Days 3–30, document production, interviews, negotiating search limits

Expect requests for pricing data, algorithm documentation, competitor‑feed records and internal communications. Work with counsel to respond fully but proportionately, negotiating the scope and search terms of any document request so that production is targeted rather than open‑ended. Prepare custodians for interviews and ensure factual accuracy in every submission.

Interim measures and fines, what to expect and how to engage

The CPC may impose interim measures to halt conduct while it investigates; where that risk exists, be ready to propose voluntary commitments, such as reinstating price caps, that address the concern without a formal infringement decision. On penalties, the regime provides for significant fines (up to 10% of annual turnover for the most serious infringements), so the goal throughout is to demonstrate good faith, prompt remediation and a functioning compliance programme, all of which can bear on the outcome.

When to involve outside counsel

Engage specialist competition counsel as soon as a dawn raid occurs, an interim measure is threatened, or the conduct in question touches a market where you may hold significant market power. Cost varies with complexity and duration; see the FAQ below for how to think about budgeting a CPC response.

Implementation checklist and sample templates

Use the compact checklist below to drive the programme to completion. Treat the referenced templates as samples requiring legal review before deployment.

Complete the Step 0 inventory of every pricing algorithm and data feed.
Populate the risk register (sample template) and rank exposures.
Implement default price caps and the human override escalation flow.
Switch on comprehensive logging of prices, inputs, feed snapshots and interventions.
Adopt a retention policy for pricing logs and model documentation.
Deploy the pricing algorithm governance policy (sample template).
Insert the supplier compliance clause (sample) into new and renewing contracts.
Run competition‑awareness training and collect attestations.
Document the internal escalation, pause and preservation procedures.
Schedule a counsel‑led review of the whole programme.

Three sample templates support this checklist: an audit checklist covering the inventory and risk‑register stages; a minimal pricing algorithm governance policy setting out caps, override rules and logging obligations; and a supplier compliance clause for contracts and onboarding. Each should be adapted to your business and reviewed by counsel. For a bespoke package, request a tailored review.

Decision framework, choose the right posture for your business

Rather than hedging, this framework tells you which posture to adopt. Most businesses should do both A and B in proportion, but the balance depends on your profile.

Choose A, rapid remediation and governance, when you operate live algorithmic or surge pricing, you lack human caps or comprehensive logging, you use third‑party competitor feeds, or you have any indication that prices have spiked without justification. Fix the architecture first; it is the cheapest and most effective risk reduction available.
Choose B, defensive posture and forensic readiness, when your systems are already well‑governed but you operate in a concentrated market, hold potential market power, or have received a complaint or informal contact. Here the priority is preservation discipline, privilege management and a rehearsed CPC response.

Resource allocation heuristic: a small marketplace or seller should concentrate limited resources on caps, logging and contract clauses (Posture A). A large platform should invest in full governance plus forensic readiness across both postures. A dominant seller must treat excessive‑pricing exposure as a board‑level risk and fund both remediation and a standing response capability, because market power sharply increases the likelihood and severity of enforcement.

Key takeaways and recommended next steps

Managing algorithmic pricing bulgaria risk comes down to five priorities:

Run a priority audit of every pricing algorithm and data feed now.
Implement default caps, a human override flow and comprehensive logging.
Update supplier and marketplace contracts to remove RPM clauses and secure audit rights.
Train the product and data teams that actually build and run the systems.
Schedule a counsel‑led review and keep it on a recurring calendar.

Need Legal Advice?

This article was produced by Global Law Experts. For specialist advice on this topic, contact Ivelina Cherneva at Dinova Rusev & Partners, a member of the Global Law Experts network.

Sources

FAQs

What counts as an “excessively high price” under Bulgarian competition law?

It is a price imposed by a dominant undertaking that lacks a reasonable relationship to the economic value of the product or service and is objectively excessive by reference to cost, comparable markets or benchmarks. The concept mirrors the two‑limb test from the EU’s United Brands case and the unfair‑pricing prohibition in the Protection of Competition Act and Article 102 TFEU. The consolidated text of the Act is published in the State Gazette.

Can a pricing algorithm alone trigger CPC liability?

Yes, where the undertaking is dominant. In the context of algorithmic pricing bulgaria, the fact that a price was set automatically is no defence. If the outcome is objectively excessive and you cannot show justification, liability can follow. Mitigate this with caps, a human override flow, and logging that records the rationale for material price movements.

What evidence will the CPC expect during an investigation?

Expect requests for timestamped price series, algorithm and model documentation, competitor‑feed snapshots, logs of human interventions, and internal communications. Procedural guidance is published by the Commission for Protection of Competition.

How long should we retain pricing logs and training data?

Retain logs and model documentation for contested pricing events for a period proportionate to competition limitation and review timelines, several years is a common baseline, subject to confirmation with counsel and to data‑protection obligations. Where personalised pricing is involved, consumer‑protection oversight rests in part with the Commission for Consumer Protection.

How much does a competition lawyer cost in Bulgaria for a CPC response?

Fees vary with the complexity, whether a dawn raid or interim measure is involved, and the volume of data to be produced. A short advisory review is modest; a full response to a contested investigation is a more substantial engagement. Budget for an initial scoping assessment and request a tailored estimate before committing to the full programme.