REMIT and MAR, part 1: who has to monitor and what the software must do

Leggi in italiano


Note. I’m a software architect, not a lawyer. This article looks at REMIT and MAR from the point of view of someone who designs and builds monitoring systems, and it is not meant to be a complete legal analysis. The regulatory interpretations I give are based on public sources and may be incomplete or out of date. Compliance decisions need advice from legal counsel.

Intro – REMIT and MAR: the rules against market abuse in energy

In Europe, wholesale electricity and gas markets are fairly young. They were born with the liberalisation started by Directives 96/92/EC for electricity and 98/30/EC for gas, which opened to competition sectors that until then had been run as monopolies. Italy transposed the electricity directive with Legislative Decree 79/1999 (the Bersani decree), and the power exchange run by GME started trading on 1 April 2004. For gas, after transposition with Legislative Decree 164/2000, trading concentrated on the Virtual Trading Point (PSV), and in 2010 GME launched its own trading platforms.

Market abuse is usually associated with equity markets: insider trading, spreading false news, manipulating share prices. The same risks apply to energy too. Today electricity and gas are traded every day on organised exchanges such as EPEX SPOT, EEX or the GME markets and, to a large extent, over the counter. Here too a market participant can push price formation in its own favour. To deal with these risks, the European Union adopted two regulations: REMIT and MAR.

REMIT (Regulation (EU) 1227/2011, Regulation on Wholesale Energy Market Integrity and Transparency) covers wholesale energy markets: electricity and gas supply contracts, transportation capacity and derivatives on these products. It prohibits insider trading and market manipulation, and requires market participants to disclose inside information in a timely manner. Participants must also register and report their transactions to ACER, the EU agency of energy regulators. Oversight is shared between ACER and the national regulators, which in Italy is ARERA. In 2024 the regulation was substantially revised (the so-called REMIT II, Regulation (EU) 2024/1106) to bring it closer to financial market rules and strengthen investigative powers and penalties.

MAR (Regulation (EU) 596/2014, Market Abuse Regulation) is the market abuse framework for financial instruments. It matters for energy too, because many contracts, such as power futures, are financial instruments in every respect. In Italy the competent authority is Consob.

Why should anyone working in the sector care? A typical energy company trades both physical products and financial derivatives, so it falls under both regimes. For many participants, REMIT II also turned monitoring their own trading from good practice into an explicit obligation. They need systems and procedures to detect suspicious orders and transactions and report them to the authorities. The 2022 energy crisis showed how fragile these markets are and how much their failures hit households and businesses. This is where the analysis and monitoring software I cover in this two-part series comes in: what it must do, what it must analyse and how it must be built to comply with the rules.

Series outline

The focus is on the analysis and monitoring tools used by compliance officers at energy market participants. The series has two articles.

First article (this one): obligations and features

  1. Who has to monitor, and why. We start with the obligations: who has to monitor their own trading under REMIT and MAR, and how the picture changes for those who also trade financial instruments compared with those who only trade physical products.
  2. What the software must do. We look at the functions authorities expect from a surveillance system, from generating alerts to reporting suspicious cases.
  3. What it must analyse. We go through the behaviours to detect: insider trading, the main forms of manipulation and algorithmic trading.

Second article: privacy, security and deployment model

  1. Privacy and NIS2. A tool that analyses traders’ activity processes employees’ personal data and becomes a critical supplier for the market participant. We look at what that implies.
  2. SaaS or on-premises. We close with the criteria for choosing the deployment model and with how responsibilities are split between the participant and the vendor.

1. Who has to monitor, and why

Not every wholesale energy market participant has the same surveillance obligation. Two factors matter: whether it also trades financial instruments, and whether it owns assets.

Prohibitions for everyone, surveillance for some

REMIT prohibits insider trading (Art. 3) and market manipulation (Art. 5) for all market participants, and requires disclosure of inside information (Art. 4). These obligations apply to anyone active in the market, whatever its size.

The obligation to have systems and procedures in place to detect and report suspicious orders and transactions is set by Art. 15 instead. It only applies to persons professionally arranging or executing transactions, the so-called PPAETs.

PPATs and PPETs

REMIT II distinguishes two roles:

  • PPAT (Persons Professionally Arranging Transactions): exchanges, brokers and platforms that organise trading. The obligation has existed since the first version of REMIT. Since 7 May 2024 it also covers suspicious orders, not just transactions.
  • PPET (Persons Professionally Executing Transactions): those who professionally execute transactions. “Executing” might suggest only those who execute orders on behalf of clients, such as an intermediary. ACER has clarified, however, that it also includes those dealing on own account, so an energy company buying and selling for itself is in scope too (Open Letter of 25/9/2024 and ACER report of May 2025). The obligation applies from 8 November 2024.

The dividing line: financial instruments

The PPET obligation does not apply to everyone. Art. 15(2) only applies to PPETs already subject to Art. 16(2) MAR, that is, those who also professionally trade financial instruments.

According to ESMA (MAR Q&A 6.1), this category also includes non-financial companies dealing on own account, even for hedging purposes. Having a dedicated desk or executing your own orders directly on a trading venue are indicators. An energy company that buys and sells power futures, for example on EEX, is therefore typically a PPET.

For these participants REMIT II does not create a new system: it asks them to extend the surveillance already required by MAR to physical energy products and to breaches of Arts. 3, 4 and 5 REMIT. Reports go to ACER and to the national regulator.

The “physical-only” participant is out of scope: it buys and sells electricity or gas for actual delivery, for example on spot markets or through forward contracts with physical delivery, and does not trade financial derivatives such as futures. Among those active in the market, it is the only profile not subject to Art. 15(2). Be careful, though: whether a contract is a financial instrument is decided by MiFID II, the EU directive on markets in financial instruments, not by whether it provides for physical delivery, so the classification has to be checked contract by contract. All the prohibitions still apply. Anyone using algorithmic trading must also comply with Art. 5a, which requires specific systems and controls on algorithms.

The role of assets

Owning generation, storage or consumption assets does not change whether Art. 15 applies, but it widens the range of risks.

  • Disclosure of inside information. The Art. 4 obligation covers information about the assets the participant owns, controls or operates, including unavailabilities. Since 2021, disclosure must go through one of the Inside Information Platforms (IIPs) recognised by ACER. In Italy the most widely used is GME’s PIP.
  • Insider trading on your own information. The typical case is trading while knowing about an outage that has not yet been published. Art. 3(4)(b) provides an exception only for transactions that cover the immediate physical losses of an unplanned outage.
  • Capacity withholding. Withholding generation or transportation capacity to influence prices is a form of manipulation that only those who control the assets can commit.

Is software mandatory?

No rule requires a specific piece of software. Delegated Regulation (EU) 2016/957 asks for systems and procedures that are appropriate and proportionate to the scale, complexity and nature of the business. Automation is not mandatory for everyone, but the participant must be able to show the authority that its choices are appropriate, including the level of automation.

In practice, with non-trivial volumes and with orders as well as transactions to analyse, manual checks do not hold up. For those who only trade physical products, monitoring remains a risk management choice. In an investigation, though, it is the strongest way to show due diligence.

Participant profileSurveillance obligation (Art. 15 REMIT)What else to monitor
Exchange, broker, platform (PPAT)Yes; orders included since 7/5/2024Clients’ orders and transactions
Trader with financial derivatives, no assetsYes, as a PPET, from 8/11/2024Own orders and transactions, physical and financial
Trader with financial derivatives and assetsYes, as a PPET, from 8/11/2024Also: IIP disclosures, asset availability, capacity withholding
Physical-only participant, no assetsNoOptional; algorithm controls if trading algorithmically
Physical-only participant, with assetsNoOptional, but with insider risk on outages and the disclosure obligation (Art. 4)

2. What the software must do

The rules do not describe a piece of software but a process: monitor all orders and transactions, generate alerts, have them assessed by competent people, decide whether to report and keep a record of every step. The software is there to make this process effective and, above all, demonstrable to an authority.

Where the requirements come from

Art. 15 REMIT sets the obligation but does not go into operational detail. The more precise requirements are in Delegated Regulation (EU) 2016/957, which implements Art. 16 MAR, and in ACER’s guidance. Since ACER asks PPETs to extend the systems already required by MAR to energy products, Regulation 2016/957 is in practice the reference for REMIT too.

The five-step process

  1. Data collection and quality. The system must cover all orders and transactions, including amended, cancelled or rejected ones, both on and off trading venues. Typical sources are trading platforms and the ETRM, exchange data, the reports sent to ACER, IIP disclosures and market data. ACER explicitly asks for data quality checks before analysis.
  2. Detection. The software analyses each order and transaction on its own and against the others, and generates alerts on activity that deserves a closer look. Detection scenarios and their thresholds must be calibrated, documented, tested and reviewed periodically.
  3. Human analysis. No alert becomes a report automatically. Regulation 2016/957 requires an appropriate level of human analysis, and ACER asks for each alert to be enriched with context: market conditions, news, IIP disclosures, fundamental data. The software must therefore offer real case management, not just a list of alerts.
  4. Decision and reporting. If there is reasonable suspicion, a STOR is sent to ACER and to the national regulator through the notification platform, without further delay and in any case within four weeks. The report must be adequate, complete and informative.
  5. Record keeping. Every analysis must be documented, including the decision not to report and its reasons. Records must be kept for five years, kept confidential and provided to the authority on request.

The governance the software must support

A good tool is not enough if the organisation around it does not hold up. The software must, however, make some safeguards possible and verifiable:

  • Independence. Surveillance staff must be free of conflicts of interest, and management must not be able to block a report. Access profiles and an audit trail are how you prove it.
  • Periodic review. Regulation 2016/957 requires systems and procedures to be reviewed at least once a year, with a documented audit and internal review.
  • Training. Staff involved in monitoring must receive effective and regular training.
  • Transparency on automation. On request, the participant must be able to explain to the authority how automated its systems are.
  • Delegation. Data analysis and alert generation can be outsourced to an external provider under a written agreement. The participant, however, remains fully responsible, must keep the expertise to oversee the provider and must have direct access to the information. We will come back to this in the second article of the series, on SaaS versus on-premises.

What practice shows

The ACER report of May 2026 gives a picture of PPATs, that is, exchanges, brokers and platforms, not of participants dealing on own account. It still offers useful pointers:

  • about 70% use a surveillance system, but only 21% use professional third-party software;
  • about half rely on in-house tools and 19% on general-purpose tools such as Microsoft Office;
  • those using professional systems report shorter times between detection and reporting, although ACER notes this is a correlation, not a proven causal link;
  • automatic alerts sent as STORs without further analysis increase the volume but give little value to the authorities;
  • coverage of Arts. 3 and 4 (insider trading and disclosure of inside information) still lags behind that of manipulation (Art. 5).

One finding matters for anyone choosing a tool: calibration. With in-house systems, surveillance staff change the parameters themselves. With professional software, the vendor often has to step in. Being able to calibrate scenarios independently, with tracked changes, is therefore a selection criterion not to underestimate.

3. What it must analyse

A surveillance system has to cover three families of risk: insider trading together with the inside information obligations, market manipulation and, for those who use it, algorithmic trading. Each needs different data and detection scenarios, and much of that data does not live in the trading system.

Insider trading and inside information

In the energy market, the typical piece of inside information is about assets: an outage, an extended maintenance, a capacity reduction. The risk arises when someone who knows it, for example dispatching, passes it to the trading desk before it is published.

The Engie case is a textbook example. In 2017 a member of the dispatching team told the trading desk that a power plant’s unavailability had been extended before it was published, and the desk concluded five trades on the French intraday market of EPEX SPOT. In 2022 the CRE imposed a fine of 80,000 euros (ACER).

To catch these cases, the software must match the timing of orders and transactions against IIP disclosures and, where possible, against asset availability data. The same comparison is used to check that Art. 4 disclosures are timely and complete.

Market manipulation

Forms of manipulation fall into transactions and orders that give false or misleading signals, trades that set prices at artificial levels and the dissemination of false information. ACER has published specific guidance notes on three practices: wash trades (2017), transmission capacity hoarding (2018) and layering and spoofing (2019).

Some practices are shared with financial markets, others are specific to energy because they require control of assets or transportation capacity. For layering and spoofing, ACER lists order characteristics as indicators: size, price, duration, status, pattern and repetition. That is why the analysis must be based on the full order book, that is, the record of all buy and sell orders entered and not yet executed, including cancelled ones, and not just on completed transactions.

Market makers are a special case: by trade they keep orders on both sides of the book and constantly amend and cancel them. Their legitimate behaviour can look like layering, so scenarios must be tuned carefully to avoid flooding surveillance with false positives. We will go deeper into this in a dedicated article.

Errors count too. A wrong order entered near the close or near a price assessment window can amount to manipulation even if it is later cancelled.

Algorithmic trading

With Art. 5a, REMIT II introduced specific obligations for those who trade with algorithms. They need effective systems and risk controls, resilient systems with sufficient capacity, compliant with trading venue rules, tested, monitored and covered by a business continuity plan. Since 7 May 2024 the use of algorithms must also be notified to the national regulator and to ACER.

For surveillance, this means being able to tell algorithm-generated orders from manual ones and trace them back to the algorithm that produced them. The Dutch regulator ACM also points out that data on algorithmic systems and controls must be kept for at least five years.

From behaviours to data

The table sums up, for each behaviour, what the software must do and what data it needs.

BehaviourWhat it isWhat the software needs
Insider trading on outagesTrading while aware of an outage not yet publishedComparison of order and transaction timing with IIP disclosures and asset data
Late or incomplete disclosure (Art. 4)Publishing inside information late or only in partComparison of asset events with published messages
Wash tradeTrades that involve no real change in beneficial ownership or market risk, for example between companies in the same groupCounterparty and corporate group reference data
Layering and spoofingNon-genuine orders placed on one side of the book, at several price levels (layering) or in large size at a single level (spoofing), to trade on the opposite sideFull order book with amendments and cancellations, analysis of size, duration and repetition
Marking the closeTrading in the session’s reference period to raise, lower or hold the reference priceAnalysis by time window around closes and price assessment periods
Capacity withholdingNot offering, without justification, available generation capacityComparison of bids with asset availability
Transmission capacity hoardingAcquiring all or part of the available transmission capacity without actually using itComparison of allocated capacity with nominations
Erroneous orders and algorithmsWrong orders or orders generated by algorithms out of controlIdentification of algorithmic orders, pre-trade controls, algorithm history

Not every row applies to everyone. A trader without assets has no capacity to withhold and no outages to disclose, while for a generator these are the main risks. As we saw in section 1, the scope of the analysis must be defined according to the participant’s profile.

In short

The surveillance obligation does not apply equally to everyone: it depends above all on whether the participant also trades financial instruments, while owning assets widens the risks to cover. The rules do not mandate software but an effective and demonstrable process: complete, good-quality data, calibrated alerts, human analysis, timely reports and records kept for five years. What to analyse depends on the participant’s profile, from insider trading on outages to layering and algorithmic trading.

In the next article we will look at what surrounds the software. A system that analyses traders’ activity processes personal data and holds critical information: we will see what GDPR, Italian labour law (the Statuto dei Lavoratori) and NIS2 imply, and how to choose between SaaS and on-premises.

Appendix – Glossary

TermMeaning
ACEREuropean Union Agency for the Cooperation of Energy Regulators. Oversees the application of REMIT together with the national regulators.
Algorithmic tradingTrading in which an algorithm automatically decides order parameters such as price, quantity and timing.
ARERAItalian Regulatory Authority for Energy, Networks and Environment, the national regulator competent for REMIT in Italy.
Capacity withholdingNot offering available generation capacity on the market without justification.
ConsobItalian financial markets authority, competent for MAR.
EEX, EPEX SPOTEuropean energy exchanges: EEX mainly for derivatives such as futures, EPEX SPOT for electricity spot markets.
ESMAEuropean Securities and Markets Authority.
ETRMEnergy Trading and Risk Management: the system where a participant records trades, positions and risks.
Financial instrumentA contract falling within the MiFID II categories, such as shares, bonds and many derivatives. Trading in these instruments falls under MAR.
ForwardContract to buy or sell energy at a future date, often bilateral and with physical delivery.
FuturesStandardised derivative contract, traded on an exchange, to buy or sell energy at a future date at a price fixed today. It is a financial instrument.
GMEGestore dei Mercati Energetici: runs the Italian power exchange, the gas markets and the PIP platform.
IIP, PIPInside Information Platform: platform recognised by ACER for disclosing inside information. The PIP is the one run by GME.
Inside informationPrecise, non-public information that, if made public, would be likely to significantly affect prices, such as a power plant outage.
Insider tradingTrading, or getting others to trade, on the basis of inside information.
Layering, spoofingPlacing non-genuine orders on one side of the book to influence prices and trade on the opposite side.
MARMarket Abuse Regulation, Regulation (EU) 596/2014: the EU market abuse framework for financial instruments.
Market manipulationTrades, orders or dissemination of information that give false signals or set prices at artificial levels.
Marking the closeTrading in a session’s reference period to raise, lower or hold the reference price.
MiFID IIDirective 2014/65/EU on markets in financial instruments. Among other things, it defines which contracts are financial instruments.
Order book (book)The real-time record of buy (bid) and sell (ask) orders entered on a platform and not yet executed, sorted by price.
OTCOver-the-counter: bilateral trading between the parties, off exchange.
Physical-only participantA participant that buys and sells energy for actual delivery and does not trade financial derivatives. Not subject to the surveillance obligation of Art. 15(2) REMIT.
PPAET, PPAT, PPETPersons professionally arranging (PPAT) or executing (PPET) transactions; PPAET covers both.
REMIT, REMIT IIRegulation (EU) 1227/2011 on wholesale energy market integrity and transparency. REMIT II is the 2024 revision (Regulation (EU) 2024/1106).
SpotMarkets for very short-term delivery: the next day (day-ahead) or the same day (intraday).
STORSuspicious Transaction and Order Report: the report of a suspicious order or transaction to ACER and the national regulator.
Trading venueExchange or organised platform where supply and demand meet.
Wash tradeA trade that involves no real change in ownership or market risk.

Sources