Energy-market surveillance needs to look beyond traditional exchange trading
In June this year, extreme heat placed the UK’s electricity grid under significant stress. A system-wide event on 23 June was followed by Electricity Margin Notices for the evening peaks on 24 and 26 June, as high demand coincided with lower generation capability and energy availability. Ofgem subsequently commissioned a formal review of the drivers of the electricity grid system event, including system conditions, operational decisions and market actions.
For power traders, responding to changing conditions does not happen in a single market. A position may first be established in the day-ahead auction, adjusted as forecasts, demand, or asset availability change through intraday trading and intraday auctions, and ultimately interact with activity in the balancing market as the system approaches physical delivery.
For surveillance teams, these are not necessarily independent events. A participant’s day-ahead auction position, subsequent intraday activity and eventual balancing-market behaviour can form part of the same economic strategy. Understanding that activity therefore requires visibility across the different stages of the trading lifecycle and the ability to evidence how those behaviours were assessed.
This matters because firms need to be able to demonstrate the effectiveness of their surveillance to national regulators and, across EU wholesale energy markets, within the REMIT framework overseen by ACER (Agency for the Cooperation of Energy Regulators) and the relevant NRAs (National Regulatory Authorities).
Auctions create a different surveillance risk
Traditional exchange surveillance is typically centred on a continuous order book where orders are entered, amended, cancelled and executed sequentially, allowing surveillance to identify patterns such as layering, spoofing or marking the close.
An auction works differently because, unlike continuous intraday trading, where orders can execute whenever matching liquidity is available, day-ahead and intraday auctions concentrate liquidity around defined auction windows. Participants submit bids and offers before those points, after which a market algorithm aggregates and matches orders to establish clearing prices and volumes for the relevant delivery periods. In coupled European markets, the process can also incorporate available cross-border capacity between bidding zones.
This creates a distinct set of surveillance risks, including:
- Economic withholding, where available generation is offered at prices designed to reduce the likelihood of clearing;
- Physical withholding, where otherwise available capacity is not offered into the market;
- Transmission capacity hoarding, where capacity is acquired but not effectively used.
Auction-specific risks can also arise from:
- Large or strategically timed orders that influence indicative or clearing prices;
- Orders that create a misleading impression of supply or demand before being amended or cancelled;
- Attempts to influence auction-derived reference prices;
- Erroneous or late-stage order activity close to gate closure.
Other behaviours may only become significant when viewed across adjacent delivery periods, related products, multiple venues or subsequent intraday and balancing activity.
Effective auction surveillance needs to reconstruct the auction context.
This includes market phases, indicative and final clearing prices, available capacity and the evolution of participant orders up to gate closure – rather than assessing only the final executed outcome.
From day-ahead to balancing: seeing the whole position
We already support surveillance across key auction venues including Nord Pool and EPEX, with our onboarding roadmap extending into additional European auction markets. Broader venue coverage will allow the same surveillance principles to be applied more consistently across different bidding zones, delivery periods and market structures, particularly where significant behaviour only emerges when activity is compared across venues or adjacent markets.
For cross-border transmission capacity, planned JAO and PRISMA connectivity will add a complementary surveillance dimension. Their data can help surveillance teams assess risks that cannot be understood from auction order books alone, including capacity hoarding or inefficient use of acquired transmission rights, while providing additional context around a participant’s wider market position.
Returning to the June heat event, a trader’s day-ahead position, intraday adjustments and subsequent balancing-market activity should not necessarily be viewed as separate events. As weather forecasts, demand expectations and system conditions evolved, so too did the economics of the position.
A participant may establish a significant position in a day-ahead auction on a venue such as Nord Pool or EPEX, adjust that position through the intraday market as forecasts, demand expectations or generation availability change, and later submit bids or offers into the relevant balancing market.
Effective surveillance therefore needs to reconstruct the participant’s position across the trading lifecycle and place individual transactions in their wider market and physical context.
That means linking economically related activity across day-ahead, intraday and balancing markets, while taking account of different delivery periods and instruments.
Physical information is equally important because positions, available capacity, outages, Urgent Market Messages (UMMs) and other fundamental information may materially change the interpretation of trading behaviour and help distinguish genuinely unusual conduct from activity explained by changing market or operational conditions.
Context is king
The technology challenge is therefore not simply to generate more alerts, but to give surveillance teams enough context to understand how a position developed, what changed as delivery approached, and whether behaviour across one market created consequences or an economic benefit in another. That context allows investigators to determine whether unusual activity has a legitimate economic or physical explanation, or whether it warrants further investigation and potential regulatory escalation.
As power systems incorporate greater volumes of wind and solar generation, differences between day-ahead forecasts and actual generation closer to delivery become increasingly important. This increases the role of intraday and balancing markets as participants adjust positions in response to changing forecasts, demand and asset availability.
The Scila advantage
Scila already provides many of the core building blocks for this approach, combining energy-specific auction surveillance with linked order books, position aggregation and UMM context. As connectivity expands across additional auction and balancing/capacity data sources, that analytical framework can be applied across more of the power-trading lifecycle.
The direction of travel is clear: the next generation of energy surveillance needs to understand not just the transaction, but the market, physical context and evolving position behind it – from auction through to physical delivery.
