UMM assessment

Article 4 of REMIT requires a market participant to publish the inside information it holds about its own facilities, effectively and in a timely manner. Inside information is defined by its effect: information that, if made public, would be likely to significantly affect wholesale energy prices. The regulation sets no fixed volume above which that is true, so every compliance function has to decide, case by case, whether an outage, a derating or a position is large enough to matter.

The enforcement record shows the cost of getting it wrong. Ofgem fined SSE for waiting eight days to publish that units it had said would close were likely to stay open; RME fined Statnett for taking six hours to report an interconnector outage, against a one-hour benchmark; CoRDiS fined EDF for not publishing a request for additional reactor controls in time. Read the inside information guide.

It divides what your firm cleared in an auction by what each delivery period actually traded, and marks the periods that reach the share your policy treats as material. It is a sizing aid: whether an event is inside information, and whether it must be published, remain your firm’s decisions, and the assessment produces no determination either way.

A share is only as honest as what it is divided by, and there is no single market volume for a delivery period. Depending on what the publisher served, the denominator is the buy or sell side of one auction, one exchange’s traded volume, the crossing of the aggregated curves for a coupled area, or a book shared by several zones. Each is wrong in a stateable direction, and the console names which one answered and which way it biases your percentage. Where the bias runs towards under-flagging, a period just under your threshold deserves a second look. Read the four denominators.

Share alone can mislead: a large share that moved nothing and a small share that moved the clearing price are different facts. Where a published aggregated curve exists, the assessment re-clears the auction without your volume and reports the price that would have cleared. The counterfactual assumes everything else would have been submitted unchanged and cannot represent block orders, and a reconstructed result carries its out-of-sample model error beside the price shift. Read the counterfactual’s assumptions.

A position is the most sensitive thing a desk holds. The volume and threshold you enter stay in the browser tab: they are not sent to the server, stored or written to the URL. The assessment file you can download is built in the browser too.