January trading on SEEPEX looked, on the surface, like a familiar winter month: the day-ahead market cleared 404,970.3 MWh with an average daily baseload of €118.13/MWh and peak of €136.27/MWh, while the month’s most stressed day printed €293.84/MWh in peak and €228.29/MWh in baseload terms. What made January structurally different was not the averages, but the way risk migrated away from the day-ahead screen into two adjacent layers that Serbia now cannot treat as secondary: the balancing/imbalance layer run by EMS, and the CBAM-driven “carbon add-on” layer that will increasingly attach itself to cross-border electricity economics and to contracts that claim low-carbon delivery.
In Serbia’s case, the balancing layer is the hidden setter of scarcity value because it is where physical system stress is finally priced, not just implied. The SEEPEX hourly pattern already tells you where that stress concentrates. The average hourly price profile rises from the morning shoulder into the evening ramp, with the month’s highest average hours clustering around the late afternoon and early evening, with average hourly prices around €166.55/MWh at hour 17 and €178.58/MWh at hour 18 before easing. This is a classic flexibility premium, and in Serbia it is structurally tied to the depth of controllable upward regulation available to EMS at the same time of day. Nuclear is not present domestically, lignite is slow and operationally constrained, and variable renewables add forecast error; the balancing stack becomes steep precisely in those hours when schedules are hardest to follow.
EMS publishes the imbalance settlement logic in a way that matters for trading strategy. Settlement price is determined as a weighted price of activated resources and is administratively bounded, including a rule that settlement cannot be negative and that it can be capped relative to the most expensive upward activation in the interval. That design pushes Serbia toward a regime where imbalance is not merely a “small penalty,” but a convex exposure: once the system starts activating expensive upward energy, every BRP that is short in the same direction is pulled into the same pricing gravity well.
Even without pulling a complete January time series of imbalance settlement prices in this session, there is a useful Serbia-only quantification that can be done using hard anchors. EMS discloses that the reference “balancing energy price” used for risk value and unbalanced schedule financial calculation is 154.174 EUR/MWh for 2026. That number is not a January average; it is a regulatory-weighted reference. But it frames the core point: Serbia’s system-level balancing value is already structurally above the €118.13/MWh January baseload day-ahead average.
From there, the economic envelope for January is driven by one variable: how much energy ends up being settled in imbalance rather than as scheduled. A conservative trading-grade scenario for a well-functioning control area is that net activated balancing/imbalance exposure sits in the low single digits of the day-ahead cleared volume, not because the system is failing, but because forecast error, outages, ramp limits, and cross-border adjustments are unavoidable. If Serbia’s net balancing/imbalance energy that is effectively repriced is 2% of January’s SEEPEX-cleared volume, that is ~8,099 MWh against the 404,970.3 MWh monthly DAM turnover. If the average “all-in” settlement level for those stressed hours sits only €50/MWh above the day-ahead baseload reference, the implied monthly transfer is ~€0.40 million. If the premium is €100/MWh, the transfer is ~€0.81 million. If the premium is €150/MWh, it becomes ~€1.21 million. The point is not the exact figure; it is the scale: in Serbia, a seemingly small share of energy repriced in the balancing layer can move seven figures of value in a single winter month, and it concentrates into the same evening hours that already dominate day-ahead tail risk.
This is why, for Serbia-only trading strategy, “hedging the month” is increasingly the wrong mental model. The operational truth is that the portfolio is marked in the last hours of the day, where schedule error, reserve scarcity, and congestion align. January’s market shape shows that Serbia has already entered a regime where a trader can be directionally right on the month and still be wrong in the settlement layer if they carry systematic short exposure into the evening ramp.
The second Serbia-only layer is CBAM-adjusted electricity economics. CBAM moved into its definitive regime from 1 January 2026, and the certificate price is linked to the EU ETS allowance price (expressed as €/tCO₂, with quarterly averaging in 2026). For electricity, reporting and default-value logic has been evolving toward a framework where default emission values are based on the overall emission intensity of the exporting country’s power system rather than assuming only fossil-based generation, and the use of verifiable actual emissions is intended to be easier in practice. That change matters for Serbia because it effectively converts “national grid carbon intensity” into a tradable basis risk on every cross-border electricity sale into the EU that is not backed by recognised, auditable low-carbon documentation.
Quantification for Serbia starts with the carbon price anchor. EU carbon permits were reported around €81.35/tCO₂ on 9 February 2026. The second anchor is Serbia’s grid emissions intensity, which different public trackers and studies place in a high range consistent with lignite-heavy generation. A reasonable working band for stress testing Serbia-only cross-border exposure is ~0.67–1.05 tCO₂/MWh, reflecting the fact that Serbia’s power mix remains coal-dominant in recent years while hydro provides a meaningful but insufficient counterweight.
Multiplying those two anchors gives the CBAM-style carbon value embedded in 1 MWh exported from Serbia under a default/system-average logic. At 0.67 tCO₂/MWh and €81.35/t, the carbon value is ~€54.5/MWh. At 1.05 tCO₂/MWh, it is ~€85.4/MWh. This is the magnitude Serbia must now treat as a potential add-on to cross-border trading economics into the EU when attribution is conservative or documentation is weak, and it is large enough to flip trades.
To see what this means against January price reality, place that carbon value next to Serbia’s observed SEEPEX economics. January baseload at €118.13/MWh becomes an EU-delivered “CBAM-adjusted” economic level of roughly €173–203/MWh once you add €54–85/MWh of carbon value. In the very hours Serbia most wants to monetise exports—those evening ramps where January averaged €166.55–178.58/MWh—the CBAM-adjusted economic level becomes roughly €221–264/MWh before congestion, losses, and any corridor-specific spread. That changes the whole intuition of “export when prices are high.” Under CBAM-style attribution, the export economics can become least attractive precisely when the Serbian physical system is tightest, because those same hours often coincide with higher marginal emissions and higher risk of documentation gaps.
This creates a very Serbia-specific tradeoff that will define 2026 behaviour. If Serbian exporters can prove low-carbon intensity through accepted, auditable instruments and credible measurement, they can sell into the EU on something closer to the physical price spread and keep the carbon basis tight. If they cannot, then the carbon basis becomes a structural haircut in the €55–85/MWh range at current EUA levels, which is large enough to re-route flows, change contracting terms, and shift value away from short-term arbitrage toward long-term structured delivery with embedded certification and traceability.
Put the two layers together and January’s Serbia-only risk trend becomes clear. The balancing layer is where Serbia pays for being wrong in time, with a steep penalty concentrated into evening ramps and capable of moving high six to low seven figures of value in a single winter month from a few percent of volume. The CBAM layer is where Serbia pays for being wrong in attribution, with a potential €54–85/MWh economic drag on EU-facing trades unless the carbon profile is credibly documented. January’s day-ahead numbers were the visible surface; the 2026 trading game in Serbia is increasingly decided underneath it, in settlement and in carbon-adjusted cross-border economics.
By virtu.energy
