January exposed a structural reality of South-East Europe’s power system: day-ahead prices are no longer the marginal signal of system stress. The true marginal signal increasingly comes from balancing energy and ancillary services, where scarcity emerges earlier, clears at higher implied prices, and redistributes value more asymmetrically than the headline market.
Across Serbia, Romania, Bulgaria, and Croatia, January’s repeated evening ramps showed a consistent pattern. Day-ahead markets on SEEPEX, OPCOM, and CROPEX cleared at high but finite peak prices, while balancing markets operated close to their technical limits. This is visible indirectly in price behaviour: when DAM peaks reach €170–300/MWh repeatedly without fuel shortage, it indicates that the system is already consuming a large share of its available balancing headroom.
The underlying issue is depth. In most SEE systems, less than 5–10% of installed capacity is effectively available as fast, fully controllable balancing reserve during winter peak hours. Hydro provides the bulk of this flexibility, followed by a small number of gas and CHP units. Wind and solar do not meaningfully contribute to upward reserves in January conditions, while demand response remains marginal in most markets. This creates a situation where balancing scarcity emerges before energy scarcity, and where the marginal MWh in real time is far more expensive than the marginal MWh in the day-ahead auction.
The economic consequence is that imbalance prices frequently exceed DAM prices by wide margins during stress intervals. While published imbalance prices differ by jurisdiction and settlement design, January conditions imply effective imbalance costs in the €300–500/MWh range during the tightest hours, even when DAM peaks remain below those levels. For traders and suppliers short in real time, this transforms what looks like a manageable DAM exposure into a materially loss-making position after settlement.
Romania provides a clear illustration. Despite OPCOM clearing January at an average baseload of €150.51/MWh and peak of €176.60/MWh, the system relied heavily on a narrow pool of hydro and thermal balancing providers during evening ramps. Nuclear, while critical for baseload, contributes almost nothing to short-term balancing. As a result, the effective marginal price of system security was far above the published DAM peak, with balancing markets absorbing the true scarcity rent.
Serbia shows a similar structure with different optics. SEEPEX peak days approached €294/MWh, already signalling stress, yet the balancing system is even thinner than Romania’s. Serbia’s hydro fleet is large, but its balancing capability is highly concentrated. When hydro is withheld for water management or already dispatched into DAM, the residual balancing stack becomes extremely steep. This creates a convex risk profile for suppliers: profits earned during high DAM prices can be erased by a handful of imbalance hours.
In Croatia, the interaction between balancing and cross-border constraints is decisive. When imports are available, balancing pressure is muted. When interconnectors bind, Croatia’s domestic balancing resources must absorb the full ramp. January’s €165.66/MWh average peak price on CROPEX therefore understates the true cost of flexibility, which is reflected instead in reserve activation and imbalance settlement.
For asset owners, this environment has created a silent repricing of value. Balancing-capable assets now earn a premium that is not visible in day-ahead prices. Hydro units with fast response, gas units with flexible dispatch rights, and any asset able to offer aFRR or mFRR are monetising scarcity in a structurally superior way to pure energy trading. In January, the risk-adjusted return on balancing participation often exceeded that of DAM arbitrage, particularly in smaller systems where reserve depth is minimal.
For buyers and traders, the implication is stark. A portfolio that appears hedged in the day-ahead market can still be structurally short in the balancing market. January confirms that balancing risk is now the dominant hidden risk in SEE, and that managing it requires physical optionality, not financial hedges. Markets that fail to deepen balancing participation will continue to experience extreme DAM peaks not because energy is scarce, but because the system is priced off the fear of losing control in real time.
By virtu.energy
