The Southeast European Power Exchange enabled negative prices on 5 May 2026. Exactly one month later, the day-ahead market printed nine consecutive negative hours, the lowest at −€45.50/MWh - and Bulgaria's grid was charging.
The most useful frame for the next 24 months of SEE storage is not a forecast. It is the UK BESS revenue stack between 2022 and 2024.
In 2022, frequency response was 80% of typical UK BESS revenue. By 2024, it was 20%. The opportunity didn't disappear - it migrated. Market saturation compressed ancillary prices; arbitrage and capacity markets absorbed the difference. The operators who built early captured the peak of the stack. Those who came later built into a structurally different return profile, on the same CAPEX.
Southeast Europe is running the same play, faster. Bulgaria has allocated 14+ GWh of standalone storage under RESTORE - a volume that will saturate the BG ancillary market well inside 24 months. Romania's ancillary prices today (~€9/MW/hr aFRR, ~€70/MW/hr FCR) look generous because the market is shallow. Each new 100 MW of operational BESS compresses them.
The data from this month is the inflection. SEEPEX enabled negative prices on 5 May. Within a month, the day-ahead printed nine consecutive negative hours. Nova Zagora and St. George - commissioned a few months earlier - were charging into that negative-priced window. The arbitrage thesis is no longer projection; it is this quarter's revenue.
Three implications follow.
One - the timing premium is the single largest IRR contributor available to SEE BESS in 2026. Projects reaching commercial operation before mid-2027 will operate at the top of the revenue stack. Projects commissioning in 2028 will operate in an arbitrage-dominant, compressed market. Same equipment, same CAPEX, materially different returns. The window is open and visible.
Two - the conventional bankability framework needs revision. Lenders modelling Romanian or Bulgarian BESS on today's ancillary headlines are pricing yesterday's market. A 2027 base case should assume UK-2024 composition: roughly 60–70% arbitrage, 20% ancillary, 10–20% capacity. The Nova Zagora / St. George structure - financial PPA floor with merchant arbitrage tail - is already the template the next financings should follow.
Three - Romania CfD II at €40.35/MWh is a structural warning, not a celebration. Greek CfD solar lost 60% of April revenue against negative prices and curtailment. Romanian developers who bid the cap without an in-built storage strategy will discover the same arithmetic by 2028. The Greek government's zero-price ruling is a fiscal patch; it does not change the underlying physics. Every new SEE PV megawatt without a credible storage line is a future stranded cash flow.
What we are watching next month: the first ENTSO-E balancing data from the Nova Zagora and St. George BESS, Romania's Q3 CfD Round 3 framework, and whether Croatia's draft HERA fees actually clear the regulatory backlog or stall again.