Carbon convergence has become the most consequential timing risk in South-East European power trading. Direction is broadly agreed: carbon costs will rise, coal and lignite will exit, and market coupling will deepen. What remains deeply uncertain—and now decisively priced by markets—is when these forces will align, how fast they will propagate through cross-border trade, and whether grids and flexibility will arrive in time to prevent a disorderly repricing. Seasonal system assessments by ENTSO-E frame adequacy under current assumptions; trading desks price the risk that those assumptions shift abruptly under carbon convergence.
The mechanics of carbon convergence in SEE differ from those in fully ETS-integrated markets. Several systems still operate dispatchable coal and lignite outside the full cost of EU carbon pricing, anchoring marginal costs in the €25–35/MWh range for energy while neighbouring markets internalise carbon costs that push gas- or coal-set prices to €70–120/MWh in normal conditions. This asymmetry has supported cross-border arbitrage and price dampening during stress. Convergence—via CBAM exposure, market coupling, or domestic carbon pricing—erodes that asymmetry. The trading question is not whether convergence happens, but whether it precedes or follows the commissioning of replacement capacity and grid reinforcement.
Timing matters because carbon convergence removes a safety valve before it replaces it. Coal and lignite units provide coincident availability during winter stress—exactly when demand spikes and renewables underperform. When carbon costs price these units out of the merit order or accelerate closures, the system loses dispatchable depth immediately. Replacement assets—storage, pumped hydro upgrades, grid-forming inverters, new interconnectors—arrive on multi-year timelines. If convergence outruns build-out, markets face a gap period characterised by higher volatility, wider spreads, and more frequent scarcity pricing.
Forward curves already reflect this risk. Across SEE, winter peak products show persistent premiums of €40–60/MWh over baseload, not solely explained by demand patterns. Beyond Y+2, bid–ask spreads widen materially, and liquidity thins, signalling disagreement over convergence pace. Traders price a distribution of outcomes rather than a single trajectory: a slow convergence path with lingering lignite anchors, and a fast convergence path with abrupt repricing. The convexity embedded in these curves is the market’s way of insuring against policy-driven shocks.
Carbon convergence also reshapes congestion economics. As carbon costs rise unevenly, power flows reorient toward zones with lower effective marginal costs—until corridors bind. In early convergence phases, this can increase congestion frequency as markets exploit remaining differentials. Later, as coal exits accelerate, congestion flips sign: scarcity propagates upstream, and previously exporting zones import during stress. Each phase produces distinct trading patterns, and desks that fail to distinguish them misprice corridor risk.
Quantitatively, the impact is visible in stress outcomes. During winter events under partial convergence, deficit zones have cleared at €250–400/MWh day-ahead, with intraday and balancing prices exceeding €500–600/MWh when response is scarce. Under faster convergence scenarios—where coal availability is economically constrained—models show higher activation volumes (+30–50 %) and more frequent corridor saturation (commercial capacity compressing from 1.5–2.0 GW to 500–700 MW on key interfaces). Markets are not waiting for these scenarios to materialise; they price their probability.
CBAM adds a non-linear trigger. Even without domestic ETS alignment, export exposure to carbon costs can change dispatch behaviour abruptly. Assets that were marginally profitable become loss-making at the border, altering availability during peak hours. The effect is lumpy: a policy threshold crossed in one year can remove hundreds of megawatts from winter availability overnight. Trading books exposed to such cliffs carry asymmetric downside if they assume smooth adjustment.
The interaction with inertia deepens the risk. Carbon convergence tends to retire synchronous units first, accelerating inertia decline before grid-forming alternatives scale. As inertia falls, balancing prices spike and intraday volatility increases. Empirically, days with low synchronous online capacity exhibit 2–3× higher intraday variance than comparable demand days five years ago. Carbon convergence therefore raises not just energy prices, but the cost of risk itself—through imbalance charges, higher margins, and reduced liquidity.
Investment sequencing determines whether convergence is disruptive or manageable. Grid reinforcement at €0.8–1.2 million per kilometre for new 400 kV lines, storage at €500–700 thousand per MWh, and pumped hydro modernisation at €1.5–2.5 million per MW can compress volatility materially—but only if commissioned ahead of or alongside convergence. Where approvals lag, markets price the gap. Congestion rents of €30–70 million per year on key interfaces are the monetary expression of that lag.
From a trading strategy perspective, carbon convergence is best treated as a calendar risk. Near-term books benefit from residual low-carbon-cost baseload and dampened volatility; mid-term books face the highest convexity as policy timing uncertainty peaks; longer-term books hinge on infrastructure delivery. Successful desks layer exposure accordingly—protecting winter peaks, diversifying corridor exposure, and valuing flexibility as a hedge against policy shocks.
For industrial offtakers, convergence reframes procurement. Fixed-price baseload contracts that look attractive under average conditions can become punitive when peak exposure is unhedged. Many buyers are shifting toward structures that cap peak risk, index balancing exposure, or secure access to flexibility. The premium paid for such structures is, in effect, an insurance premium against convergence timing risk.
Systemically, the risk is not high prices per se, but misaligned transitions. If carbon costs rise faster than flexibility and grids, volatility becomes chronic; if infrastructure leads policy, markets absorb convergence smoothly. The former path entrenches scarcity trading; the latter compresses spreads and stabilises prices. Markets signal which path they fear by how much convexity they embed in winter products and how wide bid–ask spreads run beyond Y+2.
The structural mismatch in remuneration compounds the challenge. Assets that stabilise the system—by reducing peak scarcity and balancing needs—earn less as they succeed. Without explicit recognition of stability value, investment lags, and convergence shocks hit harder. Markets respond rationally by pricing risk rather than resolution. Until frameworks evolve to pay for stability, carbon convergence will remain a timing risk with tradable consequences.
In South-East Europe, carbon convergence is not a distant endpoint; it is a sequence of thresholds that markets approach with caution. Each policy announcement, each closure schedule, each grid delay shifts probabilities and reprices curves. Traders who frame convergence as a smooth glide path will be surprised. Those who frame it as a series of discrete jumps—some anticipated, some not—will be prepared.
The cold logic of the market is this: carbon will converge, but physics does not wait for policy. If dispatchable depth exits before alternatives arrive, prices will spike, corridors will bind, and volatility will pay. If investment and policy align, volatility will compress and premiums will fade. Until that alignment is credible, South-East European power trading will continue to price when more than what.