Panagiotis G. Papaioannou

We investigate the key factors that shape the dynamic evolution of Day-Ahead spot prices of seven European interconnected electricity markets of the Core Capacity Calculation Region, Core CCR (Austria AT, Hungary HU, Slovenia SI, Romania RO), the Southeast CCR (Bulgaria BG, Greece GR) and the Greece-Italy CCR (GRIT CCR), with emphasis on price surges and discrepancies observed in SEE CCR markets, during the period 2022–2024. The high differences in the prices of the two groups have generated political reactions from the countries that ‘suffer’ from these price discrepancies. By applying Machine Learning (ML) approaches, as Markov Blanket (MB) and Local, causal structures learning (LCSL), we are able of ‘revealing’ the entire path of volatility spillover of both spot price and the Cross-Border Transfer Availabilities (CBTA) between the countries involved, from north to south, thus uncovering i.e., ‘lifting the blanket’, to discover the ‘true’ structure’ of the path of causalities, responsible for the price disparity. The above methods are supported by the ‘mainstream’ approach of computing the correlation of the spot price and CBTA’s volatility curves of all markets, to detect volatility spillover effects across markets. The main findings of this hybrid approach are (a) the volatility of some Core CCRs (AT, HU, RO) markets’ spot price and CBTAs with neighboring countries, ‘uncovered’ to be pivotal, operating as a ‘transmitter’ of volatility ‘disturbances’, over its entire connection and causal path from Core CCR to SEE CCR markets, partially contributing to their price surge, (b) reduced available capacity for cross-border trading of some Core and SEE CCRs (they have not satisfied the minimum 70% requirement margin available for cross-zonal trade, MACZT), combined with local weather and geopolitical conditions, could have exacerbated the impact of the Flow-based Market coupling method (FBMC) used in the Core CCRs, on the prices’ surge of SEE CCR’s countries, e.g., via induced non-intuitive flows. This phenomenon questions the efficiency and reliability of the European Target’s model (TM) in securing ‘homogeneous’ power prices across all interconnected markets, core and peripheral.

We investigate the key factors that shape the dynamic evolution of Day-Ahead spot prices of seven European interconnected electricity markets of the Core Capacity Calculation Region, Core CCR (Austria AT, Hungary HU, Slovenia SI, Romania RO), the Southeast CCR (Bulgaria BG, Greece GR) and the Greece-Italy CCR , GRITR (Italy), with emphasis on price surges and discrepancies observed in SEE CCR markets, during the period 2022-2024. The high differences in the prices of the two groups, has generated political reactions from the countries that ‘suffer’ from these price discrepancies, as shown by the intense reactions of the governments of the affected countries and other institutions, as the sending of letters of Prime Ministers European Commission President). By applying Machine Learning (ML) approaches, as Markov Blanket (MB) and Local, causal structures learning (LCSL), we are able of ‘revealing’ the entire path of volatility spillover of spot price as well as the Cross-Border Transfer Availabilities (CBTA) between the countries involved, from north to south, thus uncovering i.e. ‘lifting the blanket’, to discover the ‘true’ structure’ of the path of causalities, responsible for the price disparity. The above methods are supported by the ‘mainstream’ approach of computing the correlation of the spot price and CBTA’s volatility curves of all markets, to detect volatility spillover effects across markets. The main findings of this hybrid approach are: a) the volatility of Austrian market’s spot price and its CBTAs with DE, CZ, and SI, ‘uncovered’ to be a pivotal market, behaving as a ‘transmitter’ of spot price and cross-border activity volatility, over its entire connection path with SEE CCR markets, which finally ‘receive’ the volatility disturbances, causing their price surge, b) the combination of weather and geopolitical factors with the limited interconnectivity of SEE markets, seem to have exacerbated the impact of the Flow-based Market coupling method (FBMC) used in the Core CCRs, on the prices of SEE CCR’s countries that rely on the Net Transfer Capacity -NTC- mechanism, by inducing non-intuitive flows, thus challenging the reliability of European Target’s model (based on FBMC) in protecting SEE markets from ‘unexpected-unfair-irrational’ price surge.