Integrating renewables in the electricity system in a cost-efficient way requires massive transmission system investments and the efficient use of available transmission capacity. Markets are pivotal in the latter, especially in coordinating flows between countries. In the European Union, flow-based market coupling (FBMC) arose as the preferred market-based cross-border capacity allocation method, which has recently been extended to the CORE region, involving 12 countries of the EU. While the expansion of the flow-based methodology brings the EU closer to a single internal electricity market, its complexity and scale hinders analytical efforts of market participants, system operators and regulators. They conduct analysis to obtain insights into price formation, to enhance coordination as well as a more efficient use of assets. The limited and fragmented data available on flow-based domain strategies of TSOs and cost structures of participants obstructs the analysis.

This thesis attempts to bridge the gap between stylized academic models on FBMC and real-world market outcomes, to be able to reason about operational day-ahead markets via these simplified models. To this end, a multi-step modelling process is carried out to forecast day-ahead zonal market prices and cross-border flows. Inverse optimisation is utilised to recover cost functions on a bidding zone and technology level. This is followed by a spatial reconstruction of the static grid, which is subsequently used to infer the flow-based domain based on historical observations. The model-based approach has the added benefit of being interpretable, and can be adjusted for structural and regulatory market changes.

Inverse optimisation has proven to be able to recover aggregated technology cost functions with which real-world market outcomes can be forecasted in a tractable way. The model developed in this work is shown to outperform a commercially available, machine-learning-based algorithm in forecasting day-ahead prices. The limitations of reconstructing flow-based domains using publicly available flowbased market data are identified. Analysis concludes that while the ability to recover cross-border flows is sensitive to the shape and size of the inferred domains, the performance of price forecasts is robust against the quality of domain inference. The delivered work is argued to yield valuable insights to both market participants optimising their assets, and regulators structurally assessing the effects of flow-based domain configurations on welfare outcomes of real-world day-ahead markets.