Liquidity Optimization in Gross Settlement Systems with Quantum Reordering: Application to TARGET2

9 March 2026

Banca d'Italia today publishes 'Liquidity Optimization in Gross Settlement Systems with Quantum Reordering: Application to TARGET2', the new issue of the series 'Markets, infrastructures, payment systems'.

In High-Value Payment Systems (HVPSs) with gross settlement, which require high levels of liquidity, optimizing the processing order could provide significant savings in the liquidity allocation needed for institutions to stay solvent. Building on McMahon et al. (2024), who showed that a hybrid quantum solver can improve intraday liquidity efficiency in Canada's HVPS, we apply a similar technique to payments between Italian institutions in TARGET2, using a Constrained Quadratic Model (CQM) Solver. Our application optimizes payment batches, yielding average daily liquidity savings between EUR 23 million and EUR 38 million, over a 35-day sample. The use of machine learning techniques also allows the identification of batch features that enhance these savings. Finally, we benchmark the results against those obtained using a Simulated Annealing Algorithm (SAA), finding comparable liquidity savings. The SAA is extended to handle larger batch sizes, which are still challenging for current quantum hardware, demonstrating an increase in liquidity savings more than proportional to the growth in batch size.