Researchers from the Laboratory for financial and risks analytics (LAFRA) have been working to answer this question in cooperation with the European Investment Bank (EIB) as a part of the research project "Predicting cash flow patterns and timing in international bank accounts".
Project team members from FER include Bruno Gašperov, MSc, Stjepan Begušić, PhD, and Associate Professor Zvonko Kostanjčar, PhD. The project is funded by the EIB Institute under the STAREBEI (STAges de REcherche BEI-EIB research internships) programme and is the first such research grant with a Croatian university.
More information on the project and the developed solution can be found in the detailed news content.
The EIB performs its cash management based on cashflows expected from the lending, funding and treasury activities on the international markets. Better predicting the timing of intraday incoming payments from its multiple counterparties would improve its liquidity management and optimise prefunding decisions on EIB accounts thus reducing the associated costs.
Based on historical cash flow data provided by the EIB Finance Directorate, the researchers designed and developed machine learning models for inflow timing predictions. In addition, the models provide information on time intervals within which the inflows are expected, thus providing additional certainty for the EIB experts when managing liquidity. Within the project, a workshop on AI applications has also been organized, as well as a working visit to the EIB in Luxembourg, where the potential of the developed models was demonstrated.