Skip to main content

AI in Treasury

Purpose​

An example of applying AI Conveyor in treasury: liquidity forecasting, placement optimization, alerting, and scenario analysis.

Key ideas​

  • Initiatives involve time series, scenarios, and fast decisions; the data is often sensitive, and requirements for accuracy and explainability are high.
  • Impact includes improved forecasts, lower funding costs, and more precise management of liquidity and risk.
  • Governance accounts for data confidentiality and alignment with the boundaries of responsibility between treasury and risk.

How it works​

Treasury ideas enter the common portfolio of the AI office; experiments test data quality and the predictive power of models. When moving to prototype and production, integration with treasury systems and near-real-time monitoring are ensured. Impact is captured using treasury metrics (for example, forecast accuracy, savings on interest).