Barclays, Citi, Deutsche Bank and Standard Chartered Adopt Ant International's FalconTST 2.0 for FX and Cashflow Forecasting
Barclays, Citi, Deutsche Bank and Standard Chartered Adopt Ant International’s FalconTST 2.0 for FX and Cashflow Forecasting #
Ant International has unveiled the second generation of its Falcon Time-Series Transformer (FalconTST) AI model, with Barclays, Citi, Deutsche Bank and Standard Chartered already integrating it into their foreign-exchange and liquidity management operations.
The Singapore-based fintech said FalconTST 2.0 improves the accuracy of cashflow and FX exposure forecasting for cross-border payments. Ant International said the model achieved a Mean Absolute Scaled Error (MASE) score of 0.666 on a leading public benchmark for time-series foundational models, which the company says places it at the top of the leaderboard ahead of comparable models from other major technology firms. The company said the model delivers forecast accuracy consistently above 93 percent.
Each bank has integrated the model within its own existing infrastructure. Barclays has connected FalconTST 2.0 into its BARX NetFX hedging platform, while Citi has paired the model with its Fixed FX Rates product. Standard Chartered is running the technology alongside its SCALE FX system as part of both companies’ participation in the Monetary Authority of Singapore’s PathFin.ai programme. Deutsche Bank has adopted the model for its broader foreign-exchange operations.
Before extending the technology to banking partners, Ant International first deployed the earlier Falcon model internally to manage its own cashflow and currency exposure across hourly, daily and weekly time horizons. The company said the model now handles more than 60 percent of its own FX forecasting workload.
Jiang-Ming Yang, chief innovation officer at Ant International, said the model’s value lay in converting forecasts into actionable liquidity and currency decisions rather than simply generating projections. Standard Chartered has publicly stated that the integration has delivered more than 90 percent forecasting accuracy for its operations and a significant reduction in liquidity management costs.
Ant International said FalconTST 2.0, a transformer-based model with close to two billion parameters, is designed to work across sectors. The company plans to extend it into supply-chain demand forecasting for e-commerce platforms and predictive operations management in aviation, beyond its current financial services applications.
Copy link