RealTime Enterprise Liquidity Maximizing margins reducing risk advancing

Real-Time Enterprise Liquidity Maximizing margins, reducing risk, advancing business objectives Jerry Baulier CTO

Intraday Liquidity Management § Liquidity has to be available: – at the right time – in the right place – in the correct currency 2

Intraday Liquidity Management: Situation § Many disparate systems § Manual processes § Time delays § This lack of real-time information & accurate forecasting results in – excess liquidity (i. e. , sub-optimal yields) – excess borrowing (i. e. , increased costs) 3

Benefits of Liquidity Management § Maximize margins – Dynamically forecast cash with precision – Minimize uninvested cash – Minimize funding cost – Maximize yield on investments / assets § Reduce risk – Provide integrated view of liquidity / exposures – Avoid surprises, regulatory fines, and reputation risk § Advance business objectives – Deliver integrated view of customer positions – Adapt rapidly to customer behavior and events 4

Desired State § Systems connected in real-time § Sophisticated predictive knowledge for intraday forecasting § Forecasted receipts and payments tracked and replaced with actual § Tight connection between funding desk and payments Internal DDA Debt Trading ALCO Custody INFLOW OUTFLOW FORECASTS XXX NET XXX ACTUAL XXX XXX XXX NET 5 External NOSTRO WEB SWIFT

The Aleri Liquidity Management System (ALM) • 30+ different systems • Streaming and static • Real-time position management • Actual and forecast positions with drill-down • Payment flow control • Highly customizable Accounting Systems Authoring Environment RTGS Gateway SWIFT Gateway Lending Systems INPUT STREAMS Payment Systems Aleri Streaming Platform Streaming Output On-Demand Queries Data Migration OLAP Database Settlement Systems Clearing Data 6 • Real-time OLAP • Queries across operational and historic data

Aleri Liquidity Management – 85 “streams” at Barclays 7

Why Event Processing Platform for ALM? § Traditional RDBMS technology can’t deliver the performance: – Complex data model, transformations, correlations, and analysis – Maintain all “Materialized Views” (e. g. , account positions) in near-zero latency § Like RDBMSs, stream processing platform must also deliver: – Retention of large data sets (5 days of transactions) – On-demand SQL query access to retained data – Security (authentication, encryption, access control, audits) § Plus data integrity & flexible/fast data state recovery – Persistent source/sink queues – Synchronous data commits with minimal performance overhead – Data retention with optional (local) persistence 8

Barclays Giving them a real-time consolidated view of liquidity Situation • Barclays has large volumes of payments and receipts across business units, currencies, and geographies • They didn’t know where they stood and had large buffers in place to ensure all obligations could be fulfilled • There was risk from manual processes and time delays of not knowing their positions in real-time • They had over 30 systems where this information resided and it was not an option to rip out and replace those Risk Requirements • Prior to the solution there was a significant potential for reputation risk should there ever be situation where obligations could not met due to poor liquidity management • To access information in real-time from the 30 plus systems • To persist the data • To hold the data for analytic analysis to improve predictive forecasting • To track payments and receipts throughout the lifecycle from forecast, to processing, to final settlement Solution System that can access the different inputs (30+ systems and 80+ actual streams of data), apply normalization and filtering logic to that information, provide real time data with ad-hoc query capability, and a historical database for trend analytics 9 Key Features Benefits • Consolidated single access point to monitor and forecast intraday liquidity positions • Ad-hoc query analysis for drill down research • Historical database with OLAP • Reduced funding costs • Reduced risk from timing delays and manual processes • More value add understanding of how customers use cash on an intraday basis

Jerry Baulier, CTO jerry. baulier@alerilabs. com