Bank Balance Sheet Optimisation

  • We gathered data from over 30 different sources, some of them in the Big Data category. We were able to track down the data and negotiate with stakeholders to use the data. We then built efficient ETL software to consolidate the data into a SQL Server database. It was important to significantly reduce the data size by normalising data on import, resulting in our ability to hold daily data over a 2 year period. The importance of this is that for the first time we were able to look at trends, time-series analysis and apply regression techniques to find key drivers.
  • We were then able to recreate over 90% of the customer's balance sheet on a daily basis. This enabled significant netting opportunities to be uncovered.
  • The data required significant enhancement to enrich data for missing information from other sources as well as running allocation and optimisation algorithms.
  • We built an application to allow users to interrogate this data down to trade level.
  • Users could see what the key drivers of the balance sheet are and they could easily spot inefficient trades. Metrics analysed were IFRS Assets and Liabilities, Risk Weighted Assets (RWA), Leveraged Balance Sheet (LBS) and Bank Levy.
    Tier 1 Global Bank
    June 2017 - February 2020
    SQL Server, C#, WPF Desk Application, C# Windows Services