Customer behaviour is changing in retail banking as customers switch their current accounts, mortgages and credit cards more frequently.
Savvy customers are now more likely to shop around for the best rates, meaning teaser rates no longer necessarily lead to standard rates and predictable income recognition.As a result, retail banks and building societies are required to be more dynamic in recording their income, so changes can happen daily rather than monthly.
This places greater demand on the methods of recording data and requires greater accuracy by staff at the bank, especially under the new regulatory requirements under Senior Managers & Certification Regime (SM&CR). The old spreadsheet-based ways of recording income recognition are becoming insufficient for the task.
Saffron Walden Building Society’s Maurice Mills (MM) and David Webber (DW) from its platform provider Whistlebrook discuss the practical issues in greater detail.
What were the financial instrument complexities that the Saffron Walden Building Society wanted to address?
MM: When customers switch from a teaser rate to an alternative product, considerable rigour needs to be applied to the recognition of interest income and the crystallisation of fee costs and income. Whereas two years ago it may have been sufficient for us to consider customers behaviour in reversion in the context of months, this has changed as market sophistication has evolved. Brokers are recommending product switches to their clients very quickly after the end of a special terms period on a loan and our customers themselves are using comparison sites and other market information to switch to the best deal for them.
What are other organisations experiencing in regard to their spreadsheet models?
DW: We have seen a marked increase in concern about the operational risk inherent in using spreadsheets. In addition, and in part driven by the Senior Managers & Certification Regime Audit Committee, Chairs were challenging the C-Level executives as to their confidence over the judgments and control of this material and highly judgmental area.
Effective interest rate calculations have long been spreadsheet-based for many banks, but the PRA’s 2018 Dear CEO letter brought attention to the challenges around income recognition of interest on fixed rate mortgages and the accounting for possible early redemption charges. Scrutiny of the judgments and calculations related to interest income, fee income and fee cost recognition by internal and external auditors as stepped up on response.
What were the concerns in relation to spreadsheet-based models?
MM: Even in organisations of scale typically spreadsheets are created by one individual and only that individual truly understands the spreadsheet. Very often documentation of the spreadsheet is scant and an audit trail of changes to the spreadsheet non-existent. The Society decided that that if we would not consider it best practice to carry out our monthly management accounting on a spreadsheet that it was logical to reach that same conclusion for our income recognition calculations.
DW: KPMG published a survey that reported 70% of spreadsheets for financial modelling contained errors that were material to results and 75% of models had no formal quality assurance processes of any kind. Even for the smallest lender the deferred interest income and fee income asset can be highly material. There is now a consciousness across Boards that such a material asset should have as much process and control rigour as would be in place for other material assets.
MM: As an organisation grows, year on year new product types with new features and terms are adding to the size and complexity of a spreadsheet platform. However at Saffron our other concern was that with the PRA’s expectation that EIR income recognition be given a separate focus within our Internal Capital Adequacy Assessment Plan; would a spreadsheet based approach allow us to test different scenarios with the rigour that we required and how long would it take to run scenarios in a spreadsheet model?
DW: With input from the Saffron we added significant granularity to our platform to enable customer behaviour to be modelled in days rather than months. We also added a broader range of configuration for the modelling of fee income and cost recognition particularly on product switches, even before a teaser rate had come to a conclusion, and when considering possible charges on early redemption.
Any last thoughts?
DW: Whistlebrook believes that whether you are providing regulatory returns to the FCA, Bank of England or PRA, or a monthly Asset-Liability Committee (ALCO) report or statutory reporting that an organisation is applying a different lens to the same data set. The regulators expect organisations to be able to exercise consistent control and insight over their business and spreadsheets alone are not sufficient to achieve this.
About the authors
Brian Cantwell, Commissioning Editor, ICAEW Financial Services Faculty and Philippa Kelly, Head of ICAEW Financial Services