Case study: Banco Santander

facebooktwittergoogle_plusredditlinkedin rss

With fraud losses on the rise across the banking industry, Brazil’s third largest privately-held bank is beefing up its anti-fraud efforts by replacing manual processes with technology from SAS.

With fraud losses on the rise across the banking industry, Brazil’s third largest privately-held bank is beefing up its anti-fraud efforts by replacing manual processes with technology from SAS.

Carlos Sovegni, SASBanco Santander Brazil, the country’s third largest privately-held bank, accounts for 50 per cent of the total profit of the Santander group. It started business in 1982 with just one branch in São Paulo and since then it has not only expanded its range of services but also its branch network (which now stands at around 5000 offices across the country). Due to this development, says Carlos Sovegni, fraud specialist at SAS Brazil, ‘Banco Santander Brazil needed to completely understand the objectives of the organisation and establish the framework it needed, not only to progress further, but to protect itself from fraudulent activities.’

The bank has opted for SAS Fraud Network Analysis, to flag and address all types of fraudulent activity, SAS Visual Analytics, an application that helps banks analyse reports, providing a point-and-click environment to explore data graphically, and SAS Grid Manager, which offers a shared, managed analytics environment with workload balancing and parallel processing. The decision to choose this product set was formally agreed at the end of October 2013.

According to Sovegni, the bank’s objective is to reduce fraud in three core areas – credit cards, current accounts and loans. It also hopes to improve efficiency when creating new retail credit policies and models.

To prevent card-based fraud, Sovegni says that Brazilian banks have been migrating their credit and debit card portfolios to chip technology (EMV) since 2008. Almost 90 per cent on average have been migrated. However, the challenge now with fraud in the country, he says, is on the consumer side, mainly when a fraudster applies for credit or opens a bank account with a falsified or stolen identity. This type of activity is ‘hard to detect’, Sovegni remarks, and Banco Santander Brazil loses a ‘huge amount’ of money each year through this.

The bank currently does not have any anti-fraud tools in place to prevent this kind of fraud. Instead it relies on manual processes and spreadsheets.

There was a two-year selection process in which the bank undertook internal assessments and invited a number of market players to analyse its IT set-up and data around losses. Banco Santander Brazil narrowed down its choice to three suppliers, which were SAS, Experian and FICO. NICE Actimize was also evaluated but lost out earlier in the process. The final stage of the selection involved detailed product walkthroughs and demonstrations of the solutions on offer.

Main requirements included ‘strong analytics performance’ and that the vendor could provide ‘three or four ways to do the analytics’. The bank also wanted to reduce internal operating costs and be able to provide automatic processes across every channel. SAS’s product set ‘covered all of the bank’s requirements’, Sovegni remarks.

The implementation work started in December last year. Sovegni says the assessment of the project and processes has now concluded. The project has been split into five phases because ‘there are a huge amount of disparate processes and solutions that need to be integrated as one solution’.

The first part of the project will focus on credit cards and bank accounts. ‘We basically want to deploy the SAS tools and concentrate on the weakest channels in the first stage,’ comments Sovegni. In the second stage, the SAS team will finish the credit card environment, before pulling together all the bank account channels in phase three, he adds. The first part is expected to be completed in late June, the second by the end of this year and the third in Q1 2015.

The efforts in phases four and five will move onto consumer finance channels, such as personal loans and auto loans. These areas are provisioned to be finished during the course of 2015. A team, made up of six SAS staff and 15 people from Banco Santander Brazil, is working on the implementation. SAS’s team will provide the bank’s staff with training sessions explaining their knowledge of the systems throughout the project.

Sovegni says the data warehouse component of the SAS Fraud Network Analysis tool has been installed, so system testing is now underway. ‘With the data warehouse component we can see if our systems work and how many fraudulent risks can be flagged,’ he says.

Sovegni believes the bank will in future be able to benefit from reduced operational costs and fraud losses. Using the SAS tools Banco Santander Brazil expects to shrink fraud-related losses by around 90 to 95 per cent, he states.

{loadposition reprints}

{loadposition suppamllist}

{loadposition rribsj}

{loadposition rrbossrisk}

© Copyright IBS Intelligence, a trading name of IBS Publishing Ltd. Other than downloading a small proportion of content, strictly for your personal non-commercial use, or where this amounts to “fair dealing” no part of the Website may be reproduced, stored in a retrieval system or transmitted, in any form or by any means (electronic, mechanical, photocopying, recording and/or otherwise) without our prior written consent.