underpin the franchise growth strategy, which was to recruit cheque account customers,
and then nurture this base to cross sell other products. The underlying assumption for
using these systems was that customers hold a full financial relationship with the bank
and that the internal data collected, along with any external information, was sufficient
to establish the customer’s financial profile. As a consequence, current and historic data
is, therefore, collected on a monthly cycle, to provide a snapshot view of the customer’s
financial situation, accurate at the time it is prepared. This data consists of internal data i.e.
information derived from the Bank’s own account records, and external data i.e. information
recorded at Credit Reference Agencies, which is consolidated to create a single set of
customer information. This consolidated data forms the basis of a monthly credit behaviour
score, which is refreshed on a monthly cycle, indicating whether a customer is potentially
eligible for credit products in the future, based on their previous performance.
However, in today’s environment, some of the key assumptions that underpinned this
traditional system have changed:
• Customers now tend to hold multi-product, multi-banking relationships and internal
data is not enough to provide a full customer risk profile.
• Credit Risk is not the only measure that a bank should use to determine appropriateness
of lending.
• Financial transactions occur and are recorded on a daily basis but are not factored into
the monthly batch credit risk cycle.
In the current economic environment, a bank needs to adopt a credit decision making
system that is able to provide effective credit decisions for existing franchise customers with
‘primary’ and ‘secondary’ relationships, as well as non-franchise customers. The key areas
where banks could improve their infrastructure are:
• Positive External Data: banks in many countries with advanced financial services markets
(e.g. the UK) are now sharing full data as part of information exchange reciprocity
process. This data will help to strengthen the credit risk scoring models currently used to
make both account management and account acquisition decisions.
• The information relating to customers’ balances and income available through Credit
Reference Agencies can better inform the bank of each customer’s affordability. This will
help to determine shadow limits that are accurate and in line with the customers ability
to repay.
• The availability of data will help to considerably strengthen the internal overindebtedness
segmentation that currently determines the customers’ indebtedness
levels, based on internal transactions and coupled with information gathered from
Direct Debit and Standing Order data.
• The data will further provide the means to evaluate some generic tools that are
available for purchase from the Credit Reference Agencies. In particular, the Consumer
Indebtedness Index (CII) may help to aid the identification of over-indebted customers
that do not have an existing banking relationship with the bank. Also available from
3rd party vendors, such as Experian, is the Affordability Index (AI) that may provide an
alternative to the current method of determining shadow limits. These tools are only
available for suppliers of positive data.
• Collection of Point-of-Sale characteristics: the credit decision making infrastructure
in most banks should be expanded. This would enable the collection of appropriate
information at the time of sales interview for specific segments of the customer base.
This process should be integrated seamlessly with current application processing, to
minimise its impact on the sales process, controlled robustly to ensure compliance, and,
where possible, corroborated with internal information or external bureau data to verify
the information. This should be automated so that all the information captured flows
through into the existing databases and is available for monitoring and management
information.
• Transaction scoring: Current system limitations mean that the data used for credit
scoring for franchised customers is updated on a monthly basis. This may mean that a
customer decision can be up to a month out of date. In an environment where millions
of financial transactions are occurring on a daily basis, banks must seek to have a scoring
system that refreshes on a more frequent basis and is able to use relevant bureau,
transaction and customer information. This can be facilitated by adopting the most up
to date front-end processing system which provides:
• Full strategy control in the hands of the business user
• Ultimate flexibility in designing and testing strategies
• Multi-user and change management
• Simulation and analytics
• A controlled number of new variables, generated ‘on-the-fly’ without the need for IT
involvement
• Further to this, the banks should invest in the latest releases of the revolving account
management systems which provide many enhancements. These can help to drive
further improvements in marketing response rates, account acquisition, revenue, cost
control and profit management through cross-selling, customer and debt management.
The latest versions of Fair Isaac’s TRIAD system allow management access, both at the
account and at the customer level, from a single platform, to have decision strategies
that initiate other strategies, to configure new and unique decision areas e.g. for early
warning, and removes restrictions on decision keys and scores. They also come with
improved data analysis, ‘what if’ estimators and reporting. This will further improve the
quality of decisions adding to the virtuous circle of retention, revenue and profit for the
best customers and risk management and cost control for others.
The three pillars of Basel II are each based on a foundation of data. These are: risk weighted
assets and capital requirements, supervisory processes and market discipline and disclosure.
All three pillars stand or fall in relation to the way that the bank utilises data with respect
to speed, quality, accuracy, completeness and transparency. Retail banks need to expand
and integrate their current data management systems to meet the demands of Basel II and
provide a group-level credit risk management system. In particular, data needs to be:
• Scalable: The complexity and granularity of account level data needs to be stored and
accessed in scalable databases.
• Available: Databases and mining tools need to be reliable, complete and accurate.
• Secure: Utilising data to the degree of detail and granularity required by Basel II requires
a high level of integrity and confidentiality, not least because data will be consolidated
from multiple locations and subsidiaries.
This kind of model allows banks to capture and store data in a secure environment. It also
allows for the business to use information across the whole organization that is consistent
across all functions, including credit and finance, and allows for risk management to be
integral to everything the entire organization does. Not only will this solution meet the
requirements of Basel II, but it will also provide the architecture to achieve considerable
business leverage. There will only be ‘a single version of truth’, supported by standardized
data management process, and available through an infrastructure that provides the means
to interrogate mass account level data consistently and quickly.
Integration
Figure 12 below represents a ‘blueprint’ for bringing together all the key technology
components of the customer management system and the credit risk management system.
This will create an integrated enterprise IT environment for managing risk, compliance and
performance within a retail bank.
