Experian, the global information solutions company, has released the latest version of Tallyman, its specialist debt management system, to help its customers tackle the problems caused by growing numbers of consumers encountering financial difficulties as economic conditions become more difficult.

Tallyman, part of the suite of products offered by Experian's Decision Analytics business, is used to manage customers who have fallen into arrears with their payments by organisations that grant credit to consumers, including banks, building societies, loan providers, credit card companies and debt collection agencies, as well as organisations that provide consumer credit as part of their business model, such as retailers, utilities and communications companies.

New features for Tallyman 3.4 include:

Brian Dewis, Managing Director of Experian's Tallyman business, said: “We are continuously improving Tallyman in response to the needs and demands of existing customers and prospects. The Tallyman solution meets the challenge of ever increasing levels of consumer debt, as our growing customer base attests”.

Companies using Tallyman include British Gas, Orange , Vodafone, ABN AMRO, Severn Trent Water, Welsh Water, EDF Energy, Scottish Power and npower.

	Experian Decision Analytics, part of Experian, the global information solutions company, has signed an agreement with Garanti Bank, the third largest private bank in Turkey, to supply a set of solutions that will enable the bank to automate its new customer acquisition process.

With these solutions Garanti Bank will be able to accurately define and control its credit risk management strategy from the point of application. Garanti Bank will use the Experian solutions to accurately assess the financial risk of SMEs in its portfolio and make fast, consistent decisions about their potential value as customers, By
strengthening its current application processing systems for credit cards, consumer loans and overdraft, the bank will improve its decisioning process across the different portfolios and have the speed and flexibility to respond to applications within seconds.

The solutions will initially be implemented in Turkey and Romania, followed by a second phase rollout to other countries in Eastern and Central Europe. The simulation capability built into Experian Decision Analytics’ solutions will enable Garanti Bank to evaluate business scenarios and calculate the impact of potential changes to its policies, as well as assess the effectiveness of its lending strategies. This will help the company to quickly modify and adapt its strategies when required.
The solutions to be implemented by Garanti Bank will collate the different sources of information within the business to provide high level reports with summarised statistics and indices, and also highlight divergences from its business goals, with detailed analysis down to individual record level if necessary.

Ali Temel, Executive Vice-President of Garanti Bank, commented: “We have chosen Experian Decision Analytics as it is a proven provider of decisioning technology and its solutions’ functionality, ease of use and advanced analytics capabilities will ensure that we deliver consistent decisions across our business. Moreover, its reporting modules will enable us to understand the impact that our business strategies and processes have on profitability and forecast this into the future."

Mr. Hüsnü Erel, Executive Vice-President of Garanti Bank, added: “Having worked with Experian Decision Analytics’ local team in Turkey for a number of years, we are aware of its strong experience and commitment to our marketplace, which made them our preferred choice.”

Cenker Ozhelvaci, Country Manager for Turkey & Middle East, Experian Decision Analytics, said: “We have been working with Garanti Bank since 1999 and we look forward to continuing to contribute our business expertise and technical support to assist further developments with Garanti Bank and the entire Turkish credit market.”

This article introduces the first of a series of articles that will outline the basic concepts of scoring and will focuses on the different types of scoring models. Future topics will include, Masterfile Analysis, building and testing the scorecard, reject inference and deliverables. What is a scorecard? A table listing the characteristics which provide the most predictive information, together with the associated attributes and points.

Types of scorecards:

•  Application Scorecard

A statistical technique to predict, at the time of application, the probability of repayment.

This is the most obvious use for credit scoring which is used to assess the risk of customers when they apply for credit. The lower the score, the higher the risk of the customer, ie. the higher the chance that the loan will not be repaid.

Application scorecards are used by establishing a cut-off score, below which applicants are deemed to be too high a risk and hence are rejected. The score represents a trade-off between the volume of accounts which will be accepted and the ultimate bad rate (proportion of customers who will not repay). When implementing a new application scorecard, the cut-off is chosen either to maintain current acceptance rates whilst reducing bad rates, or to maintain current bad rates whilst increasing the volume of accounts accepted, or somewhere between the two.

In addition to making straight accept/reject decisions, the score may be used to rank applicants according to their risk and differing policy rules for each score band established. For example, some low scoring applicants for a current bank account may be offered restricted facilities, such as only being allowed a cheque-book, rather than a credit card.

•  Behavioural Scoring

A statistical technique to provide a complete evaluation of an existing customer's past performance which can then be used to predict the future performance.

The objective of behavioural scoring is to measure the ongoing risk of a customer. Behavioural scoring can be applied to the company's accounts each month, as part of the normal billing cycle during which time each account is examined, so that any changes in the risk position of any given account can be detected.

2.1 Collections

Once the customer has failed to meet one contractual payment then he is in breach of contract and the account passes into the collections department. Although collection strategies vary between organisations, the primary emphasis is always on the continuing relationship with the customer.

The first step following the signs of early delinquency should be the customer service approach where the willingness and ability to pay is determined.

Behavioural scoring is fundamental in the collections area of credit scoring. Previous research proves that we are creatures of habit, and are likely to pay on the same day of each month over long periods of time. The advantage of a behavioural score for collections is that it enables the organisation to approach customers as soon as they are seen to be a risk and to prioritise collection activity.

Customers that always pay late may not be judged to be a risk until later than those customers who habitually pay on time. This is due to the fact that they habitually pay late and are therefore not immediately considered a risk every month on paying late, whereas a customer who normally pays on time might fall into a riskier category when he suddenly starts paying later.

The idea of scoring in collections is to reduce the number of early reminders with no increase in the rate of delinquency. It follows therefore, that the number of customers suffering the aggravation and attrition as a result of early warning letters decreases as well.

•  Authorisation

Instant authorisations - required when a customer wishes to make a purchase, causing them to exceed their current limit.

Authorisation scores provide access to the relevant personal data, including the spend and payment data on the account right down to the transaction level and going back over a period of years. If the card forms part of further banking facilities under the customers name, it then becomes necessary to summarise this data. In addition, the customer is always fully aware of his personal status and is rarely forgiving if the entire relationship with the lender is not considered.

Through the use of behavioural models, the risk of the customer can be ascertained, and the leeway set accordingly. A high risk customer will be given little grace whilst a low risk customer will be given more credit than the level originally set.

2.3 Credit Limits

The higher an organisation sets its credit limits, the more the risk of delinquency. However, where an organisation is able to do this, the risk of delinquency should not increase more than profitability. Where a good balance is found, the organisation stands to be extremely profitable. Hence the importance of monitoring all accounts, as this provides the organisation with invaluable information on its lending policies and its customers.

Credit limits can be set by applying a fixed level to all new accounts, or by rating the credit risk of the applicant and setting the credit limit in accordance to the score achieved. The limit may then be altered several months later, once the customer's payment behaviour has been assessed.

Credit modifications might be made depending on the current balance of the account, the fraction of the current account limit that has been used, the utilisation level of the limit ever used, and the time since the last change.

•  Response Scoring

Response scoring is a ranking mechanism which predicts the likelihood of a person responding to an offer via mail. This system is based upon analysis of statistically valid samples of those who have and have not responded to the same offer.

Response scoring should therefore be used in conjunction with risk scoring to predict both the likelihood of customers responding favourably to mailings, and the likelihood of them defaulting on a subsequent advance. ie. a response model should be used in the first instance, in order to identify those most likely to respond, and in the second instance, this group should be scored with a credit risk model so as to remove those who are seen to be the highest risk. This then leaves a target market who, whilst being credit worthy, are also likely to respond.

•  Litigation Scoring

It is costly to pursue all bad accounts, especially if the likelihood of them being able to repay their debts at the end of legal action is low.

Litigation scoring assists the organisation in identifying those accounts in a portfolio which are more likely to repay an outstanding balance on receiving legal demand for payment, and prior to entering judgement.

For example:

•  A fast track of customers directly to judgement stage, thereby reducing the number of letters and telephone calls.

•  Early identification of cases where litigation is inappropriate.

•  Churn/Attrition

Churn is the level of customers either defecting to competitors or closing accounts with an organisation voluntarily. This is particularly expensive where commissions have been paid to dealer intermediaries or brokers in the initial acquisition of the business and where no claw-back arrangements exist.

One is able to determine both at application point and later on, more effectively, through behavioural scoring, the likelihood that a customer will churn. Certain characteristics will be predictive in this and the applicant can then be scored accordingly.

Read next months Credinews for further information on Scorecards and the Masterfile Analysis.

Written by Ben Maseko, Scoring Analyst, Experian Decision Analytics.

Bridging the digital divide through credit management in South Africa

	Fraud is a growing problem with investment in fraud systems on the up and yet few realise the true role that collections systems can play in reducing fraud. True risk accuracy is determined by combining the rich real time transactional and behavioural data of a fraud management system with that held in the collections system.

It is estimated that as much as 50% to 70% of bad debt is fraud.1 A figure that indicates a significant increase in fraudulent activity, which is having a considerable impact across the telecom and financial sectors.

In the credit card industry, where initiatives such as the introduction of EMV chip and Pin has had a considerable impact in reducing Point of Sale (POS) fraud, fraudsters are now shifting their focus on to other types of fraud.

The telecom industry is also under pressure. Although telcos are often successful in tackling technical fraud, subscription fraud is on the rise and the proliferation of new IP based services are creating more fraudulent opportunities. Recent research by Telecom group Analysys estimates that as much as 2.9% of revenue is currently being lost to fraud and predict that such fraud will rise significantly as 3G services are rolled out to the mass market.

To combat this growing trend, telcos and banks have been heavily investing in fraud and risk management systems based on real time transactional and behavioural analysis, and are beginning to see returns on their investment. However, recent reports show that in many cases fraud is still not picked up until late in the collections process.

Thus, it is often best practice collections systems - through the segmentation and profiling of accounts for levels of risk - that manage to identify fraud at a relatively early stage in the arrears management cycle.

This white paper was written by Martin Aldridge, Business Consultancy Manager, Tallyman. To request the full white paper, click here

Thousands are ‘debt hospital cases' 

Posted on the Businessday website on 10 October 2007 CAPE TOWN — There are more than 300000 people in SA who are so indebted as to constitute “hospital cases” who require external help, says Gabriel Davel , chief executive of the national credit regulator .

The monthly financial commitments of these people exceeded their income to such an extent that some arrangement had to be made with creditors such as banks and retailers to rescue them, he said yesterday.

Davel told Parliament's trade and industry committee that this group needed debt counselling but argued that the implementation of this section of the National Credit Act remained a big challenge for the regulator.

He said that there were a million more people who were “debt-stressed” but would be able to address their problems if they exercised self-discipline. However, he said, this group represented only about 5% of all those with some form of debt.

The most chronically indebted — who numbered between 50000 and 100000 — were faced with a net income three to five times lower than their monthly financial commitments.

Democratic Alliance trade and industry spokesman Pierre Rabie described this group as “basically bankrupt”. Davel said the fees for debt counsellors needed to be regulated if more people were to be encouraged to enter the sector as the fees now were too low.

Debt counsellors were able to get court-approved schemes for debt restructuring with the agreement of creditors.

Davel said there had been a “huge, huge” increase of R391bn, or 135%, to R681bn in household credit by banks in the past four years.

However, he did not believe there was a “systemic problem”. The bulk of this growth (R270bn) was in mortgages. Credit card debt had grown from R13bn to R37,5bn in the period.

“Consumer credit extension grew faster than household consumption, gross domestic product or growth in formal employment,” he said.

Important structural changes had occurred in the consumer credit market as the credit criteria used by banks had become more inclusive and the lowering of the income thresholds for credit cards had provided low-income individuals with cheaper sources of finance than micro-loans or furniture finance.

Davel said the recent reduction in credit extension was not so much due to the implementation of the act as to monetary and fiscal policy which, during the past two years, had been geared to achieving this result.

The act also had a significant effect on motor vehicle financing .

He estimated that about 6-million consumers had benefited from the full or partial cleansing of data from their credit records held by credit bureaus as required by the act. Of these, 200000 had received clean reports.

The industry had reported the cleansing exercise had not undermined its ability to undertake credit assessments of clients.

The following career opportunities are available at Experian: Head of Scoring - This position requires strong analytical skills in the realms of Credit Scoring and a participative leader to achieve quality, service and productivity levels in order to maintain excellent delivery standards and client satisfaction. You will be expected to manage a team of Scoring Analysts and assist in the consultancy, development and monitoring of scoring models. Click here for more information.

Head of Business Change Enablement - This person will report to the MD: Bureau and is responsible to devise and implement a strategy to co-ordinate operational roles within Bureau and to ensure Experian Bureau's ability to deliver products to a consistently high service standard by: Selecting a standards and quality assurance methodology which will enable delivery and service excellence; implementing and monitoring the standards and quality assurance methodology; ensuring business requirements are documented and presented clearly; management of IT testing functions; quality assurance of Experian Bureau processes including but not limited to Experian Bureau Contact Centre, Experian Bureau administration; Experian Bureau IT Bridge and Experian Bureau support email boxes. Click here for more information.

Business Analyst - Two positions available, one for Experian Decision Analytics and one for Experian Credit Services. The applicant is expected to respond to the client's business needs in the production/customization of software to the required level of quality. The applicant is also required to undertake and oversee thorough system testing to ensure that the end product meets the agreed specification. Click here for more information.