SME Conference
Lahore, Pakistan - May 2005
David Yong
GRM Consumer & Programme Lending
1

 Introduction
 Scope
 Overview major ABN AMRO SME businesses
 Business Models
 Credit Process & Product Programmes
 Portfolio MIS
 Use of scoring
 Key Challenges
2

 1948: First foreign bank in Pakistan
 1993: Expansion of branch network (7 branches in 3 cities)
 1998: Introduction LCY deposits & penetration into Affluent segment
 1999: Introduction Personal Loans & Mortgages
 2002: Launch Commercial Banking focusing on Mid-sized Corporates
 2005: Launch Credit Cards
3

(1)
 On 29 March 1824, King Willem I issued a royal decree creating the
Nederlandsche Handel-Maatschappij
(NHM) with the aim of reviving trade between the Netherlands and the
Dutch East Indies.
 In 1964, NHM merged with De Twentsche Bank to form
Algemene Bank Nederland (ABN) while Amsterdamsche
Bank and Rotterdamsche Bank joined to become
Amsterdam-Rotterdam (Amro) Bank.
 In 1991, these two banks merged as ABN AMRO.
4

Strong Balance Sheet *
Total assets
Group capital
Risk weighted assets
EUR (bln)
742.9
35.6
245.6
Global Resources *
Employees worldwide
Branches and offices
Countries/territories
97,000
3,000 +
60 +
Solid Credit Rating
Moody’s
Standard & Poor
Fitch IBCA
Long-Term
Aa3
AA-
AA-
Well- Positioned Bank **
Ranking
Total assets
- Worldwide
- European
18
11
* as at 31 March 2005 ** The Banker, July 2004
 conducts banking, fund management & insurance.
 ranks 11th in Europe and 18th in the world based on total assets.
 has over 3,000 branches in more than
60 countries, a staff of about 97,000
FTEs and total assets of EUR 742.9 billion (as at 31 March 2005).
 is listed on the Euronext and the New
York Stock Exchange.
5

 Backbone of economy
 Family owned/dependent (informal)
 Various types of legal entity
 Importance of personal relationship with client
 Quality of financials
 Understated revenues for taxation purposes
 Level of Owner’s commitment and equity

Part of (small) community network

Influential community
6

Target market definitions across our sample varied widely across markets and within markets
North
America
(10 banks)
Benelux
(3 banks)
Main SB segment definition by geography
2.3
Average
5.3
7.9
2.5
4.2
6.0
Specialist segmentations observed
Specific segments served according to website (e.g. 70% of the 23 banks had a defined
‘Agriculture’ offer)
Franchises
32%
41%
Minorities
UK
(4 banks)
0.7
1.6
2.9
57%
Europe
(12 banks)
1.0
2.9
7.5
Non-profit
Other
(3 banks)
1.2
2.8
4.3
Agriculture
70%
0% 20% 40% 60% 80% 100%
0 2 4 6
Cut-off turnover (

MM)
8
Mercer Oliver Wyman
Note: n = 32 banks
7

Players have positioned themselves as having a few
12
10
8
6
4
2
0 products in each family – ‘simple and standard’
Simplicity of external offering
(Products per family – according to web-site)
# prods
14
Observations
 Often standardised ‘core’ offer with segment variation injected via the communications mix
 However, most players interviewed said they had many more products internally
– Poor record of product management and rationalisation
 Most players applied a variety of standard customer insight tools into SB product design
– Market sizing and projected profit pools research
– Competitive research
– ‘Mystery shopping’
– Networking with local influencers or trade associations
 Best practices would be banks that conducted SB-specific quantitative research around needs More specialist products such as leasing, factoring normally outsourced to sister companies
 Only a few offered external value-added products
– e.g. advisory services
Key: Max
Avg
Min
Mercer Oliver Wyman
Families n = 18 banks
8

 Home Market BU’s (NL, US and Brazil) adopted Product
Programmes for SME’s
 Identify potential for further migration (NL & Brazil)
 Products across markets are similar and ‘traditional’
 Emerging markets: cautious start by other BU’s (India,
Indonesia, Pakistan & Taiwan)
 Knowledge-sharing across the Group
 Cut-off point program lending & non- program lending to be determined per market
 Overseeing 17 Product Programmes with ENR of EUR 2,8
Billion, with transition portfolios
9

Total Portfolio (December 2004)
 Total portfolio: EUR 2,8
Billion (Dec 2004)
 Largest portfolio: BU US with USD 1,3 Billion
 Portfolio consist of traditional and simple products (4Q2004: 84%)
 Rapid growth in NL
(organic) and Brazil
(organic + acquisition)
 Transition portfolio BU
NL earmarked >EUR 2,2
Billion
 Cautious start in India &
Indonesia
 Development of plans for VGPB Business
Owners in NGM Asia
3500
3000
2500
2000
1500
1000
500
0
Br azil
37%
SME Programme Lending Portfolio
(December 2004)
India
1%
US
36%
SME Program m e Lending Portfolio
(Dec 01 - Dec 04)
Dec 2001 Dec 2002
Indonesia
1%
Dec 2003
NL
25%
Dec 2004
NL
US
Brazil
India
Indonesia
NL
US
Brazil
India
Indonesia
Total
10


– large client base
– traditional domestic standardised products
– relatively larger distribution network

– start from (‘almost’) zero
– competitive edge
– limited distribution network
– limited reliable data for credit assessment
11

Return to Local Intimacy
 Aus/NZ Bank had lost Small Business (SB) share after attempting to build low-cost remote channel model
 Re-launched a named local point of contact (RM or branch manager, direct dial number) with letter campaign and outbound follow-up
 Saw +50% increase in customer satisfaction, 4-5 points of share gain and now back in top 2 share position, in less than 18 months
Information-based Strategy
 Capital One entered SB market in ’99-’00 applying IBS techniques
 Combined marketing and risk analytics, with predictive modelling to target asset-based products with a direct marketing model
 39% 3-yr CAGR in asset growth, now #2 in SB cards and #3 in SBA loan origination
Exploit Retail Branch Coverage
 HBOS’s merger had given it a strong SB business with the BoS franchise but locked into small branch footprint
 Rapidly hired network of RMs to tackle the £1-10MM turnover segment – ‘hunter’ force to source acquisitions – and doubled UK share from 3% to 6%
 Now re-branding 400 branches (out of 800) in England in 2005 and creating in-branch SB specialists to exploit full retail presence and go after £0-1MM businesses
Mercer Oliver Wyman
Multiple Business Model Focus
 US bank had reduced branch network and lapsed in service quality – suffering SB attrition rates of 30%
 Segmented business into 3 distinct propositions: national direct, ‘affluent’/upper-end SB and branchbased micro-SB, combining personal/business accounts
 Expanded SB franchise beyond branch footprint, now in Top 3 in US SB lenders. Grew profits at 30% CAGR over 5 years, vs. bank growth of 20% CAGR
12

Transaction characteristics
Low volume
Large heterogeneous
Approval process Deal-based
Credit data Audited financial statements
Handling and checking
Hierarchy of approval
Commercial Loans Consumer & SME Loans
High volume
Small heterogeneous
Volume-based
Limited consumer financial data
Account management
Review loan-by-loan
Portfolio management by statistics
Senior managers approve large/complex loans
Senior managers approve credit product programmes
13

Traditional dealbased underwriting don’t work in high volume lending
The Boss
Planning Underwriting Operations Collections
14

Portfolio Management is in the core of Risk Management function
Product Planning
Dynamic
Write-Off
Collections
Management Information Systems
Portfolio Management
Risk Management
Credit Acquisition
Interactive
Account Maintenance
15

 Standardised and ‘simple’ products
 Critical mass/scale
 Risk – reward balance
 Predictability
 Factory - style
 Management by exceptions
Portfolio Management done through approved Product Programmes
16

 A standardized set of rules of credit extension for a group of customer with similar characteristics or product needs
 A sign-off on a Product Program is considered an approval of the complete risk/reward characteristics of the product and the credit cycle process
 Demonstrates that portfolio performance will be predictable in terms of revenues, delinquencies, and losses
17

 Product Description
 Target Market
 Economic and Competitive Environment
 Eligibility Criteria (Terms and Conditions)
 Account Initiation
 Account Maintenance
 Collection and Write-Off Policy
 Treasury, Funding, and Pricing Considerations
 Support Systems and MIS reports
 Product Profitability and Stress Testing
The Product Programme process is strongly supported through MIS
18

 Continuously process and transforms data into information, which will be used in the decision making process directed towards optimizing results:
Risk-Return Trade-Off
 Serves as a base for effective credit cycle and portfolio management: controlling risks and assessing opportunities
19

Brazil - Total
Brazil - Overdrafts
Brazil - Loans
Brazil - Leasing
Brazil - Mortgages
Brazil - Rewrites
754.322
5,60% 3,78% 22,49%
74.593
9,10% 9,33% 81,52%
591.027
2,90% 2,07% 16,72%
45.751
0,80% 0,28% 5,08%
8.707
12,40% 1,38% 7,81%
33.975
49,90% 23,38% 13,91%
20

• An incentive to place as much of the SME portfolio into a standardised Program-Lending format, as is practical & prudent. (GRM limit EUR 5MM)
• Whilst the advanced approach has distinct advantages, the standardised approach is also beneficial with risk weights at 75%.
• Both Capital relief and cost efficiencies can be expected from standardisation.
• As more SME portfolios become standardised to take advantage of Basel II, the product will in turn become easier to securitize, so the product will become more attractive to banks, who may choose to keep it on or off balance sheet based on capital need.
21

 Sponsorship of Senior Management and Line support
 Available credit data to assess clients and prospects
 Human resource and Organization issues
– Provide adequate and on-going training in credit cycle management and program lending
– Ensure an organization capable of meeting the need for greater coordination and teamwork
 Redesign of business process
– Identify a project team of individuals who represent functional areas of the credit cycle
– Be ready to supplement internal resources with external consultants
22

 Sufficient IT systems and data base resource
– Resolve gaps on data source, data quality, data integration and data usage
– Ensure an automated front-end Loan Origination System (LOS) and
Collection System in place

Effective and consistent MIS reporting
– Establish one single standard or common language reporting system
– Provide end-to-end MIS consistent with Product Program definition
– Ensure that the MIS has the ability to “peel the onion” or disaggregate performance data at various level of detail
23

 Implementation of scoring methods
– Scoring systems make sense only for high volume business
– Application Scoring
• increase underwriting process efficiency
• improve portfolio quality thru statistical control
– Behaviour Scoring
• increase profitable customer relationship
24

Super
Bad
Scorecard Maximum Benefit
Super
Good
25

 Internal Application Risk Scorecard
– Based on applicants past performance
– New accounts
 Internal Behaviour Risk Scorecard (Behaviour and
Collections Scoring)
– Existing Business
– Based on account holders past performance
 External Credit Bureau Scorecard
– Based on account holders external credit activity
26

40%
35%
30%
25%
30%
 Rank All Applications By Default Risk

Set Cutoff To Achieve Approval Objectives
Default Rate
Reject High Risk Apps
To Lower Bad Debt
Simplified Underwriting
On Low Risk Applications
20%
15%
15%
Manual Review
Marginal Cases
8%
10%
5%
0%
Low Score
6%
5%
4%
3%
1% 0,5%
0,2%
High Score
27

 Senior Management commitment
 Skill-set & human resources
 Overhaul of existing credit process
 Investments in IT
 Commence Data collection
 Rationalization of Product Management
 Alignment of service concept to cost/benefit
 Availability of credit data
28

david.yong@nl.abnamro.com
29