UBS 19th Annual Financial Services Conference: The technology frontier
Big Data in banking
UBS 19th Annual Financial Services Conference | 13 October 2016 | slide no. 2
Content
Modern Credit, Marketing, Financial Crime, CVM and Operations
Big data enablers
The evolution of Retail and Commercial Banking
Key challenges and what’s next
UBS 19th Annual Financial Services Conference | 13 October 2016 | slide no. 3
Content
Modern Credit, Marketing, Financial Crime, CVM and Operations
Big data enablers
The evolution of Retail and Commercial Banking
Key challenges and what’s next
UBS 19th Annual Financial Services Conference | 13 October 2016 | slide no. 4
The evolution of Retail/Commercial Banking
Branch Online
Product Customer
Automated Scoring
Judgmental
UBS 19th Annual Financial Services Conference | 13 October 2016 | slide no. 5
Evolution from branch to e-channels Number of interactions Millions
0
500
1 000
1 500
2 000
2 500
3 000
3 500
4 000
4 500
2010 2011 2012 2013 2014 2015 2016
Online Mobile App ATM POS Branch
UBS 19th Annual Financial Services Conference | 13 October 2016 | slide no. 6
Evolution – from judgemental to automated credit
0
10
20
30
40
50
60
70
80
90
100
1999 2001 2003 2005 2007 2009 2011 2013 2015
Judgemental decisions Accuracy
% Gini for
‘banked’ customers,
non-banked much lower
UBS 19th Annual Financial Services Conference | 13 October 2016 | slide no. 7
Evolution from product centric to customer centric view
99.5 95
Customers on CIS* (%)
Individual Juristic
200 220
260 300
375
2012 2013 2014 2015 2016
Number of customer documents on Documentum** (Million)
* Customer Information System ** Central electronic document repository
Customer information management has been
centralised outside of product systems
Customer and product records have been centralised
outside of product systems
UBS 19th Annual Financial Services Conference | 13 October 2016 | slide no. 8
Content
Modern Credit, Marketing, Financial Crime, CVM and Operations
Big Data enablers
The evolution of Retail and Commercial Banking
Key challenges and what’s next
UBS 19th Annual Financial Services Conference | 13 October 2016 | slide no. 9
Big Data enablers
Big Data
Comprehensive single view of
customer
Enterprise data warehouse/lake
Analytical capability
Operational decision making
UBS 19th Annual Financial Services Conference | 13 October 2016 | slide no. 10
Comprehensive single view of customer
FirstRand Citizen
UBS 19th Annual Financial Services Conference | 13 October 2016 | slide no. 11 First National Bank – a division of FirstRand Bank Limited. An Authorised Financial Services and Credit Provider (NCRCP20).
Risk Marketing & Sales
Financial Crime CVM Operations
Enterprise Data Warehouse/Lake
UBS 19th Annual Financial Services Conference | 13 October 2016 | slide no. 12
More sophisticated analytical capabilities required to extract value
EXPERT SYSTEMS SIMPLE STATISTICAL MODELS
MORE ADVANCED METHODS
Logistic Regressions Decision Trees Neural Networks
Random Forest
Σ
UBS 19th Annual Financial Services Conference | 13 October 2016 | slide no. 13
Operationalising decisions Br
anch
Onl
ine
App
Mob
ile
ATM
Integration layer
CIS
IDS
ILP
CAM
S Channels
Customer and Product Systems DATA STREAMS
Actions
Decisions
Estimates
“Analytics without decisions is meaningless”
Ability to deliver Big Data Analytical Decisions
into production systems is key
UBS 19th Annual Financial Services Conference | 13 October 2016 | slide no. 14
Content
Modern Credit, Marketing, Financial Crime, CVM and Operations
Big Data enablers
The evolution of Retail and Commercial Banking
Key challenges and what’s next
UBS 19th Annual Financial Services Conference | 13 October 2016 | slide no. 15
Big Data use cases
Risk Financial crime CVM Operations Marketing
and Sales
UBS 19th Annual Financial Services Conference | 13 October 2016 | slide no. 16
Big Data use cases - Risk
• Enhanced Credit Models
• Integrated view across products, channels, enablers
• Sophisticated behavioural analysis
• Augmented/Derived characteristics
• Risk appetite – optimised at every level
• ICAAP models
• Regulatory Risk/Compliance & Market Conduct
• Insurance
Risk
UBS 19th Annual Financial Services Conference | 13 October 2016 | slide no. 17
Big Data use cases – Marketing and Sales
• Facilitated by integrated view of customer across products/
accounts/ transactions/ channels and external data
• Needs-based modelling across borrow, transact, deposit and
insure
• Behaviour-based propensity modelling at product level
• Life-event and other specialized models deployed on
experimental basis
• Significant lift in customer product take-up
Marketing and Sales
UBS 19th Annual Financial Services Conference | 13 October 2016 | slide no. 18
Digital ‘Big-3’ platform instrumental in sales growth
0
500 000
1 000 000
1 500 000
2 000 000
2 500 000
FY13 FY14 FY15 FY16
Borrow
Transact
Investment
Insurance
UBS 19th Annual Financial Services Conference | 13 October 2016 | slide no. 19
Big Data use cases – Financial crime
• Money-laundering risk modelling pioneered use of Big Data
analytics platforms and combination models
• Fraud modelling combining product, channel, geography, sector
and device behavioural information
• Network analysis used for complex investigative efforts (AML and
Fraud)
Financial crime
UBS 19th Annual Financial Services Conference | 13 October 2016 | slide no. 20
Big Data use cases - CVM
• Customer profitability modelling allows customer-level
decisioning and strategy guidance across levels
• Retention modelling retains good customers
• Pricing optimisation across multiple dimensions
• Rewards designed to influence customer behaviour
CVM
UBS 19th Annual Financial Services Conference | 13 October 2016 | slide no. 21
CVM – Deep understanding of profitability
0
5
10
15
20
25
30
35
40
45
Monthly profit per risk grade
NIACC Cost of capital Tax Opex TTC imp
Risk grade Low risk High risk
Million
UBS 19th Annual Financial Services Conference | 13 October 2016 | slide no. 22
Big Data use cases - Operations
• Operational implementation has taken place to automate a wide
range of credit, marketing, sales, fraud, record keeping, validation
and other tasks
• Call-centres and collections large focus for efficiency and
effectiveness optimisation
• Major focus for channel optimisation going forward and perhaps
area of greatest future opportunity
Operations
UBS 19th Annual Financial Services Conference | 13 October 2016 | slide no. 23
Modern operations – Contactability is one of many key drivers of efficiency
Aim is to contact the right customer using the right method at the right time with the right message
NPV: TBC
Relevant Contact
Contact language/ message
Contact frequency & priority
Best Time To Contact
Contact Method CONTACTABILITY
01
02
03
04
05
UBS 19th Annual Financial Services Conference | 13 October 2016 | slide no. 24
Content
Modern Credit, Marketing, Financial Crime, CVM and Operations
Big Data enablers
The evolution of Retail and Commercial Banking
Key challenges and what’s next
UBS 19th Annual Financial Services Conference | 13 October 2016 | slide no. 25
The modern FI requires new infrastructure and skills
FROM TO
• Silo product approach and systems • Integrated customer approach and systems
• Overnight batch • Real-time streaming
• BI people, IT developers, DB administrators • Data scientists, architects, engineers
• Separate RDB and analytic platforms • Combined data-analytic platforms (e.g. Hadoop)
• Accounting orientation (historical, incurred basis, top-down)
• Analytic orientation (forward-looking, predictive, bottom-up)
• Traditional marketing • One-on-one marketing
UBS 19th Annual Financial Services Conference | 13 October 2016 | slide no. 26
What’s next
Intelligent agents - robo-advisors, service and sales agents
GAAF (Google, Apple, Amazon & Facebook) & others entering financial services
Integrated Financial services (Lend, Deposit/Invest, Transact, Insure, Telco, …)
Thank you