2014 AN EVEREST GROUP REPORT

Analytics in Banking
Eric Simonson, Managing Partner
Conquering the Challenges Posed By Data Anupam Jain, Practice Director

Integration, Technology Infrastructure, and Right

Copyright 2014, Everest Global, Inc. All rights reserved.
Talent to Operationalize Analytics in Banking

This report has been licensed for exclusive use and distribution by Genpact

research.everestgrp.com

EGR-2014-11-V-1189
 ANALYTICS IN BANKING

Executive Summary

Analytics is helping the banking industry become smarter in managing the

Benefits
Better understanding of
myriad challenges it faces. While basic reporting and descriptive analytics
consumer behaviour, and continues to be a must-have for banks, advanced predictive and prescriptive
responding to changes in analytics are now starting to generate powerful insights, resulting in significant
consumer tastes quicker business impact.
Meeting regulatory
requirements
Improved product design and
Advanced analytics-backed solutions are enabling banks to not only manage
overall product portfolio the increasing cost of compliance, but also the risk (both monetary and
optimization reputational) of non-compliance. Product and portfolio optimization modeling
is helping banks achieve profitable growth in an environment with significant
volatility across asset classes and rising losses in traditional banking products.
Sophisticated risk modeling presents a powerful way to understand short- and
long-term profitability and capital adequacy or, in other words, chances of
banks survival in future. Fraud and AML/KYC analytics are helping banks stay
ahead of fraudsters, drug cartels, terrorists, organized mafia, and others in
Keys to operationalize analytics preventing money laundering and associated potential losses. Consumer
in banking behavior and marketing analytics are driving sustainable competitive advantage
Setting up of required data
and technology infrastructure
in an era with eroding product differentiation, waning customer loyalty, and
Breaking down internal silos exploding volume, velocity, and variety of data.
Internalizing analytics-driven
decision making As a result, adoption of third-party analytics business services in banking is
Leveraging third-party
growing rapidly and is expected to quadruple by 2020.
providers

However, the overall penetration of analytics services in banking is still in the

nascent stage and penetration levels are in low single digits. This implies
significant untapped value creation potential. Multiple competing priorities,
functional silos, talent crunch, and inadequate data and systems infrastructure
are key challenges in effectively operationalizing analytics in financial
Fast reference institutions.

Executive summary..................2 This paper describes the art of the possible in analytics, and within the context
Key drivers..............................3 of how it adds value to the banking industry. The paper focuses on:
Banking industry challenges and opportunities where analytics can play a
Impact of analytics..................5 role
Operationalizing analytics.......8
Range of analytics leveraged in banking and examples of how analytics
creates value for business
Conclusion............................12 Critical challenges and emerging best practices in operationalizing analytics
in banking

research.everestgrp.com 2

EGR-2014-11-V-1189
 ANALYTICS IN BANKING

Banking Industry Challenges Requiring Greater Insight

The economic crisis of 2008 changed the face of the banking industry.
Regulatory oversight expanded dramatically, increasing the cost of compliance as
well as the risk of non-compliance. Achieving profitable growth, while ensuring
long-term solvency, became challenging with greater volatility across most asset
classes and traditional products losing money. Managing enterprise risk, as well
as the increasing incidence of fraud, became a strategic priority. Advancements
in technology are significantly improving the speed-to-market, thereby eroding
product differentiation and customer loyalty. Analytics is helping banks become
smarter in managing these challenges (see Exhibit 1).

EXHIBIT 1 Eroding product Evolving and more

differentiation and customer stringent regulatory
loyalty environment
Drivers of analytics in banking Explosion in volume,
Competitive Compliance
Increasing cost of
advantage
velocity, and variety of data compliance
Faster response time to Significant risk of non-
Drivers of
changing macroeconomic compliance
analytics in
variables
banking

Profitable growth
and solvency

Greater volatility across asset classes

Traditional retail banking products losing money
Increasing incidence of fraud
Integrated risk management at enterprise level

1. Rising cost of compliance and risk of non-compliance

The aftermath of 2008 financial The financial crisis of 2008 exposed the inter-linkages between credit risk,
crisis market risk, and liquidity. This unleashed a wave of newer and stricter regulations
From January 1, 2008 to such as Basel II/III, Dodd-Frank Wall Street Reform, Consumer Protection Act,
April 15, 2011, the FDIC closed Credit Card Accountability, Responsibility, and Disclosure Act, and the Durbin
356 banks that failed to manage Amendment. Central banks are acknowledging the fact that some financial
the risks building up in their institutions may be a systemic risk and are demanding greater say and
residential and commercial transparency in adherence to risk norms. Several state and federal government
mortgage exposures consumer protection laws (such as servicer consent orders and servicer alignment
initiatives, amongst others) have also been passed.

New regulatory demands for managing AML/KYC1 and fraud have also emerged
including FATCA2, FCPA3, FINRA4 rules, BSA5/AML amendments, MiFID6, global
PEP7 lists, and several others. These regulations continue to be refined, changed,
and made more stringent. Consequently, compliance budgets have increased
significantly over the last few years. Yet, most banks feel their compliance

1 AML Anti-Money Laundering; KYC Know Your Customer

2 FATCA Foreign Account Tax Compliance
3 FCPA Foreign Corrupt Practices Act
4 FINRA Financial Industry Regulatory Authority
5 BSA Bank Secrecy Act
6 MiFID Markets in Financial Instruments Directive
7 PEP Politically Exposed Person

research.everestgrp.com 3

EGR-2014-11-V-1189
 ANALYTICS IN BANKING

departments are under-staffed and over-burdened, and the risk of non-

compliance is greater than ever before. Penalties for non-compliance have
risen in a tough and uncertain economic environment where funds are limited.

2. Profitable growth and long-term solvency challenges

Declining profitability, rising Acquiring and retaining profitable customers is more challenging than ever in
losses todays hyper-competitive financial services market. Traditional banking
On an average, each
products (such as checking accounts) are losing money as a result of changing
customer preferences. Incidences of fraud and money laundering are also
checking account in the
increasing. The mortgage meltdown over the past decade has also shown that
United States is losing
most financial asset classes can be volatile; hence managing market volatility,
US$196.5 (2012)
credit risks, and liquidity risks becomes imperative for financial institutions.
In the year 2012, fraud
losses on cards in the United Historically, all these risks were managed by individual Lines of Businesses
States reached US$5.33 (LoBs) somewhat separately. However, post the economic crisis of 2008, the
billion, up 14.5% need for integrated risk management at the enterprise level has increased
significantly to understand short-term and long-term profitability and capital
adequacy or chances of a banks future survival. By quickly determining
exposure, portfolio value at risk, and liquidity coverage, a bank can determine
products to take to market, or markets to exit, much faster. It can also fine-tune
responses to changes in interest rates, exchange rates, and counterparty risk to
ensure profitability and long-term solvency.

Effective risk management can also lead to significant business advantage. For
instance, the outputs from credit risk models help banks in risk-based pricing,
exposure and concentration limits setting, Risk Adjusted Return on Capital
(RAROC), managing portfolio-return profile, setting loss reserves, and
economic capital calculation. The stress testing requirements, mandated by the
U.S. Federal Reserve led to design and implementation of such models, and
now these models are being utilized as potential sources of input for designing
features of new and existing products.

3. Creating a sustainable competitive advantage

Advancements in technology are significantly improving speed-to-market,

thereby eroding product differentiation and customer loyalty. Customer
demographics, buying behavior, and needs are also changing and evolving.
Today, banks need a 360-degree view of each customer to target the right
products, cross- and up-sell, and adapt to customers changing needs.

Banks generate huge volumes of internal data (customer accounts, credit

The data explosion
scoring, payments, assets, etc.) and now need to understand its linkages to

external data (interest rates, macroeconomic variables, and customer
By 2020, the worlds computers preferences). The velocity of this data creation is also increasing exponentially.
are expected to hold 35 This is compounded by the variety of non-traditional or digital touch-points that
zettabytes (1021) of data. have emerged ATMs, Internet, IVR systems, social media, and mobile, among
IBM Corp. others. The explosion in volume, velocity, and variety of data is also forcing
banks to leverage advanced analytics to make sense of the huge and complex
information sources, and make near real-time decisions to stay competitive.

research.everestgrp.com 4

EGR-2014-11-V-1189
 ANALYTICS IN BANKING

The Role and Impact of Analytics in Banking

Analytics is helping banks become smarter in managing the myriad challenges

they face. Consequently, adoption of third-party analytics business services is
growing rapidly, and is expected to quadruple its current size by 2020 (see
Exhibit 2). The overall penetration of analytics services in banking is still in a
preliminary stage and penetration levels are in low single digits implying
significant untapped value creation potential.
Analytics BPS1 adoption by industry Analytics BPS1 industry size and growth
EXHIBIT 2 2013; Revenue in US$ billion in banking
Annual revenue in US$ million
Size and growth of analytics BPS 100% = 2.0-2.5
900-1,000
CPG & retail
in banking
38%

CA 5%
GR
Source: Everest Group

2
20-
Others2 7% 10% Banking
Media &
6%
entertainment 10%
Professional 6% Insurance 200-250
services 7% 7% 8%
Manufacturing Hi-tech & 2013 2020E
Healthcare telecom

1 Analytics Business Process Services (BPS) represents third-party services of the analytics industry and does not
include size of internal analytics initiatives and/or revenue of analytics products from companies such as SAS,
Oracle, SAP, and Microsoft
2 Include public sector, travel & logistics, and energy & utilities

Analytical solutions have grown tremendously over the last decade, in terms of
their sophistication and the resulting business impact they create. There is a
range of analytics that banks are deploying today (see Exhibit 3). While basic
reporting continues to be a must-have for banks, advanced predictive and
prescriptive analytics are now starting to generate powerful insights.
Relative maturity of analytics solutions
EXHIBIT 3 Percentage of third-party offshore analytics FTEs

Range of analytics leveraged in

10%
retail banking
Emerging
areas of 4. Prescriptive
interest analytics
20%
Business impact

3. Predictive
analytics

40% 2. Descriptive
30% analytics
Most prevalent
analytics
solutions
1. Reporting

Sophistication of solution

1. Reporting. Basic version of analytics solution that focuses on building data

repositories and reporting the current situation using simple and uni- or bi-
variate data. Typical examples in banking include suspicious activity

research.everestgrp.com 5

EGR-2014-11-V-1189
 ANALYTICS IN BANKING

reporting and account validation against watch lists

2. Descriptive analytics. Generating actionable insights on the current situation
using complex and multi-variate data. Typical examples in banking include
customer segmentation and profitability, campaign analytics, and parametric
Value at Risk (VaR) calculations
3. Predictive analytics. Predicting the likely future outcome of events often
leveraging structured and unstructured data from a variety of sources.
Typical examples in banking include pattern recognition and machine
learning to predict fraud, generating risk alerts at customer/product/
geography level, designing personalized and next-best offers, and trigger-
based cross-sell campaigns
4. Prescriptive analytics. Prescribing action items required to deal with predicted
future events using big data from a variety of sources, often associated with
simulations in various business scenarios. Typical examples in banking
include behavioral PD1, LGD2, and EAD3 modeling, channel mix modeling,
real-time offer models, next-best offer models, and stress testing for
mandated and custom scenarios

Significant technological advancements over the last decade have made this
Everest Groups recent enterprise
possible. Traditional data storage and processing technologies could not
cloud survey indicates that
handle unstructured data; handling of large data sets was time consuming and
analytics emerged as a key
prohibitive; response time was too large; and the systems were not flexible and
driver for cloud services in scalable. Several technological advancements have helped overcome these
banking challenges:
The NoSQL movement created alternatives to relational databases that were
unable to handle unstructured data
Hardening of Hadoop framework enabled parallel processing, thus enabling
faster response time and ability to handle larger data volumes at a cheaper
price
Cloud-based utility computing provides virtual shared servers reducing
upfront capital expenditure and increasing accessibility

There are three key areas in banking where analytics has created maximum
impact: 1) Consumer and marketing analytics 2) Risk, fraud, and AML/KYC
analytics, and 3) Product and portfolio optimization modeling. Typical analytics
in each of these areas is summarized in Exhibit 4 on next page.

1 PD Probability of Default
2 LGD Loss Given Default
3 EAD Exposure At Default

research.everestgrp.com 6

EGR-2014-11-V-1189
 ANALYTICS IN BANKING

EXHIBIT 4 Banking segments with

significant use of analytics

Types of analytics in banking

Consumer behavior and Product and portfolio
Risk, fraud, and AML/KYC
marketing optimization

Customer Lifetime Value Value at Risk (VaR) Detailed asset level

descriptive analytics
(CLV) calculation (historical / reporting

Reporting and
Customer profitability non-parametric) Portfolio dashboards
dashboards Suspicious activity Static analysis of portfolio
Drill-down reporting by reporting and customer to estimate capital
customer risk scoring requirements
Campaign analytics Account validation against Collateral analysis
watch-lists Collections delinquency
Risk alerts at customer/
geography/product level

Customer segmentation Value at Risk (VaR) Simulations to predict

prescriptive analytics

Channel mix modeling calculation (variance- default or repayment risk

Predictive and

Conversion and covariance and Monte based on lagging and

engagement by channel Carlo simulation) leading indicators
and segment Behavioral PD, LGD, and Stress testing for
Silent and proactive EAD modeling mandated or customized
churn Stress testing against scenarios
Next-best offer multiple macroeconomic Determining regulatory
Trigger-based cross sell scenarios and economic capital
Bundled pricing Pattern recognition and based on credit portfolio
Social media listening machine learning to predict model
and measurement fraud Central limits
Test and review current risk management
thresholds

1. Consumer behavior and marketing analytics. Advanced analytics now offers

Examples of marketing analytics
banks the power to study their customers and prospects like never before.
impact
Banks that leverage analytics to study customer behavior have been able to
A 15% increase in assets by
significantly improve marketing outcomes (greater topline impact, ability to
designing unique offers for
leverage digital channels, and faster time-to-market) without a
customers proportionate increase in marketing budgets. Key benefits reported include
Improve time-to-market by increased ability to identify profitable customers, expand wallet share with
~25% profitable customers, identify relevant cross- and up-sell opportunities,
Cut marketing costs by 20% migrate customers from less profitable relationships to more profitable ones,
acquire new profitable customers by targeted marketing campaigns, and
launch new product offerings that are in line with customer expectations

2. Risk, fraud, and AML/KYC analytics. Risk modeling and analytics allow
financial institutions to analyze any/all portfolios (of assets as well as
liabilities) to forecast likely losses, and make provisions for those adequately.
Analytics also enables banks to understand risk dimensions faster, without
expanding the pool of human resources. Advanced analytics solutions also
help reduce the complex and expensive burden of compliance on AML and
KYC departments.

Effective use of analytics to fight fraud helps improve profitability, reduce

payouts and legal hassles, and most importantly, improve customer
satisfaction. Analytics bolsters the ability of existing fraud experts to focus on
real threats more efficiently and effectively (by expanding monitored
transactions and reducing false alerts). Automated alerts can also be sent to
the customer directly. Advanced analytics also helps recognize patterns of
fraudulent transactions, and then use these to be one-step ahead of
fraudsters, predict the next such fraud in progress, and recommend

research.everestgrp.com 7

EGR-2014-11-V-1189
 ANALYTICS IN BANKING

preventive action for saving both the bank and the customer. This also helps
banks in protecting themselves against potential fallout (non-compliance
fines and reputation loss risk, amongst others) of AML incidents

Examples of product and

3. Product and portfolio optimization modeling. Advanced portfolio analytical
portfolio modeling impact solutions not only help determine asset pool quality, they also help to
determine prepayments, delinquencies, defaults, and cash-flows. Analytics
Reduce delinquencies from
allow firms to adjust LTV ratios, in accordance with regulations, to meeting
147 to 37 BIPS
capital requirements. If the mortgage portfolio is used for trading/investment,
Using PD/LGD models to
it can also be used to calculate various portfolio risk measures (for instance
reduce the time needed to
VaR)
identify problem loans from
>100 hours to <5 min
Reduce time-to-market of Operationalizing Analytics in Banking Challenges and
new offerings by 99% by Emerging Best Practices
calculating multiple risk
factors automatically Analytics in banking promises significant value creation potential if the solutions
are operationalized, keeping the following three success factors in mind:
1. Business insights are created using data
2. Insights are used to take decisive actions
3. Results from business decisions are fed back to improve data and analytics

This data-insight-action loop is summarized in Exhibit 5 below.

Challenge: Multiple Challenge: Inadequate Challenge: Talent Challenge: Functional
EXHIBIT 5
competing priorities data and systems crunch silos
infrastructure
Typical challenges across the
Data-Insight-Action loop in 1. Objective 2. Data 3. Insight 4. Action

banking Problem Extraction Data Enable decision

definition / Collection contextualizing making
hypothesis Filtering Predictive and Embed in business
creation Categorization prescriptive rules
Set boundary Integration modeling Modify business
conditions Integrity Output visualization processes
Dashboards and
scorecards
Custom reporting

5. Feedback

Learn from actions

Improve data
collection sources
Refine/adjust
objective

Challenge: Ongoing
relevance, validity, and
improvement

Operationalizing analytics in banking requires significant investments of time

and money across people, process, organization, and technology. This section
summarizes key challenges that financial institutions face across each element of
the data-insight-action loop, as well as provides some of the emerging best
practices.

research.everestgrp.com 8

EGR-2014-11-V-1189
 ANALYTICS IN BANKING

Phase of Data-Insight-Action Critical challenge in operationalizing

A strategic mindset is required to loop analytics
manage multiple competing Multiple competing priorities
1. Objective
priorities

Staying competitive by meeting evolving customer demands and remaining
Our risk organization and compliant amidst increasing regulatory requirements often pulls banks in
functions were established to different directions. Increasing competition strains a banks ability to exercise
support and enable our more stringent fraud prevention. For instance, customers want their cards to be
organization to achieve strategic accepted widely, but due to the demands of regulatory compliance, the usage
goals such as sustainable growth is restricted and care advised. Banks have to walk the tightrope between these
and profitability, competitive forces to succeed.
advantages, and capital
management. Put simply, we The emerging best practice. Banks need to change their mindset to consider
recognize risk as a part of the
compliance as a source of competitive advantage instead of treating it as a
strategic agenda.
An Asia-Pacific bank
burden and cost. This is increasingly witnessed in the area of risk management.
Enterprise risk strategy should not be to just comply with, but to use risk
management for creating competitive advantage. Risk modeling and analytics
should be directly linked to business outcomes.

Phase of Data-Insight-Action Critical challenge in operationalizing

The challenge of fragmented loop analytics
and outdated data and systems Inadequate data and systems infra-
2. Data structure

We have 22 different definitions

of what is a customer.
Effective analytics requires a robust and integrated security master, account
master, and customer master. Without strong master data management, it is
hard to analyze the flow of money, the structure of the instrument/business deal,
concentration of liquidity or exposure, and how one component relates to
A loan valuation process
typically results in about 500
another. However, the underlying data fragmentation makes it hard to get an
billion Monte Carlo simulations, integrated and reliable internal data stream.
composed of multiple
calculations that need to be Beyond integrated internal data streams, advanced analytics requires significant
carried out for each amount of external data. For instance, banks need to track online behavior,
macroeconomic scenario. social media habits, etc., to build a forward looking assessment of Customer
Lifetime Value (CLV). Since large parts of data gathered and used for
Following a merger, we faced understanding customer behavior are unstructured, it is hard to work with
the challenge of integrating two traditional and legacy databases and applications. Rapidly changing customer
vastly different credit platforms... tastes, preferences, and macroeconomic environment also pose a significant
and we had to ensure challenge in modeling predictive/prescriptive analytics.
compliance with Basel II/III
requirements. Predictive and prescriptive analytics involve various internal and external data
A European retail bank sources across lagging and leading indicators. Processing huge volumes of
data requires state-of-the-art IT infrastructure. Regular and clean data inputs
are also required to ensure that the model generates valid and meaningful
output. As a result, data cleansing is also a major challenge. Most of the
existing technologies work with traditional channels but are often rendered
ineffective with digital media. Moreover, the cost of system modernization is
prohibitive.

research.everestgrp.com 9

EGR-2014-11-V-1189
 ANALYTICS IN BANKING

The emerging best practice

Leveraging offshore delivery for data management. Increasingly, banks are
leveraging offshore resources (Global In-house Centers (GICs) and/or third-
party outsourcing) for data integration and data cleansing. Beyond cost
arbitrage, offshoring offers access to new talent pools, ability to leverage
different time zones, and manage volume fluctuations for data management
Long-term planning for overall application & system integration and an
agile organizational structure that can adapt and change quickly. Server
virtualization and other cloud solutions are increasingly being leveraged

Phase of Data-Insight-Action Critical challenge in operationalizing

Leveraging global sourcing to loop analytics
manage analytics talent crunch Talent crunch
3. Insight

Around 320 third-party

outsourcing deals with an
analytics component were
Implementing advanced analytics is not just about hiring statisticians with
advanced degrees. Banks require analytics professionals with business context,
data management, technology understanding, and knowledge of existing and
emerging regulatory requirements. The problem is further exacerbated, given
announced between 2005 to
that several regulatory requirements (such as PD, LGD, and EAD modeling)
2007 by banks. This nearly
emerged less than five years ago. Such talent is scarce and expensive.
doubled to 610 deals during
2011 to 2013.
Everest Group (2013) The emerging best practices
Centralization of analytics resources, via creation of Center of Excellence
(CoE). A centralized team works on analytics across different Lines of
Businesses (LoBs) and/or across asset classes (mortgages and equity). A
CoE-based approach, as opposed to resources embedded within individual
lines of business, improves overall utilization of these scarce and expensive
resources. It also helps in setting standard processes, driving common
governance standards and process rigor, as well as improving ease of
accessibility
Leveraging global sourcing. Using GICs and/or third-party outsourcing
service providers enables banks to tap into a global pool of analytics
resources. The business case for global sourcing of analytics is less about
cost savings and more about expertise
Intra-industry collaboration. Banks are starting to realize the power of
working together, especially in areas such as fraud analytics. Building
individual models is costly and time consuming. The solution is for banks to
move to network-based models, where fraud detection algorithms from
different banks work together for the group, and help achieve more
extensive and better results, while also keeping the costs low
Inter-industry collaboration. Customer behavior in banking often follows
similar patterns as in hospitality and travel industries, amongst others. Banks
are taking advantage of this fact to create teams of resources having
exposure and experience of multiple industries to create non-linear learning
curves

research.everestgrp.com 10

EGR-2014-11-V-1189
 ANALYTICS IN BANKING

Phase of Data-Insight-Action Critical challenge in operationalizing

Inefficiencies due to functional
loop analytics
silos
Functional silos
4. Action

Due to significant dependencies

on branch managers, a single
KYC can take up to 20 days.
Banks today offer a variety of products and services (retail banking, corporate
banking, cards, lending, asset management, and wealth management) but are
poorly integrated internally from an organization and systems perspective. For
example, the view of a customer from the CRM system would not typically
Large Germany-based global
bank
incorporate a risk profile, performance history, or regulatory data associated
with KYC requirements yielding an incomplete and possibly inaccurate view of
a customer. An end-to-end process view is hard to build, as back-, mid-, and
front-office processes are disjointed. This often delays flow of information
required to make decisions based on smart analytics. Moreover, it increases the
response time when warning signals are generated.

The emerging best practices

Need for greater collaboration
Common analytics platform. Banks are increasingly centralizing certain
to overcome challenges from
functions (such as AML) across all geographies and LoBs into a single
existing functional silos
platform to establish global systems and processes, capture/prevent

suspicious transactions, and report to relevant regulatory bodies
In todays banking world, it is Collaboration between analytics and business units. Marketing experts
the balancebetween creativity across different product lines and analytics SMEs need to collaborate to
and discipline, between art and ensure that analytics insights are taken to action, and the insights themselves
sciencethat we need to strike. are actionable
CMO of a U.S. bank Real-time actions on market insights. Banks are starting to realize the power
of cutting time elapsed between understanding customer behavior and
acting upon it. Not only that, they are also realizing the immediate effect of
their strategies in shaping customer behavior
Global knowledge sharing. Digital revolution is fast erasing distinction
between customers in different geographies. It is therefore imperative, that
different business units within banks today collaborate more between
themselves, and make sure that lessons learnt in one geography for
customer behavior modeling are utilized in others

Phase of Data-Insight-Action Critical challenge in operationalizing

The need for ongoing relevance, loop analytics
validity, and improvement of Ongoing relevance, validity, and
5. Feedback
analytical models improvement

The fraudsters are getting better

at knowing where the data has
been compromised
Predictive and prescriptive analytics need to be fluid, dynamic, and open to
learning and improving. They are not a one-time exercise and need regular
updates and refinements. A continuous feedback mechanism is required from
frontline systems, which is hard to implement.
and staying within that particular
footprint of that customer,
making the detection that much Most variables within an analytics model are customer demographics,
more difficult. financial data, economic and regulatory environment related data, and the
Senior executive, of a mid-sized analytical models that need to be designed to accept these data sets. The
card-issuing financial institution analytical model, of course, needs to stay relevant and stand the test of time.
Making sure that the predicted and actual values stay within a zone of
acceptable error is a significant challenge.

research.everestgrp.com 11

EGR-2014-11-V-1189
 ANALYTICS IN BANKING

Post the 2008 economic crisis, model management needs to be independent of

model development. Most models need validation and certification before their
release for usage, every 6 to 12 months.

The emerging best practices

Self-learning and flexible solutions. Analytical and simulation models need
to be designed, keeping in mind future adjustment and refinements. For
instance, sophisticated fraud models are emerging that have the ability to
learn from their own functioning and positive identifications made, in order
to evolve and improve, reduce false alerts over time, and help banks stay
ahead of fraudsters. Also, enabling sharing of results between algorithms
helps in triangulation of results
Offshore third-party service providers can also be utilized for testing and
ongoing maintenance of models. This enables banks to keep model
development and validation separate, as well as improve scalability and
speed-to-market

Conclusion

Sophisticated predictive and prescriptive analytic solutions exist today that can
improve a banks probability of ensuring survival, compliance, profitability &
growth mandates, and competitiveness. However, institutionalizing and
operationalizing analytics to take smart business decisions is a challenge given
the ground realities of functional silos, talent crunch, competing priorities, and
outdated data/system infrastructure. The critical question today is not why
analytics? or which analytics? but how to operationalize analytics?

There is no other value-creation lever available today for the banking industry
that is as powerful as analytics. But, it needs to be elevated from an IT- or LoB-
level discussion to a C-level strategic agenda to really unleash its true potential.

This study was funded, in part, by support from Genpact.

research.everestgrp.com 12

EGR-2014-11-V-1189
 ANALYTICS IN BANKING

About Everest Group

Everest Group is an advisor to business leaders on next generation global

services with a worldwide reputation for helping Global 1000 firms dramatically
improve their performance by optimizing their back- and middle-office business
services. With a fact-based approach driving outcomes, Everest Group counsels
organizations with complex challenges related to the use and delivery of global
services in their pursuits to balance short-term needs with long-term goals.
Through its practical consulting, original research and industry resource
services, Everest Group helps clients maximize value from delivery strategies,
talent and sourcing models, technologies and management approaches.
Established in 1991, Everest Group serves users of global services, providers of
services, country organizations, and private equity firms, in six continents across
all industry categories. For more information, please visit www.everestgrp.com
and research.everestgrp.com.

For more information about Everest Group, please contact:

+1-214-451-3110
[email protected]

For more information about this topic please contact the author(s):

Eric Simonson, Managing Partner

[email protected]

Rajesh Ranjan, Partner

[email protected]

Anupam Jain, Practice Director

[email protected]

research.everestgrp.com 13

EGR-2014-11-V-1189