WWW.WIPRO.COM Arvind TiwaryAdvisor Insurance Vertical

Point of View - Wipro Insurance Practice

ANALYTICS IN THE LARGE & SOLVENCY II Key Drivers for Businesses Towards a Sustainable Future

Table of Contents Abstract 3

Analytics in the Small 4

Solvency II (SII) 5

The SII Architecture (Work in Progress) 5

Analytics in the Large 7

Risk and Uncertainty : Levels of Uncertainty 11

Analytics Everywhere 12

nsurers have been using analytics at Iworkgroup or departmental level and for

silos of the value chain or ‘Analytics in the Small’.

Solvency II requires large scale modeling and

imposes a level of precision and complexity

that will be a challenge for the Industry.

Solvency II is an ‘Analytics in the Large’ project

and will exceed anything done by central

bankers or the IMF. Solvency II is also being

seen as a gold standard in insurance regulation

and is influencing US and Asia Pac regulations.

The massive investment in IT and Analytics, as

well as senior management time, needs to be

harnessed for more than just compliance.

Building an ‘Enterprise Analytics’ platform and

capability will provide the ability to recoup the

SII investments and make insurers more agile

and able to sense and respond in a

exceptionally volatile and complex world.

nsurance companies have a long tradition of analyzing Idata and understanding risk and pricing it. Yet, most

informed observers seldom regarded insurers as smart

users of analytics until recent years. Analytics was rarely

used for insurance consumer segmentation, predicting

business outcomes and developing products and

processes to manage them. But the industry approach

has changed. Today, analytics is used by almost all

insurance companies, the most common forms being

performance dashboard and business intelligence. Some

insurers use predictive analytics, thus segmentation and

pricing have become more fine-tuned and granular. With

the adoption of Solvency II (SII) norms, demand and use

of analytics in the insurance business will grow

exponentially.

03

Analytics in the Small

Analytics is the processes, technologies, and best

practices that turns data into information and

knowledge that drives business decisions and actions

Thomas H Davenport“Competing on Analytics: The New Science of Winning”

Sales & Marketing ? - Analysis of the current book of business to cluster policyholders based on common demographics

? - Analyze profitability, retention and business mix across a segment to develop new products

? - Extrapolation of existing sales data to predict the likely outcome of targeted campaigns

? - Success rate in conversion of quotations to customer policies

Customer Segmentation

Segment Analysis

Campaign Simulation

Conversion Rate Analysis

Underwriting ? - Evaluate written policies across geography and demography of customers with identified risks

? - Monitor and revise prices based on previous loss experience and risk distribution

? for business growth vis-à-vis expenses and exposure, pricing, renewal, etc.

? - Measure revenues after deducting loss payouts and overhead expenses across various lines of business, geography, agents, underwriters, etc.

Risk spread analysis

Price Modeling

Predictive Modeling and Multivariate Analysis

Product Profitability Analysis

Servicing & Claims ? - Evaluation of variance in initial reserves set and claims paid out

? - Workload distribution across lines, notification of recovery potential and claim types and turnaround time for claim settlement

? - Development of losses based on historical catastrophe information in terms of claim numbers and severity

? - Rules for fraud detection based on investigation agencies and adjuster experience. Measurement of accuracy of applied rules and refining criteria for detection

Reserve Analysis

Adjuster Performance Analysis

Catastrophe Loss Model

Fraud Detection

Customer ? - Computation of net business is capable of delivering

? - Sorting customers into logical clusters based on behavioral and lifestyle characteristics such as preferred channel, payment methods, products purchased, etc.

? - Identify customer profiles to be targeted for cross sales and up sales based on lifetime event occurrence like marriage, child birth, etc.

Customer Lifetime Value

Customer Profiling

Cross Sell / Up Sell

revenue, the customer

Agent/Intermediary ? - Agents are measured on their retention, commissions earned, qualification and other parameters for their performance measurement.

Some sample parameters include:

? - Average time taken by agent to receive response to a quote

? - Figures achieved by agent with respect to targets fixed for the business

Agency Management

Turnaround Time

Performance Management

04

1 International Financial Reporting Standards promulgated by IASB (International Accounting Standards board) and being adopted by Europe as well as many countries. 2 The US FASB (Financial Accounting Standards Board) has circulated a discussion paper which will align US GAAP for Insurers to IFRS-4.

Solvency II (SII) II is the updated set of regulatory requirements pertaining to solvency capital for insurance firms that operate in the 27-member European Union. Once the Omnibus II Directive is S

approved by the European Parliament, SII will come into effect on 1 January 2013. These developments could also set the tone for future insurance regulation in major economies like the US, Japan and China.

SII is being seen as the model for Risk Based Capital (RBC) and has influenced Basel III, a new global regulatory standard on bank capital adequacy and liquidity. SII has a three-pillar architecture, similar to the banking regulation Basel II, as given below.

The SII Architecture (Work in Progress)

II has been in the making since 2004 and is still undergoing changes. The regulatory norms are likely to be finalized by March 2012. SII will bring in uniform regulation of various lines S

of insurance business covering life, non-life and health, newer hybrid instruments like secondary guarantee unit-linked life insurance and all types of reinsurance. It will harmonize the various patchworks of regulation (Solvency I) across the European Union including the UK.

1SII builds on IFRS accounts of an insurer and requires the risk margin or Solvency Capital Requirement (SCR) to be at the 99.5% confidence level. It may be noted that IFRS-4, specific

2accounting standards for insurance contracts, will likely come into force by 2012 .

IFRS-4 uses a balance sheet approach to accounting for long-term insurance contracts and requires a best estimate of discounted future profit (income like premium and fees less expected expenses like claims) to be carried on the balance sheet and released to comprehensive income over the contract life. The best estimate differs from the past GAAP approach. Therefore, a probability weighted cash flow (expected value) will have to be used.

05

DISCLOSURE REQUIREMENTS

Current disclosure requirements:?National GAAP?National regulatory reporting?IFRS 4?IFRS 7Future disclosure requirements:?IFRS (4 Phase 2 and IFRS 7)?IAIS?EU legislation

GREAT UNIFYING THEORY

SUPERVISORY REVIEW PROCESS

?Internal control?Risk management?Corporate governance?Stress testing?Continuity testing

?Eligible capital?Technical provisions?Capital requirements?Asset valuation?Risks to be included?Risk measures and assumptions?Risk dependencies?Calculation formula?Internal model approach

MEASUREMENT OF ASSETS, LIABILITIES AND CAPITAL

3 European Insurance and Occupational Pension Authority replaces CEIOPS. 4 70% insurers have participated in the QIS 5 study

https://eiopa.europa.eu/

Coming Complexity

SII builds the capital required through a series of modules that are decomposed into sub-modules or components. The capital required by each sub-module is rolled up to a higher level, not additively but with due regard to the correlation between the risk factors.

3European Insurance and Occupational Pension Authority (EIOPA) through its predecessor (CEIOPS) has conducted five rounds of Quantitative Impact Studies (QIS) with more and more

4insurers participating in it to calibrate the factors used in the calculations under the Standard Model. These calibrations cover several macro-economic variables and correlations across equity, bond, currency, and property asset classes as well as quantitative scenarios for stress tests. More than 200 individual stress tests and multiple paired and integrated scenarios have been defined and several more are in the works.

SII is more a business model change than a risk capital calculation exercise as it requires insurers to demonstrate that their managers understand the available economic capital and use the capital charge in their decision making. SII is expected to impact all aspects of the insurance business from product design to operations.

Solvency Capital Requirements

(SCR)

Basic SCR Operational Risk Ajustments*

Market Health Default Life Non-Life Intangibles

Interest Rate

Enquiry

Property

Spread

Currency

Concentration

Illiquidity

Revision

Expenses

Lapse

DisabilityMorbidity

Longevity

Mortality

HealthSLT

HealthNon SLT

HealthCAT

PremiumReserve

Lapse

Mortality

Longevity

DisabilityMorbidity

Lapse

Expenses

Revision

CAT

PremiumReserve

Lapse

CAT

* loss-absorbing capacity of technical provisions and deferred taxes

06

II is a large Analytics project taking insurers into a new territory. They are required to 5Sconsider various qualitative scenarios , convert them into quantified economic scenarios

(and estimate a probability for their occurrence) and project their balance sheet using the quantitative models. Insurers are thus required to project their balance sheet with a great deal of rigor. Further, the SII balance sheet will have to be reconciled with the IFRS balance sheet, which will then be reported to investors.

The 99.5% confidence requirement under SCR places unprecedented emphasis on historical data. Hence, insurers would need to demonstrate the use of the highest quality data and processes in their financial reporting. The 99.5% confidence level can be demonstrated using two methods:

1) Fit the observed time series (historical data) for the variable to an assumed probability distribution, derive its parameters and use a mathematical method to compute

6the level for the 99.5% confidence . For example, if the distribution is a normal (bell) curve, then the second moment of the distribution (volatility or standard deviation) can be used to compute 3 standard deviations as the 99% percentile.

2) Organize many observations (1,000+) and order the frequency of observed values to establish a cut-off for the 0.5%. This requires adequacy of coverage for extreme events and

7Through The Cycle (TTC) rather than making Point-In-Time (PIT) estimates.

8As time series data extending back a 100 years or more is a rarely available, there would be problems in establishing frequency data that is robust enough. As a proxy, stochastic simulations can be used to produce a number of values and that sample is then used to derive confidence intervals.

SII requires validation of the random simulations as adequate and comprehensive. The models and algorithms used are required to be disclosed for third party review. SII also specifies tests for adequacy and the number of outliers or failures needs to be within specified limits. The projections tend to be non-linear. Therefore, the correlation between variables have to be modeled and the assumptions of future consumer behavior (for example, propensity to claim losing renewal discount, lapse and take-up rates of options under insurance contracts) documented.

Insurers who deal with rare but extreme events have used Extreme Value Theory (EVT) to develop estimates of confidence. However, lack of data of the right level of granularity is seen to

9be particularly acute for operational risk and so expert estimates or Bayesian techniques are likely to dominate.

These projections have to be done by breaking a book of business into Homogenous Risk Groups (HRGs) for the risk factor being projected. For instance, a terrorist risk assessment would need accumulation of policies within 150 meters of a landmark, whereas a flood risk projection may be made on the basis of accumulated policies across several states in a flood plain.

Insurance groups need to do this projection at the group level but given the nature of liability stress projection they would have to start at a level lower than the business unit balance sheet. It is not a mere addition or aggregation of individual legal entities but a consolidated projection of the underlying data at the HRG level. It is not unusual to have 2,000-10,000 HRGs. It is expected

10that each simulation or projection will use 100,000 - 300,000 (Stochastic) iterations . It is also not unusual to see these projections taking weeks on farms of 2,000 or more servers and generating terabytes of data. This has been taxing the IT infrastructure of most insurers.

Some insurers who have been ahead of the curve on SII have looked to alternatives and optimizations like mean variance reduction or replicating portfolio approaches to manage the SII reporting timeline and the need to crunch the timelines by a third or more.

Analytics in the Large

07

5 The scenarios use macro economic variables like Equity market Index, Property price Index and currency exchange rates6 This calibration may need to repeat periodically as economies are rarely stationary as relationships between variables (slowly) change. 7 Bear and bull markets and recession and wars if possible.99.5% implies 1 in 200 years event. 8 For instance, interest rates beyond 1970s are hard to come9 See http://en.wikipedia.org/wiki/Bayesian_probability10 Simulations use random number to sample variables from a specified probability distribution and advance the simulation to the next iteration. Sometimes internal simulation of 1000-10,000 iterations is

needed to value derivatives.

Advancing the Art and Science of Large Scale Projections

Forecasting Complexity

Traditionally, modeling in the Large has been done by central bankers, IMF, among others. However, they were not called upon to demonstrate the efficacy of the models to the regulators as under SII. As such, life insurers are required to project cash flow 40-60 years out. That being so, larger insurers expect to get significant relief in the capital that they need to hold by showing a well managed book with ample diversification. Since the results of the model impact them directly and immediately, they are pushing for the art and science of macro-economic forecasting. A small diversion on some of the considerations will help.

Macro Economic Variables: There are few publicly available actual models of validation used by insurers or banks but the academic literature is full of discussions on methods.

Most models attempt to explain the following observed properties of price returns:

1) Daily returns are log normal

2) Price returns follow a trend (mean) over medium time-frames and random variations around this mean drift are volatile in nature. A Ornstein-Uhlenbeck process.

The most common approach is to use Generalized Auto Regressive Conditional 11Heteroskedasticity (GARCH ) . However, all models fail to adequately match at extremes; since

fat tails are common (i.e. extreme events are much more frequently observed than implied by the assumed probability distribution). As compensation, SII tends to favor Conditional Tail Event (CTE) CVaR, or Tail VaR. The required capital can be 3-6 times more than under a VaR approach. Much more work is expected in this area, especially with the wealth of data coming out on the recent global economic crisis or ‘the Great Unwind’ caused by the US sub-prime mortgage

12crisis .

Calculating the macro-economic variables would entail the following:

13Inflation: The Maturity Guarantee Working Party (MGWP) in 1980 used a multivariate Wilkie model using short-term interest rates, long-term interest rates, share dividends and share yields for long-term (100 years) projections. This remains the most discussed model to date.

Interest Rates: Insurers hold significant debt in their books. Their investments are much more affected by interest rates movements as they have cash flow going out 50-80 years. Forecasting these is crucial to valuing the long-term liabilities and assets of insurers. Short and long-term rates quite often move differently and the term structure of yields can take different shapes over

14the full economic cycle. There are numerous issues and the CFO Forum has been working to standardize this risk-free interest rate be used for discounting future assets and liabilities.

Equity Prices: Banks need to model the Value at Risk (VaR) for their (trading) books and have been at the forefront of developing models for market and credit risk. Banks typically model daily to annual forecasts but insurers need to view a much longer period. The American Society of

15Actuaries Life Capital Adequacy Subcommittee (LCAS) has recommended the three methods that follow.

1. Independent Log Normal (ILN), where (continuous) stock returns are normally distributed. This is adequate for short term (daily) forecasts and used by banks for their trading book.

2. Two-state Regime Switching Log Normal (RSLN2). This switches between a bull and a bear regime. Within each regime the prices are modeled following a trend (mean reverting) with random variations around the trend line or the Ornstein-Uhlenbeck process.

3. Stochastic Log Volatility (SLV), where (log) volatility is a mean-reverting process with constant variance.

08

11 Robert Engle, GARCH 101: The Use of ARCH/GARCH Models in Applied Econometrics Journal of Economic Perspectives—Volume 15, Number 4—Fall 2001—Pages 157–16812 See http://www.nytimes.com/2009/06/12/opinion/12brooks.html13 A joint initiative of The Institute of Actuaries and the Faculty of Actuaries, UK14 15 AAA 2005 See A Comparison of Actuarial Financial Scenario Generators, by Kevin C. Ahlgrim, Stephen P. D'Arcy, and Richard W. Gorvett, VOLUME 2/ISSUE 1 CASUALTY ACTUARIAL SOCIETY

http://www.cfoforum.eu/solvency.html

Property Prices: Insurers invest significantly in real estate but there is lack of a continuous and broad-based price index either of the whole market or of sub-markets like commercial and retail property - at a country or Pan-European level. These indices will be developed over time and may provide greater insight.

Correlation

SII specifies correlation matrices covering different asset classes like property and equity. Insurers are expected to continuously validate these for their own book and develop them appropriately for their risk factors. “The Great Unwind” showed that historical correlations have changed. Gold, property and equity prices behaved differently than in the past, so too the correlations. These suggest the use of more sophisticated models like Copulas.

Credit or Default Risk

16SII adopts the Basel II/III method (IRBA) of assessing default risk. This is a reduced form model which ignores micro-structure or cause-effect modeling and uses correlation between ratings assigned by (unspecified) process and observed defaults. An alternative approach recognizes interconnections and uses network models. Scale-free networks developed by Hungarian

17scientist Albert-László Barabási have given significantly better results .

Agent Interactions or Feedback

Physical systems seem to obey invariant laws expressed mathematically and can be modeled with great precision. Social systems and economic systems have actors who modify their behavior based on internal state (fear, anger) and by observing what others are doing (herd behavior). There are successful approaches to modeling these systems at the small (micro structural) level

18by Agent-Based Modeling (ABM) .

Expert Judgment

Enterprise Risk management is not a new topic and marquee names like AIG, Enron and Fannie Mae failed despite their much praised risk management systems. Even rating agencies like Moody and S&P failed to see the risk in the mortgage market and rated CDO and MBS at lower risk than they ultimately proved to be.

It is important to accept that precise numbers produced by excel or other numerical models do not ensure accuracy, and imprecise expert judgement may provide a better estimate for a

19changing situation. Adam Borison and Gregory Hamm in their well argued article “How to Manage Risk (After Risk Management Has Failed)” have documented the problem of frequentist (statistical) approaches and argue for enhancing these with expert judgement. This is probably an area where there is rapid development as the Bayesian approach is suitable in many areas of quantifying uncertainty.

Nassim Nicholas Taleb has captured the popular imagination with his metaphor of the Black Swan for high-impact, hard-to-predict, rare events. Traditional techniques anchor on projecting the past to the future and fail to see a tipping point that produces rapid, radical change. Physicists have been lucky to develop a measure of the large-scale statistical stage of matter called temperature, thus being able to correlate with the rapid change of phase from solid (ice) to liquid (water) or gas (vapor). Inductive analytics, which has the ability to tease out weak trends or signals and the ability to build scenarios where these trends may break-out, is an important skill

20to be developed for building more robust forecasts .

Inductive Analytics for Black Swan

09

16 A Reduced Form model is a top down model which simplifies detailed (casual) relationship and uses a small simpler variables and relationships. 17 Albert, R and A.L Barabasi, 2002 “Statistical Mechanics of Complex Netwroks”, Review of Modern Physics, 74(1), pp 47-9718 Agent-based model (ABM) is a class of computational models for simulating the actions and interactions of autonomous agents. It combines elements of game theory, complex systems, emergence,

computational sociology, multi-agent systems, and evolutionary programming. Monte Carlo Methods are used to introduce randomness. The process is one of emergence from the lower (micro) level of systems to a higher (macro) level.

19 Adam Borison and Gregory Hamm, How to Manage Risk (After Risk Management Has Failed), MIT Sloan Management review October 1, 2010. They rail against the fantasy of precision versus the realty of judgement

20 Life Sciences have used innovative techniques. Genes detection requires scanning for 1-2% signal in 95-98% noise as they proteins marking a gene or a cluster are a small fraction of the bases in the DNA.

Early Warning of Changing Trends

The ability to spot changes in consumer behaviour and the economic environment will be a major competitive advantage as this will allow the insurer to conduct research early and build up strategy for taking advantage of emerging trends. Insurers have historically had a utility business model and waited till trends became mainstream before adapting them. This may not be the smartest way to cope with a more volatileworld; inductive intelligence to get early warning could provide sustainable advantage.

The goal of precision may prove counter-productive to the need for robustness. SII may end up encouraging alternative models to be used for critical forecasts to make sure one does not get blindsided by the apparent precision of the numbers emerging from quantitative models. As noted by Lotif Zadeh "As complexity rises, precise statements loose meaning and meaningful

statements loose precision...”

We expect the state of Science and Art of Forecasting to rapidly improve as regulators encourage the use of sophisticated methods. Insurers are looking at a multi-year, steep learning curve as the bumblebee of regulatory nudge (review) will cross-pollinate the industry and slowly but surely raise the level of Analytics in the Large.

As of now, insurance groups across Europe are racing to get their SII projects shipshape given that most of them are not in good shape. Their focus is taking the shortest way to become compliant. However, once the current crunch is overcome there will be fresh thinking on how to maximize or leverage the huge investments in SII analytics infrastructure.

Wipro believes the best way is to link the Analytics in the Small and Analytics in the Large into one common platform with shared processes. For example, this would help a CFO to take a macro qualitative scenario generated by a long-term planning group to a quantitative macro scenario (like a stress test) and use that to understand the buying behaviour and renewal rates for different channels and LOBs. It would also be very insightful to play with various managerial actions (incentive schemes, pricing actions) for different customer segments and see how they ripple through to the balance sheet.

Precision or Robustness in the Face of Complexity

Enterprise Analytics

10

Risk and Uncertainty: Levels of Uncertainty

11

21 Monto Carlo Markov Chain models (MCMC)

Few insurers today, if any, have the “Enterprise Analytics” capability to quantify the impact of superior claim fulfilment or reduction in billing errors to improve stickiness on renewals or premium pricing power. Similarly insight on the underwriting of policy and setting a risk appetite to enable an optimal reinsurance program to maximize the risk adjusted return on capital would be worth their weight in gold. An enterprise analytics platform with detailed customer, policy and intermediary data for many years and LOBs and channels enhanced with cross-linked external data on the economy and demographics of the participants would allow one to use longitudinal studies to derive this information. The methods of econometric modelling in the

21large combined with agent-based models and MCMC models would be very illuminating.

These would provide the following benefits:?A better segmentation of customers?Better matching of risk factors (HRG aligned to products/channels/market segments)?Better understanding of the interplay of risk capital and product and channel features?Optimization of hedging and reinsurance programs, improving margins through financial

engineering?A more agile response to catastrophic events

A major objection to this completion of the circle comes from the complexity and cost. Current SII projects use summary data (model points and HRG) to develop projections. The Enterprise Analytics would require detailed data to be kept and used in a common consistent manner. This would tax the IT infrastructure of most insurers. However, one may need to look at the likes of Google, Amazon, and Facebook which have invested in managing and exploiting the intelligence embedded in their data to build successful business.

here is a continuum of risk and uncertainty. As discussed by Frank Knight (Knight F.H T1921, 'Risk, Uncertainty and Profit'; Boston, Houghton Miffin), one side is amenable to (formal) quantitative analysis, especially using statistical methods (game of chance like Poker), while the other side is not amenable to such approaches (meteor crashing into Manhattan). Much of SII assumes that uncertainty can be modeled with statistical approaches. However, there is a continuum as seen in physical systems to social systems and economic systems. Lo and Mueller describe partially or fully reducible uncertainty that is the mainstay in SII modeling.

Level 1 Complete Certainty: Simple physical phenomena. For example, time to travel a fixed distance by light rays.

Level 2 Risk Without Uncertainty: Known and static probability distribution. For example, card games. Risk can be fully valued and the expected value can be used for decision making.

Level 3 Fully Reducible Uncertainty: Unknown but static probability distribution. A large data sample is used to estimate the distribution. Actuaries have developed a prior expectation on the type of distribution for different events. For example it is popular to use Poisson or Inverse Binomial distributions for modeling P&C claims.

Level 4 Partially reducible uncertainty: Unknown dynamic probability distributions are fairly common in social and economic systems such as RSLN2 for equity price in global markets, regime switches in operational risk assessment. Concepts like these break the central dogma of statistics. The 'law of large numbers' and the 'central limit theorem' which posit a Gaussian or bell-shaped distributions is the approximate form. Experts and informed participants have been able to provide good 'best estimates' to such complex concepts in the past. Hence, it is vital to research and develop some cause effect models of these concepts, with an inherent ability to nullify the bias of the human brains. The current evolution of topics such as crowd sourcing are indictors that there is some kernel of truth in this.

Level 5 Irreducible Uncertainty: This touches upon topics like philosophical enquiries, mystical truths, and existence of God. These need a major improvement in science before one can make a judgment on them. Till that time they remain in an irreducible uncertainty zone. Global warming is a topic transitioning from Level 5 to Level 4 as more research progresses.

ince the seminal article ‘Competing on Analytics’ in the 2005 Harvard Business Review by Thomas H Davenport, analytics has received great attention of most businesses. A new class S

of business based on extracting information and insight from data has grown to be very successful. Think Google and Amazon!

Insurers have adopted Comparative Analytics, and Business Intelligence solutions and dashboards are now widely used. Much less work has been done on Predictive Analytics. It is rare to see Inductive Analytics being used. This can be teasing out weak signals which grow and provide early mover advantage.

SII provides a good platform for insurers to migrate to Enterprise Analytics and complete the circle of Analytics in the Large and the Small and broaden their toolkit to include gaming, agent based models, inductive analytics as well as forecasting and balance sheet projection to develop a much more agile and forward-looking business.

Analytics Everywhere

12

rvind brings 25 years of Industry experience. He has a multi-disciplinary experience in helping financial institution, exploit Information Technology. He has worked in Sales & A

Marketing, managed software projects and acted as Product Manager and Chief Architect to large enterprise scale applications over three generations. He has also run operations with Board level responsibility in Hong Kong, Singapore and India. Arvind is a popular speaker and has been adjudged the best speaker in many conferences and has participated in strategic planning (Business and IT) at the board level for many insurers.

Arvind currently advises companies on innovative business models thru business IT Architecture at SangEnnovate.

Arvind is an alumnus of the Wharton Business School, Indian Institute of Technology, Kanpur and Indian Institute of Management, Ahmadabad. Arvind has taught IT to insurers and insurance to IT professionals.

Wipro Insurance Practice works with 35+ global insurers including many among Top 500 organizations. Our customers include 4 of the top 6 P&C carriers in the world, 2 of the top 5 health insurance and services providers globally, and 4 of the top life annuity & pension carriers in the world. Our offerings cover the entire spectrum of the insurance value chain - from Sales & Distribution, Policy Administration and Claims - straddling across Life and P&C markets, and delivered by over 6000+ dedicated resources. Our expertise in Business Advisory Services, and Solutions & Centres of Excellence reflect our commitment towards building the future of insurance.

Arvind TiwaryAdvisor Insurance Vertical

Wipro Technologies, the global IT business of Wipro Limited (NYSE:WIT) is a leading Information Technology, Consulting and Outsourcing company, that delivers solutions to enable its clients do business better. Wipro Technologies delivers winning business outcomes through its deep industry experience and a 360 degree view of “Business through Technology”– helping clients create successful and adaptive businesses. A company recognized globally for its comprehensive portfolio of services, a practitioner’s approach to delivering innovation and an organization wide commitment to sustainability, Wipro Technologies has 120,000 employees and clients across 54 countries.

About the Author

Wipro Insurance Practice

About Wipro

13

DO BUSINESS BETTER

WWW.WIPRO.COM NYSE:WIT | OVER 120,000 EMPLOYEES | 54 COUNTRIES | CONSULTING | SYSTEM INTEGRATION | OUTSOURCING

WIPRO TECHNOLOGIES, DODDAKANNELLI, SARJAPUR ROAD, BANGALORE - 560 035, INDIA TEL : +91 (80) 2844 0011, FAX : +91 (80) 2844 0256

North America Canada Germany Switzerland Austria Finland Portugal Japan Singapore AustraliaSouth America United Kingdom France Poland Sweden Benelux Romania Philippines Malaysia

©Copyright 2011. Wipro Technologies. All rights reserved. No part of this document may be reproduced, stored in a retrieval system, transmitted in any form or by any means, electronic, mechanical, photocopying, recording or otherwise, without express written permission from Wipro Technologies. All other trademarks mentioned herein are the property of their respective owners. Specifications subject to change without notice.