Insights

Financial modelling
Topical articles, tutorials, case studies and news Welcome
It gives me great pleasure to be welcoming back our readers after a lay-off of the best part of a year. 2010 has proved to be challenging as well as exciting from both an industry Joel Fox and Towers Watson perspective. From an industry position, 2010 saw substantially more activity as companies have geared up for Solvency II, with increasing amounts of time spent on the design and implementation of internal models. Significant consideration has also been given to exploring how Solvency II solutions integrate with other emerging requirements, such as IFRS Phase II and BAS standards in the UK. 2011 is guaranteed to be a busy year! On a more internally focused note, this is our first edition of this bulletin since the Towers Perrin and Watson Wyatt merger on 1 January 2010. The merger has seen the amalgamation of two organisations that were very much at the forefront of developing and implementing actuarial software solutions. The decision to sell the VIPitech business in June 2010, the sale of one of our softwares being a condition of the merger, hopefully served to demonstrate our commitment to MoSes and RiskAgility and their future roadmaps. The merger has also seen the rebranding of this publication from the Financial Modelling Update to Towers Watsons Insights: Financial Modelling. I hope that you find this issue both interesting and informative and welcome any comments that you have regarding future editions.
Joel Fox Director, Towers Watson

February 2011

Internal models
As life insurers focus on preparing for Solvency II, many are devoting significant resources to developing their own internal models. Neil Chapman, senior consultant within the Risk Consulting and Software business of Towers Watson, discusses why they are making the effort.
Across the EU, life insurance companies are hard at work preparing for the upcoming Solvency II regime. Among the many challenging decisions they face, a key one is whether to use the standard approach to calculate the amount of capital they need to hold, or build their own tailored internal model.

In this issue
1 Internal models Why develop an internal model? 4 Success factors for model implementation Nationale-Nederlanden case study. An example of how to use the V-diagram in practice

7 In-depth tutorial: The standard approach is the simpler default option Building dedicated and will be applied to companies across Europe. Our tutorial article in this issue gives an example of a risk reporting models capital aggregation calculation, based on a correlation in MoSes How to use MoSes matrix approach, similar to the Solvency II standard to minimise manual approach. As the name standard approach implies, operations during it is a more generic approach which does not reporting necessarily reflect the companys individual risk
profile. However, insurers can overcome this deficiency by developing their own internal model, although they will require supervisory approval to use it for regulatory capital calculations.

What is an internal model?

An internal model can be best described as a risk management system developed by an insurer to analyse the overall risk position, to quantify risks and to determine the amount of capital needed to meet those risks 1. It is far more than a complex number crunching programme. It is the key process that leading companies use to quantify, monitor and manage risk and is central to asset liability management, capital management, strategic decision taking and in some cases executive remuneration.

12 MoSes user group What happened at the European and Korean MoSes user groups? 13 The back page News from the world of financial modelling

Source: International Association of Insurance Supervisors

Insights | February 2011

Figure 01. The relationship between risk, capital and value

Risk
io n

s Ri

a qu

i nt

t ca

Ec

on

om

ic

va

lu

Capital

Capital utilisation

Value

With the Internal Model methodology providing a robust relationship between risk, capital and value, the natural next step is for insurers to use this framework for competitive advantage.

Why develop an internal model?

Life insurers are in the business of creating value, which typically involves taking and managing risks. As a companys capital base is typically constrained, in the short term at least, the main goal is usually to maximise stakeholder value within current capital resources. To do this, companies need to understand and communicate throughout the organisation the relationship between risk, capital and value (Figure 01). However, many insurance companies have evolved without a consistent view on the interaction of risk, capital and value within the firm. In practice, companies apply varied metrics based on a variety of accounting, local regulatory and economic measures. These measures can give conflicting signals making it difficult to manage the company and justify key business decisions. Solvency II has the potential to change all this as it sets out a single framework which places values on the assets and liabilities consistent with their observed market prices. In effect, it allows insurers to apply a common currency to the measurement of risk, value and capital across the entire business. The Solvency II framework is very similar to the economic framework which leading companies have developed to help steer the business. Solvency II therefore represents an opportunity for companies to apply the models and methodologies that they have developed for their own decision-making processes to the supervision of the company. This in effect will align the information used by management and supervisors and should simplify the management of the business. The challenges in developing an internal model should not be underestimated and it is likely to require significant senior management input,

close interaction with the relevant supervisor and dedicated resources from an early stage. Even in the UK, where insurers have had some prior experience with economic capital models, a significant amount of work will be required to extend and enhance these to meet the high standards expected of internal models under Solvency II.

Advantages of internal models

With the Internal Model methodology providing a robust relationship between risk, capital and value, the natural next step is for insurers to use this framework for competitive advantage. One vital advantage of using an internal model is that it more accurately reflects the amount of capital that a company needs to hold. This is because an internal model is designed specifically for the risk profile, exposures and mitigation strategy of that individual company. In order to create the proper incentive for companies to move to internal models, it is expected that the standard approach alternative will generally be calibrated to give a conservative view of the capital requirements of the business. In addition, companies can use an internal model framework to identify and implement the best risk-adjusted business strategies. At any one time, management will need to consider a range of strategic options, some of which might involve reducing risk, for example via reinsurance, hedging or securitisation arrangements. Other strategic options, such as the launch of new products or managing sales/profitability targets, might add new risks to the balance sheet. A key way in which companies can use an internal model to add value is to optimise diversification effects between different risks thereby allowing a more efficient use of capital.

2 towerswatson.com

Insights | February 2011

Internal models also provide insurers with the tools to perform more frequent and granular assessment of their risks. This means that companies can continuously monitor and fine tune their overall risk positions. In the advent of shocks such as the recent financial crisis, internal models allow the company to quickly assess the implications and to take corrective action. Overall this should lead to a more secure insurance industry with reduced risk of company failure.

plays under Solvency II in providing protection and security to policyholders it will not come as a surprise that obtaining such approval is likely to be a lengthy and involved process. Using the UK as an example, the FSA has built up a large dedicated team for the very large number of models they expect to have to approve ahead of Solvency II. This will take the form of detailed reviews of internal model documentation, analysis and benchmarking of results and numerous workshops and meetings with company staff. In particular, and as per the Solvency II Directive, the financial regulator in each country will need to satisfy themselves that the companys internal model can satisfy six key criteria. These are set out in the table below, along with our thoughts on how best to meet them.

What companies will need to do to get supervisory approval

Before a company can use its internal model to calculate capital requirements under Solvency II, it will need to have the model approved by its supervisors. Given the role that this capital

Table 01. Key requirements of internal models

Test Key requirements and how best to meet them

Use test

Companies will need to demonstrate to supervisors that their internal model is widely used and plays an important role in key risk decisions. In order to do this companies will need to have a compelling story explaining how their internal model fits their business model, is integrated into risk management systems, supports strategic decision taking and is well understood. Being able to evidence this and, in particular, the involvement and understanding of the Board and senior management will be critical. Companies documentation will need to provide a detailed outline of the theory, assumptions, and empirical basis for the internal model as well as highlighting any circumstances under which it may not work effectively. A documentation hierarchy with clear and distinct objectives will be the most effective approach. For example, Board papers should not have the same level of technical content as methodology papers detailing the inner workings of the model. Having appropriate documentation, especially at Board and senior management will be an important element in meeting the Use test described above. The methods, assumptions and data underlying the internal model will need to be approved by supervisors. Being able to demonstrate why the internal model is a material improvement relative to the standard formula in terms of capturing the companys specific risk exposures and characteristics will be essential. This criteria requires supervisors to ensure that companies internal models are generally comparable across the EU after allowing for specific characteristics. Companies are likely to be required to produce and explain benchmark results relative to results of other companies internal models and the standard formula. Companies are required to demonstrate to supervisors that they have the processes and controls to ensure that their models will remain appropriate in the future. This will require companies to have in place continuous cycles of validation. As well as updating assumptions each year, companies will need to assess whether the methodology and data sources used are still the most appropriate, whether all their risks exposures are still adequately captured and whether their model is being used appropriately. Having clear independence between the assessors and owners of the model and an appropriate level of challenge will be essential. This requires companies to have appropriate processes and tools to identify and explain all material sources of profit and loss. In practice, this would form a key part of the validation process and so is unlikely to be a significant additional requirement.

Documentation standards

Statistical quality test

Calibration standards

Validation standards

Profit and loss attribution

towerswatson.com 3

Insights | February 2011

Solvency II in a nutshell
Solvency II is an EU-wide insurance directive which will revolutionise the sector when it is implemented in 2012. The new regulatory regime is designed to provide a principles-based supervisory framework for European insurers and reinsurers, and will be a key business priority for insurers across the region over the next few years. The risk-based system will put an understanding of risk and capital at the heart of insurers business decisions, basing the amount of capital an insurer is required to hold directly on the risks to which it is exposed to. The intention is to ensure that insurance companies hold sufficient capital to withstand a loss so severe it would happen only once every 200 years. For organisations to effectively quantify this level of capital, many will have to develop a more accurate view of what their risks are and how much capital should be held against each of them. Insurers across Europe will have a significant amount of preparation to do involving investment in new systems and controls, amending governance structures and risk policies that need to be embedded into the business at every level. Also many insurance companies, particularly the more complex organisations, will want to build tailored internal models and risk reporting systems. As well as the technical side, insurers will also need to look at the expertise within their organisation, and in many cases, work to upgrade their risk management skills. Despite these challenges, Solvency II will deliver significant business benefits to the sector, which will make all the hard work worthwhile. For example, because Solvency II will demand that organisations have a much improved visibility of risk within their business, as well as the trade off between risk and reward, management teams will be able to make better-informed decisions. This could lead to improvements in areas such as pricing, reinsurance and investment strategy and product innovation. Under Solvency II, companies are required to have an enhanced understanding and management of risk, which will reduce the risk of corporate failure within the sector. Solvency II will also ensure a more consistent standard of solvency across the European insurance industry.

Success factors for model implementation NationaleNederlanden case study

Introduction
In this case study, we focus on a MoSes implementation project for the individual life insurance business of Nationale-Nederlanden, based in the Netherlands, and discuss the reasons for the success of this project.

It is a flexible model with the principle that new products or new portfolios can be added without changing the code only product and insurance parameters need to be changed. This implementation project has been successful in many aspects. These include the management and progress of the project, the confidence in the capabilities of the model and the awareness within NN of the possibilities of a projection model for all reporting purposes (replacing several different legacy systems which were used for different reporting purposes). The following factors have contributed to the success of the project, and we discuss these in more detail in this article:

Background
Nationale-Nederlanden (NN) is part of the ING Group, a worldwide supplier of financial services with its origins in the Netherlands. NN has chosen Towers Watson to jointly develop a projection model (called Forecast) for the life insurance portfolio. The model projects both the traditional and unit linked portfolios and is used for MCEV reporting and, in later phases, will serve as the basis for Solvency II and IFRS Phase 2 calculations.

Phasing of specification, development and testing. Applying the V-diagram process. Acknowledging the importance of testing during the implementation.

4 towerswatson.com

Insights | February 2011

Planning in phases
NN has chosen to split the development into several phases with different parts of the business being developed in different phases. The Annuity portfolio was developed first because this is a relatively simple portfolio without the complication of regular premiums, surrenders, paid ups, commissions and profit sharing. After completion of the first phase, other parts of the portfolio were implemented, where the complexity of the portfolio dictated the order of implementation. The second phase, for example, consisted of the traditional portfolio without profit sharing (with regular premiums, surrenders, paid up and commissions), and subsequent phases were the unit linked portfolio and the profit sharing traditional portfolio. Within each phase the implementation was further split into the components of the functionality such as premium, insurance coverage, commission, provisions, surrender functionality and so on.

The advantage of planning in phases is twofold: the model can be taken into production phase by phase (and prove its use in the day-to-day business of NN), and the testing of the model can be done per component (as described under the V-Process below).

V-process
Together with Towers Watson, NN has defined an implementation process based on the V-diagram (see Figure 01). This V-process was applied to each phase and each component of the project and has proved a solid basis for the management of the project. Elements of the V-Process are performed by several parties: NN, Towers Watson and two other consultancy firms (for project management and testing). The good cooperation between the parties adds to the smooth implementation of the Forecast model.

Figure 01. The five-step V-model

Business requirement User acceptance and deployment

Business

System design
Software development

Detailed specication

Functional and system testing

Software development

Development and unit testing

Project management

The five-step V-model provides a robust and easy-to-follow map of the development process, covering the following stages in an implementation project: Business requirements System design and detailed specification Development and unit testing Functional testing System testing, user acceptance and deployment

Business requirements and system design

The business requirements contain high-level objectives defining the project scope, global design deliverables, timeframe and roles and responsibilities. The system design converts the detailed specifications into a MoSes model design.

Business

towerswatson.com 5

Insights | February 2011

NN discussed its objectives and requirements for the model internally and subsequently asked Towers Watson to advise on the feasibility of these requirements. NN and Towers Watson jointly held several workshops to discuss these requirements and the system design of the projection model.

System testing, user acceptance testing and deployment

When the model is ready to take into production, the last stages of testing are performed. System testing verifies that the model is stable and performance is as required. It also checks end-to-end processing such as interfaces. User acceptance testing verifies that the model meets expectations and requirements as set out in the requirements document (performed by the end-user). Deployment delivers the model into a production environment including final training, finalising the model, documentation and operation and support procedures. User acceptance testing was performed by NN on the actual portfolio policy data (whereas the functional testing was performed using fictional policies) and against the legacy model. This was done at the end of each phase of development.

Detailed specifications
The detailed specifications explain the required calculations for the model including policy data, assumptions, output and products specifications. For each of the components of the model (probability system, claims, and so on), NN delivered a requirements document. Towers Watson then analysed this document, discussed with NN and finally delivered a detailed specification. This detailed specification is a translation of the requirements document to the model and describes the functionality and the modelling decisions made (use of smartarrays, scalars, approximations if needed, and so on). Separate detailed specifications were produced for policy data, assumptions, functionality and output.

Advantages of different stages of testing

Functional and user acceptance testing of the model are both important phases of the project as they verify that the model meets the functional and user requirements, respectively. Detailed functional testing, especially with a shadow tool as NN has done, and user acceptance testing has the following advantages: Internal It results in users having strong confidence in the model and its results. In our experience, this encourages and accelerates the timeframes involved in deploying the new models into the production environment. Confidence in the model results will eventually lead to shorter periods needed for analysis. Time needed for analysis of results has been moved away from the production phase to the testing phase. Analysis of results will still need to be performed in production runs, but no time needs to be spent on proving that the model is correct. External Test results, test documentation and the model documentation result in a smooth model validation process. The model validation will help convince third parties, for example auditors, of the use of the results of the model. As models like the Forecast model are the basis for more and more reports (MCEV, profit testing, in later phases IFRS and Solvency II) this is an increasingly important part of deployment of the model.

Model development, documentation and unit testing

Upon approval from NN of the detailed specifications, Towers Watson developed the functionality in the MoSes environment following our Implementation Process and Standards Guide (see Financial Modelling Update Issue 7 January 2008). This stage of the implementation included documentation of the model and unit testing, for instance, code testing during the development of specific functionality.

Functional testing
The second stage of testing, functional testing, verifies that the model meets functional requirements, including reasonableness of the results. NN carried out their functional testing against the requirements as set out in the agreed detailed specification documents, and against the testing plans and test sets that they produced once the development of a component had finished. NN recognised the importance of the testing process and have developed a shadow test tool (in VBA) with which they test the functional requirements of the MoSes model. For each of the phases and components, a large set of policies is tested against the test tool. This ensures the functionality is thoroughly tested and NN is confident of the results when taking the model into production.

6 towerswatson.com

Insights | February 2011

Building dedicated reporting models in MoSes

Introduction
This article focuses on how MoSes can be used to build dedicated reporting models and to minimise manual operations outside of MoSes (for example, in Excel). The method discussed uses a projection task to consolidate selected items of pre-calculated Column Output files into meaningful reports that are then presented in User Defined View (UDV) worksheets. The projection task performs the following steps which are explained later: Extract and aggregate selected columns from a selection of Column Output Files into FoxPro Tables. Read these FoxPro tables into a submodel of our reporting model to present results in a UDV. Use the top model to derive additional results and to present overarching result overviews. To illustrate these steps we have developed a simple example using our C-ALM application. We have set up a batch run for a stochastic full- company calculation with varying assumptions. The sets of assumptions consist of a base set and several sets where selected risk parameters have been shocked. In our example, we have set up a UDV report with a market value balance sheet for each shocked calculation and an overarching report where a risk capital aggregation is performed, based on a correlation matrix approach, similar to the Solvency II standard approach.

Structure of the reporting model

The reporting model consists of a top model named export and a submodel with type Array User Defined named pv_report. See Financial Modelling Update (FMU) issue 11 and 12 (March 2009 and June 2009) for more details on submodel types.

Figure 01. Model tree

The following variables are required:

Name of the variable

report_files_tbl field_names_tbl output_path output_path_avg

Type
Table of type Row Names(strings) Table of type Row Names(strings) String String

Description
List of Column Output Files to be included in the report List of field names to be extracted from the Column output files Name of the file path which includes the Column Output files Name of the file path where the files with averaged results are stored

When a projection is started, the code in the initialise column of the top model checks in the table report_files_tbl how many result files need to be included and resizes the submodel array accordingly.

Step 1:
Extract and aggregate columns from a selection of Column Output Files into FoxPro tables.

towerswatson.com 7

Insights | February 2011

The code in the initialise column in the top model loops through the table report_files_tbl to read the names and the descriptions of the Column Output Files and to assign them to the submodel instances.

Figure 02. report_files_tbl

Dedicated reporting models can easily be built and customised in MoSes. As a result, manual operations outside of MoSes can be significantly reduced.

Figure 03. initialise(t)

In the same loop the External Function read_column_output() is called to build output files containing averages of the results and to filter all periods that are different from t=0 (see Figure 04). All columns of the FoxPro tables with column number greater than eight are averaged. The first eight columns include information about the product, purpose, group and time step and are not required for our purpose.

Figure 04. read_column_output(file_name)

8 towerswatson.com

Insights | February 2011

After performing the initialise step in the top model we have now created seven FoxPro tables with results averaged across iterations, under the same names as the original Column Output Files but in a different folder. The FoxPro tables query in read_column_output depends on the format of the Column Output Files. For the code given above, the Column Output Files must sorted by group.

Step 2:
Read these FoxPro tables into a submodel of our reporting model to present results in a UDV At the end of the initialise column of the top model, the initialise columns of the submodels are called. Here the required present values that have been listed in the table field_names_tbl are read into a SmartArray called pv_matrix. pv_matrix is class variable that has been defined in the CLASS section of the startup column of the pv_report submodel.

Figure 05. read_pv_values(void)

To be able to read the present values into the model the header names are used.

Figure 06. List of Headers to be read in (field_names_tbl)

towerswatson.com 9

Insights | February 2011

For each item that is stored in the SmartArray, the model includes a corresponding column. A UDV is then linked to that Column Output. The syntax is as follows:

PV_CLM_DTH is a placeholder that we have defined in the startup column using the #define compiler directive. We have done this to make the code easy to read. In our example, we only read in results at t=0. Figure 07 shows a UDV with the Market Value Balance Sheet. For each shock, the Market Value Balance Sheet can be compared with the Market Value Balance Sheet of the base case.

Figure 07. UDV with Market Value Balance Sheet

The UDV includes two Column Output objects which are both linked to period t=0 only. The left column is linked with a group selector to be able to choose the different shock results and the right column is linked to the base case results.

Step 3:
Use the top model to derive additional results and to present overarching result overviews In the UDV, the available capital is calculated as market value assets less market value liabilities. Furthermore, the standalone risk capital SCRi is defined as the difference between the available capital of the base and shock results. To calculate the total risk capital SCRTotal a correlation matrix Pij is applied. Altogether, we can derive a result overview with the risk capital contribution of each risk factor.
9

10 towerswatson.com

Insights | February 2011

To be able to present the results in a UDV, we pull up the available capital column from the submodels, but not as a sum of the underlying submodels. We do not interpret t as time period anymore. Instead, we numerate the shock results (in our case from one to seven) and pull up the results for the corresponding system variable t. t=0 is used for totals.

Figure 08. risk_capital_standalone(t)

For example, Equity_down has the calculation number 5 (see also Figure 02). This means that in the top model risk_capital_Standalone(5) = risk_capital_standalone(Equity_down). Applying this approach we can now choose the relevant results and show them in an aggregated result overview.

Figure 09. Risk Capital aggregation in a UDV

The UDV in Figure 09 shows the values from the columns risk_capital_standalone(t), diversification(t) and risk_capital_contribution(t) for t=1,,6. In addition the column risk_capital_total(t) for t=0. Hereby the diversification DFi and the risk capital contribution CONTRi for the risk i are calculated as:

Conclusion
Dedicated reporting models can easily be built and customised in MoSes. The method described in this article is one possible example. Similar techniques can also be used for movement analysis reports or other purposes. As a result, manual operations outside of MoSes can significantly be reduced. Ideally, there is no manual operation required anymore between pressing the button to start the projections and receiving a final (aggregated) report directly from MoSes.

towerswatson.com 11

Insights | February 2011

MoSes User Groups 2010

What happened at the European and Korean User Groups?
Two MoSes User Groups were held recently across the European/APAC region and we provide here a brief summary of both events. Many of the presentations highlighted the modelling implications of changes in the regulatory environment, such as, capital requirement, liability adequacy testing and reserving for O&G: Required Capital: The Korean regulator (FSS: Financial Supervisory Service) announced in 2009 that it plans to replace Koreas solvency measure (EU solvency I type) with a risk-based capital (RBC) measure. The system is now moving forward from standard models to an internal model approach. LAT for Korean International Financial Reporting Standards (K-IFRS): The IFRS adopted in Korea will be the word-for-word IFRS as issued by Internal Accounting Standards Board without modification, which implies requirement on LAT. Reserving for Options and Guarantees of Variable Products: The most commonly used methodology to calculate reserves for cost of options and guarantees (O&Gs) is generally an accumulation of O&G fees collected. Reserves from end 2010 are required to be calculated using stochastic approach; maximum of 70% of Conditional Tail Expectation (CTE) of net loss and the standard O&G reserve. We highlighted to the participants that MoSes is an excellent tool to support a RiskCapitalValue (R-C-V) framework, an overarching framework to not only prepare for Korean RBC, but also for wider usage in fair valuation and economic capital, in line with global best practice. Feedback from the event was very positive with 75% of the respondents rating the contents of the overall programme as good or excellent and 65% having an improved understanding of how to make best use of MoSes after attending the event. O&G Stochastic Reserves and LAT were selected as the two key areas of immediate modelling needs.

European MoSes user group

The 2010 European MoSes User Group took place in London on 20 and 21 October and was attended by over 75 clients and Towers Watson consultants from across Europe and further afield. In developing the agenda, we sought to include a varied range of topics including those covering both business and technical issues. Solvency II was a recurring theme throughout many of the presentations. We were pleased to be joined by three external speakers who all presented interesting and engaging talks Guy Shepherd from Prudential, Morena DellAgio from Akhela and Peter Foy from XChanging. The feedback from the event was very positive with 87% rating the overall conference as good or excellent. Particularly well received sessions included MoSes and the Application of Grid Computing to Actuarial Modelling in Prudential, Solvency II: The State of Play and a MoSes tutorial on Advanced Reporting Techniques. Due to the positive feedback received on the Advanced Reporting session, we have provided a more detailed write-up in this issues tutorial. We have now started planning our 2011 conference and will update you with more details in due course. Please get in touch if you have any suggestions for future presentations.

Korean MoSes User Group

The Korean MoSes User Group was held on 18 November 2010 at Towers Watsons Seoul office. The meeting focused on the theme of advanced modelling using MoSes to meet the requirements of Koreas new regulatory regime. Topics covered include Risk-Based Capital (RBC) and Valuation Models, Liability Adequacy Test (LAT), Option and Guarantee Stochastic Reserves and Whats new with MoSes?. It was well received throughout the meeting that MoSes is an efficient tool to prepare and lead the way in the changing environment. We had 100% participation from all active MoSes users in Korea. A total of 34 people attended the meeting, representing 11 insurance bodies including both life and non-life insurance companies and the Korean Insurance Development Institute.

In developing the agenda, we sought to include a varied range of topics including those covering both business and technical issues. Solvency II was a recurring theme throughout many of the presentations.

12 towerswatson.com

Insights | February 2011

News from the world of financial modelling

Provisional training course dates
Below is a list of training courses scheduled in our London office for 2011. The course dates will be finally confirmed based upon registration.

Dates
Q1 2011 Q2 2011 Q3 2011 Q4 2011

Users
7 8 March 13 14 June 5 6 September 5 6 December

Foundation developer
10 11 March 16 17 June 8 9 September 8 9 December

Intermediate developer
14 15 March 20 21 June 12 13 September 12 13 December

Further information
For more information contact [email protected] to register your interest or request further details. Alternatively, if you would like to speak to someone please contact: Merryl John +44 20 7170 2537 [email protected] Courses may be also be arranged in other territories by speaking to your local MoSes representative.

Recent Towers Watson papers

2010 saw the publication of many new Insights papers, Emphasis newsletters and other publications by Towers Watson. The following gives a flavour of those relevant to financial modelling. Emphasis Lessons from the Industrial Revolution: New Actuarial Reporting Systems and Processes As the actuarial profession implements new processes and develops new tools in preparation for Solvency II and other imminent regulatory changes, its members would do well to remember the important lessons of the industrial revolution. To access the paper click the icon, which will take you to the newsletter. Insights: Measuring Risk Culture under Solvency II, December 2010 This document sets out our solution to enable companies to make a quick start on measuring and improving risk culture and associated risk management, and discusses, our experience of doing this in practice.

towerswatson.com 13

Insights | February 2011

Contacts
Regional support centres
Europe, Middle East and Africa
+44 20 7170 3000 [email protected] [email protected]

Asia-Pacific
+61 2 8198 9008

FMU Europe/Asia-Pacific editorial team

Lucy Vickers Julie Have +44 20 7170 2561 +44 20 7170 2078 [email protected] [email protected]

European/EMEA contacts
Belgium
Valentin Bauwens +32 2 678 1537 +33 1 53 93 1403 +49 221 9212 3426 +353 1 614 6855 +39 02 7787 2329 +31 88 543 3053 +46 8 5064 1785 +34 91 590 3075 +41 43 488 4483 +44 20 7170 5221 [email protected] [email protected] [email protected] [email protected] [email protected] [email protected] [email protected] [email protected] [email protected] [email protected]

France
Guillaume Beneteau

Germany/Austria
Christian Naecker

Ireland
Orla Walsh

Italy
Simona Parise

Netherlands
Juriaan Borst

Scandinavia
Roland Kristen

Spain/Portugal
Ana Escudero

Switzerland
Bernhard Gose

UK, Middle East and South Africa

David Pond

Asia-Pacific contacts
Sherry Du +852 2820 0027 [email protected]

14 towerswatson.com

Insights | February 2011

towerswatson.com 15

Insights | February 2011

About Towers Watson

Towers Watson is a leading global professional services company that helps organisations improve performance through effective people, risk and financial management. With 14,000 associates around the world, we offer solutions in the areas of employee benefits, talent management, rewards, and risk and capital management.

Towers Watson Limited, Towers Watson UK Limited and Towers Watson Capital Markets Limited are authorised and regulated by the Financial Services Authority. The information in this publication is of general interest and guidance. Action should not be taken on the basis of any article without seeking specific advice. To unsubscribe, email [email protected] with the publication name as the subject and include your name, title and company address. Copyright 2011 Towers Watson. All rights reserved. TW-EU-2010-19046. February 2011.

towerswatson.com