Conférence – Speakers et abstracts

MODELLING IN LIFE INSURANCE: A MANAGEMENT PERSPECTIVE
6 et 7 octobre 2015

Guillaume ALABERGERE
Head of the internal model unit, ACPR

Shortbio
Head of the internal model unit at ACPR since 2014 After one year and a half of experience as line Supervisor in ACAM – former name of the French insurance supervisory Authority , Guillaume Alabergère joined the new insurance internal model unit once the French supervisor renamed ACPR in 2010. Guillaume Alabergère is also a member of the Internal Model Committee within EIOPA since 2010.

Validation in life modelling, a supervisory point of view

Abstract
After reviewing the Solvency 2 requirements for best estimate calculation models and internal models in terms of validation, we will share some examples of good practices and, then, we will broaden the debate on what could be improved and expected in the future.

Michel BERA
professor, chair of statistical risk Modeling, CNAM, France

Shortbio
Michel Bera has more than  20 years of experience in  developing algorithms for Matra Corporation and Caisse des Depots et Consignations. At KXEN, he was responsible for developing  the company’s algorithms  and business plan. Dr. Bera is a highly respected leader  in the knowledge management field. As an acclaimed author of many published  works, he has received numerous prestigious awards  and is a highly sought after  speaker. Since 2010, he is Full Professor at the Conservatoire National des Arts et Métiers in charge  of the chair Modélisation statistique du risque.

Participated in a roundtable on the topic :
Making sense of big data in insurance

Fabrice BOREL-MATHURIN
Research economist, ACPR

Shortbio
Fabrice Borel-Mathurin is a research economist at ACPR. He also chairs the stress test sub group of the Financial Stability Committee of EIOPA, the European Commission supervisory authority for insurance, reinsurance and occupational pensions. He also teaches stochastic processes and random modelling at École Centrale Paris. Prior to his position at ACPR,he was a front office quantitative analyst in Fixed Income modelling and statistical arbitrage within corporate and investment banks and later became a consultant in financial engineering for small caps or local authorities. While at ACPR he worked in both bank and insurance supervision departments. He holds MSc’s in both pure mathematics (Université Paris 6) and applied mathematics (Université Paris 7). His research focuses on financial modelling and regulatory framework designs with respect to stress testing and balance sheets deformations for insurances and banks.

Top down analysis of the low yield environment for the largest EU insurance companies

Abstract
The downward trend in interest rates, reinforced by accommodating monetary policies in the Eurozone, motivated the introduction of a “satellite exercise” on low yield environments within the 2014 European insurance stress tests. This bottom up stress test was complemented by a top down analysis to conduct some in-depth investigations on the market with the available dataset. Several risk metrics were introduced as a way to score the effects of the different risk factors on insurance undertakings.
The presentation gives some background information on the 2014 insurance stress test exercise and outlines the various methodological approaches we used for this analysis. Eventually, we present our main conclusions at the aggregated level.

Bernard BOLLE-REDDAT
chief risk officer at BNP Paribas Cardif

Shortbio
Bernard Bolle-Reddat graduated from the French Business School Ecole Supérieure des Sciences economiques et Commerciales (ESSEC) and the French Institute of Chartered Accountant. He is Cardif (the insurance sub-group of BNP Paribas) Chief Risk Officer. From 1984 to 1994, he worked as auditor at Ernst & Young, where he specialized himself in the audit of the insurance industry. In 1994, he joined Cardif as Chief Accountant Officer. From 2005 to 2009, he also represented Cardif in the lobbying groups at the French & European levels (CFO Forum) on matters linked to the Embedded Value or IFRS 4 Phase 1 then Phase 2. During the same period, he was appointed member of the insurance Group of the European Financial Reporting Advisory Group (EFRAG). In 2009, he was appointed CRO of Cardif & sponsor of the program of projects aiming at shifting Cardif & its subsidiaries from Solvency II to Solvency II. His responsibilities cover Cardif modelling issues and, in that capacity he is a member of the steering committee of the chair of excellence “Management of modelling” supported by the Laboratoire des Sciences Actuarielles et Financières (SAF) & BNP Paribas Cardif.

Participated in a roundtable on the topic :
The role of models in management decision making

Heidi DELOBELLE
CRO Managing Director, AG Insurance

Shortbio
Heidi Delobelle serves as Chief Risk Officer and Director of AG Insurance SA She is graduate from the Catholic University of Leuven (mathematics and actuarial sciences). From 1994 to 1998, she was consultant for Bio/Mibis (at present Praxis Belgium) for the Belgian & Dutch insurance companies. She joined AG Insurance in 1998 and was appointed Chief Risk Officer in 2012. Heidi was elected as the first female President of «l’Institut des Actuaires en Belgique» in 2011 and is a member of the Council of the International Actuarial Association.

Participated in a roundtable on the topic :
The role of models in management decision making

Thomas BREUER
Professor of statistics and financial mathematics
Director, Research Centre PPE, FH Vorarlberg, Austria

Shortbio
Prof. (FH) Dr. Thomas Breuer is head of the Research Center Process and Product Engineering at the FH Vorarlberg. Previously Breuer was among others in the Oesterreichischen Nationalbank, worked at the Austrian Academy of Sciences and at the University of Cologne at theInstitute for Theoretical Physics. Breuer studied physics at the ETH Zurich and subsequently mathematics at Princeton University in America. Breuer completed his doctorate at the University of Cambridge and completed his habilitation at the University of Salzburg. Since 1999 he is working at the FH Vorarlberg as professorof statistics and financial mathematics.

Systemic stress testing and model risk

Sami FAYE-CHELLALI
Global head of strategy and offer for individual protection, AXA

Shortbio
Sami Faye Chellali has been global head of strategy and offer for individual protection at AXA Global Life for 1.5 years. With 13 years of experience overall in the insurance industry, he spent all his career at AXA taking multiple roles in different companies of the Group: AXA France, AXA Belgium, Group Headquarters. Member of French Institute of Actuaries, Sami received a Master degree in Mathematical Engineering from ISFA, University Lyon 1 in 2001 and a Master degree in Actuarial Science from CNAM in 2007.

Participated in a roundtable on the topic :
Making sense of big data in insurance

Georges DIONNE
Professor in Finance, Head of the Canada Research Chair in Risk Management, HEC Montreal, Canada

Shortbio
Georges Dionne holds the Canada Research Chair in Risk Management. He is past president of EGRIE. Georges awards include the Les Échos (2001) and Kulp-Wright awards (2002) for the Handbook of Insurance. Georges has published 5 books and more than 165 articles peer reviewed. He has been the Editor of the Journal of Risk and Insurance from 2007-01 to 2012-12. One of his main achievements was to propose (with M. Boyer) a new model for automobile insurance pricing that was implemented in Quebec.

Governance of risk management

Abstract
This paper tests the effects of the independence and financial knowledge of directors on risk management and firm value in the gold mining industry. Our original hand-collected database on directors’ financial education, accounting background, and financial experience allows us to test the effect of each dimension of financial knowledge on risk management activities. We show that directors’ financial knowledge increases firm value through the risk management channel. This effect is strengthened by the independence of the directors on the board and on the audit committee. Extending the dimension of education, we show that, following unexpected shocks to gold prices, educated hedgers are more effective than average hedgers in the industry. As a policy implication, our results suggest adding the experience and education dimensions to the 2002 Sarbanes–Oxley Act and New York Stock Exchange requirements for financial literacy.

Michaël DE TOLDI
Head of Data & Analytics, BNP Paribas Cardif

Shortbio
Michael de Toldi is Head of Data & Analytics for BNP Paribas Cardif since 2013. His first mission is to push Analytics and Predictive Modelling techniques everywhere it makes sense in all BNP Paribas Cardif’s countries and businesses: Risk, Finance, Marketing, Ops, Fraud & Money laundering; indeed, each time there is a process, associated with costs and data, there is room for improvement through analytical approaches. His second task is to prepare BNP Paribas Cardif to the (big) data revolution that is currently occurring in its various dimensions: HR, IT, organisation, data driven culture.

Governance for Data & Analytics in insurance

Abstract
For Life and P&C insurance, the systematic use of Data & Analytics will become a strategic growth lever. Profits will materialise through an increased policyholders’ satisfaction thanks to more personalised offers and services and accelerated decision processes. In parallel, insurers will better manage their risks and will gain productivity.
But these benefits will arise only if:
1) Data Science is understood and internalised,
2) Internal & external Data is freed, shared, controlled and secured,
3) IT framework is Data Science friendly.
And woe to those that are late in understanding what is at stake!

Dieter HENDRICKX
Prudential Policy & Financial Stability Adviser, Actuary, National Bank of Belgium

Shortbio
Dieter Hendrickx is currently prudential policy & financial stability adviser at the National Bank of Belgium. He has an MSc degree in business engineering and an MSc degree in financial and actuarial engineering from the Catholic University of Leuven. He joined the operational supervisory team of the former CBFA in April 2010, after having worked for close to four years in the risk management department of a large bancinsurrance group in Belgium. In January 2012 he moved to the policy department of the NBB, where is currently mainly involved in the fields of internal models and financial stability. He is, amongst others, member of the Internal Model Committee and the Financial Stability Committee of EIOPA.

Participated in a roundtable on the topic :
Roles of models in regulation & financial reporting

Pierre-Olivier GIBERT
Chief Executive Officer, Digital & Ethics

Shortbio
Paul is CEO and founder of  Digital & Ethics, which project is to make confortable  with data its clients.  http://www.digital-ethics.com Digital & Ethics is in charge the ethical and regulatory issue of several big data projects for Insurance, Health Care and Gouvernemental Bodies. Previously, he was chief security and compliance officer of AG2R LAMONDIALE, a French pension and life insurance company.

Participated in a roundtable on the topic :
Making sense of big data in insurance

Christophe GEISSLER
Chief Executive Officer, Advestis

Shortbio

Christophe is graduate from Ecole Normale Supérieure Ulm (mathematics and computer science), and member of the French Actuaries Institute. He has occupied various positions in quantitative finance since 1986 in banks (Lazard, CPR, BNP Paribas, Société Générale) before launching Advestis in 2011. His fields of interest include quantitative finance, asset allocation models and machine learning. Christophe is also partner of Quinten, a french company created in 2008, applying data science to healthcare and insurance.

Participated in a roundtable on the topic :
Making sense of big data in insurance

Clément PETIT
Life specialist, internal model unit, ACPR

Shortbio
Clément Petit joined the internal model unit of the French insurance supervisory Authority (ACPR) in 2011 and is specialized in life modelling under Solvency 2. He led some reviews on internal models, best estimate and SCR computation and he is also in charge of the Solvency 2 ACPR working group on life technical provision and life/market SCR. He had former experiences in audit and consulting at Ernst&Young and in a saving products team and then life modelling at BNP Paribas Cardif.

Validation in life modelling, a supervisory point of view

Abstract
After reviewing the Solvency 2 requirements for best estimate calculation models and internal models in terms of validation, we will share some examples of good practices and, then, we will broaden the debate on what could be improved and expected in the future.

Antoon PELSSER
Full Professor of Finance and Actuarial Science, Maastricht University, The Netherlands

Shortbio
Antoon Pelsser is working at Kleynen Consultants and is also a Full Professor of Finance and Actuarial Science at Maastricht University and a research fellow at Netspar. His academic research interests focus on pricing models for interest rate derivatives, the pricing of insurance contracts and Asset-Liability Management (ALM) for insurance companies. He has published in leading academic journals including Journal of Economic Theory, Mathematical Finance, Finance and Stochastics, Journal of Derivatives, Insurance: Mathematics and Economics. He is an Honorary Fellow of the Institute of Actuaries. From 2004 until 2007 he worked at ING Group’s staff department Corporate Insurance Risk Management. There he was involved in implementing a new internal model for measuring EconomicCapital for ING-Insurance. From 2000 until 2004 heworked as Head of ALM for Nationale-Nederlanden. Before that, he worked 7 years in the dealing-room of ABN-Amro Bank in Amsterdam, where he was responsible for the development of pricing models for derivatives.

The difference between LSMC and Replicating Portfolio in Insurance Liability Modelling

Abstract
Solvency II requires insurers to calculate the one-year Value at Risk (VaR) of their balance sheet. This involves the valuation of the balance sheet in one year’s time. As for insurance liabilities closed-form solutions to their value are generally not available, insurers turn to estimation procedures. While pure Monte Carlo simulation set-ups are theoretically sound, they are often infeasible in practice. Therefore, approximation methods are exploited. Among these Least Squares Monte Carlo (LSMC) and portfolio replication are prominent and widely applied in practice. In this paper we show that while both are variants of regression-based Monte Carlo methods, they differ in one significant aspect. While the replicating portfolio approach only contains an approximation error, which converges to zero in the limit, in LSMC additionally a projection error is present, which cannot be eliminated. It is revealed that the replicating portfolio technique enjoys numerous advantages and is therefore an attractive model choice.

David INGRAM
CERA, FRM, PRM, FSA, MAAA Executive Vice President, Willis Re

Shortbio
Dave works with Willis Re’s insurance company clients to develop and improve their Enterprise Risk Management practices. He was previously in the Insurance Ratings Group of Standard and Poor’s (S&P) where he led their initiative to incorporate ERM into insurance ratings. Dave has also held executive positions within insurance companies. Dave is a frequent writer and speaker on ERM. He is currently
the Chair of the International Actuarial Association’s Enterprise and Financial Risk Committee and Chair of the US Actuarial Standards Board ERM Committee.

Bridging the gap between Managers and Models

Abstract
Most managers have achieved their success, not by the adept use of models, but by the accuracy of their intuitions. They are now faced with a situation where the success of the insurer is determined, at least in part, by convincing a regulator that management is actually using a very complex model to help to make the most important decisions about the insurer’s future. Modelers can learn more about how that intuition works so that we can communicate better about why the manager should attribute some credence to the model. In addition, modelers and managers can have a more robust dialog to talk about the ways that the models are more or less reliable in different situations, including a discussion of whether “now” is one of those “different situations”. This session will focus on the different perspectives that often prevent modelers and managers from communicating.
Those different perspectives have great potential value to an insurer, so suggestions for improving communications will be focused not on getting (or forcing) everyone to agree but on accepting the differences in perspective.

Michael SCHMUTZ
Lecturer, University of Bern, Senior Risk Manager, Swiss Financial Market Supervisory Authority (FINMA)

Shortbio
Michael Schmutz is a Senior Risk Manager at the Swiss Financial Supervisory Authority (FINMA) and a Lecturer at the Universities of Bern, Basel and at EPFL. He has several years of experience in risk management and financial accounting. Michael holds a Ph.D. in Mathematics from the University of Bern and a Diploma in Business Administration from Bern University of Applied Sciences.

Group solvency tests, intragroup transfers and intragroup diversification: a set-valued perspective

Abstract
The aim of risk-based solvency frameworks, such as Solvency II to be introduced in the EU and the Swiss Solvency Test (SST) that has been in force in Switzerland since 2011, is to assess the financialhealth of insurance companies. This is achieved by quantifying capital adequacy, calculated from the solvency capital requirement (SCR). These calculations are based on scalar risk measures. The assessment of the financial health of insurance groups (of several connected companies) is an even more challenging task, since e.g. fungibility constraints usually exist that have an impact on diversification.
We discuss transfer constraints and intragroup diversification based on a set-valued perspective.

Andreas TSANAKAS
Reader in Actuarial Science, Cass Business School, City University, UK

Shortbio
Dr Andreas Tsanakas is a Reader in Actuarial Science at Cass Business School, which he joined in 2006. Prior to this, he was at Lloyd’s, mainly involved in capital modelling projects. His research interests are in quantitative risk management, with particular focus on measuring portfolio risks, capital allocation, and dealing with model uncertainty. Andreas often speaks at insurance industry events on his areas of expertise, particularly economic capital modelling. Andreas studied Electrical and Computer Engineering at the University of Patras, Greece, and carried out his doctoral research on risk management at Imperial College London. He also holds an MSc in Control Systems from Imperial College London and an MA in Modern German Studies from Birkbeck College.

Model Risk Culture

Abstract
Model risk can be understood as the risk to a decision maker arising from specific ways of (not) using a quantitative model. Model risks range from basing decisions on flawed calculations, to impacts on perception, behaviour and accountability. Building on anthropological theories of risk, we stipulate that there are four distinct perceptions of models and their legitimate use. Each such perception is justified by, but also generates, concern for a different type of uncertainty. Furthermore, each distinct way of perceiving and (not) using models, when deployed on its own, generates a different sort of risk. This argument has practical implications for risk management and governance. First, it highlights the required pluralism of stakeholders involved in model governance and validation. Second, it expands the space of legitimate challenges to quantitative models from a purely technical critique, in order to also reflect operational and commercial criteria. In particular, accepting that the concerns of a variety of stakeholders are legitimate, supports both the scientific integrity of modellers and the accountability of management. Joint work with Bruce Beck, Michael Thompson and the Institute & Faculty of Actuaries Working Party on Model Risk.

Pierre THEROND
Associate Lecturer, ISFA, University Lyon 1, Member of SAF Laboratory
Partner, GALEA & Associés

Shortbio
Pierre Thérond holds a PhD in management sciences from University of Lyon and is a former student of ISFA. He is fellow of the Institut des Actuaires. He is now partner of the consulting company Galea & Associés and associate Professor at ISFA, Université Lyon 1. Pierre main’s research interests include measurement of insurance and financial risks (especially for solvency, accounting and financial reporting purposes), insurance risk management and policyholder behavior. He received the SCOR PhD award in 2007. Pierre is a member of the editorial Board of the French actuarial magazine l’actuariel. He chairs the accounting committee of the French Institut des Actuaires.

Participated in a roundtable on the topic :
Roles of models in regulation & financial reporting