Research Projects

 Sponsors  

RESEARCH contracts 

 Applied research contract

2014-2018

SAS Partnership

The SAS software is one of the most used statistical software in the world.
Since several years, there exist a partenariat between SAS and Institut de Statistique, Biostatistique et Sciences Actuarielles (ISBA)
through which courses of programming in SAS and data mining techniques are organized.
These courses are open to all master students as well as to PhD students and to all researchers of the UCL.
Within the context of this partenariat, SAS also support (financially and logistically) the organisation of short courses within ISBA.
 

Promoter: Catherine Legrand

BSP (Belgian Science Policy)

2016-2021

BRAIN-be
Valorisation de 70 ans d'observations soclaires de l'Observatoire Royal de Belgique

This is an interdisciplinary project (VAL-U-SUN) between the Solar Influence Data Analysis Center of the Royal Observatory Belgium (Uccle, Brussels), Drs Laure Lefevre et Veronique Delouille, and l’ISBA, on the analysis of the international Sunspot Index. In a four years’ PhD project, the task is to statistically characterise the sunpot data set, by developing a model for its short and long term behaviour over time, including its statistical uncertainties. The overall goal is to come up with an automated quality control algorithm that allows (almost) online supervision of the evolution of the data reported by a collection of sunspot observation station around the world that contribute to establishing a statistically “clean” index. Assistance to this project is given by the SMCS (Christian Ritter).
 
 

Promoter: Rainer von Sachs, Researcher: Sophie Mathieu

FNRS
CDR

2019-

Modeling of jump clustering for managing interest rates and creditrisks

Our research focuses on the modelling of jump clustering for interest rates and default risks with Hawkes processes. A team of two PhD students and me explores two axis. The first axis focuses on the modelling of jump clustering in interest rates. Our main purpose is to explain the spread between swap rates of same maturity but of different tenors, arising from the risk of liquidity breakdowns in the interbank market. The second axis focuses on the actuarial evaluation of credit risk. Insurance companies propose a wide variety of credit coverages e.g. trade credit insurance, business credit insurance, export credit insurance. Most of stakeholders in these contracts are non-listed small or medium companies and there is no alternative to diversification for hedging the insurer’s exposure to default risk. We will focus on the valuation of these contracts with actuarial techniques when the intensity of default is a jump diffusion process with and without jump clustering. Both axis are related through the fact that we aim to model events of bankruptcy. For axis 1, we study the bankruptcy of the interbank market while we analyse the default risk of a firm in the second project.
 

Promoter: Donatien Hainaut

 

FNRS
CDR

2018-2019

Optimal transport in nonparametric statics: Copulas for non-Euclidena data and multivariate tail quantile countrous

The aim is to explore the potential of concepts and methods from the theory of optimal measure transport for statistical modelling and inference. We will extend the classical probability integral and quantile transforms on the real line to more general state spaces thanks to cost-minimizing mappings pushing a reference measure towards a target measure. Based on these transforms, we will seek to extend Sklar's celebrated copula theorem to more general metric spaces. The unit circle will serve as test case, and will allow us to model dependence between directional data. Further, we will study properties of tail quantile contours of multivariate regularly varying distributions defined via cyclically monotone mappings. Such contours are expected to satisfy a shape constraint involving a compact convex body, and we will look to exploit this information to construct efficient estimators.
 
 

FNRS
PDR

2016-2019

Measuring inequality from trade data

 

Promoter: Pierre Devolder

 

FNRS
PDR

2016-2020

Semiparametric inference for multi-state models

 

Promoter: Anouar El Ghouch, Researcher: Kassu Mehari Beyene

FNRS
PDR

2016-2020

Risk management and pricing in finance and insurance

In this project we will develop new risk management and pricing tools for several risks in finance and insurance.
A) A first goal of this research is to develop a multiple curve interest rate model that combines tractable model dynamics and semi-analytic pricing formulae with positive interest rates based on positive multiplicative spreads and by including regime switching. In this model we will first price interest rate derivatives. Afterwards we will study the influence of future cash flow evaluation by multiple curve models for the valuation and solvency requirements of life insurance and pension liabilities.
B) A second aim is to obtain a multivariate stochastic skew model by introducing first a tractable new time-changed Lévy model and afterwards its multivariate extension. We will focus on Index1Equity options markets as well as on foreign exchange markets. We will derive fast and accurate pricing formulae for Vanilla options through the use of FFT techniques. All the qualities of both the univariate and multivariate model will be illustrated by several numerical examples first mainly in derivatives pricing but later also in solvency risk measurement for a bank or insurance portfolio.
C) A third goal concerns the valuation and solvency requirements of life insurance and pension liabilities. Since the Solvency II regulation is based on a one year time horizon for the computation of the solvency capital we will first focus upon the long term aspect of the life insurance and pension liabilities. Therefore we will develop alternative risk measurements to Solvency II based on a time consistent framework inspired by dynamic risk measures and iteration techniques. In terms of valuation taking into account the joined presence of actuarial and financial risks we will develop a universal pricing method coherent with the risk neutral pricing in finance but also in line with the law of large numbers and the classical premium principles used in insurance.
Dans ce projet nous allons développer de nouveaux outils de gestion et d'évaluation de risques rencontrés en finance et assurance.
 
 

Promoter: Pierre Devolder

 

Generali
Chair

Le financement des pensions

The purpose of the research is to develop actuarial and financial models in order to estimate the impact of various scenarios of reform of the pension systems in Belgium.
In particular, stochastic models will be considered in order to establish actuarial balance sheets of the social security pension system.  The longevity risk in particular will be deeply analyzed and modeled.
 

Promoter: Pierre Devolder

 

AG Insurance
Chair

2016-2020

 

Pension valuation and solvency

Development of a coherent and universal model of valuation and solvency requirement of pension liabilities for pension funds and insurance companies in a stochastic environment.
 

Promoter: Pierre Devolder

DKV Belgium
Chair

2015-2019

Health insurance and longevity

Development of actuarial and financial techniques for the pricing, hedging and reserving of health, disability and long term care insurances ; analysis of the influence of the longevity risk on these products.
 

Promoter: Pierre Devolder

AXA Research
Fund

2016-2020

 

Actuarial dynamic approach of customer in P&C

While actuaries carry their models’ calculations for insurance products considering each product in isolation, the consumers tend to view all the products bought in a global way.
This research project conducted with AXA Belgium aims at reconciling the two points of view, allowing insurers to offer the most appropriate damage insurance covers together with optimal premiums.
 

Promoter: Michel Denuit