Programme des cours 2016-2017
FINA0064-1  
Financial Risk Modeling
Durée :
30h Th
Nombre de crédits :
Master en sciences économiques, orientation générale, à finalité5
Master en ingénieur de gestion, à finalité5
Master en sciences mathématiques, à finalité5
Nom du professeur :
Fabien Boniver, Marie Lambert
Langue(s) de l'unité d'enseignement :
Langue anglaise
Organisation et évaluation :
Enseignement au premier quadrimestre, examen en janvier
Unités d'enseignement prérequises et corequises :
Les unités prérequises ou corequises sont présentées au sein de chaque programme
Contenus de l'unité d'enseignement :
"Financial Risk Modeling" introduces students to a set of techniques for measuring risks in financial portfolios. The course is not about risk management but about risk measurement and modeling. Lectures are divided into two main parts. The first part focuses on measuring market risks. The second part presents a few real life, practical issues in financial risk management and demonstrates how stochastic processes can help model these.

Part I: Modeling market risk (M. Lambert)
As initially emphasized by Markowitz, the two relevant characteristics of an asset portfolio are its expected return and the dispersion of possible returns around the expected return, i.e. the standard deviation of returns. Presuming risk aversion, rational investors will choose to hold efficient portfolios, i.e. those that maximize the expected return for a given degree of risk or, alternatively, minimize risk for a given level of expected return.

Modules 1 and 2 on "How to model market risk" reviews Modern Portfolio Theory by emphasizing the limits of a Markowitz analysis. We first focus on the challenge in estimating the input parameters (Module 1) and the need for higher-moment measures (i.e. beyond the mean-variance) (Module 2). 
Keywords: shrinkage estimators, constant correlation, robust variance-covariance matrix, VaR, Expected Shortfall, Cornish Fisher VaR, Higher-moments and co-moments (co-skewness, co-kurtosis).

Module 3 on "How to achieve risk diversification" challenges the hypothesis of the value-weighted portfolio as a proxy for the efficient portfolios. It introduces students to fundamental indexing and smart beta strategies.
Keywords: smart beta strategies, equally-weighted portfolio versus value weighted portfolio, risk parity, maximum diversification, diversity.


Part II: Stochastic processes and financial risk modeling (F. Boniver)
In this part, we present a few real life issues in financial risk management and show how stochastic processes can help model those. The focus is on the link between mathematical ingredients and business needs, rather than on theory or applications only.
 
As a starting point, we observe that the quantitative studies involving stochastic processes that are carried out nowadays in financial institutions relate to (at least) one of the two paradigms made up by the "real-world" and the "risk-neutral" measures. We thus focus on three practical sample cases (making up course modules) that enable us to discuss relevant features of stochastic processes in a financial setting:






  • Valuation of assets according to a risk-neutral probability
  • Strategic Asset Allocation simulations under the real-world probability
  • The mixed case: direct (internal model) measure of market risk under the Solvency II regulations
 
The first case lets us present two fundamental theorems of asset pricing involving the equivalent martingale measure, and illustrate it on the very classical but informative pricing of a European call option in both discrete and continuous time, in the models of Cox, Ross, and Rubinstein, and of Black, Scholes, and Merton. The second case links with the first part of the course by showing how stochastic processes can support an optimization quest for the risk/return combination. The third one introduces students to a state of the art application of financial risk modeling combining both paradigms: the estimation of a high-order quantile (Value-at-Risk) of some value-in-the-future distribution.
Keywords: asset pricing, call option, derivatives, martingale, equivalent martingale measure, strategic asset allocation, Monte Carlo simulations, Value-at-Risk (VaR), Solvency II, (insurance company) solvency. 
 
Acquis d'apprentissage (objectifs d'apprentissage) de l'unité d'enseignement :
Consistently with the Key Learning Outcomes, students will acquire the following capacities at the end of the course:
 
(1) They will strengthen their knowledge and understanding of financial risk management and rely on their knowledge to perform a rigorous analysis of a management situation. They will design optimal and creative solutions (through modeling methods).
 
(2) They will gain knowledge and understanding of financial engineering and be able to mobilize them in order to implement solutions to concrete management problems or cases.
 
(3) They will communicate about financial risk management problems in English and develop team work abilities.
 
(4) They will adapt their managerial practice and research autonomously and methodically the information needed to solve a complex and transversal management problem.
 
 
 
Specific skills and competencies are trained during this course.
 
At the end of part I, students will be able to:
 
- measure robust variance-covariance matrices for asset allocation and risk measurement;
 
- measure the potential for diversification of a portfolio taking into account its volatility, downside and extreme risks;
 
- measure the VaR of a stock portfolio using GARCH model;
 
- define alternative efficient portfolio specification (using heuristic and statistic methods) and test their outperformance over the common commercial (value-weighted) indexes.
 
 
 
At the end of part II, students will be able to:
 
- understand and explain how stochastic processes contribute to the modeling of typical financial risk management problems;
 
- exemplify this contribution on concrete cases;
 
- describe in business terms the mathematical setups needed for asset pricing, asset value prospective simulation, and VaR estimation of a random financial variable in the future;
 
- perform simulations and valuations associated to stochastic processes by implementing elementary models in a high-level programming language.
Savoirs et compétences prérequis :
Students attending this course are expected to have a good background in investment and portfolio management and have a good understanding of asset pricing models. For the second part of the course, students should be familiar with basic probability, statistics, and linear algebra concepts and methods.
Activités d'apprentissage prévues et méthodes d'enseignement :
The course is lecture-style with active discussions about practical examples.
 
Each theoretical session will be followed by a "computer lab" session for applying the studied concepts on practical case-studies.
 
The course has been developed to effectively combine each theoretical session with real-business case-studies.
 
For each part of the course, students will work on a group project. They are invited to work in groups of 2 or 3 students.
Mode d'enseignement (présentiel ; enseignement à distance) :
The course is face-to-face lecture and group-meeting style. 
Lectures recommandées ou obligatoires et notes de cours :
The recommended textbook for Part I is:
Market Risk Analysis: Practical Financial Econometrics by Carol Alexander ISBN-10: 0470998016 | ISBN-13: 978-0470998014 | Edition: Volume II


Online references for part II are provided in the accompanying material. Yet, interested students can find further reference material in




  • Martin Haugh's works (http://www.columbia.edu/~mh2078)
  • Brandimarte, Paolo, Handbook in Monte Carlo Simulation, Wiley & Sons, 2014 (ISBN-10: 0470531118 ISBN-13: 978-0470531112)
  • Glassermann, Paul, Monte Carlo Methods in Financial Engineering, Springer, 2003 (ISBN-10: 0387004513, ISBN-13: 978-0387004518)
Slides and information about the courses can be found on the lol@ platform http://lola.hec.ulg.ac.be
Modalités d'évaluation et critères :
The final grade will be determined by :
 
1. Case studies and group projects: 30%
2. Active class participation (attending class and actively discussing/applying material in class) - individual grade: 30%
3. Individual Oral exam: 40%
Stage(s) :
none
Remarques organisationnelles :
Attendance to lectures and computer labs are mandatory and will be controlled and graded (individual class participation).
Absences from exams are allowed only for justified medical reasons. Unexcused absences from exams will lead to a zero score in the calculation of the final grade.
Contacts :
Part I: Modeling market risk Prof. Marie Lambert - email: marie.lambert@ulg.ac.be 
Boris Fays - tutor - boris.fays@ulg.ac.be
N1 (office 109): Please schedule an appointment by email!


Part II: Stochastic processes and financial risk modeling
 
Affiliate Prof. Fabien Boniver - email: f.boniver@ulg.ac.be

 
Notes en ligne :
online notes
The core materials for the course consist of the required textbook readings. Lecture notes will be available on the course web page (on lol@). Other items such as problem sets will also be available on the course web page. Some additional readings on materials related to the course over the term may be provided throughout the course via the course web page.