The Implementation of the Value at Risk in the Tunisian Foreign
Transcription
The Implementation of the Value at Risk in the Tunisian Foreign
LES JEUDI D’ECONOMIE & DE FINANCE (JEF) Université Université Université de Carthage Tunis El Manar de Tunis (IHEC) (FSEG) (ISG) L.E.F.A La.R.E.Qua.D I.F.G.T U.A.Qu.Ap Laboratoire d'Economie & Finance Appliquée Laboratoire de Recherche en Econonomie Quantitative du Développement International Financial Group Tunisia Unité d'Analyse Quantitative Appliquée ORGANISENT UN SEMINAIRE DE RECHERCHE N°12/03 The Implementation of the Value at Risk in the Tunisian Foreign Exchange Market Nidhal Misbahi (LEFA, ISCCB, University of Carthage) Jeudi 15 mars 2012 à 14h30 à la salle des Thèses (FSEGT-Campus) Abstract: The objective of this study was the development and implementation of the Value at Risk in the Tunisian foreign exchange market, using various approaches (Parametric approaches, Monte Carlo simulation, Historical simulation, copulas methods) for the univariate and the bivariate analysis (the spot FX rates USD/TND and EUR/TND). We evaluated them using the recent mathematical, econometric and numerical extensions that can be applied to Tunisian’s FX characteristics. Advantages and disadvantages of each method are confirmed based on both literature review and empirical study. The efficacy of the methods was tested and compared using the backtesting methods. The results confirmed that these recent extensions showed adequacy for the Tunisian FX market in the estimation of the VaR. Indeed, we demonstrated for the parametric models, that the asymmetric Garch models confirmed efficiency, using Generalized Error Distribution (GED). Then, for the Monte Carlo simulation, we proved that the Quasi Monte Carlo have better estimations, with less numbers of iterations, than the crude method. In addition, the historical simulation demonstrated that the bootstrapping methods with replacement (or without replacement) could enhance the results, and finally the Nelson 12 and 13 produced more acceptable results than the other Archimedean Copulas. Finally, we expect that this study will provide more clarifications for different members of the Tunisian FX market to estimate the VaR in the context of managed flexible exchange regime. Responsables du Séminaire : Med Safouane Ben Aïssa ([email protected]), Med Ayadi ([email protected]), Med Goaïed ([email protected]) & Chaker Aloui ([email protected]) Avec le concours des Economètres Tunisiens pour l’organisation logistique