Skip to Main content Skip to Navigation
Theses

Modélisation des risques en présence de valeurs extrêmes : Une approche Gini

Abstract : This thesis proposes a new approach to the treatment of available financial information by robust data analysis tools, using the Gini index. It aims at keeping all the information available, even rare events, for the modeling of returns and risk. A bad appreciation of risk by the agent distorts his expectations. Decision-making is based on the assessment of risks and forecasts that agents are able to make in the immediate future. The expected return on investment by the agent will depend on several sources of risk according to the arbitrage pricing theory. Yitzhaki and Schechtman (2013) have laid the foundations for a new econometrics based on the Gini index. They propose to use the coGini operator rather than the covariance to study samples whose underlying statistic distribution may be a statistic distribution different from the normal distribution.This thesis has two main contributions. The first deals with traditional data analysis methods : PCA, LDA and scoring. These methods are adapted to extreme values by using the coGini operator (or covariance in the Gini sense): Gini-PCA and Gini-LDA. The second concerns the analysis of the returns / risk couple. Applications are made in financial theory, in particular the pricing of financial assets by the arbitrage pricing theory and the performance analysis of investment strategies. Beyond finance, the tools developed in this thesis can be applied to any risk assessment (climate risk, governance risk, risk related to the evaluation of rare events such as earthquakes, coronavirus, etc.).
Document type :
Theses
Complete list of metadata

https://tel.archives-ouvertes.fr/tel-03184784
Contributor : Abes Star :  Contact
Submitted on : Monday, March 29, 2021 - 7:04:39 PM
Last modification on : Tuesday, March 30, 2021 - 3:38:10 PM
Long-term archiving on: : Wednesday, June 30, 2021 - 7:01:23 PM

File

82689_OURAGA_2020_archivage.pd...
Version validated by the jury (STAR)

Identifiers

  • HAL Id : tel-03184784, version 1

Collections

Citation

Téa Ouraga. Modélisation des risques en présence de valeurs extrêmes : Une approche Gini. Economies et finances. Université de Nîmes, 2020. Français. ⟨NNT : 2020NIME0007⟩. ⟨tel-03184784⟩

Share

Metrics

Record views

88

Files downloads

53