Pomiar ryzyka rynkowego za pomocą miary Value at Risk – podejście dwuetapowe

University of Warsaw Repository

pl | en
 
 

Show simple item record

dc.contributor.advisor Kokoszczyński, Ryszard
dc.contributor.author Chlebus, Marcin
dc.date.accessioned 2014-11-14T08:30:27Z
dc.date.available 2014-11-14T08:30:27Z
dc.date.issued 2014-11-14
dc.identifier.uri https://depotuw.ceon.pl/handle/item/903
dc.description.abstract Praca dotyczy oceny jakości prognoz Value at Risk uzyskiwanych na podstawie modeli EWS-GARCH. Modele klasy EWS-GARCH prognozują wartość Value at Risk w dwóch etapach. Najpierw prognozowany jest jeden z dwóch stanów, a następnie, na podstawie modelu dla prognozowanego stanu, szacowana jest wartość Value at Risk. Przeprowadzona w badaniu analiza została wykonana na podstawie stóp zwrotu wybranych aktywów z Giełdy Papierów Wartościowych w Warszawie. Uzyskane wyniki wskazują, że modele EWS-GARCH mogą podnosić jakości prognoz Value at Risk otrzymywanych na podstawie modeli benchmarkowych zarówno według kryterium konserwatywnego, jak i kryterium adekwatności. Wybór optymalnych założeń powinien być wypadkową oczekiwanego poziomu adekwatności, konserwatyzmu oraz kosztowności stosowanego modelu.
dc.description.abstract The aim of my study was to evaluate the EWS-GARCH models in terms of the quality of the Value at Risk (further also: VaR) forecasts derived from them. The EWS-GARCH models are two-step models. In the first step, a model to predict the state (state of tranquility or turbulence) of the analyzed portfolio is estimated. In the second step, two different models (for each of portfolio states) for the level of the market risk calculation are estimated. The assessment of the Value at Risk predictions quality was based on the comparison between forecasts for different EWS-GARCH and benchmark models (GARCH(1,1), EGARCH(1,1,1), GARCH-t(1,1), GARCH(1,1) with the empirical distribution of returns correction and EGARCH (1,1,1) with the empirical distribution of returns correction). Searching for the optimal model to predict the Value at Risk measure is important because of the international regulations concerning market risk management in banks (and other financial institutions), such as Basel II or CRD IV package. The study was conducted for 79 different shares listed on the Warsaw Stock Exchange. The shares were chosen randomly. The only condition to be met was that the shares had been listed on the Warsaw Stock Exchange since at least January 2006. The study covered the period from 1 January 2006 to 31 January 2012. The quality of the Value at Risk predictions obtained from the analyzed models was evaluated based on the following criteria: the Value at Risk exceedances ratio, cost functions (i.e. Lopez & Abad functions), Basel II back-testing procedure and conditional & unconditional coverage tests. Additionally, the stressed Value at Risk measure analysis were performed. The results indicate that the EWS-GARCH models can improve the quality of the Value at Risk forecasts in comparison to the Value at Risk forecasts obtained from the benchmark models. The final selection of the best assumptions for the EWS-GARCH models may be performed with respect to predictions accurateness (the relation between the ratio of VaR exeedances and assumed 1% of exeedances) or predictions safety (decreasing the ratio of VaR exeedances). The results obtained show that the use of the EWS-GARCH models can raise the conservatism of the Value at Risk predictions in comparison to the benchmark models without increasing, at the same time, costs associated with the market risk. Additionally, for the EWS-GARCH models with the GARCH(1,1) model, as the model for the state of tranquility, it is also possible to obtain more accurate Value at Risk predictions than for the benchmark models.
dc.language.iso pl
dc.rights Creative Commons Uznanie autorstwa 3.0 Polska
dc.rights info:eu-repo/semantics/openAccess
dc.rights.uri http://creativecommons.org/licenses/by/3.0/pl/legalcode
dc.subject Komitet Bazylejski ds. Nadzoru Finansowego
dc.subject ryzyko rynkowe
dc.subject modele zmiany stanu
dc.subject modele GARCH
dc.subject Value at Risk
dc.subject market risk
dc.subject switching models
dc.subject GARCH models
dc.title Pomiar ryzyka rynkowego za pomocą miary Value at Risk – podejście dwuetapowe
dc.title.alternative Market risk measuring using Value at Risk - two-step approach
dc.type info:eu-repo/semantics/doctoralThesis
dc.description.eperson Marcin Chlebus
dc.contributor.department Wydział Nauk Ekonomicznych
dc.date.defence 2014-12-03

Files in this item

This item appears in the following Collection(s)

Show simple item record

Creative Commons Uznanie autorstwa 3.0 Polska Except where otherwise noted, this item's license is described as Creative Commons Uznanie autorstwa 3.0 Polska

Search Repository


Advanced Search

Browse

My Account

Statistics