Internal models for market risk assessment based on expected shortfall

Upravljanje tržišnim rizicima u finansijskim institucijama se prevashodno zasniva na proceni parametra „vrednost pod rizikom“ (VaR). Za ovakvu praksu je, u najvećoj meri, zaslužna bazelska regulativa koja kreira smernice i preporuke u oblasti upravljanja rizicima, a koje centralne banke većine razvijenih zemalja slede. Pa ipak, svetska finansijska kriza dovela je u pitanje adekvatnost ove mere da predvidi ekstremno rizične događaje, a samim tim i njenu dalju upotrebu u regulatorne svrhe. Kao rešenje nameće se upotreba uslovnog gubitka, relativno nove mere, koja procenjuje očekivani gubitak iznad VaR-a. Sa tim u vezi je i osnovni cilj ove doktorske disertacije, a to je da obezbedi metodologiju za upravljanje tržišnim rizicima zasnovanu na uslovnom gubitku. To podrazumeva određivanje adekvatnih modela za procenu uslovnog gubitka, modela za testiranje predviđanja, kao i novih kapitalnih zahteva za pokriće tržišnih rizika.Management of market risk in financial institutions is primarily based on the estimation of the “Value-at-Risk” (VaR) parameter. The credit for this practice is largely with the Basel regulation which creates the guidelines and recommendations in the market risk management field, followed by the majority of central banks of the developed countries. However, the world economic crisis has questioned the adequacy of this measure to predict extremely risky events, and thus its further application for regulation purposes. The use of a relatively new measure, the expected shortfall, which estimates the expected loss above VaR, is imposed as a solution. The main goal of this dissertation is to provide a methodology for managing market risk based on expected shortfall. That implies determination of adequate models for expected shortfall assessment, adequate models for backtesting expected shortfall models, as well as new capital requirements for market risk based on expected shortfall