Return to search

Backtesting of simulated method for Counterparty Credit Risk

After the financial crisis of 2008 regulators found that the derivative market, where financial institutions traded OTC derivatives with each other, played a significantrole in triggering the crisis. This led to the emergence of Counterparty Credit Risk(CCR) which is used to measure the exposure banks have to their counterparties. In simple terms CCR is a mix of Market and Credit risk which defines the risk that your counter party will go into bankruptcy. CCR involves the risk factors used in market risk since all of the derivatives are based on underlying assets such as interest rate and currencies. The thesis will focus on how one can backtest individual risk factors driving the value of OTC derivatives. We will present different Monte Carlo simulation techniques that are being used to simulate and represent all possible future outcomes for the risk factors. In order to better understand the performance of a chosen model and how to adjust the calibration window for the ingoing parameters, two different approaches are presented,Quantitative Backtesting and Statistical Backtesting. As an extension to this, a portfolio of interest rate Swaps are backtested whose value are driven by the evolution of the underlying risk factors. The backtesting ofthe portfolio is done with netting. The time horizon for the backtesting procedureis 2010-2020 giving the user up to 261 independent observations with a forecast length of 14 days. Both of the backtesting methods provide the practitioner with a graphical results guiding the user to choose an appropriate model and calibration method for simulating the risk factors. We found that a combination of the two approaches provides the best result. Hence, no backtesting method is superior the other. Instead they complement each other and should be used simultaneously. Using the two backtesting methods one can find a model that perfectly fit the underlying distribution of risk factors, theoretically. However, one should be careful since there will always be uncertainty about the future and there is no guarantee that tomorrow will follow historical evolution exactly.

Identiferoai:union.ndltd.org:UPSALLA1/oai:DiVA.org:umu-173284
Date January 2020
CreatorsLundström, Love, Öhman, Oscar
PublisherUmeå universitet, Institutionen för matematik och matematisk statistik, Umeå universitet, Institutionen för matematik och matematisk statistik
Source SetsDiVA Archive at Upsalla University
LanguageEnglish
Detected LanguageEnglish
TypeStudent thesis, info:eu-repo/semantics/bachelorThesis, text
Formatapplication/pdf
Rightsinfo:eu-repo/semantics/openAccess

Page generated in 0.002 seconds