Measuring systemic risk in time-frequency domain
Název práce v češtině: | Měření systémového rizika v časově-frekvenční doméně |
---|---|
Název v anglickém jazyce: | Measuring systemic risk in time-frequency domain |
Klíčová slova: | systémové riziko, podmíněná hodnota v riziku, závislost na chvostech, DCC GARCH, waveletová analýza |
Klíčová slova anglicky: | systemic risk, conditional value at risk, tail dependence, DCC GARCH, wavelet analysis |
Akademický rok vypsání: | 2013/2014 |
Typ práce: | diplomová práce |
Jazyk práce: | angličtina |
Ústav: | Institut ekonomických studií (23-IES) |
Vedoucí / školitel: | doc. PhDr. Jozef Baruník, Ph.D. |
Řešitel: | skrytý![]() |
Datum přihlášení: | 09.02.2014 |
Datum zadání: | 09.02.2014 |
Datum a čas obhajoby: | 23.09.2015 00:00 |
Místo konání obhajoby: | IES |
Datum odevzdání elektronické podoby: | 31.07.2015 |
Datum proběhlé obhajoby: | 23.09.2015 |
Oponenti: | prof. PhDr. Michal Bauer, Ph.D. |
Seznam odborné literatury |
Acharya, V.V.,Pedersen, L.H., Philippon, T. and Richardson, M.P. (2010), “Measuring Systemic Risk”, Denver Meetings Paper.
Adrian, T. and Brunnermeier M. K. (2010), “CoVar”, Report no. 348, Federal Reserve Bank of New York. Billio, M., Getmansky, M., Lo, A. W. and Pelizzon, Y. (2010), “Econometric measures of systemic risk in the finance and insurance sectors”, NBER Working Paper 16223. Bisias, D., Flood, M., Lo, Andrew W., Valavnis, S. (2012), “A Survey of Systemic Risk Analytics”. Office of Financial Research. Engle, R.F., and Manganelli, S. (2004), “CAViaR: Conditional Autoregressive Value at Risk by Regression Quantiles”, Journal of Business and Economic Statistics 22, 367-381. Han, H.and Linton, O. B. and Oka, T and Whang, Y. (2014), “The Cross-Quantilogram: Measuring Quantile Dependence and Testing Directional Predictability between Time Series”. Koenker, R., Bassett, G. (1978). “Regression quantiles”, Econometrica 46: 33-50. Mittnik, S. (2013), “VaR-implied tail-correlation matrices”, CFS Working Paper Series 2013/05, Center for Financial Studies (CFS)“. Zikes, F., & Barunik, J. (2013), “Semiparametric conditional quantile models for financial returns and realized volatility”. |
Předběžná náplň práce |
Out of all the approaches towards measuring systemic risk, I chose the analysis of returns of financial institutions in the tail of their distribution. First of all, I would like to analyze the tail dependence measures of systemic risk that have already been proposed, particularly Adrian’s CoVar (2010), Mittnik’s tail correlation matrices (2013) and Han’s cross quantilogram (2013). I would like to apply these approaches to datasets with data of particular financial institutions and perform different robustness checks and out-of-sample forecasts to determine strengths and weaknesses of the models at hand. I consider this to be the preparatory part, where I will become familiar with both the data and the estimation techniques developed already. Should I identify any weakness of the measure which could be corrected for, I would like to design a new riskometer, either by merging different features of proposed models into a new one or by a direct adjustment of one of the existing measures.
One of the ideas is to follow the reasoning of Barunik & Zikes (2013) and include realized volatility into the quantile regression used to estimate Adrian’s CoVar. The quantile regression theory introduced by Koenker and Bassett (1978) provides a convenient tool for robust inference over observations drawn from heavy-tailed or skewed distributions. The appeal of using realized volatility to measure variation in asset returns lies in its non-parametrical nature and ease of estimation, once high-frequency data are at hand. In the first part of my work, I would like to use a dataset consisting of daily returns of financial institutions to analyze comparative performance of proposed measures. In addition, studied models use 3-month T-rate, repo rate, VIX index and CRSP index in their estimation. Later, for the purpose of realized volatility estimation, I need high frequency data of returns of the institutions. 1. Motivation: there is no consensus on usage of a particular systemic risk measure, different approaches proposed. 2. Literature review of studies on tail risk measures 3. Application and comparison of proposed measures, out-of-sample forecast 4. Designing an alternative measure – theoretical approach 5. Application, robustness checks 6. Comparison against proposed benchmarks 7. Concluding remarks |
Předběžná náplň práce v anglickém jazyce |
Out of all the approaches towards measuring systemic risk, I chose the analysis of returns of financial institutions in the tail of their distribution. First of all, I would like to analyze the tail dependence measures of systemic risk that have already been proposed, particularly Adrian’s CoVar (2010), Mittnik’s tail correlation matrices (2013) and Han’s cross quantilogram (2013). I would like to apply these approaches to datasets with data of particular financial institutions and perform different robustness checks and out-of-sample forecasts to determine strengths and weaknesses of the models at hand. I consider this to be the preparatory part, where I will become familiar with both the data and the estimation techniques developed already. Should I identify any weakness of the measure which could be corrected for, I would like to design a new riskometer, either by merging different features of proposed models into a new one or by a direct adjustment of one of the existing measures.
One of the ideas is to follow the reasoning of Barunik & Zikes (2013) and include realized volatility into the quantile regression used to estimate Adrian’s CoVar. The quantile regression theory introduced by Koenker and Bassett (1978) provides a convenient tool for robust inference over observations drawn from heavy-tailed or skewed distributions. The appeal of using realized volatility to measure variation in asset returns lies in its non-parametrical nature and ease of estimation, once high-frequency data are at hand. In the first part of my work, I would like to use a dataset consisting of daily returns of financial institutions to analyze comparative performance of proposed measures. In addition, studied models use 3-month T-rate, repo rate, VIX index and CRSP index in their estimation. Later, for the purpose of realized volatility estimation, I need high frequency data of returns of the institutions. 1. Motivation: there is no consensus on usage of a particular systemic risk measure, different approaches proposed. 2. Literature review of studies on tail risk measures 3. Application and comparison of proposed measures, out-of-sample forecast 4. Designing an alternative measure – theoretical approach 5. Application, robustness checks 6. Comparison against proposed benchmarks 7. Concluding remarks |