Econometrics of risk: applications to financial sectors

เศรษฐมิติเกี่ยวกับความเสี่ยง: การประยุกต์กับภาคการเงิน

Name: Panisara Phochanachan

ThaSH: Econometrics

Classification :.DDC: 330.015195

ThaSH: Risk management

ThaSH: Financial risk management

Abstract: Investors face risks and uncertainties when investing in financial assets. Appropriate tools to measure risks and to reduce risks or the magnitude of losses are important for investors to make investment decisions. The main objective of this thesis is to illustrate the applications of econometric tools to analyze investment risks for international stock and commodities markets. This thesis consists of three issues. In the first study, we applied the time-varying copula-GARCH and the DCC-GARCH models to estimate the Value-at-Risk (VaR) and Expected Shortfall (ES) of G7 stock markets as well as constructed the optimal portfolio. The empirical results showed that the time-varying copula-GARCH model successfully captured the tail dependences and extreme losses for G7 returns. There existed tail dependences between Germany and France, France and Italy, and France and UK. Moreover, there existed asymmetric tail dependence between Canada and US. We also found that the optimal portfolio constructed using time-varying copula-GARCH model has a more accurate and correct forecasting than that from the DCC-GARCH model in the case of G7 returns. For the optimal portfolio, Canadian index held a high weight, while the Italian index had a much smaller weight. This was because the Italian stock market showed a stronger volatility than the Canadian stock market. The optimal weights and ES that we computed based on conditional volatility and nonlinear correlation showed that theinternationally diversified portfolios can help investors to reduce risk and avoid adverse consequences from contagious crisis. The second study focuses on measuring systemic risk in G7 financial markets. We applied a new approach that is called component expected shortfall (CES). The CES method not only calculates systemic risk, but also measures the stock markets of the G7 countries’ absolute contribution to the ES. It overcomes the main drawbacks of MES and SRISK. In this study, we incorporated CES with copula-GARCH models, which allows for higher scalability for capturing asymmetric dependence. The preferred model was selected from 8 kinds of GARCH families to estimate the volatility of each G7 stock return. For the marginal distributions, the results indicated that two types of asymmetric distributions, skewed student-t distribution (SSTD) and skewed generalized error distribution (SGED), provided the lowest AIC and BIC. Moreover, we made a comparison among various kinds of multivariate copulas to capture asymmetric dependence. The mixed C-vine copula had the best performance among all multivariate copulas for forecasting the systemic risks of G7 countries. The results showed that the pre-crisis featured lower levels of risk contribution, and the risk contribution increased gradually as the crisis unfolded. In addition, the systematic risk of G7 reached a higher level after the financial crisis due to the Europe debt crisis. Lastly, this study dealt with uncertainty or risk from extreme events or rare phenomena to the crude oil price. We introduced a method to estimate the VaR using the autoregressive moving average (ARMA), the extreme value theory (EVT), and belief function. The ARMA model was applied to the mean equation of the returns on crude oil prices. The EVT was then used to model the unobserved shocks. Finally, a likelihood-based belief function was constructed to quantify the uncertainty on estimated parameters. The results indicated that the plausibility of VaRs for the positive and negative return shocks at the 95% level are 0.107 and 0.122, respectively. An additional contribution of incorporating the belief function theory into the method is that the belief function allows us to add expert opinions into the estimations.

Chiang Mai University Library

Address: CHIANG MAI

Email: cmulibref@cmu.ac.th

Name: Hung T. Nguyen

Role: Chairman

Name: Songsak Sriboonchitta

Role: Member

Name: Sompong Dhompongsa

Role: Member

Name: Jirakom Sirisrisakulchai

Role: Member

Name: Jianxu Liu

Role: Member

Created: 2016

Modified: 2017-10-30

Issued: 2017-10-30

วิทยานิพนธ์/Thesis

application/pdf

CallNumber: Th 330.015195 P192E

eng

DegreeName: Doctor of Philosophy

Level: Doctoral Degree

Descipline: Economics

Grantor: Chiang Mai University

ลำดับที่.	ชื่อแฟ้มข้อมูล	ขนาดแฟ้มข้อมูล	จำนวนเข้าถึง	วัน-เวลาเข้าถึงล่าสุด
1	econ60516ppc_full.pdf	10.65 MB	16	2025-07-12 00:12:20

ใช้เวลา

0.019867 วินาที