Three essays in time series analysis

Quantile Impulse Response;Growth-at-Risk;Value-at-Risk;Idiosyncratic Shocks

Jung, Whayoung

关键词: Quantile Impulse Response;    Growth-at-Risk;    Value-at-Risk;    Idiosyncratic Shocks;   

Others  :  https://www.ideals.illinois.edu/bitstream/handle/2142/107905/JUNG-DISSERTATION-2020.pdf?sequence=1&isAllowed=y

美国|英语

来源: The Illinois Digital Environment for Access to Learning and Scholarship

【 摘 要 】

The first essay studies quantile impulse response functions (QIRFs) and their applications in macroeconomics and finance. We build a multi-equation autoregressive conditional quantile model and propose a new construction of the QIRF. We investigate dynamic QIRFs of the US economy in response to monetary policy and financial shocks, providing some interesting results: (i) Economic activity has the most heterogeneous response across its distribution among the variables under study. The left tail of economic activity is the most responsive to monetary policy and financial stimuli. (ii) We also assess the impacts of financial and monetary policy shocks on Growth-at-Risk during the global financial crisis. Negative financial shocks during August 2007-June 2009 substantially aggravated Growth-at-Risk over 2008-2010. Unconventional monetary policy tools used during July 2009-December 2015 ameliorated Growth-at-Risk successfully over 2011-2015. (iii) When a measure of financial conditions (NFCI) stays at its right tail quantiles (tighter financial conditions), NFCI displays locally explosive behavior. As a result, the consecutive right tail events create substantial downside risks to the economy.The second essay investigates the estimation and inference of quantile impulse response functions for financial data. We propose a new estimation method using the local projections by Jorda (2005). We establish consistency and asymptotic normality of the estimator, thereby enabling asymptotic inference. We also consider the confidence interval construction based on stationary bootstrap and prove its consistency. Confirmatory simulation results and empirical practices on Value-at-Risk dynamics are provided.In the third essay, I quantitatively assess the role of stock-specific shocks on aggregate volatility in the U.S. stock market. When a power law is fitted to the upper tail market capitalization distribution in the S&P 500, the estimate of tail exponent is slightly above one, in which case the contribution of stock-specific shocks to aggregate volatility can be non-trivial. The empirical results of a variance decomposition suggest that the contribution of stock-specific shocks to aggregate returns volatility is insignificant. The volatility of the shocks is about 20% of the volatility of aggregate returns. This small role of stock-specific shocks is attributable to the size-variance relationship and the highly positive correlations among macro-sectoral shocks. However, I find that stock-specific shocks have an impact on the conditional volatility of market returns in the U.S. stock market by extending the GARCH model.