Essays on Financial Econometrics
This dissertation consists of three essays. The first essay, "Volume, volatility and Macroeconomic Announcements" studies the relationship between trading intensity and price volatility and how it is affected by investors' disagreement on a common public signal around macroeconomic announcements. Inspired by a difference-of-opinion model in which investors agree to disagree, we use high frequency data and empirically show that the volume-volatility elasticity of SP 500 ETF is uniformly below 1. Besides, the elasticity decreases with disagreement measures such as the forecast dispersion on unemployment rate and uncertainty measures, as well as a textual based tone measure constructed using FOMC statements. This paper provides new evidences on how information is processed in financial market.
The second essay, "Investor Sentiment and Volume Volatility Relationship" shows that investor sentiment plays a role on information processing in financial markets. We incorporate a one-factor asset pricing model into the difference-of-opinion model to derive the volume-volatility relationship for individual stocks. We separate the sample into high and low sentiment periods and use high frequency data to show that investors' disagreement measures only significantly reduce volume-volatility elasticity around macroeconomic announcements during high-sentiment periods, for both the S&P 500 ETF and Dow Jones
30 components. This result is consistent with
changes in the confidence level of investors when sentiment regime shifts. Our estimates of elasticity also decrease significantly with the
ratio of idiosyncratic variance, which indicates that higher idiosyncratic risks introduce larger dispersion among investors.
In the third essay, "Efficient Estimation of Integrated Functional of Variance with Irregular
Observation Time", we propose an efficient estimator of the integrated functional of the variance with irregular observation time of prices. We propose the consistency and central limit theorems, and then validate the theorems through proofs and simulations.
This work is licensed under a Creative Commons Attribution-Noncommercial-No Derivative Works 3.0 United States License.
Rights for Collection: Duke Dissertations