dc.contributor.advisor | Xu, Ke-li | |
dc.contributor.advisor | Kim, Hwagyun | |
dc.creator | Yeo, Hyosung | |
dc.date.accessioned | 2016-07-08T15:16:00Z | |
dc.date.available | 2016-07-08T15:16:00Z | |
dc.date.created | 2016-05 | |
dc.date.issued | 2016-05-10 | |
dc.date.submitted | May 2016 | |
dc.identifier.uri | https://hdl.handle.net/1969.1/157079 | |
dc.description.abstract | We measure macroeconomic uncertainty and study its link to asset returns via a consumption-based model employing recursive preferences. We introduce a stochastic volatility model with two asymptotic regimes and smooth transition. Smooth transition in regimes produces sizable equity premiums for even a small amount of consumption volatility if uncertainty unravels slowly. The relative risk aversion is estimated around two, the estimated elasticity of intertemporal substitution is greater than one, and the simulation suggests that our volatility channel matters in explaining asset returns.
Next, we propose to use the Hodrick-Prescott filter to nonparametrically extract the conditional mean and volatility process of a time series. We find an optimal smoothing parameter for HP filter by minimizing the first order sample correlation of the residuals. The process extracted from our HP filter is therefore defined as the predictable component of the given time series, while the conventional HP filter decomposes a time series into trend and cyclical components. By simulations, we show that our HP filter performs better than the local linear estimator in terms of average mean squared error for both discrete and continuous time models.
Finally, we develop a novel methodology to test for stock return predictability using multiple predictors. It has been reported that the conventional least squares approach has an unacceptable level of size distortions and over-reject the null hypothesis of no predictability. Previous literatures which tried to resolve the Endogeneity problem with a persistent predictor have failed to allow multiple covariates in the predictive regression. We propose to apply Heteroskedasticity and Endogeneity correction sequentially to tackle the issue. Our approach not only makes it possible to correctly test for the predictability of stock returns by multiple predictors but also reveals the marginal predictive power of each predictor. Using our new test, we find strong evidence for joint return predictability by dividend-price ratio, earnings-price ratio, short-term interest rates and term spread. | en |
dc.format.mimetype | application/pdf | |
dc.language.iso | en | |
dc.subject | macroeconomic uncertainty | en |
dc.subject | asset pricing | en |
dc.subject | equity premium | en |
dc.subject | regime switching volatility with smooth transitions | en |
dc.subject | volatility premium predictive regression | en |
dc.subject | multiple predictors | en |
dc.subject | persistency and nonstationarity | en |
dc.subject | stochastic volatility | en |
dc.subject | endogeniety correction | en |
dc.subject | Hodrick-Prescott Filter | en |
dc.subject | nonparametric conditional mean estimation | en |
dc.title | Three Essays in Macroeconomics and Empirical Finance | en |
dc.type | Thesis | en |
thesis.degree.department | Economics | en |
thesis.degree.discipline | Economics | en |
thesis.degree.grantor | Texas A & M University | en |
thesis.degree.name | Doctor of Philosophy | en |
thesis.degree.level | Doctoral | en |
dc.contributor.committeeMember | Sekhposyan, Tatevik | |
dc.contributor.committeeMember | Zubairy, Sarah | |
dc.type.material | text | en |
dc.date.updated | 2016-07-08T15:16:00Z | |
local.etdauthor.orcid | 0000-0001-9854-4775 | |