Essays on stochastic volatility and aggregate fluctuations
Open Access
- Author:
- Hu, Guanliang
- Graduate Program:
- Economics
- Degree:
- Doctor of Philosophy
- Document Type:
- Dissertation
- Date of Defense:
- May 10, 2022
- Committee Members:
- Neil Wallace, Major Field Member
Jingting Fan, Major Field Member
Jingzhi Huang, Outside Unit & Field Member
Marc Albert Henry, Professor in Charge/Director of Graduate Studies
Russell Cooper, Special Member
David Argente, Major Field Member
Jonathan Eaton, Chair & Dissertation Advisor
Russell Cooper, Special Member - Keywords:
- Stochastic volatility
Business cycle
Misallocation - Abstract:
- Recent papers have pointed to the role of fluctuation in the second-moment of firm productivity (or demand) in generating aggregate fluctuations. However, the commonly used model which assumes an AR(1) process of productivity has two implications that are inconsistent with data: (i) the economy recovers from a downturn induced by an increase in the second-moment much faster in the model than in SVAR; and (ii) in the model, a large increase in the second-moment leads to an increase in investment rate dispersion, which is not observed in the data. To address these inconsistencies, I propose a general information structure which allows for a rich specification of second-moment shocks and find one of them that can resolve these inconsistencies. In chapter 1, using data on US publicly traded firms, I document the stylized facts that firms experience larger changes in productivity will have larger uncertainty in future and that this association decreases over time. Motivated by these facts, I propose a general information structure, which incorporates a learning process, and embed it into a heterogeneous-firm RBC model. I then consider an economy without aggregate shocks and show that this information structure can explain well these facts both qualitatively and quantitatively. In chapter 2, based on the model framework developed in chapter 1, I study four types of second-moment shocks: shocks to the second-moment of persistent productivity (i.e., the standard formation in the literature), shocks to frequency of persistent productivity change, shocks to the second-moment of transitory productivity, and shocks to the second-moment of signal noises. Specifically, I estimate four models, each of which has one type of second-moment shocks. I find that the model that best fits data on firm-level and aggregate uncertainty is the one that has shocks to the second-moment of transitory productivity. In chapter 3, I explore the quantitative implications of each model estimated in chapter 2. I find the model with the best fit solves the aforementioned inconsistencies with the data. This result suggests a new reasonable way to model fluctuation in uncertainty.