Text

theses

This dissertation studies methodologies for hypothesis testing and forecasting in financial econometrics, and comprises two essays on these topics, respectively. The first essay mainly aims to shed light on the importance of the proper application of jump testing methods, while the second essay provides alternative methodological approaches to volatility forecasting. In Chapter 2, we examine and compare a variety of jump tests in the financial econometrics literature. Numerous tests designed to detect realized jumps over a fixed time span have been proposed and extensively studied in recent years. These tests differ from "long time span" jump tests that detect jumps by examining the magnitude of the intensity parameter in the data generating process of an asset. In this chapter, a long time span jump test thereof called the CSS test, which is a variant of Corradi et al. (2018), is compared with a variety of fixed time span jump tests, in a series of Monte Carlo experiments. The CSS test is consistent against the null hypothesis of zero jump intensity, while the fixed time span tests are not designed to detect jumps in the data generating process, and instead detect realized jumps over a fixed time span. An empirical investigation of individual, sector specific and market level stock prices is also carried out in order to contrast findings based on these different varieties of tests. The fixed time span tests examined include the higher order power variation ASJ test of Ait-Sahalia and Jacod (2009), the classic bipower variation BNS test of Barndorff-Nielsen and Shephard (2006), and the truncated power variation PZ test of Podolskij and Ziggel (2010). It is found that both the ASJ test and the CSS test exhibit good finite sample properties for time spans both short and long. The other tests suffer from finite sample distortions under long time spans. When applied to stock price and stock index data, the two aforementioned tests indicate that the prevalence of jumps is not as universal as might be expected. Various sector ETFs and individual stocks, for example, appear to exhibit no jumping behavior during a number of annual periods. In Chapter 3, we use factor-augmented heterogeneous autoregressive (HAR)-type models to predict the daily integrated volatility of asset returns. Our approach is based on a proposed two-step dimension reduction procedure designed to extract latent common volatility factors from a large dimensional and high-frequency returns dataset with 267 constituents of the S&P 500 index. In the first step, we apply either least absolute shrinkage and selection operator (LASSO) or elastic net shrinkage on estimates of integrated volatility of all constituents in the dataset, in order to select a subset of asset return series for further processing. In the second step, we utilize (sparse) principal component analysis to estimate latent common asset return factors, from which latent integrated volatility factors are extracted. Although we find limited in-sample fit improvement, relative to a benchmark HAR model, all of our proposed factor-augmented models result in substantial out-of-sample predictive accuracy improvement. In particular, forecasting gains are observed at market, sector, and individual-stock levels, with the exception of the financial sector. Further investigation of the factor structures for non-financial assets shows that industrial and technology stocks are characterized by minimal exposure to financial assets, inasmuch as forecasting gains associated with factor-augmented models for these types of assets are largely attributable to the inclusion of non-financial stock price return volatility in our latent factors.

ETD_8765

Ph.D.

Includes bibliographical references

by Mingmian Cheng

doi:10.7282/T3988BFN

ETD doctoral

The author owns the copyright to this work.