Files and links (1)
PDFOpen Access (License Unspecified), Open Access
Dissertation
Three essays on modeling stock returns: empirical analysis of the residual distribution, risk-return relation, and stock-bond dynamic correlation
Doctor of Philosophy (Ph.D.), Drexel University
Jun 2007
DOI:
https://doi.org/10.17918/etd-1784
Abstract
This dissertation studies the following issues: the presence of non-normal distribution features and the significance of higher order moments, the tradeoff between risk and return, and the dynamic conditional correlation between stock returns and bond returns. These issues are structured into three essays. Essay #1 tackles the non-normal features by employing the exponential generalized beta distribution of the second kind (EGB2) to model 30 Dow Jones industrial stock returns. The evidence suggests that the model with the EGB2 distribution assumption is capable of taking care of stock return characteristics, including fat tails, peakedness (leptokurtosis), skewness, clustered conditional variance, and leverage effect, therefore, is capable of making a good prediction on the happenings of extreme values. The goodness of fit statistic provides supporting evidence in favor of the EGB2 distribution in modeling stock returns. Evidence also suggests that the leverage effect is diminished when higher order moments are considered. Essay #2 examines the risk-return relation by applying high frequency data of 30 Dow Jones industrial stocks. I find some supportive evidence in favor of the positive relation between the expected excess return and expected risk. However, this positive relation is not revealed for all 30 stocks using a standard weighted least squares regression (WLS) method. Using a quantile regression method, I find that the risk-return relation evolves from negative to positive as the returns' quantile increases. This essay also finds interesting evidence that the intraday skewness coefficient explains a great deal of the variation in the excess returns. Essay #3 mainly focuses on the analysis of the time-varying correlations between stock and bond returns using the asymmetric dynamic conditional correlation (ADCC) model (Cappiello et al., 2004). The estimated coefficients show some volatile behavior and display some degree of persistence over time. Testing the asymmetric dynamic correlations by using a set of macroeconomic information, I find that the federal funds rate, the relative volatility between the stock and bond markets, the yield spread, and oil price shocks are the significant factors for the coefficients' time varying.
Metrics
17 File views/ downloads
32 Record Views
Details
- Title
- Three essays on modeling stock returns
- Creators
- Jiandong Li - DU
- Contributors
- Thomas Chi-Nan Chiang (Advisor) - Drexel University (1970-)
- Awarding Institution
- Drexel University
- Degree Awarded
- Doctor of Philosophy (Ph.D.)
- Publisher
- Drexel University; Philadelphia, Pennsylvania
- Resource Type
- Dissertation
- Language
- English
- Academic Unit
- Bennett S. LeBow College of Business; Finance; Drexel University
- Other Identifier
- 1784; 991014632528904721