This dissertation focuses on testing and exploring the usage of the jump-diffusion two-beta asset pricing model. Daily and monthly security returns from both NYSE and AMEX are employed to form various samples for the empirical study. The maximum likelihood estimation is employed to estimate parameters of the jump-diffusion processes. A thorough study on the existence of jump-diffusion processes is carried out with the likelihood ratio test. The probability of existence of the jump process is introduced as an indicator of "switching" between the diffusion process and the jump process. This new empirical method marks a contribution to future studies on the jump-diffusion process. It also makes the jump-diffusion two-beta asset pricing model operational for financial analyses. Hypothesis tests focus on the specifications of the new model as well as the distinction between it and the conventional capital asset pricing model. Both parametric and non-parametric tests are carried out in this study. Comparing with previous models on the risk-return relationship, such as the capital asset pricing model, the arbitrage pricing theory and various multi-factor models, the jump-diffusion two-beta asset pricing model is simple and intuitive. It possesses more explanatory power when the jump process is dominant. This characteristic makes it a better model in explaining the January effect. Extra effort is put in the study of the January Effect due to the importance of the phenomenon. Empirical findings from this study agree with the model in that the systematic risk of an asset is the weighted average of both jump and diffusion betas. It is also found that the systematic risk of the conventional CAPM does not equal the weighted average of jump and diffusion betas.
| Identifer | oai:union.ndltd.org:pdx.edu/oai:pdxscholar.library.pdx.edu:open_access_etds-2324 |
| Date | 01 January 1996 |
| Creators | Chen, Hongqing |
| Publisher | PDXScholar |
| Source Sets | Portland State University |
| Detected Language | English |
| Type | text |
| Format | application/pdf |
| Source | Dissertations and Theses |
Page generated in 0.0022 seconds