In this thesis, we investigate the relationship between the US stock returns and downside risk in a cross-sectional context. When the classic market model with a moving window approach is adopted, downside risk estimated coefficients exhibit a positive impact on stock returns. However, when two other non-linear time-varying models; the cuiic piecewise polynomial function (CPPF) and the Fourier Flexible Form (FFF) models are adopted, downside risk estimated coefficients show a negative impact on stock returns, Cross-sectinally, the reisk estimated coefficients of the town non-linear models produce a much better fit than the classic market model. The predictive power for future stock returns of downside risk estimated coefficients are found to be weak. Two more risk factors: commodityh market risk and Aruoba-Diebold-Scotti (ADS) business condition index risk (both downside and upside versions thereof), are shown to have a significant effect on stock returns.
Identifer | oai:union.ndltd.org:bl.uk/oai:ethos.bl.uk:575709 |
Date | January 2013 |
Creators | Huang, Fangzhou |
Publisher | Cardiff University |
Source Sets | Ethos UK |
Detected Language | English |
Type | Electronic Thesis or Dissertation |
Source | http://orca.cf.ac.uk/49343/ |
Page generated in 0.0018 seconds