Van Rensburg and Robertson (2003a) stated that the CAPM beta has little or no relationship
with returns generated by size and price to earnings sorted portfolios. This study intends to
demonstrate that a reformulated CAPM beta, estimated using return on equity as opposed to
share returns, unravels the size and value premium. The study proves that the “cash-flow”
generated beta partially explains the cross-sectional variation in share returns when measured
over the long run, specifically when portfolios are sorted on book to market, however the
cash flow beta is less successful when attempting to explain the small size premium. The
premise of the study is that the cash flow dynamics of share returns eventually dominate the
first and second moments and thus result in cash flow based measures of risk and return that
should succeed in explaining the cross-sectional variation in share returns. The study makes
use of vector autoregressive models in order to examine the short term effect of structural
shocks to the cash flow fundamentals of a stock or portfolio through impulse response
functions as well as quantifying a long-term relationship between cash flow fundamentals and
share returns using a VECM specification. The study further uses fixed effects, random
effects and GMM/dynamic panel data cross-sectional regressions in order to examine the
ability of the cash flow beta explaining the value and size premium. The results of the study
are mixed. The cash flow beta does well in explaining the returns of portfolios sorted on book
to market, but fails to do the same with size sorted portfolios. In the cash flow betas favour, it
performs far better than the conventionally measured CAPM beta throughout the study.
Identifer | oai:union.ndltd.org:netd.ac.za/oai:union.ndltd.org:wits/oai:wiredspace.wits.ac.za:10539/11613 |
Date | 05 July 2012 |
Creators | Page, Daniel |
Source Sets | South African National ETD Portal |
Language | English |
Detected Language | English |
Type | Thesis |
Format | application/pdf |
Page generated in 0.0014 seconds