Chapter 7 • Portfolio theory and its relevance to real investment decisions
that there are four factors that explain the risk/risk premium relationship of a par-
ticular security.
Basically, CAPM says that
E(ri) =rf+λβi
where λis the average risk premium [E(rm) −rf].
However, APM holds that
E(ri) =rf+λ 1 βi 1 +λ 2 βi 2 +λ 3 βi 3 +λ 4 βi 4
where λ 1 , λ 2 , λ 3 and λ 4 are the average risk premiums for each of the four factors
in the model, and βi 1 , βi 2 , βi 3 and βi 4 are measures of the sensitivity of the particular
security ito each of the four factors.
The four factors in the APM relate to future macroeconomic factors, including
industrial output and levels of inflation. Other versions of the APM have also been
developed that have more factors than the four in the Ross version.
Tests conducted on APM appear to show it to be superior to CAPM as an explainer
of historical security returns. However, it contains four (or more) factors rather than
one. This limits its practical usefulness in deriving a discount rate to be applied in
NPV analyses. APM seems not to be used in practice to any significant extent. It seems
likely, however, that the decline in the reputation of CAPM will encourage researchers
to revisit APM and other multi-factor models.
7.13 Portfolio theory – where are we now?
Despite the fact that CAPM has substantially been discredited as a practical decision-
making tool, portfolio theory, together with some of the other points that have come
up in this chapter, have some valuable lessons for shareholders and managers.
Lessons for shareholders
l Diversify shareholdings. The evidence (see Figure 7.1, page 188) shows that there are
clear risk reduction opportunities. Risk was defined here as the standard deviation
(or variance). Although some researchers have partly attributed the apparent
weaknesses of CAPM to the limitations of standard deviation as a risk measure in
this context, it seems likely that it is a helpful indicator.
l Expected average returns.Average returns from equity investment have been 6.5 per
cent p.a. (in real terms) over the period 1900 to 2001 (see Dimson, Marsh and
Staunton (2002), discussed on page 205). It seems, therefore, that it is reasonable to
see this level of return as the average for the future for a diversified portfolio. Over
short periods, returns may well be much less or much more than this. Riskier invest-
ments seem likely to generate higher average returns than this, but without CAPM
or something similar we cannot say what level of risk relates to what level of return.
Lessons for financial managers
l Expected average returns.Since the returns from risk-free investments have aver-
aged 2.0 per cent (in real terms) over the period 1900 to 2001 (Dimson, Marsh and