Chapter 10
Bodie
Kane Marcus Perrakis
Ryan
INVESTMENTS, Fourth Canadian Edition
Empirical
Evidence on
Security Returns
Slide 10-1
Copyright © McGraw-Hill Ryerson Limited, 2003
Bodie
Kane Marcus Perrakis
Ryan
INVESTMENTS, Fourth Canadian Edition
Chapter Summary
 Objective: To discuss the empirical
evidence in support of equilibrium
models.



Slide 10-2
Tests of the Single Factor Model
Tests of the Multifactor Model
Other Studies
Copyright © McGraw-Hill Ryerson Limited, 2003
Bodie
Kane Marcus Perrakis
Ryan
INVESTMENTS, Fourth Canadian Edition
Overview of
Investigation
 Tests of the single factor CAPM or APT
Model
 Tests of the Multifactor APT Model

Results are difficult to interpret
 Studies on volatility of returns over time
Slide 10-3
Copyright © McGraw-Hill Ryerson Limited, 2003
Bodie
Kane Marcus Perrakis
Ryan
INVESTMENTS, Fourth Canadian Edition
Tests of the Single
Factor Model
Tests of the expected return beta
relationship
 First Pass Regression

Estimate beta, average risk premiums and
unsystematic risk
 Second Pass: Using estimates from the
first pass to determine if model is
supported by the data
 Most tests do not generally support the
single factor model
Slide 10-4
Copyright © McGraw-Hill Ryerson Limited, 2003
Bodie
Kane Marcus Perrakis
Ryan
INVESTMENTS, Fourth Canadian Edition
Thin Trading
 Many Canadian securities do not trade
very frequently
 This may cause biases in the statistical
estimates
 Several techniques exist to correct these
biases
Slide 10-5
Copyright © McGraw-Hill Ryerson Limited, 2003
Bodie
Kane Marcus Perrakis
Ryan
INVESTMENTS, Fourth Canadian Edition
Single Factor Test
Results
Return %
Predicted
Actual
Beta
Slide 10-6
Copyright © McGraw-Hill Ryerson Limited, 2003
Bodie
Kane Marcus Perrakis
Ryan
INVESTMENTS, Fourth Canadian Edition
Roll’s Criticism on the
Tests
 The only testable hypothesis: the meanvariance efficiency of the market
portfolio
 All other implications are not
independently testable
 CAPM is not testable unless we use the
true market portfolio
 The benchmark error
Slide 10-7
Copyright © McGraw-Hill Ryerson Limited, 2003
Bodie
Kane Marcus Perrakis
Ryan
INVESTMENTS, Fourth Canadian Edition
Measurement Error in
Beta
Statistical property:
 If beta is measured with error in the first
stage,
 Second stage results will be biased in the
direction the tests have supported
 Test results could result from
measurement error
Slide 10-8
Copyright © McGraw-Hill Ryerson Limited, 2003
Bodie
Kane Marcus Perrakis
Ryan
INVESTMENTS, Fourth Canadian Edition
Conclusions on the Tests’
Results
 Tests proved that CAPM seems
qualitatively correct


Rates of return are linear and increase with
beta
Returns are not affected by nonsystematic risk
 But they do not entirely validate its
quantitative predictions

Slide 10-9
The expected return-beta relationship is not
fully consistent with empirical observation.
Copyright © McGraw-Hill Ryerson Limited, 2003
Bodie
Kane Marcus Perrakis
Ryan
INVESTMENTS, Fourth Canadian Edition
Summary Reminder
 Objective: To discuss the empirical
evidence in support of equilibrium
models.



Slide 10-10
Tests of the Single Factor Model
Tests of the Multifactor Model
Other Studies
Copyright © McGraw-Hill Ryerson Limited, 2003
Bodie
Kane Marcus Perrakis
Ryan
INVESTMENTS, Fourth Canadian Edition
Tests of the Multifactor
Model
Factors identified by Chen, Roll and Ross
in their 1986 study:





Growth rate in industrial production
Changes in expected inflation
Unexpected inflation
Changes in risk premiums on bonds
Unexpected changes in term premium
on bonds
Slide 10-11
Copyright © McGraw-Hill Ryerson Limited, 2003
Bodie
Kane Marcus Perrakis
Ryan
INVESTMENTS, Fourth Canadian Edition
Study Structure &
Results
 Method: Two-stage regression with
portfolios constructed by size based on
market value of equity
Findings
 Significant factors: industrial production,
risk premium on bonds and unanticipated
inflation
 Market index returns were not statistically
significant in the multifactor model
Slide 10-12
Copyright © McGraw-Hill Ryerson Limited, 2003
Bodie
Kane Marcus Perrakis
Ryan
INVESTMENTS, Fourth Canadian Edition
Anomalies Literature
Is the CAPM or APT Model Valid?
 Numerous studies show the approach is
not valid
 Why do the studies show this result


Slide 10-13
Other factors influence returns on securities
Statistical problems prohibit a good test of
the model
Copyright © McGraw-Hill Ryerson Limited, 2003
Bodie
Kane Marcus Perrakis
Ryan
INVESTMENTS, Fourth Canadian Edition
Summary Reminder
 Objective: To discuss the empirical
evidence in support of equilibrium
models.



Slide 10-14
Tests of the Single Factor Model
Tests of the Multifactor Model
Other Studies
Copyright © McGraw-Hill Ryerson Limited, 2003
Bodie
Kane Marcus Perrakis
Ryan
INVESTMENTS, Fourth Canadian Edition
Fama and French Study
(1992)
 Size and book-to-market ratios explain
returns on securities
 Beta is not a significant variable when
other variables are included
 Study results show no support for the
CAPM or APT
Slide 10-15
Copyright © McGraw-Hill Ryerson Limited, 2003
Bodie
Kane Marcus Perrakis
Ryan
INVESTMENTS, Fourth Canadian Edition
Researchers’ Responses
to Fama and French
 Utilize better econometric techniques
 Improve estimates of beta
 Reconsider the theoretical sources and
implications of the Fama and Frenchtype results
 Return to the single-index model,
accounting for non-traded assets and
cyclical behavior of betas
Slide 10-16
Copyright © McGraw-Hill Ryerson Limited, 2003
Bodie
Kane Marcus Perrakis
Ryan
INVESTMENTS, Fourth Canadian Edition
Jaganathan and Wang
Study (1996)
 Included factors for cyclical behavior of
betas and human capital
 When these factors were included the
results showed returns were a function
of beta
 Size is not an important factor when
cyclical behavior and human capital are
included
Slide 10-17
Copyright © McGraw-Hill Ryerson Limited, 2003
Bodie
Kane Marcus Perrakis
Ryan
INVESTMENTS, Fourth Canadian Edition
Stochastic Volatility
 Stock prices change primarily in reaction
to information
 New information arrival is time varying
 Volatility is therefore not constant
through time
Slide 10-18
Copyright © McGraw-Hill Ryerson Limited, 2003
Bodie
Kane Marcus Perrakis
Ryan
INVESTMENTS, Fourth Canadian Edition
Stock Volatility Studies
and Techniques
 Pagan and Schwert Study


Study of 150 years of volatility on NYSE
stocks
Volatility is not constant through time
 Improved modeling techniques should
improve results of tests of the risk-return
relationship

Slide 10-19
GARCH Models to incorporate time varying
volatility
Copyright © McGraw-Hill Ryerson Limited, 2003
Bodie
Kane Marcus Perrakis
Ryan
INVESTMENTS, Fourth Canadian Edition
Equity Premium Puzzle
 Rewards for bearing risk appear too
excessive
 Possible causes:


Unanticipated capital gains
Survivorship bias
 Survivorship bias also creates the
appearance of abnormal returns in
market efficiency studies
Slide 10-20
Copyright © McGraw-Hill Ryerson Limited, 2003