Finance For Normal People: Chapter 10: Behavioral Asset Pricing

Finance for Normal People
Chapter 10: Behavioral Asset Pricing

Behavioral Asset Pricing
Useful asset pricing models associate expected returns of

investment assets with factors or characteristics
Factors and characteristics include risk and liquidity
Investment asset pricing models are like pricing models of

meals, cars, movies and virtually every other product and
service
The expected price of a meal is a function of:

1. Wants for utilitarian benefits such as high nutrition and
great convenience
2. Wants for expressive and emotional benefits such as great

prestige and pleasing esthetics
3. Cognitive and emotional errors such as inferring wine

quality from its price
Behavioral asset pricing
Question
Think about a pricing model of cars, associating car prices with its features
and characteristics
What are some features and characteristics that account for the difference
in price between a Toyota Corolla and a Rolls Royce Phantom, and
between a Rolls Royce Phantom and a Lamborghini Veneno?
Describe the utilitarian, expressive and emotional benefits of each feature
and characteristic
What do car manufacturers do to enhance the utilitarian, expressive, and
emotional benefits of their cars?
How do car manufacturers and car sellers exploit the cognitive and
emotional errors of car buyers?
The expected return of a stock is a function of:

1. Wants for utilitarian benefits such as low risk and high liquidity
2. Wants for expressive and emotional benefits such as the virtue of

socially responsible funds, the prestige in hedge funds, and the thrill of
stock trading
3. Cognitive and emotional errors such as a belief that stocks of

admired companies are likely to yield higher returns than stock of
spurned companies, and that frequent trading is likely to yield higher
returns than rarer trading

Arbitrage in meal markets
Zagat directories
Arbitrage in investment markets
Closed-end funds
Question
Risk-free arbitrage occurs when we buy an item at one price and simultaneously
sell it or its perfect equivalent at a higher price
The binomial and Black-Scholes option pricing models are based on risk-free
arbitrage
But the term arbitrage is used also when it is not risk-free. Closed-end fund
arbitrage is one example
What are examples of arbitrage and why might it fail to eliminate the effect of
wants and errors on securities prices?
Theoretical asset pricing model
Capital Asset Pricing Model (CAPM)
Empirical asset pricing model
3-factor model
Standard Asset Pricing

Rationales for factors or characteristics account only for
wants for utilitarian benefits
Rationales for factors or characteristics also account for

wants for expressive and emotional benefits and the
occurrence of cognitive and emotional errors
The Theoretical Capital Asset Pricing Model (CAPM)
The expected return of an asset is a function of:

1. Risk related the Market factor
Figure 10-1: The capital Market Line
Expected
Returns Capital
Market
Line
M (Market
Portfolio)
Risk
Free
Rate
Standard Deviation of Returns

The theoretical CAPM
𝑅𝐺 - 𝑅𝑓 = 𝛼𝐺 + 𝛽𝐺 [𝑅𝑀 - 𝑅𝑓 ]
Where:
𝑅𝐺 is the realized return of Alphabet (Google)’s
stock
𝑅𝑀 is the realized return of the market
portfolio
𝛼𝐺 is the excess return of Alphabet (Google)’s
stock (i.e. its alpha)
𝛽𝐺 is the beta of Alphabet (Google)’s stock
𝜎𝐺
𝛽𝐺 = 𝜌𝐺𝑀
𝜎𝑀
Where:
𝛽𝐺 is Alphabet (Google) stock’s Market-
factor beta
𝜌𝐺𝑀 is the correlation between the returns
of Alphabet (Google)’s stock and the returns of
the market portfolio
𝜎𝐺 is the standard deviation of the returns of
Alphabet (Google)’s stock
𝜎𝑀 is the standard deviation of the returns of
the market portfolio
𝐸(𝑅𝐺 ) = 𝑅𝑓 + 𝛽𝐺 [𝐸 ሺ𝑅𝑀 ሻ− 𝑅𝑓 ]
Where:
𝐸 ሺ𝑅𝐺 ሻ is the expected return of Alphabet
(Google)’s stock
𝐸(𝑅𝑀 ) is the expected return of the market
portfolio
𝑅𝑓 is the risk-free rate
𝛽𝐺 is Alphabet (Google) stock’s Market-factor
beta
ൣ𝐸 ሺ𝑅𝑀 ሻ− 𝑅𝑓 ൧ is the “expected market-factor
return” or “market-risk premium” or “equity
premium”
The Empirical Three-Factor Asset Pricing Model
The expected return of an asset is a function of:

1. Risk related the Market factor;
2. Risk related to the Small-Large factor;
3. Risk related the Value-Growth factor

CAPM and Three-Factor Model
The expected return of the Barrier Fund by the CAPM:
Risk-free rate + 0.84 [Expected Market-factor return]
The expected monthly return of the Barrier Fund by the three-

factor model:
Risk-free rate
+ 0.82 [Expected Market-factor return]
+ 0.05 [Expected Small-Large-factor return]
– 0.05 [Expected Value-Growth-factor return]
Question
Excel file “factors-Students” includes monthly factor returns of 5

factors, market, small-large, value-growth, profitability, and
investment. Factor returns are returns minus Treasury-bill returns
that proxy for the risk free returns. It also includes monthly returns
minus Treasury-bill returns of four mutual funds:
VISGX - Vanguard Small Capitalization Growth Index Fund

VISVX - Vanguard Small Capitalization Value Index Fund
VUVLX - Vanguard US Value
VIGRX - Vanguard Growth Index Fund
5 Factors
Mkt-RF is the return on a portfolio of all US

stocks minus the return on Treasury bills
SMB (Small Minus Big) is the return on a small stock portfolio (small market
capitalization) minus the return on big stock portfolio (large market
capitalization
HML (High Minus Low) is the return on a value portfolio (high ratio of
book to market) minus the return on a growth portfolio (low ratio of
book to market)
RMW (Robust Minus Weak) is the return on a robust operating

profitability portfolio minus the return on a weak operating
profitability portfolio
CMA (Conservative Minus Aggressive) is the return on a conservative investment portfolio

minus the return on an aggressive investment portfolio
Examine the returns minus Treasury-bill returns of each mutual fund by

the CAPM, 3-factor model, and 5-factor model
What are the betas of each factor according to each model?
How do you interpret the betas?

Are the betas consistent with the names of the funds?
How are the betas similar or different across the 3 models?

VISGX- VISVX- VUVLX- VIGRX-

Mkt-RF SMB HML RMW CMA RF RF RF RF RF
200605 -3.57 -3.01 2.79 0.6 1.32 0.43 -6.39 -4.26 -3.48 -3.99
200606 -0.35 -0.38 1.49 1.05 -0.06 0.4 -1.20 0.69 -0.62 -0.76
200607 -0.78 -3.64 3.42 1.72 0.96 0.4 -5.70 -1.86 0.70 -2.42
200608 2.03 0.44 -1.78 -1.82 2.14 0.42 1.77 1.90 2.26 2.58
200609 1.84 -1.35 -0.5 1.63 0.53 0.41 0.48 0.47 2.20 2.43
200610 3.23 1.87 0.52 -0.95 0.23 0.41 5.25 4.41 2.20 3.03
200611 1.71 0.77 0.53 -0.16 -0.85 0.42 2.71 2.62 0.32 1.77
VISGX-RF
CAPM
Coefficients t Stat P-value
Intercept -0.07 -0.42 0.68
Mkt-RF 1.23 31.47 0.00

VISGX is a small growth fund
Based on the CAPM (1-factor model) it has a market-factor

beta of 1.23
So when the return of the stock market increases by 1

percentage point, the return of VISGX can be expected to
increase by 1.23 percentage points
3-Factor Model
Intercept -0.03 -0.37 0.71
Mkt-RF 1.09 50.37 0.00
SMB 0.78 18.87 0.00
HML -0.25 -6.46 0.00

Based on the 3-factor model, it tilts toward small

capitalization stocks and away from large capitalization stocks
(Its Small-Large (SMB) beta is positive 0.78)
It tilts away from value stocks and toward growth stocks (Its
Value-Growth (HML) beta is negative -0.25)
5-Factor Model
Intercept 0.09 1.02 0.31
Mkt-RF 1.05 48.64 0.00
SMB 0.74 18.83 0.00
HML -0.16 -3.79 0.00
RMW -0.20 -3.19 0.00
CMA -0.25 -3.52 0.00

Based on the 5-factor model, it tilts away from stocks with

robust profitability and stocks with conservative investment
Its RMW coefficient is negative -0.20, and its CMA coefficient

is negative -0.25
Empirical Evidence and Theoretical Rationales
What is the rationale for the value-growth factor?
What is the rationale for the profitability factor?

Fleeting and sustained factors in asset pricing

models
Does the association between a factor and future returns

disappear with no evidence of arbitrage?
Does the association disappear with evidence of arbitrage?
Does the association not disappear even with evidence of

arbitrage?
What is the rationale for the empirical associations between

stock returns and Small-Large and Value-Growth factors?
1. Data Mining Hypothesis:
The empirical associations are due to “data mining”

Expectations of stock returns =
3.47**
- 0.54** (Book-to-market ratio)
+ 0.31** (Market capitalization)
– 0.05 (Market-factor beta)
** Statistically significant at the 0.01 level or better
This is inconsistent with the data mining hypothesis


2. Risk Hypothesis:
The empirical associations are due to the role of the Small-

Large and Value-Growth factors as indicators of risk
Expectations of stock returns =
3.47**
- 0.54** (Book-to-market ratio)
+ 0.31** (Market capitalization)
– 0.05 (Market-factor beta)
This is inconsistent with the risk hypothesis


3. Cognitive Errors Hypothesis:
The empirical associations are due to cognitive errors, such

as representativeness, where investors extrapolate high past
sales, earnings, and other measures into the future
Table 10-2: Questionnaire where surveyed investors had only the
characteristics of companies
Past
Stock
Price-to- Market Return
Company Book Capitalization Bad Good
Company 1 Low High Average 1 2 3 4 5 6 7 8 9 10

Company 2 Average Low High 1 2 3 4 5 6 7 8 9 10
Company 3 Average High Low 1 2 3 4 5 6 7 8 9 10
Company 4 High Low High 1 2 3 4 5 6 7 8 9 10

Fortune ratings on long-term investment value =

5.93 + 0.06 (Characteristic score)
Not statistically significant
This is inconsistent with the cognitive errors hypothesis


4. Emotional Errors Hypothesis:
The empirical associations are due to emotional

errors, such as misleading affect
Expected-return score = 8.4 – 0.4** (Risk score)
This is consistent with the emotional errors hypothesis.

Table 10-3: Questionnaire where surveyed investors had only the
names of companies and their industries
Never heard of the

Company Industry Bad Good company
Apache Crude-Oil 1 2 3 4 5 6 7 8 9 10 ○
Sara Lee Food 1 2 3 4 5 6 7 8 9 10 ○
Cardinal
Health Health Care 1 2 3 4 5 6 7 8 9 10 ○
Procter
&Gamble Household 1 2 3 4 5 6 7 8 9 10 ○
Fortune ratings on long-term investment value =
2.13 + 0.6** (Affect score)

This is consistent with the emotional errors hypothesis


5. Wants for Expressive and Emotional Benefits

Hypothesis:
The empirical associations are due to wants for the high

expressive and emotional benefits delivered by stocks of
large and growth companies
Standard and behavioral asset pricing rationales
Empirical findings can consistent with more than one

possible rationale
Rationales for the momentum factor are one example

We illustrate behavioral asset pricing models by adding two

social responsibility factors to the four-factor asset pricing
model with Market, Small-Large, Value-Growth, and
Momentum factors:
The Top-Bottom factor – Reflecting cognitive errors

The Accepted-Shunned factor – Reflecting wants
The “Top-Bottom” factor consists of the difference between

the returns of stocks of companies ranked high and low on
five social responsibility criteria:
1. Community (e.g., generous giving, support for housing)

2. Diversity (e.g., promotion of women and minorities)
3. Employee relations (e.g., strong union relations)
4. Environment (e.g., pollution prevention)
5. Products (e.g., product quality and safety)
The “Accepted-Shunned” factor consists of the difference

between the returns of stocks of companies commonly
accepted by socially responsible investors and the returns of
company stocks commonly shunned by them
Shunned stocks include stocks of companies in the alcohol,

tobacco, gambling, firearms, military, and nuclear industries
Table 10-4: Comparison of the Barrier fund to the Vanguard 500 fund
by a six factor model

Barrier fund Vanguard
500 fund
Alpha (annualized) 1.48% -0.68%
Beta of Market factor 0.87 0.98
Beta of Small-Large Factor 0.15 -0.16
Beta of Value-Growth Factor -0.11 0.01
Beta of Momentum Factor 0.11 0.00
Beta of Top-Bottom Factor -0.05 0.03

Beta of Accepted-Shunned
Factor -0.40 0.01
Factors and characteristics in asset pricing models
“Smart Beta” and asset pricing models

Finance For Normal People: Chapter 10: Behavioral Asset Pricing

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Finance For Normal People: Chapter 10: Behavioral Asset Pricing

Uploaded by

Copyright:

Available Formats

Finance for Normal People

Chapter 10: Behavioral Asset Pricing

Useful asset pricing models associate expected returns of

Factors and characteristics include risk and liquidity

Investment asset pricing models are like pricing models of

The expected price of a meal is a function of:

2. Wants for expressive and emotional benefits such as great

3. Cognitive and emotional errors such as inferring wine

The expected return of a stock is a function of:

2. Wants for expressive and emotional benefits such as the virtue of

3. Cognitive and emotional errors such as a belief that stocks of

Arbitrage in meal markets

Arbitrage in investment markets

Theoretical asset pricing model

Capital Asset Pricing Model (CAPM)

Empirical asset pricing model

Standard Asset Pricing

Behavioral Asset Pricing

Rationales for factors or characteristics also account for

The Theoretical Capital Asset Pricing Model (CAPM)

The expected return of an asset is a function of:

Standard Deviation of Returns

The Empirical Three-Factor Asset Pricing Model

The expected return of an asset is a function of:

2. Risk related to the Small-Large factor;

3. Risk related the Value-Growth factor

The expected return of the Barrier Fund by the CAPM:

Risk-free rate + 0.84 [Expected Market-factor return]

The expected monthly return of the Barrier Fund by the three-

Excel file “factors-Students” includes monthly factor returns of 5

VISGX - Vanguard Small Capitalization Growth Index Fund

Mkt-RF is the return on a portfolio of all US

RMW (Robust Minus Weak) is the return on a robust operating

CMA (Conservative Minus Aggressive) is the return on a conservative investment portfolio

Examine the returns minus Treasury-bill returns of each mutual fund by

What are the betas of each factor according to each model?

How do you interpret the betas?

How are the betas similar or different across the 3 models?

VISGX- VISVX- VUVLX- VIGRX-

Coefficients t Stat P-value

Intercept -0.07 -0.42 0.68

Mkt-RF 1.23 31.47 0.00

VISGX is a small growth fund

Based on the CAPM (1-factor model) it has a market-factor

So when the return of the stock market increases by 1

Coefficients t Stat P-value

Intercept -0.03 -0.37 0.71

Mkt-RF 1.09 50.37 0.00

SMB 0.78 18.87 0.00

HML -0.25 -6.46 0.00

VISGX is a small growth fund

Based on the 3-factor model, it tilts toward small

Coefficients t Stat P-value

Intercept 0.09 1.02 0.31

Mkt-RF 1.05 48.64 0.00

SMB 0.74 18.83 0.00

HML -0.16 -3.79 0.00

RMW -0.20 -3.19 0.00

CMA -0.25 -3.52 0.00

VISGX is a small growth fund

Based on the 5-factor model, it tilts away from stocks with

Its RMW coefficient is negative -0.20, and its CMA coefficient

Empirical Evidence and Theoretical Rationales