THREE FACTOR MODEL:

FAMA AND FRENCH (1992)

Oren Hovemann
Yutong Jiang
Erhard Rathsack
Jon Tyler
Cross Section of Expected Returns

 A Firm’s Size and its Book/Market ratio combine to become a

strong predictor of the Firm’s expected return
 The value of Beta as a predictor of return is challenged
 Additional known factors used to predict stock returns are the
firm’s Leverage and Earnings/Price ratios
 Size and B/M absorb the roles of Leverage and E/P ratios in
the Three Factor Model
Pricing Data

 Pricing data includes stocks from:

 NYSE

 AMEX

 NASDAQ

 Date Ranges from 1962 to 1989

 Data collected from CRSP and COMPUSTAT
 Historical reporting of Book Value limits data range
Financial Reporting vs. Returns
 Matching of returns with accounting data has a six month minimum gap.
 December accounting values are used to calculate (t – 1)
 Book/Market

 Leverage

 Earning/Price

 June accounting values are used to calculate (t)

 Size (price factor)

 The six month gap between financial reporting and realized returns insure
the reflection of all information into the stock pricing
 Different fiscal year-ends between firms complicate the timing of matching
accounting values with returns
100 Size–Beta Portfolios

 Portfolio Assignments
 First stocks are divided into Size ranked deciles

 Then each Size based decile is sub-divided into Pre-Ranked Beta deciles

 A stock can move between portfolios over time if either its size or pre-
ranked Beta changes
 Estimated Betas for each Portfolio
 Historical monthly returns are regressed against CRSP derived market
returns to estimate Post-Ranked Betas
 Estimated Betas for portfolios based on a Size-Beta ranking magnify the
range of Beta values
 Allows tests that distinguish between the effects of size and beta upon
stock returns
Size-Beta Portfolios increase range of
estimated Betas

 Size based beta variation is 1.44 – 0.92 = 0.52

 Size-Beta Portfolio based beta variation is 1.79 – 0.53 = 1.26
 The range of variation of Beta in Size-Beta based Portfolios is 1.26 / 0.52
= 2.4 times greater than Size based portfolios
Table II
 Strong negative relation between size and average
return.
 Strong positive relation between average return
and beta.
Table II
Table II
 No relation between average return and beta
during the 1964-1979 period.
 Table III
 Firm size ln(ME) is measured in June of year t
 If earnings are positive, E(+)/P is the ratio of total earnings to market equity
and E/P dummy is 0.
 If earnings are negative, E(+)/P is 0, E/P dummy is 1.
 T-statistic is the average slope divided by its time-series standard error.
Fama-Macbeth Regressions
 The Fama-Macbeth regression is a method used to estimate
parameters for asset pricing models.
 The method estimates the betas and risk premium for any risk
factors that are expected to determine asset prices.
 The method works with multiple assets across time (panel data).
 The parameters are estimated in two steps:
 First regress each asset against the proposed risk factors to
determine that asset's beta for that risk factor.
 Then regress all asset returns for a fixed time period against
the estimated betas to determine the risk premium for each
factor.
Table III
 Size helps explain the cross-section of average
stock returns.
 Market beta does not help explain average stock
returns for 1963-1990
 Table IV
 Portfolios ranked by values of book-to-market
equity (BE/ME) and earnings-price ratio (E/P)
 13 portfolios formed with the lowest and

highest portfolios split and stocks with negative

E/P in a separate portfolio (only for E/P)
 Negative BE firms not included (on average,
there are about 50/2317 per year)
 Table IV
Properties of Portfolios Formed on Book-to-Market Equity
(BE/ME)
 Table IV
Properties of Portfolios Formed on Earnings-Price Ratio
(E/P)
 Variables in Table IV
 Return – Time-series average of the monthly equal-weighted
portfolio returns (%)
 β – Time-series average of the monthly portfolio Bs
 Ln(ME) – Market equity representing firm size (outstanding shares
x share price)
 Ln(BE/ME) – Book equity divided by market equity
 Ln(A/ME) - Book assets divided by market equity
 Ln(A/BE) – Book assets divided by book equity
 E/P dummy – Dummy variable used to distinguish between
positive and negative earnings
 E(+)/P – Positive earnings to price ratio
 Firms – average number of stocks in the portfolio each month
 Average Returns
 Average returns sorted by BE/ME
 Strong positive relationship
 Difference of 1.53% from lowest to highest portfolios
 Unlikely a β effect
 Negative BE and high BE/ME have similar returns as a result
of capturing relative distress
 Average returns sorted by E/P
 Returns have a U shape
 Portfolio 0 (negative earnings) has higher than average
returns
 Returns increase as positive E/P portfolios increases
 Table IV
Properties of Portfolios Formed on Book-to-Market Equity
(BE/ME)
 Table IV
Properties of Portfolios Formed on Earnings-Price Ratio
(E/P)
 BE/ME
 Monthly regressions of returns on book-to-market
equity has strong relationship
 More significant than the size effect
 Book-to-market equity does not replace size
 Monthly returns of regressions on book-to-market equity
and size:
 Size has a slope of -.11 and a t-statistic of -1.99
 Book-to-market equity has a slope of .35 and a t-statistic of
4.44
 Table III
Average Slopes (T-Statistics) from Month-by-Month Regressions of Stock
Returns on β, Size, Book-to-Market Equity, Leverage, and E/P
 Leverage
 Two leverage ratios are used
 A/ME (book assets to market equity) - Measure of
market leverage
 A/BE (book assets to book equity) - Measure of book
leverage
 Both leverage ratios are related to average returns,
with opposite signs but similar absolute values
 The difference between these ratios is what helps
explain average returns
 Table III
Average Slopes (T-Statistics) from Month-by-Month Regressions of Stock
Returns on β, Size, Book-to-Market Equity, Leverage, and E/P
 Leverage & Book-to-Market
ln(BE/ME) = ln(A/ME) – ln(A/BE)
 Close link between leverage and BE/ME
 Two interpretations:
 High book-to-market ratio could be low stock price
compared to book value
 High book-to-market ratio could be a firms market
leverage is high relative to its book leverage
 Relative distress (captured by BE/ME) can also be
viewed as a leverage effect (captured by the
difference between A/ME and A/BE)
 E/P
 It is believed that earnings are a proxy for future
earnings
 E/P dummy is used because negative earnings are not a
proxy for future earnings
 E/P dummy (negative earnings) has a strong
relationship with returns
 Add size to the regression and the relationship becomes
insignificant
 This shows that the high returns for negative E/P is better
explained by size
 E(+)/P has a strong relationship with returns
 Table III
Average Slopes (T-Statistics) from Month-by-Month Regressions of Stock
Returns on β, Size, Book-to-Market Equity, Leverage, and E/P
 Table IV
Properties of Portfolios Formed on Earnings-Price Ratio
(E/P)
 E/P & Book-to-market
 Regressions of returns on ME, BE/ME and E/P gives
insignificant results for E/P
 Regressions of returns using ME, BE/ME and E/P
produce very similar results to regressions using just ME
and BE/ME for ME and BE/ME
 Suggests that E/P is insignificant in explaining returns when
book-to-market ratios are used
 Results suggest that the relationship between E(+)/P
and average return is mostly due to the positive
correlation between E/P and BE/ME
 Firms with high E/P have high book-to-market ratios
IV. A Parsimonious Model For Average
Returns
1) When we allow for variation in β that is unrelated to size,
there is no reliable relation between β and average return
2) The opposite roles of market leverage and book leverage in
average returns are captured well by book-to-market equity
3) The relation between E/P and average return seems to be
absorbed by the combination of size and book-to-market
equity.

Do not use β
Size and book-to-market equity are better indicators
A. Average Returns, Size and Book-
to-Market Equity
 A) Controlling for size, book-to-market equity
captures substantial variation in average returns
 B) Controlling for BE/ME leaves a size effect in
average returns.
A. Average Returns, Size, and Book-to Market Equity

Table V: Average Monthly Returns on Portfolios Formed on Size

and Book-to-Market Equity; Stocks Sorted by ME (Down) and then
BE/ME (Across): July 1963 to December 1990

0.58% per month average spread of returns

B. The interaction Between Size and
Book-to-Market Equity
 Low Market Equity
 Low stock prices
 High book-to-market equity
Table III
Correlation between ln(ME) and ln(BE/ME) = -0.26
C. Subperiod Averages of FM Slope

 Table III
 Sizehas a negative premium
 Book-to-Market has a positive premium

 Market β has a neutral 0 premium

 Table VI
 Subgroups created and tested with FM Slope
 β weak and inconsistent

 Size Effect lacks power

 Book-to-Market consistently reliable

 January Effect also found to be significant

D. Β and the Market Factor:
Caveats
 Average premiums for β, size, and book-to-market
equity depend on the definitions of the variables
used in the regressions.
 UsingB/E will change slope
 SLB model
 Overturns simple relationship between return and β being
flat
 Leaves β as the only variable
V. Conclusions and Implications
 Sharpe-Linter-Black (SLB) Model
 Positive
simple relation between average return and
market β (1926-1968)
 Reinganum (1981) and Lakonishok and Shapiro
(1986)
 (1963-1990)
V. Conclusions and Implications
 What variables can explain return?
 Banz (1981) Strong Negative Relationship between
return and firm size
 Bhandari (1988) Positive Relationship between return
and leverage
 Basu (1982) Positive Relationship between return and
E/P
 Rosenberg, Reid and Lanstein (1985) Positive
Relationship between return and book-to-market equity
 Chan, Hamao, and Lakonishok (1992) find that BE/ME
is powerful for predicting returns
A. Rational Asset-Pricing Stories
 What is the economic explanation for the roles of size and book-to-market
equity in average returns?
 Regression on returns in ln(ME) and Ln(BE/ME) are returns on portfolios
that mimic the underlying common risk factors in returns proxied by size
and book-to-market equity.
 Relation between size and average return proxies for a more
fundamental relation between expected returns and economic risk
factors.
 Relation between size and average return is a relative-prospects effect.
 More distressed firms are more sensitive to economic conditions
 BE/ME should be a direct indicator of the relative prospects of a firm
 Low BE/ME strong performance
B. Irrational Asset-Pricing Stories
 Asset pricing effects are not always rational
 Market overreaction to the prospects of the firm
C. Applications
 Size and Book-to-market equity describe the cross-
section of average stock returns.
 Willit persist?
 Does it result from rational or irrational asset-pricing?

 Explanatory power does not deteriorate over time

 Long-term average returns

 Form portfolios and measure success
 Alternate investment strategies
 Measure expected returns and evaluate performance