Fama-French 3-Factor Model: Theoretical and Conceptual Underpinnings

1
Fama-French 3-Factor Model Theoretical and
Conceptual Underpinnings

• Richard A. Michelfelder, Ph.D.
• Rutgers University
• School of Business Camden
• April 17, 2009

2
Disclaimer

• The views and methods reflected herein are those
  of the presenter and may not be consistent with
  those of Rutgers University, AUS Consultants, or
  any other affiliations of the presenter.

3
Todays Discussion

• 1. Introduction
• 2. Evolution from the Capital Asset Pricing Model
  
• 3. The Fama-French 3 Factor Model
• 4. A 4th Factor?
• 5. The Data
• 6. Estimation of the Factor Betas
• 7. Pros and Cons
• 8. Conclusion
• References

4
1. Introduction

• More evidence the better estimating cost of
  common equity
• DCF, CAPM, Risk Premium are the workhorses
• Wide range of estimates a lot of uncertainty
• Empirical theoretical issues with all methods
• Methods exist that are barely being used or not
  at all
• Fama-French 3 Factor Model (todays discussion)
• COMING SOON! Consumption Asset Pricing Model
  (maybe next years discussion completed research
  project with AUS Consultants)

5
2.Evolution from the CAPM

• CAPM
• Max return min risk by diversification
• Diversified-market portfolio return to explain
  stock return
• Other factors that explain returns?
• Fama-French (1992, 1993)
• Small size financial distress (high
  book/market) stocks have return premiums due to
  added non-diversifiable risk
• Combine these with market return to explain stock
  return
• Not proxies for unidentified state variables
• R-Squares are higher (10?50 rather than 1?5)

6
3. The Fama-French 3 Factor Model

• Fama-French 3 Factors
• RPM difference between returns on a diversified
  market portfolio (value-weighted CRSP returns)
  and a risk-free return
• SMB (small minus big) difference between
  returns on diversified portfolios of small and
  large capitalization stocks
• HML (high minus low) difference between returns
  on diversified portfolios of high (distressed
  firms) and low B/M (not distressed firms)
  stocks

7
3. The Fama-French 3 Factor

• The model in estimation form is
• Ri,t is the utility stock return during period
  t
• Rf,t if the return on the risk-free asset
• a should be zero
• ßs are the betas for each factor
• ßi,m is the CAPM beta
• ei,t is the regression error term

8
3. The Fama-French 3 Factor Model

• For application the model is (illustration
  purposes only!)
• or
• RFPL is an initial estimate of the FPL cost of
  common equity
• Rf,t is the return on the risk-free asset
• ßs are FPLs factor betas

9
4. A 4th Factor?

• 4th Factor (developed by Carhart (1997))
• MOM (momentum) difference between returns on
  diversified portfolios of stocks that perform
  well and poorly in the short-term (less than one
  year)
• Momentum is short-lived and therefore not useful
  to estimate the cost of capital although it does
  explain stock returns
• For cost of common equity capital estimates lets
  forget MOM.
• Doesnt explain utility stock returns anyway,
  based on our forthcoming research

10
5. The Data

• Web Site Source
• mba.tuck.dartmouth.edu/faculty/ken.french/data_lib
  rary.html
• Data is free and available to the public
• Daily, weekly, monthly, quarterly, annual
  frequencies beginning at 7/1926

11
5. The Data

• Factors are highly volatile ? volatile
  estimates of the cost of common equity capital
• Descriptive Statistics of 3 Factors
• 1Based on annual data ranging from 1926 to 2008
• 2Statistically significant at 1 level

12
6. Estimation of Factor Betas FPL

• 3 Factor
• CAPM

13
7. Pros and Cons

• Pros
• Recently developed addition to the toolkit for
  estimating the cost of common equity capital
• Explains a greater proportion of the
  non-diversifiable volatility of stocks returns
  relative to CAPM
• Cons
• Not based on theory, just a brute force way to
  explain more of the volatility in returns
• Factors are highly volatile as are cost of
  capital estimates
• Adding any variable to a regression increases
  R-square

14
8. Concluding Remarks

• Cannot ignore any additional evidence on cost of
  common equity capital estimation
• Fama-French 3 Factor Model is a newer tool that
  offers such evidence
• Starting to be adopted for utility ratemaking
• As we understand better the application issues,
  benchmark the results over time under different
  economic financial states, its adoption will
  grow
• Now we will hear more about application
  benchmarking issues for utility ratemaking

15
References

• Fama, E. and K. French, 1992. The Cross-Section
  of Expected Stock Returns, Journal of Finance
  32, 427-465
• Fama, E. and K. French, 1993. Common Risk
  Factors in the Returns on Stocks and Bonds,
  Journal of Financial Economics 39, 3-56
• Carhart, M. M., 1997. On Persistence in Mutual
  Fund Performance, Journal of Finance 52, 57-82
•