Successful Fama/French 3 Factor Analysis

[rokid]

I just completed a 3 factor analysis of Fidelity Low Price Stock (Flpsx). I've always wondered why it had high returns, low risk, and a low correlation with the market, i.e. a perfect diversifier.

Per Fama/French, up to 95% of a fund's return can be explained by beta, value, and size. Value and size have historically provided higher returns than predicted by CAPM, i.e. higher returns than their risk level would suggest.

Instructions for performing a 3 factor analysis with Excel are shown here:

http://www.efficientfrontier.com/ef/101/roll101.htm

Based on my analysis:

Flpsx

Alpha - 3.36% per year (1990-2006) attributed to good stock picking
Beta - .86 - lower risk and lower expected return than the market
Value loading .5 (this represents a value tilt)
Small factor loading .44 (not really small cap)

The 3 factors (beta, small, and value) account for 93% of the Flpsx return. The rest is alpha attributed to Joel Tillinghast's stock picking and timing expertise.

 

[dell]

Based on your analysis: What?

 

[Birchtree]

FLPSX is currently closed to the public for purchase. Don't know when if ever they will open it up - too much money wanting in to be managed.

 

[rokid]

Based on your analysis: What?

Basically, so what?! Right?

Eugene Fama and Kenneth French's research indicates that 95% of any fund's returns are the result of three factors: 1.) beta (is the fund more or less risky than the market), 2.) does the fund overweight/underweight small stocks (SmL), and 3.) does the fund overweight/underweight value stocks (HmL)?

The rule of CAPM is that additional returns are only possible by assuming additional risks. However, the Fama/French research shows that small stocks and value stocks (and particularly small value stocks) provide higher returns than would be expected based on risk (as measured by volatility). Consequently, a fund manager can beat the market (Wilshire 5000) by over weighting small value. There's probably some other kind of risk involved, but it is not volatility.

Alpha is a measure of the return that can be attributed to a manager's expertise in stock picking and/or market timing. Alpha is the holy grail of active fund managers. However, by including SmL and HmL factors, the alpha of many active managers disappears.

Consequently, since Flpsx has done so well over the last 16 years and I hold a significant amount of it, I wanted to determine whether or not Joel Tillinghast's success was due to over weighting small, over weighting value, excess alpha, and/or luck.

As I expected, it's due to a value tilt and alpha. Consequently, I will continue to hold a significant amount of Flpsx (low volatility, high return, and low correlation with the total market).

In addition, I can now perform the same calculations for any domestic equity fund. I also plan to run the calculations against the top TSPer returns to see if I can draw any conclusions.

Finally, as Birch states, Flpsx is closed to anyone who doesn't already have access to it through a 401K. In addition, most analysts think it is too big. However, in spite of the massive amounts of money he has to manage, Mr. Tillinghast continues to produce superior performance (17.76% vs 15.79% for the C Fund in 2006).

SmL = Small minus Large
HmL = High Book Value minus Low Book Value

 

[rokid]

Here’s an article about the famous Fama/French collaboration on 3 Factor research. Apparently, French is the very junior partner.:cheesy:

http://www.tuckprofit.com/0611_vol2_iss3/0611famafrench.htm