|
|
|
Skeptical 2,
chinwhisker
06-03-2007, 10:43 AM | Post #2395202
| 0 |
  |
|
Forgive the title 'tato; just having a little fun.
First off, I would like to say that I do recognize Fama and French to be brilliant, do agree DFA funds offer higher focus toward small and value, which can add to a portfolio if the expenses can be warranted, and DFA has proven to be pretty decent at handling their funds as far as capturing value and creating positive benchmark error. The only major problem I have with DFA is expenses.
As far as F/F and FF3F, this is a grand recognition of factors that go into risk; you pay less for higher risk assets compared to expected future cash flows. This is a brilliant observance from the data that coincides with the factors of CAPM which is still used (to the best of my knowledge) to teach these young MBAs equity valuation. Dr. Damodaran of Stern University, NY, tells his students 85% of the return you would expect from an individual equity comes from the market in which you invest in.
That said, I think Jack Bogle hit the nail square on the head when he came up with the Cost Matters Hypothesis (CMH) in explaining indexing's superiority to managed funds. This would also explain the success of FF3F in explaining the returns and alpha/beta in those markets such as small and value, you might find the same thing with say healthcare and technology. The average return/alpha for all sectors or the market is going to be determined by the average return/alpha of all funds that invest in those markets, as Dr. D also said, it is the institutional investor who sets the prices on Wall Street. Small company stocks do offer higher returns because the future cash flows are less certain, and the same holds true with value business stocks. A shoe manufacturer is going to have lower return and lower beta than a software manufacturer.
Alpha (by way of regression beta) may have very well been the cause of the fact FF3F didn't work during the 90s. The use of regression beta meant that those stocks that followed the market had a beta of the market; in other words, no higher risk assessed in the CAPM model. The large cap stocks are the market, so the large cap stocks, in particular, the large cap growth stocks of this period created the beta of the market.
So, so far I have pretty much agreed with FF3F, huh?
Where my disagreement lies, is with using FF3F + historical data to predict the future. If the CAPM has changed (the model used to value equities), then the future predictability from historic data has also changed.
At least from Dr. D's "Bottom up beta" (Google it), the CAPM model has changed in that the discount rates for small caps will be smaller than historical discount rates using regression data. Also in the way he uses CAPM, equity debt is considered as well as financial debt when setting the Weighted Average Cost of Capital (WACC -- a factor in the discount rate). So the discount rate for large cap growth may actually be higher now -- may have been higher when investors got away from beta in the CAPM. The volatility measure for value stocks would also have changed. The CAPM may have caused investors to miss value due to recent dips in price (regression beta) in the 90s, but overall, the value type businesses would have had a lower beta than growth, as they are more stable businesses.
I don't know what this means. It could mean you will see lower returns from small caps going forward as the discount rate will be lower and higher returns for value as the discount rate will be higher or lower, lower -- who knows?
The CAPM was not accepted until the 90s, so it is different than the valuation metrics used prior, and the CAPM is going to be different going forward. Also the valuation metrics used in the earlier periods looked nothing like the valuation metrics used in the even earlier periods as well as the later period prior to the 90s. If FF3F is ever accepted as a valuation metric, this would wipe out the small/value premium.
The future depends on what valuation metrics we use, but yes, back-testing will still show FF3F to work just not to what degree.
Originally posted in thread: 59183
Topics
benchmark
cash flow
Gene Fama
historical data
John Bogle
View Complete Thread
|
|
|
