Challenge to the Fama French Three Factor Model

Discuss all general (i.e. non-personal) investing questions and issues, investing news, and theory.
User avatar
Rick Ferri
Posts: 8201
Joined: Mon Feb 26, 2007 11:40 am
Location: Vagabond, USA, Twitter: @Rick_Ferri
Contact:

Challenge to the Fama French Three Factor Model

Post by Rick Ferri » Sat Oct 06, 2012 9:52 am

The Fama French Three Factor Model is a topic of frequent discussion on this forum. It's used to explain differences in the returns of diversified equity portfolios. The model compares a diversified portfolio to three distinctive types of risk found in the equity market; beta, size and BtM (book value-to-market value). Prior to the three-factor model, the Capital Asset Pricing Model (CAPM) was used as a "single factor" way to explain portfolio returns.

Since its inception in the early 1990s, the FF 3-factor model has become ingrained in academic community and by investors as a way to structure portfolios along the three risk axis. DFA was the first investment firm to specifically use this model in portfolio management and promote the concept among advisers. Morningstar help popularize the idea in the late 1990s with their 9 equiity style boxes. FF 3-factor investing become mainstream in the adviser community during the mid-2000s after the demise of growth stocks and large outperformance of small cap and value stocks.

Something as big and successful as the FF 3-factor model is going to attract competition from other academics, and there have been several alternative pricing models since its introduction. The latest challenge is from by Kewei Hou, Ohio State University, Chen Xue, University of Cincinnati, and Lu Zhang, Ohio State University. Here is a link to their paper:

Digesting Anomalies: An Investment Approach

The authors propose a new multifactor model, which they call the "q-factor model". The authors claim the new model is a significant improvement over the FF model in explaining asset returns. In short, the q-factor model replaces BtM in the FF 3-factor model with two other factors, one relating to the gain in asset value and the other relating to the return on equity (ROE).

In the q-factor model, the expected return of an asset in excess of the riskless rate is described by the sensitivity of its return to four factors: (i) the market excess return (MKT), (ii) the difference between the return on a portfolio of small-market equity stocks and the return on a portfolio of big-market equity stocks, (iii) the difference between the return on a portfolio of low-investment stocks and the return on a portfolio of high-investment stocks, and (iv) the difference between the return on a portfolio of high return on equity (ROE) stocks and the return on a portfolio of low return on equity stocks.

I found the q-factor interesting and know that many FF follower on this forum will find the reseach paper interesting also. I don't see this new model taking over the FF model any time soon, but papers that offer new insight into asset pricing is something that many people want to be aware of.

Rick Ferri

.
Last edited by Rick Ferri on Sat Oct 06, 2012 1:21 pm, edited 2 times in total.
The views expressed by Rick Ferri are strictly his own as a private investor and author and do not reflect the views of any entity or other persons.

User avatar
LadyGeek
Site Admin
Posts: 39966
Joined: Sat Dec 20, 2008 5:34 pm
Location: Philadelphia
Contact:

Re: Challenge to the Fama French Three Factor Model

Post by LadyGeek » Sat Oct 06, 2012 10:14 am

Thanks for the info, Rick. There's a new wiki article under development that describes how the Fama-French Three-factor model is used: Fama-French three-factor model analysis

The intent is to show how members can do the regression analysis (curve fitting) themselves. It's near completion, but I thought it was important in the context discussed here.

(To the new investors: This is an advanced investing topic and can be safely ignored.)
To some, the glass is half full. To others, the glass is half empty. To an engineer, it's twice the size it needs to be.

Jerry_lee
Posts: 915
Joined: Tue Mar 06, 2012 1:55 pm

Re: Challenge to the Fama French Three Factor Model

Post by Jerry_lee » Sat Oct 06, 2012 11:13 am

Rick,

Thanks for the article. A couple things: DFA was using the FF 3 F model in their business almost as it was being published. 1993 or before. It took the rest of the industry till the late 90s, as M* transitioned to their 9 style box model (which is just a commercialized FF effort) and Lipper went from "Growth", "Equity Income", to LG, LV, etc.

Next, the paper you post lists 4 factors, 2 are FF factors (market and size), one is a variation on value, and the 4th is just a "profitability" factor. Only the 4th is significant, and its already been well researched ny Robert Novy-Marks as you can read about in his paper "The Other Side of Value" (It mentions that Fama assisted at the bottom of the title page).

So it does appear that PmU (profitable minus unprofitable) is a distict return dimension worthy of inclusion in the 3F model. No need to start over, just evolve as our understanding of markets improves.

The implication of this is two-fold:

1) structured value portfolios should consider avoiding highly unprofitable stocks measured by gross revenue to assets (DFA has started doing this which may (?) explain why their US LV fund is the #1 ranked fund in the category of 600+ funds over the last few months--something you should never see from a diversified index)

2) instead of a pure growth fund (LG index, S&P 500, etc.), a portfolio of the most profitable companies (gross revenue to assets) than avoids low BtM stocks is a better "growth" compliment to a value fund--or, as Marks says, is "the other side of value". DFA has US/int'l large and small profitability funds in registration and we'll see them live in a month or two

As they say: the final chapter will never be written. And as investors, we all benefit from academic efforts to understand market behavior and price unique dimensions of risk/return.
The most disciplined investor in the world.

stlutz
Posts: 3921
Joined: Fri Jan 02, 2009 1:08 am

Re: Challenge to the Fama French Three Factor Model

Post by stlutz » Sat Oct 06, 2012 12:08 pm

Thanks for the link, Rick.

One thing I'll note is that the ROE factor was based on Compustat quarterly data, which is restated. In other words it assumes company reporting changes that would not have been known at the time the portfolios were theoretically constructed. Using non-restated annual data is better for these long-term backtests in that it's not restated, but worse in that the backtest is then based on stale data. Using point-in-time datasets that various providers have (which tell you exactly what information was known the market at any time) is even better, although the data doesn't go back as far.

In the real world, doing a backtest using data that would not have been known at the time is a big no-no, much more so than in the academic community.

Jerry_lee
Posts: 915
Joined: Tue Mar 06, 2012 1:55 pm

Re: Challenge to the Fama French Three Factor Model

Post by Jerry_lee » Sat Oct 06, 2012 12:13 pm

LadyGeek wrote:Thanks for the info, Rick. There's a new wiki article under development that describes how the Fama-French Three-factor model is used: Fama-French three-factor model analysis

The intent is to show how members can do the regression analysis (curve fitting) themselves. It's near completion, but I thought it was important in the context discussed here.

(To the new investors: This is an advanced investing topic and can be safely ignored.)


Lady Geek,

Regression analysis is a complex topic. Thankfully you don't need it to invest according to the FF3F model, which I think all new investors need to know about. After all, the FF model explains to us almost everything we need to know about risk and return, why wouldn't newbies want to know this?

Imagine we "hide" the size/value info from someone afraid it is too complex for them to grasp (they just couldn't get their head around adding Vanguard SV to TSM), and after 12 years of 0 real returns on TSM, they notice every active manager over this period who buys smaller/more value-oriented stocks has done better and they decide to give up on indexing and go the active route. Wouldn't we feel bad for not telling hem the truth and having them index S/V instead?
The most disciplined investor in the world.

Jerry_lee
Posts: 915
Joined: Tue Mar 06, 2012 1:55 pm

Re: Challenge to the Fama French Three Factor Model

Post by Jerry_lee » Sat Oct 06, 2012 12:23 pm

stlutz wrote:Thanks for the link, Rick.

One thing I'll note is that the ROE factor was based on Compustat quarterly data, which is restated. In other words it assumes company reporting changes that would not have been known at the time the portfolios were theoretically constructed. Using non-restated annual data is better for these long-term backtests in that it's not restated, but worse in that the backtest is then based on stale data. Using point-in-time datasets that various providers have (which tell you exactly what information was known the market at any time) is even better, although the data doesn't go back as far.

In the real world, doing a backtest using data that would not have been known at the time is a big no-no, much more so than in the academic community.


This isn't an issue. Most research portfolios (including FF 2x3 portfolios and the profitability portfolios in the Marks paper I referenced) use data as of year end and reconstitute only once a year, a full 6 months later in June. That is more than enough time in the real world to update metrics to include restatements.

Further, restatements that would have any material impact even without this lag in a broadly diversified portfolio are minuscule.

The amount of return variation that we cannot explain if we restrict our risk/return factors to just bonds vs cap-weighted TSM, on the other hand, is enormous.
The most disciplined investor in the world.

Bill Bernstein
Posts: 536
Joined: Sat Jun 23, 2007 12:47 am

Four Factor

Post by Bill Bernstein » Sat Oct 06, 2012 1:00 pm

Most academic studies these days add in the momentum factor, and so do a four-factor analysis; the momentum factor is downloadable from Ken French's site.

Adding it really does change things; for example, the tilted DFA portfolios have a percent or two negative alpha for year with 3-factor, almost none with four-factor. (Momentum having significantly negative returns in Mar-May of '09 during the "snap back.")

How to use this is another question; when it comes to passive portfolios, I've learned to pay more attention to fund family corporate culture and discipline than to the statistical analysis. Vanguard and DFA both score high in those areas; the other families, in my opinion, still have to prove themselves.

Best,

Bill

User avatar
grayfox
Posts: 4552
Joined: Sat Sep 15, 2007 4:30 am
Location: Earth

Re: Challenge to the Fama French Three Factor Model

Post by grayfox » Sat Oct 06, 2012 1:08 pm

From what I have read in papers by Cochrane, the current thinking is that there is a whole panoply of factors beyond the 3 factors of Fama and French.

This news should be welcomed by advisors, because it complicates investing, requiring the help of professionals.
Gott mit uns.

User avatar
LadyGeek
Site Admin
Posts: 39966
Joined: Sat Dec 20, 2008 5:34 pm
Location: Philadelphia
Contact:

Re: Challenge to the Fama French Three Factor Model

Post by LadyGeek » Sat Oct 06, 2012 1:10 pm

Jerry_lee wrote:
LadyGeek wrote:(To the new investors: This is an advanced investing topic and can be safely ignored.)

...Regression analysis is a complex topic. Thankfully you don't need it to invest according to the FF3F model, which I think all new investors need to know about. After all, the FF model explains to us almost everything we need to know about risk and return, why wouldn't newbies want to know this?

I wanted to give guidance on what's considered "basic" and "advanced" things new investors need to know. My dividing line is that anything which goes beyond a stocks/bonds total market approach, e.g. 3-fund portfolio level, is considered "advanced" - slice-n-dice, tilting, etc. When they are ready, and want to start deviating from a total market approach, certainly come back and learn about these models.

BTW, wbern is William Bernstein.
To some, the glass is half full. To others, the glass is half empty. To an engineer, it's twice the size it needs to be.

User avatar
Rick Ferri
Posts: 8201
Joined: Mon Feb 26, 2007 11:40 am
Location: Vagabond, USA, Twitter: @Rick_Ferri
Contact:

Re: Challenge to the Fama French Three Factor Model

Post by Rick Ferri » Sat Oct 06, 2012 1:16 pm

Jerry_lee wrote:Rick, thanks for the article. A couple things: DFA was using the FF 3 F model in their business almost as it was being published. 1993 or before. It took the rest of the industry till the late 90s, as M* transitioned to their 9 style box model (which is just a commercialized FF effort) and Lipper went from "Growth", "Equity Income", to LG, LV, etc.


Jerry_Lee,

Thanks for the clarification. I've changed the wording of my original post slightly to reflect these facts.

Rick Ferri
The views expressed by Rick Ferri are strictly his own as a private investor and author and do not reflect the views of any entity or other persons.

stlutz
Posts: 3921
Joined: Fri Jan 02, 2009 1:08 am

Re: Challenge to the Fama French Three Factor Model

Post by stlutz » Sat Oct 06, 2012 1:23 pm

This isn't an issue. Most research portfolios (including FF 2x3 portfolios and the profitability portfolios in the Marks paper I referenced) use data as of year end and reconstitute only once a year, a full 6 months later in June. That is more than enough time in the real world to update metrics to include restatements.


That's not how financial reporting works.

There are multiple opportunities for a company to restate quarterly data. Any 2012 quarterly numbers that are reported as of now can be restated any time through 2013 reporting. You won't be certain that the 2012 numbers won't change again until 2013 reporting is closed out in early 2014. If you ran a backtest 10 years from now and you used Compustat quarterly data (not Compustat point-in-time), you'd be basing your backtest on data known in 2014, not 2012.

Between 1/4 and 1/2 of all companies will restate net income, and large percentage of the restatements are large. That's often because of accounting rules. If they have discontinued operations, for example, companies have to restate prior quarters as if those operations had been discontinued then.

One reason I'm sure Fama/French chose price to book as their "value" metric is because such numbers are much less likely to change in the database than income-statement based data.

Any time your backtest has look-ahead bias, even in a subtle form like this, you are going to "explain" more.

Further, restatements that would have any material impact even without this lag in a broadly diversified portfolio are minuscule.


Not when you are searching for anomalies or ways to "beat the market". For example, academic papers we've discussed here in the past show that the entire value premium is due to <1% of the companies. In other words, the average value company doesn't outperform, there are simply a small number that outperform to such a large extent that a "premium" is created. Data problems with a small number of companies can and do affect backtest results.

The amount of return variation that we cannot explain if we restrict our risk/return factors to just bonds vs cap-weighted TSM, on the other hand, is enormous.


No argument from me--as I've argued in other threads, beta has no predictive ability for comparing stock returns. I'm generally in favor for using more factors to evaluate the variability in returns, and not just picking a couple of favorites and claiming that all of the mysteries of the world have been solved. For that reason, I still like the paper, it just has a problem which I was pointing out.

hafius500
Posts: 217
Joined: Sat Sep 10, 2011 4:31 pm

Re: Challenge to the Fama French Three Factor Model

Post by hafius500 » Sat Oct 06, 2012 1:48 pm

Jerry_lee wrote:...
So it does appear that PmU (profitable minus unprofitable) is a distict return dimension worthy of inclusion in the 3F model. No need to start over, just evolve as our understanding of markets improves.
...
The implication of this is two-fold:

1) structured value portfolios should consider avoiding highly unprofitable stocks ...

2) instead of a pure growth fund (LG index, S&P 500, etc.), a portfolio of the most profitable companies (gross revenue to assets) than avoids low BtM stocks is a better "growth" compliment to a value fund-...

As they say: the final chapter will never be written. And as investors, we all benefit from academic efforts to understand market behavior and price unique dimensions of risk/return.


So you say science progresses in its efforts to prove how we (=all investors!) can beat the market without possessing superior trading skills or superior (insider) knowledge, just by taking more "risk". This reminds me of Falkenstein's thoughts on risk premiums and alpha:

That is, if you find a currency, or low book equity/market equity stock that has a high value loading, these assets don't generate higher-than-average returns as theory suggests they should. So, if you magically discovered the risk that underlies risk premiums, it would be in your best interest to keep it to yourself, not give it away to investors, and call it alpha, because no one would know. It simply isn't plausible people are finding risk premiums and giving this to passive investors..,


It seems if we need sophisticated trading strategies or updated academic theories to capture (risk) premiums we need alpha.

I recall a sociologist (I forgot the name) wrote the notion that a higher return should be<=>is the reward for taking higher risk originates in Medieval Christian theology.
prior username: hafis50

bschultheis
Posts: 200
Joined: Mon Jul 23, 2007 10:15 am

Re: Challenge to the Fama French Three Factor Model

Post by bschultheis » Sat Oct 06, 2012 2:17 pm

LadyGeek wrote:
(To the new investors: This is an advanced investing topic and can be safely ignored.)


Thanks for including this statement LadyGeek.

I would only add that, in my opinion it can be ingored not only by new investors, but by investors who are successful investors, have been around the block a few times and consider themselves "sophisticated" investors.

For folks who like discussing this stuff, it serves for interesting debate, and this board is a good place to discuss.

What does it have to do with reaching long term financial goals? Nothing - or at least not much.

99.999% of investors would be better off picking a simple passive portfolio and focusing their attention, not on models, but on saving rates and burn rates.

I only point this out because when someone reads this thread, there is a tendency to think that somehow this is important.

It isn't. As Lady Geek says, it can be safely ignored.

Bill

Allan
Posts: 763
Joined: Wed Feb 21, 2007 9:15 pm
Location: Houston

Re: Challenge to the Fama French Three Factor Model

Post by Allan » Sat Oct 06, 2012 2:22 pm

bschultheis wrote:
LadyGeek wrote:
(To the new investors: This is an advanced investing topic and can be safely ignored.)



I only point this out because when someone reads this thread, there is a tendency to think that somehow this is important.

It isn't. As Lady Geek says, it can be safely ignored.

Bill


Thanks. I needed that!

User avatar
Taylor Larimore
Advisory Board
Posts: 25866
Joined: Tue Feb 27, 2007 8:09 pm
Location: Miami FL

Keep investing simple

Post by Taylor Larimore » Sat Oct 06, 2012 2:32 pm

To the new investors: This is an advanced investing topic and can be safely ignored.


Lady Geek:

I agree with Bill Schultheis and others that your message, above, to new investors is very important. Successful investing is very simple as these experts attest:
Scott Adams, author of Dilbert: "I once tried to write a book about personal investing. - After extensive research I realized I could describe everything that a young first-time investor needs to know on one page."

Christine Benz, Morningstar Director of Personal Finance: "Simplicity is one of the greatest--but in my view, woefully underrated--virtues when managing a portfolio."

Bill Bernstein, author of Four Pillars of Investing: "If, over the past 10 or 20 years, you had simply held a portfolio consisting of one quarter each of indexes of large U.S. stocks; small U.S. stocks; foreign stocks; and high quality U.S. bonds, you would have beaten over 90% of all professional money managers and with considerably less risk."

Richard Bernstein, Merrill Lynch strategist: "Investors find it hard to believe that ignoring the vast majority of investment noise might actually improve their performance."

Jack Bogle, Vanguard founder: "Simplicity is the master key to financial success. When there are multiple solutions to a problem, choose the simplest one."

Jack Brennan, Vanguard CEO and author of Straight Talk on Investing: "It's in the interest of many financial service companies to make you think that investing is difficult.--It's really quite simple."

Warren Buffet, one of the world's most successful investors: "There seems to be some perverse human characteristic that likes to make easy things difficult."

Scott Burns, Columnist: "Concentrate on a simple portfolio with two basic assets--a stock index fund and a bond index fund. Do that and you'll enjoy superior performance with less risk."

Andrew Clarke, author of "Wealth of Experience": "In investing, simple is usually more productive than complex."

Jonathan Clements, author of "You've Lost It. Now What?": "Investing is simple. To be sure, you can make it ludicrously complicated."

Paul Crafter, author of "Investment Guide": "After doing it all, I now feel I've come around in a complete circle, ending up with this: The more I learn, the less I really need to know."

Michael Edesess, author of The Big Invesment Lie: "As a mathematician I know when mathematical-sounding analyses are little more than elaborate sales pitches, designed to thoroughly obscure the simple fact that smart investing is non-mathematical and accessible to everyone."

Charles Ellis, author of "Winning the Loser's Game": "Investment advice doesn't have to be complicated to be good."

Rick Ferri, CFA, author of "All About Index Funds": "It does not take much to outperform the average investor. All you have to do is put half your money in the Vanguard Total Stock Market and the other half in an intermediate-term bond index fund. Then rebalance your account once per year. By keeping it simple, you will achieve all the benefits the markets have to offer."

Gensler & Baer, authors of "The Great Mutual Fund Trap": "If you simply buy and hold you don't need to read investing magazines, watch financial news networks, subscribe to newsletters, or pay a broker to execute new trades."

Albert Einstein: "The five ascending levels of intellect: "Smart; Intelligent; Brilliant, Genius, Simple."

Rick Ferri, CFA, author & advisor: "Most of the stuff you see about optimal allocation is garbage. -- So, choose a few low cost index funds in different asset classes, rebalance occasionally, and forgetaboutit.""

Future Metrics looked at the performance of 224 pension plans over about 14 years compared with the performance of 60% S&P 500 index and 40% aggregate bond index benchmark. Of those 224 plans, only 19 beat that simple benchmark.

Benjamin Graham: "In the stock market, the more elaborate and abstruse the mathematics, the more uncertain and speculative are the conclusions we draw therefrom."

Alan Greenspan, former chairman of the Federal Reserve: "This decade is strewn with examples of bright people who thought they built a better mousetrap that could consistently extract abnormal returns from financial markets. Some succeed for a time. But while there may occasionally be misconfigurations among market prices that allow abnormal returns, they do not persist."

Daniel Kahneman, Nobel Laurete: "All of us would be better investors if we just made fewer decisions"

Edmund Kean: "Complexity is easy. Simplicity is hard."

"Michael LeBoeuf, author of "The Millionaire in You": "The master key to wealth can be summed up in just one word: Simplicity."

MIT study: "The less well-informed group did far better than the group that was given all the financial news."

Joe Maglia, CEO TD Ameritrade: "Wall Street goes out of its way to make investing incredibly sophisticated and complex because they can make a tremendous amount of money by doing so."

Burton Malkiel, author of "Random Walk Down Wall Street": "The overarching rule for achieving financial security: Keep it simple."

John Markese, CEO of American Association of Individual Investors: "If you have more than eight funds you should slap yourself."

Wm McNabb, Vanguard CEO: "If you can't understand an investment product in five minutes, walk away."

James Montier, author: "Never underestimate the value of doing nothing."

Morningstar Guide to Mutual Funds: "Good investing doesn't have to be complicated. In fact, simplification may lead to better investment results."

Suze Orman: "We make investing so complicated and it really is not. -- A total market index fund is a great one-stop-shopping choice that provides you instant diversification among different types of stocks."

Mike Piper, financial author: "There's an entire industry built on convincing us that investing is complicated."

Jane Bryant Quinn, author of "Smart and Simple Financial Strategies": "You shouldn't buy anything too complex to explain to the average 12-year old."

John Rekenthaler, Morningstar Research Director: "How many funds should you have? Four to six should do."

Paul Samuelson, Nobel Laurete: "Investing should be like watching paint dry or watching grass grow. If you want excitement, take $800 and go to Las Vegas."

Bills Schulthies, author of "The Coffeehouse Investor": "When you simplify your investment decisions, not only do you enrich your life by spending more time on families, friends and careers, but you enhance portfolio returns in the process."

Chandon Sengupta, author of "The Only Proven Road to Investment Success": "There is overwhelming evidence that the simplest possible investment method works much better than all the other more complex ones."

Larry Swedroe, author of "The Successful Investor Today": "The more complex the investment, the faster you should run away."

David Swensen, Yale Chief Investment Officer: "As a general rule of thumb, the more complexity that exists in a Wall Street creation, the faster and farther investors should run."

Andrew Tobias, author: "I believe in selecting the most straighforward and easiest-to-implement strategy for achieving our goals."

Tweddell and Pierce, authors of "Winning with Index Mutual Funds": "Keep it simple. Investment success depends on asset allocation, diversification, and risk management, not on complexity."

Walter Updegrave, editor of MONEY magazine: "Simpler is better. Ignore the siren song of sophisticated investments"

Richard Young, author of "The Intelligence Report": "If you can't run your portfolio taking 60 minutes a month, it's too complicated."

Jason Zweig, author of "The Intelligent Investor": "The less you fool with your portfolio, the less often you'll play the fool."


Best wishes.
Taylor
"Simplicity is the master key to financial success." -- Jack Bogle

User avatar
magellan
Posts: 3446
Joined: Fri Mar 09, 2007 4:12 pm

Re: Challenge to the Fama French Three Factor Model

Post by magellan » Sat Oct 06, 2012 3:38 pm

Since we're discussing follow ons to the three factor model, Some folks might find this paper by Gerakos and Linnainmaa at Chicago Booth interesting. It asserts that recent equity returns may offer a better sort than book-to-market for isolating the value component.
Abstract:
We show that all pricing-relevant variation in book-to-market ratio-sorted portfolios is due to changes in the market value of equity. Although the remaining variation across these portfolios explains comovement among stocks, it does not spread returns. The HML factor therefore consists of a priced and unpriced risk component. Because every factor has just one price of risk, the three-factor model gives the appearance of high risk-adjusted returns for strategies that covary negatively with the unpriced component. This finding explains why the three-factor model appears to price anomalies associated with earnings-to-price and cashflow-to-price ratios. When we adjust these ratios to not overlap with the unpriced part of the HML factor, these anomalies resurface.


This is interesting for a bunch of reasons. If it holds up, it would 'clean up' the value factor to get rid of some extraneous unpriced stuff that book-to-market seems to pull in, which maybe can then be explained with other factors.

As an aside, I've always been uncomfortable with the three factor model's reliance on book-to-market, which is something of an accounting fiction. In the grand scheme of things, my quibble is probably meaningless, but it'd be nice to see book-to-market out of there.

Jim

(edited to fix typo)
Last edited by magellan on Sat Oct 06, 2012 4:00 pm, edited 1 time in total.

User avatar
LadyGeek
Site Admin
Posts: 39966
Joined: Sat Dec 20, 2008 5:34 pm
Location: Philadelphia
Contact:

Re: Challenge to the Fama French Three Factor Model

Post by LadyGeek » Sat Oct 06, 2012 3:51 pm

bschultheis wrote:
LadyGeek wrote:
(To the new investors: This is an advanced investing topic and can be safely ignored.)

Thanks for including this statement LadyGeek. I would only add that, in my opinion it can be ingored not only by new investors, but by investors who are successful investors, have been around the block a few times and consider themselves "sophisticated" investors.

For folks who like discussing this stuff, it serves for interesting debate, and this board is a good place to discuss. What does it have to do with reaching long term financial goals? Nothing - or at least not much.

99.999% of investors would be better off picking a simple passive portfolio and focusing their attention, not on models, but on saving rates and burn rates. I only point this out because when someone reads this thread, there is a tendency to think that somehow this is important. It isn't. As Lady Geek says, it can be safely ignored. Bill

Thanks, Bill. (bschultheis is Bill Schultheis )

I just realized there is something worthwhile to new investors. Take a look under the Classifying funds into style buckets section of the wiki article. Look familiar? The Fama-French Three-factor model is the basis for the Morningstar Style Boxes.
To some, the glass is half full. To others, the glass is half empty. To an engineer, it's twice the size it needs to be.

richard
Posts: 7961
Joined: Tue Feb 20, 2007 3:38 pm
Contact:

Re: Four Factor

Post by richard » Sat Oct 06, 2012 4:08 pm

wbern wrote:Most academic studies these days add in the momentum factor, and so do a four-factor analysis; the momentum factor is downloadable from Ken French's site.

Adding it really does change things; for example, the tilted DFA portfolios have a percent or two negative alpha for year with 3-factor, almost none with four-factor. (Momentum having significantly negative returns in Mar-May of '09 during the "snap back.")

It's the nature of regressions that adding factors can improve results. A question is always whether we're finding genuine economic factors (or at least persistent behavioral factors) or just showing our facility with statistics.

wbern wrote:How to use this is another question; when it comes to passive portfolios, I've learned to pay more attention to fund family corporate culture and discipline than to the statistical analysis. Vanguard and DFA both score high in those areas; the other families, in my opinion, still have to prove themselves.

How to use the models is the big question. The models are descriptive, not prescriptive.

User avatar
tetractys
Posts: 4526
Joined: Sat Mar 17, 2007 3:30 pm
Location: Along the Salish Sea

Re: Challenge to the Fama French Three Factor Model

Post by tetractys » Sat Oct 06, 2012 4:12 pm

Can anyone summarize how PmU and MOM correlate? Thanks up front. -- Tet

richard
Posts: 7961
Joined: Tue Feb 20, 2007 3:38 pm
Contact:

Re: Challenge to the Fama French Three Factor Model

Post by richard » Sat Oct 06, 2012 4:31 pm

magellan wrote:As an aside, I've always been uncomfortable with the three factor model's reliance on book-to-market, which is something of an accounting fiction. In the grand scheme of things, my quibble is probably meaningless, but it'd be nice to see book-to-market out of there.

Book is an accounting fiction. There's no necessary relation between book value and economic reality.

The claim is that book-to-market proxies for some underlying risk. A lot of work has gone into finding that risk. At the moment the best answer is some hand-waving about poor financial condition or cost of capital or some such.

I'd certainly prefer a model with risk factors that are real risks.

larryswedroe
Posts: 15597
Joined: Thu Feb 22, 2007 8:28 am
Location: St Louis MO

Re: Challenge to the Fama French Three Factor Model

Post by larryswedroe » Sat Oct 06, 2012 4:33 pm

It's interesting that the literature seems to be moving to some type of "quality" factor to help provide more explanatory power. ROE certainly could be part of that.
The series of papers on this seem likely to change the way we think about investing, advancing the science just as FF did when their paper was released. I know AQR is working on a paper with the quality factor in it.
Thanks for posting Rick

BTW-DFA I understand has new paper/research on quality/profitability, though I have not yet seen it.
Larry
Last edited by larryswedroe on Sat Oct 06, 2012 5:14 pm, edited 1 time in total.

User avatar
Rick Ferri
Posts: 8201
Joined: Mon Feb 26, 2007 11:40 am
Location: Vagabond, USA, Twitter: @Rick_Ferri
Contact:

Re: Challenge to the Fama French Three Factor Model

Post by Rick Ferri » Sat Oct 06, 2012 4:41 pm

LadyGeek wrote:I just realized there is something worthwhile to new investors. Take a look under the Classifying funds into style buckets section of the wiki article. Look familiar? The Fama-French Three-factor model is the basis for the Morningstar Style Boxes.


That is correct. There are many links between Morningstar and the University of Chicago Booth School of Business:

In the early 1970s, Roger G. Ibbotson, Ph.D. University of Chicago, along with Rex Sinquefield, MBA University of Chicago and co-founder of Dimensional Fund Advisors (DFA) assembled the annual returns for several asset classes dating back to 1926. This price history for Stocks, Bonds, Bills, and Inflation® became known as SBBI®. It lead Roger Ibbotson to found Ibbotson and Associates in 1977. His firm was eventually aquired by Morningstar in 2006. Ibbotson is a professor at Yale University and a Director at DFA mutual funds.

Paul Kaplan from Morningstar is widely considered the brains behind M* style boxes. Currently, Kaplan is currently the Director of Research, Morningstar Canada. Prior to assuming this role, he was vice president of quantitative research. Before joining Morningstar in 1999, Kaplan worked at Ibbotson and Associates.

According to Kaplan's bio on Morningstar, he "led the development of the quantitative methodologies behind the Morningstar Rating for funds, the Morningstar Style Box, and the Morningstar family of indexes....is a member of the Chicago Quantitative Alliance and the review board of the Research Foundation of the CFA Institute. Kaplan holds the Chartered Financial Analyst (CFA) designation."

Rick Ferri

PS. Bill, many people on this forum are interested in furthering their education and understanding, even "sophisticated" investors.
The views expressed by Rick Ferri are strictly his own as a private investor and author and do not reflect the views of any entity or other persons.

grok87
Posts: 7480
Joined: Tue Feb 27, 2007 9:00 pm

Re: Challenge to the Fama French Three Factor Model

Post by grok87 » Sat Oct 06, 2012 7:39 pm

Rick,
Thanks for posting.

All,
I'm trying to understand the author's explanation of the intuitive basis for the "investment" and ROE factors. I've read this paragraph from the paper a couple of times but still don't get it.
As noted, we motivate the q-factor model from investment-based asset pricing. Intuitively, in- vestment predicts returns because given expected cash flows, high costs of capital mean low net present values of new capital and low investment, whereas low costs of capital mean high net present values of new capital and high investment. ROE predicts returns because high expected ROE relative to low investment means high discount rates. The high discount rates are necessary to counteract the high expected ROE to induce low net present values of net capital and subse- quently low investment. If the discount rates are not high enough to offset the high expected ROE, firms would instead observe high net present values of new capital and invest more. Similarly, low expected ROE relative to high investment (such as small-growth firms in the late 1990s) means low discount rates. If the discount rates are not low enough to counteract the low expected ROE, the firms would instead observe low net present values of new capital and invest less. Finally, we include the size factor primarily to reduce the average magnitude of the alphas across size-related portfolios. As such, the size factor plays only a secondary role in the q-factor model, whereas the investment and the ROE factors are more prominent.

Does anyone "get" this? If so can you take a stab at explaining it in plainer english or by way of an example?
thanks
cheers,
"...people always live for ever when there is any annuity to be paid them"- Jane Austen

User avatar
Rick Ferri
Posts: 8201
Joined: Mon Feb 26, 2007 11:40 am
Location: Vagabond, USA, Twitter: @Rick_Ferri
Contact:

Re: Challenge to the Fama French Three Factor Model

Post by Rick Ferri » Sat Oct 06, 2012 7:56 pm

If the expected returns on invested capital going forward is high, then risks are low and new capital is expensive. If expected returns on invested capital are low, then risk is high and new capital is cheap. That authors are saying that investing in cheap capital companies yield a better return than investing in high ROE companies because what people expect is not what always happens.

Rick Ferri
The views expressed by Rick Ferri are strictly his own as a private investor and author and do not reflect the views of any entity or other persons.

natureexplorer
Posts: 4191
Joined: Thu Sep 03, 2009 10:52 am
Location: Houston

Re: Challenge to the Fama French Three Factor Model

Post by natureexplorer » Sat Oct 06, 2012 8:20 pm

Would that be a reason to total-market invest rather than to style-invest?

User avatar
Rick Ferri
Posts: 8201
Joined: Mon Feb 26, 2007 11:40 am
Location: Vagabond, USA, Twitter: @Rick_Ferri
Contact:

Re: Challenge to the Fama French Three Factor Model

Post by Rick Ferri » Sat Oct 06, 2012 8:25 pm

Total market investing is a perfectly fine strategy. This stuff is less important than the flavor of the icing on the cake.

Rick Ferri
The views expressed by Rick Ferri are strictly his own as a private investor and author and do not reflect the views of any entity or other persons.

User avatar
Robert T
Posts: 2432
Joined: Tue Feb 27, 2007 9:40 pm
Location: 1, 0.2, 0.4, 0.5
Contact:

Re: Challenge to the Fama French Three Factor Model

Post by Robert T » Sat Oct 06, 2012 9:36 pm

.
When I saw the article it reminded me of these earlier articles which take a similar investment/return on asset/return on equity approach:

Interestingly one of the articles finds the Fama-French model performs better in Non-US developed and emerging markets, while the other favors the alternative model for 10 countries in the European Monetary Union.

Joel Greenblaat uses a ROE/ROA approach to the construction of his mutual funds. Comparing the back test results which only to go back to May 2009, the annualized returns since then to June 2012:

    Greenblatt US Strategy Portfolio – 18.97%
    DFA US Large Value – 18.21%
So not a huge difference, and the Greenblatt sort results in a lower average market cap portfolio. But stock turnover in his funds have been huge. For example the turnover for the Formula Investing US Value 1000 Fund for the year through to end April 2012 was 183%! FWIW, out of interest I have been tracking a portfolio of Greenblatt’s funds – will be interesting to see how they do.

On a ‘quality factor’ – Grantham seems to focus a lot on this, and also had a quibble about the FF3F model related to this – Grantham v. Fama-French on risk and reward.

In assessing alternatives, both the accuracy of the model and implementation matter. I think the FF model does fairly well on both.

Robert
.

grok87
Posts: 7480
Joined: Tue Feb 27, 2007 9:00 pm

Re: Challenge to the Fama French Three Factor Model

Post by grok87 » Sat Oct 06, 2012 9:51 pm

Rick Ferri wrote:If the expected returns on invested capital going forward is high, then risks are low and new capital is expensive. If expected returns on invested capital are low, then risk is high and new capital is cheap. That authors are saying that investing in cheap capital companies yield a better return than investing in high ROE companies because what people expect is not what always happens.

Rick Ferri

Rick,
Thanks for responding.
I'm still struggling to parse the "investment" factor. I actually think they are saying the reverse. That investing in "expensive capital" companies yields a better return. I think they are saying that we can't really observe a companies cost of capital directly. But you can observe the amount of "investment" that is going on at a company by looking at the change in assets etc. If there is little investment going on then that is a pointer towards the company perhaps having a high cost of capital. And that in turn should mean that you could expect a better total return by investing in that company's stock.

THe other reason I think this is by looking at their "factors". The small cap factor is "mall minus big" so small companies outperform. The "investment" factor is "low invesment companies - high investment companies" So low investment companies aka expensive capital companies outperform.

Do I have this right?
If so the next stop for me is to try to figure out the ROE factor.
cheers,
"...people always live for ever when there is any annuity to be paid them"- Jane Austen

bschultheis
Posts: 200
Joined: Mon Jul 23, 2007 10:15 am

Re: Challenge to the Fama French Three Factor Model

Post by bschultheis » Sat Oct 06, 2012 11:08 pm

Rick Ferri wrote:

PS. Bill, many people on this forum are interested in furthering their education and understanding, even "sophisticated" investors.



I agree with you Rick, this stuff can be interesting, and as I mentioned in my post, this is a good place to discuss it. I also appreciate your comment that it is less important than the flavor of icing on the cake.

Put another way, it is largely irrelevant to one's investing success. On this particular thread I feel it is important to point this out.

There are a lot of people who read certain threads on this board, and can't decipher what is important and what isn't.

I have enormous respect for many of the research papers generated by folks in the ivory towers of academia, but sometimes I think this "Science of Investing" thing gets a little out of hand when discussed here.

Interesting? Yes.
Important? No. ;-)

User avatar
ClosetIndexer
Posts: 277
Joined: Mon Mar 19, 2012 11:00 pm
Contact:

Re: Challenge to the Fama French Three Factor Model

Post by ClosetIndexer » Sun Oct 07, 2012 2:35 am

Thanks for posting Rick. For us DIYers it's handy to have these sorts of updates on where things are at in the academic community. As Larry and Jerry both point out, this sounds like more of an evolution than a challenge, and it's great to see the potential of even better models being developed.

Edit: This also reminds me of Profitability, investment and average returns, where Fama and French show that returns vary based on profitability and cash-flows, as well as BTM. (So controlling for two of the three, returns vary on the third.) If you control for BTM and profitability, cash-flows are inversely related to investment, so this appears to deal with the same additional factors as the q-factor model.

YDNAL
Posts: 13774
Joined: Tue Apr 10, 2007 4:04 pm
Location: Biscayne Bay

Re: Challenge to the Fama French Three Factor Model

Post by YDNAL » Sun Oct 07, 2012 6:45 am

bschultheis wrote:I have enormous respect for many of the research papers generated by folks in the ivory towers of academia, but sometimes I think this "Science of Investing" thing gets a little out of hand when discussed here.

Interesting? Yes.
Important? No. ;-)

Bravo!

IMO, if academia doesn't write something "interesting" going forward, how much can they re-write what is known about the past? Now, really "important" would be writing something related to 2012-2042 that could be somewhat* guaranteed.

* I would take less than 100% guarantee, maybe 90% :D
Landy | Be yourself, everyone else is already taken -- Oscar Wilde

User avatar
Rick Ferri
Posts: 8201
Joined: Mon Feb 26, 2007 11:40 am
Location: Vagabond, USA, Twitter: @Rick_Ferri
Contact:

Re: Challenge to the Fama French Three Factor Model

Post by Rick Ferri » Sun Oct 07, 2012 8:58 am

Maybe there should be a way to categorize posts based on who they are for; beginner, intermediate, advanced, and PhD level. This would be a PhD conversation.

Rick Ferri
The views expressed by Rick Ferri are strictly his own as a private investor and author and do not reflect the views of any entity or other persons.

bschultheis
Posts: 200
Joined: Mon Jul 23, 2007 10:15 am

Re: Challenge to the Fama French Three Factor Model

Post by bschultheis » Sun Oct 07, 2012 9:41 am

Rick Ferri wrote:Maybe there should be a way to categorize posts based on who they are for; beginner, intermediate, advanced, and PhD level. This would be a PhD conversation.

Rick Ferri


In making my original post, I am not critical of discussing this as a topic, and in fact I find this stuff interesting myself. I have probably learned more from your posts than anyone else on the board (thank you!)

In thinking about categorizing posts (I realize this a meaningless idea - would never happen) a better way would be important, semi-important, not important.

Otherwise, too many investors would mistakenly think that PhD level topics = "important."

The reason I bring this up, (and maybe we can discuss a little in Philly next week), is that the whole idea of "passive" investing is to equal a benchmark and get on with your life. We go to great lengths to put down the active, stockpicking crowd as counterproductive to long term returns.

I see this type of Wall Street mentality that we shun creeping (leaping?) into the passive arena. We might not be obsessed with active management but we are becoming obsessed with which indices to track, how to load portfolios with which factors, and now discussing whether 5 factors are better than 3 factors?

Sometimes I think I need to write a Coffeehouse Investor book for Coffeehouse Investors, just to remind them that whether you own DFA, Vanguard, iShares or American Funds and how you load your portfolio and what index you benchmark against isn't nearly as important as staying the course with what you "do" own, and focusing on your financial plan.

Jerry_lee
Posts: 915
Joined: Tue Mar 06, 2012 1:55 pm

Re: Challenge to the Fama French Three Factor Model

Post by Jerry_lee » Sun Oct 07, 2012 10:38 am

I agree this is an advanced discussion. I disagree that using the 3F model (or 5F model) to make allocation decisions for new investors is unnecessarily complex. I actually think that a one-size-fits-all allocation (TSM/TISM/TBM) is a bit insulting to investors who have different goals, preferences, etc. This is really no different than the auto industry, where some people want a mini-van, others a sedan, others a wagon, small car, etc. Are any of those choices wrong if it meets the owner's objectives?

In that sense, if we think about stock ---> bond, value ---> growth, small ---> large, long-term ---> short-term bond, and corporate ---> government as a series of decisions, it allows for more appropriate asset allocations that are more intentionally working with markets instead of ignoring prominent behavior and sources of return or risk reduction. TSM and TBM portfolios still "work" within this framework, but they represent rather LG oriented (equities) and medium term government heavy (bonds) allocations. Wouldn't we be better to lay out the 3 primary options like this:

STOCK v BOND decision
80% equity/20% fixed ---> 20% equity/80% fixed

SIZE & PRICE decision
100% TSM ---> 30% TSM/70% small value

TERM & CREDIT decision
100% ST Bond Index ---> 100% LT Investment Grade

Where one investor decides on a 56% TSM, 24% TISM, 20% LT Investment Grade; another lands on 28% TSM, 14% SV, 18% TISM, 40% TBM, and yet another determines 10% TSM, 18% SV, 12% TISM, 60% ST Bond Index as the right allocation for them based on desired risk, expected and diversification of return, cash-flow, inflation sensitivity, liquidity, etc.

I mean, when you contrast this with the amount of time spent on deciding between TIPS and TBM, or TIPS vs iBonds, or whatever, I see no more complexity--and these take place among relatively novice investors. Obviously, the 3 decisions I mention above really matter, many of these other considerations don't.
The most disciplined investor in the world.

User avatar
Simplegift
Posts: 2553
Joined: Tue Feb 08, 2011 3:45 pm
Location: Central Oregon

Re: Challenge to the Fama French Three Factor Model

Post by Simplegift » Sun Oct 07, 2012 12:22 pm

bschultheis wrote:Sometimes I think I need to write a Coffeehouse Investor book for Coffeehouse Investors, just to remind them that whether you own DFA, Vanguard, iShares or American Funds and how you load your portfolio and what index you benchmark against isn't nearly as important as staying the course with what you "do" own, and focusing on your financial plan.

Excellent point. The essential ingredient for all do-it-yourself, long-term investors IMO — whether you subscribe to the three fund portfolio, or you tilt to small value, or you like REITs or high-yield bonds, or whatever — is that you BELIEVE in the approach you've chosen. Success in passive investing comes primarily from confidence and persistence.

Would it go too far to characterize many of these esoteric investment discussions on the Forum as "confidence-building exercises" for the participants? Some folks can read one of Mr. Schultheis's Coffeehouse books and they have all they need. Others need to hear chapter and verse from Mr. Ferri or Mr. Bernstein to settle their doubts. Still others need to see historical data, run regressions and pore over academic studies to gain confidence in their plan.

Whatever it takes for each investor to find their way. Thank goodness there is a Bogleheads Forum to enable this valuable process.
Cordially, Todd

User avatar
Rick Ferri
Posts: 8201
Joined: Mon Feb 26, 2007 11:40 am
Location: Vagabond, USA, Twitter: @Rick_Ferri
Contact:

Re: Challenge to the Fama French Three Factor Model

Post by Rick Ferri » Sun Oct 07, 2012 1:03 pm

Bill,

I agree that how you load your portfolio and what index you benchmark against isn't nearly as important as staying the course with what you "do" own, and focusing on your financial plan. That philosophical topic is the core of MANY conversations on this forum. It's specifically why I didn't title this post "How to invest using factors" or anything of the sort. I made it clear early on that this conversation is about high-level portfolio strategy, not basic investment philosophy. LadyGeek hammered that point home in her first post.

You can continue to repeat your CoffeeHouse message in this conversation if you wish, but I see no reason to do so . The folks who are reading and posting in this discussion have been in the indexing church for many, many years. They get low-fee, low-turnover, broadly diversified, stay-the-course indexing. This conversation is a discussion about strategy, not philosophy. I’ll be elaborating on the difference between philosophy and strategy next week during our panel discussion. See you then!

Rick Ferri
The views expressed by Rick Ferri are strictly his own as a private investor and author and do not reflect the views of any entity or other persons.

bschultheis
Posts: 200
Joined: Mon Jul 23, 2007 10:15 am

Re: Challenge to the Fama French Three Factor Model

Post by bschultheis » Sun Oct 07, 2012 7:29 pm

Rick Ferri wrote:The folks who are reading and posting in this discussion have been in the indexing church for many, many years.

Rick Ferri


Rick, while I agreet with you that posters on this thread might have a grasp on this topic, and understand that, while interesting, isn't all that important, I can't agree with you that "readers" of this thread have the same background.

My guess is that of the over 1500 folks who have viewed this thread, there are many, many readers, lots of new investors, and yes, some who index, who mistakenly think that the discussion on this thread is important.

I am looking forward to a discussion on this in Philly, as I see an unfortunate trend unfolding in the financial services industry that includes folks in the passive arena who are spending way too much time talking about investing aspects (call it strategy if you will), as seemingly important, and they aren't. I will explain why in our discussion.


Bill

User avatar
Jerilynn
Posts: 1929
Joined: Tue Sep 06, 2011 12:49 pm
Location: USA, Earth

Re: Challenge to the Fama French Three Factor Model

Post by Jerilynn » Sun Oct 07, 2012 7:33 pm

Rick Ferri wrote:Maybe there should be a way to categorize posts based on who they are for; beginner, intermediate, advanced, and PhD level. This would be a PhD conversation.

Rick Ferri


Well, if this is PhD level, what will the 9 factor model be?
Cordially, Jeri . . . 100% all natural asset allocation. (no supernatural methods used)

User avatar
Taylor Larimore
Advisory Board
Posts: 25866
Joined: Tue Feb 27, 2007 8:09 pm
Location: Miami FL

Keep investing simple

Post by Taylor Larimore » Sun Oct 07, 2012 8:30 pm

Bogleheads:

When we visited Mr. Bogle in his office during our second Boglehead Conference, Mr. Bogle demonstrated how he used his slide-rule when designing his first index fund. In my opinion, the easy availability of past performance numbers, spread-sheets and vast computer power are often used for complex schemes that are unnecessary and often counter-productive for ordinary investors like myself.

Keep investing simple.

Best wishes.
Taylor
"Simplicity is the master key to financial success." -- Jack Bogle

User avatar
LadyGeek
Site Admin
Posts: 39966
Joined: Sat Dec 20, 2008 5:34 pm
Location: Philadelphia
Contact:

Re: Challenge to the Fama French Three Factor Model

Post by LadyGeek » Sun Oct 07, 2012 8:31 pm

I just posted a new thread: For New investors - Important Investing Concepts

(I'll be demonstrating the wiki at the BH conference.)
To some, the glass is half full. To others, the glass is half empty. To an engineer, it's twice the size it needs to be.

User avatar
Rick Ferri
Posts: 8201
Joined: Mon Feb 26, 2007 11:40 am
Location: Vagabond, USA, Twitter: @Rick_Ferri
Contact:

Re: Challenge to the Fama French Three Factor Model

Post by Rick Ferri » Sun Oct 07, 2012 8:41 pm

Jerilynn wrote:
Rick Ferri wrote:Maybe there should be a way to categorize posts based on who they are for; beginner, intermediate, advanced, and PhD level. This would be a PhD conversation.

Rick Ferri


Well, if this is PhD level, what will the 9 factor model be?


Multi-factor investing is not new. It's just new people who wish to educate themselves beyond total stock market investing. BTW, there are 15 factor models already in existance and have been for some time.

The very idea that we should not be discussing advanced topics on Bogheads slams the brakes on what this forum is all about. Index investing was an advanced topic in 1975 when John Bogle and the Vanguard Group first launched index funds, and it's a good thing that John Bogle decided to discuss it.

Rick Ferri
Last edited by Rick Ferri on Sun Oct 07, 2012 8:50 pm, edited 1 time in total.
The views expressed by Rick Ferri are strictly his own as a private investor and author and do not reflect the views of any entity or other persons.

dbr
Posts: 22935
Joined: Sun Mar 04, 2007 9:50 am

Re: Challenge to the Fama French Three Factor Model

Post by dbr » Sun Oct 07, 2012 8:49 pm

Rick Ferri wrote:The very idea that we should not be discussing advanced topics on Bogheads slams the brakes on what this forum is all about. Index investing was an advanced topic in 1975 when John Bogle and the Vanguard Group first launched index funds, and it's a good thing that John Bogle decided to discuss it.

Rick Ferri


Agreed. See my reply in the thread linked just above.

User avatar
LH
Posts: 5490
Joined: Wed Mar 14, 2007 2:54 am

Re: Challenge to the Fama French Three Factor Model

Post by LH » Mon Oct 08, 2012 9:26 am

Three factors and such is nice. I would not invest by it alone anymore than I would use portfolio insurance technique (which was also nice theory). Or LTCM methodology. I think one needs exposures to real world asset classes, not just factors chosen by FF math. A two asset class strategy, like TIPS and SCV, that has great 5 factor exposure by the math(?), is still a two asset class strategy, and is relatively undiversified real world. Now like portfolio insurance, it looks great on the paper of the day, by the math of the day. But, the market could do well, and ones investment could do poorly in ones horizon.

So FF is great stuff, but will it be used by people 50 years hence for investment application like some are applying it today? Or will it be more of a heuristic footnote? Are the chosen factors the end all and be all, or just one of several factors one can model to fit? Having been fit to a small nonstatistically significant slice of the past, does that entail in a complex chaotic system with behavioral feeback loops (lets say one of the brownian particles, when looked at by Robert Brown, notices, and says: "HEY!, someones looking at us, everyone stop moving!", and about 30 percent do, then 70 percent do, then they all go ah lets move again, rinse repeat.... what would that do to the model?) that its going to persist?

What are the odds? I posit not 100 percent, not 90 percent. I would put it at 50 percent at best people will try and use it 50 years hence much like they seem to be considering using it today.

Its like trying to invest by the Gordon equation. the Gordon equation is nice, great stuff, but where was the Gordon equation 2007 in peoples mind? Did the Gordon equation help in 1981 predict the boom?

How many people used the portfolio insurance method tens of years ago? How many use it today?

The list goes on and on.

User avatar
ClosetIndexer
Posts: 277
Joined: Mon Mar 19, 2012 11:00 pm
Contact:

Re: Challenge to the Fama French Three Factor Model

Post by ClosetIndexer » Mon Oct 08, 2012 4:06 pm

LH wrote:What are the odds? I posit not 100 percent, not 90 percent. I would put it at 50 percent at best people will try and use it 50 years hence much like they seem to be considering using it today.


I put the odds at near 100% that the FF3F model will not be commonly used in 50 years. However, I also expect the odds are very high that an evolved model, ultimately originating from the 3F model, will be. (Or just as well say, "ultimately originating from the CAPM model.") No model will ever be perfect, so it is expected that over time people will come up with better ones. (Of course, 'better' doesn't just mean adding more factors. Part of the beauty of the 3F model is the level of explanation it achieves with "parsimonious" use of factors.)

LH wrote: A two asset class strategy, like TIPS and SCV, that has great 5 factor exposure by the math(?), is still a two asset class strategy, and is relatively undiversified real world.


SCV is not really an asset class. It is simply a descriptive term for a portfolio that has a certain range of 3F loadings. In other words, the concept of SCV is based on the 3F model.

Personally I don't hold all SCV, but my reasoning isn't that I want to avoid a single "asset class". Rather it's that I want to hedge against the risk of lower small and value premiums in the future by choosing HML and SMB loadings that are expected to capture as much of the premiums and factor diversification benefits as possible without being too extreme. (Alternatively, one could just as well choose higher small and value loadings and lower beta (lower overall equities allocation) to hedge against a lower ERP in the future, but I personally hedge against what I see as the greater uncertainty.)

User avatar
Rick Ferri
Posts: 8201
Joined: Mon Feb 26, 2007 11:40 am
Location: Vagabond, USA, Twitter: @Rick_Ferri
Contact:

Re: Challenge to the Fama French Three Factor Model

Post by Rick Ferri » Mon Oct 08, 2012 5:12 pm

SCV is not really an asset class. It is simply a descriptive term for a portfolio that has a certain range of 3F loadings. In other words, the concept of SCV is based on the 3F model.


This is a good quote because so often we hear people refer to factor investing as adding more diversification. It does not. US stocks are "the" asset class of which SCV is a tiny subset. Magnifying SCV in a portfolio will magnify the risks in this sub-asset class and perhaps increase portfolio return, but magnifying SCV does not increase diversification. It is not exclusive to the rest of the market.

Rick Ferri
The views expressed by Rick Ferri are strictly his own as a private investor and author and do not reflect the views of any entity or other persons.

User avatar
ClosetIndexer
Posts: 277
Joined: Mon Mar 19, 2012 11:00 pm
Contact:

Re: Challenge to the Fama French Three Factor Model

Post by ClosetIndexer » Mon Oct 08, 2012 5:39 pm

Rick Ferri wrote:
SCV is not really an asset class. It is simply a descriptive term for a portfolio that has a certain range of 3F loadings. In other words, the concept of SCV is based on the 3F model.


This is a good quote because so often we hear people refer to factor investing as adding more diversification. It does not. US stocks are "the" asset class of which SCV is a tiny subset. Magnifying SCV in a portfolio will magnify the risks in this sub-asset class and perhaps increase portfolio return, but magnifying SCV does not increase diversification. It is not exclusive to the rest of the market.

Rick Ferri


I think the problem is that the word "diversification" has multiple meanings. Since HML and SMB are not perfectly correlated with RM-RF, there is a "diversification benefit" in adding those factors (at least historically), meaning that greater standard deviation-adjusted return can be achieved by adding them to the market portfolio (ie: by tilting). However, as you say, doing so concentrates your holdings in a subset of the market, which reduces diversification in another, more commonly understood sense.

That said, I would expect that a meaningful lack of diversification in the second sense would show up as reduced R^2s in 3F regressions (due to increased exposure to the idiosyncrasies of specific sectors and/or companies). We see that happen with highly concentrated "pure value" funds, but not so much with broad SCV funds. Of course, there's also the 'tracking error regret' issue, but IMO that's separate from diversification concerns.

richard
Posts: 7961
Joined: Tue Feb 20, 2007 3:38 pm
Contact:

Re: Challenge to the Fama French Three Factor Model

Post by richard » Mon Oct 08, 2012 5:58 pm

ClosetIndexer wrote:I think the problem is that the word "diversification" has multiple meanings. Since HML and SMB are not perfectly correlated with RM-RF, there is a "diversification benefit" in adding those factors (at least historically), meaning that greater standard deviation-adjusted return can be achieved by adding them to the market portfolio (ie: by tilting). However, as you say, doing so concentrates your holdings in a subset of the market, which reduces diversification in another, more commonly understood sense.

That said, I would expect that a meaningful lack of diversification in the second sense would show up as reduced R^2s in 3F regressions (due to increased exposure to the idiosyncrasies of specific sectors and/or companies). We see that happen with highly concentrated "pure value" funds, but not so much with broad SCV funds. Of course, there's also the 'tracking error regret' issue, but IMO that's separate from diversification concerns.

That captures some of the essences of many of the lengthy discussions here regarding tilting.

The easy one first: the claim that tilting is better because it does better in mean-variance space. I find that claim a bit silly. The idea of the 3F model is that variance or standard deviation does not properly capture risk. If so, doing better in a model which only looks at standard deviation seems to miss the point.

The other claim is that tilting results in a more diversified portfolio. The 3F model says you can explain a portfolio by its exposure to the 3 factors. It does not include any bonus for diversification or lack of correlation between factors. Nonetheless, it's very frequently cited by fans of tilting.

I'm clearly not in that camp. I'm with Fama, who says TSM is always on the efficient frontier, French, who says cutting out mid caps does not increase diversification (tilting has to come at the expense of something) and your commonly understood sense of the word diversification.

There's also the behavioral argument for tilting, which is essentially that the market overvalues SG and undervalues SV. Could be true. Proponents often argue that the market is efficient with respect to stock selection and timing, but not efficient with respect to asset classes.

Jerry_lee
Posts: 915
Joined: Tue Mar 06, 2012 1:55 pm

Re: Challenge to the Fama French Three Factor Model

Post by Jerry_lee » Tue Oct 09, 2012 11:46 am

richard wrote:
ClosetIndexer wrote:I think the problem is that the word "diversification" has multiple meanings. Since HML and SMB are not perfectly correlated with RM-RF, there is a "diversification benefit" in adding those factors (at least historically), meaning that greater standard deviation-adjusted return can be achieved by adding them to the market portfolio (ie: by tilting). However, as you say, doing so concentrates your holdings in a subset of the market, which reduces diversification in another, more commonly understood sense.

That said, I would expect that a meaningful lack of diversification in the second sense would show up as reduced R^2s in 3F regressions (due to increased exposure to the idiosyncrasies of specific sectors and/or companies). We see that happen with highly concentrated "pure value" funds, but not so much with broad SCV funds. Of course, there's also the 'tracking error regret' issue, but IMO that's separate from diversification concerns.

That captures some of the essences of many of the lengthy discussions here regarding tilting.

The easy one first: the claim that tilting is better because it does better in mean-variance space. I find that claim a bit silly. The idea of the 3F model is that variance or standard deviation does not properly capture risk. If so, doing better in a model which only looks at standard deviation seems to miss the point.

The other claim is that tilting results in a more diversified portfolio. The 3F model says you can explain a portfolio by its exposure to the 3 factors. It does not include any bonus for diversification or lack of correlation between factors. Nonetheless, it's very frequently cited by fans of tilting.

I'm clearly not in that camp. I'm with Fama, who says TSM is always on the efficient frontier, French, who says cutting out mid caps does not increase diversification (tilting has to come at the expense of something) and your commonly understood sense of the word diversification.

There's also the behavioral argument for tilting, which is essentially that the market overvalues SG and undervalues SV. Could be true. Proponents often argue that the market is efficient with respect to stock selection and timing, but not efficient with respect to asset classes.


Richard,

1. You are right that SD is an incomplete measure of risk in a multifactor world. But investors still care about SD, and some care almost exclusively about it (for them, tilted portfolios are ideal). Further, there are plenty of discussions about additional small/value risks (timing of losses matter as much as magnitude), and most of us have lived through those in '07-'08. Multi-factor models simply give us a richer appreciation of all the ways markets reward risk, allowing us to pick and choose amongst them as our circumstances dictate. Larry might be focused on minimizing one-in-a-generation downturns with low beta/high tilt allocations, I might prefer a balance of large/small/growth/value and shorter term debt without the low beta emphasis, while you might which to hold mostly economically healthy companies (TSM) and take additional interest rate and credit risk. Each of these decisions has pros/cons, and we should seek to understand them and choose the avenue that is best for us.

2. When I think of diversification, I look to its multiple meanings. Security diversification is once consideration, asset class diversification is another (and you may restrict your definition of "asset classes" as distinct portfolios such as stocks/bonds/TIPS/cash with maybe regional considerations), and diversification by sources of risk/return is yet another. You are right, nothing in the FF research tells us HOW to invest according to the 3F or 5F model. No such paper such as Portfolio Selection exists in the multi-factor world. So it is up to investors to read the research, and apply it according to their circumstances and interpretations.

I would be very careful to take interview or video snipits out of context or depend on them exclusively. In the case of French, I believe he was comparing a TSM portfolio (almost no small/micro cap exposure) to S&P/Micro Cap portfolio (almost no mid cap exposure) held in the same factor adjusted allocations (TSM = 85% S&P/15% Micro). In this case, focusing on security diversification, TSM is more diversified (3,000 vs 2,500 names), and you might think excluding mid-caps is more problematic than small/micro given the relatively larger % of the former on a cap weighted market basis.

But, it is no great stretch to understand that there are multiple risks that drive returns, and these risks are not perfectly correlated. Actually, over intermediate/long periods, pure long/short return factors have negative correlation's and positive expected returns, so holding asset classes that target those risk/return dimensions fully does provide a multi-factor diversification benefit. We cannot deny that size risk or value risk has historically provided a return diversification during periods when the equity risk premium is negative, nor that when the ERP is above average, we have also seen zero or negative value and size returns. We don't need academic studies to understand that there is a portfolio benefit to holding multiple distinct risks (with positive expected returns) vs isolating one or two.

Not everyone, however, will chose to hold all of them. A small business owner may choose a larger cap equity portfolio, an employee of an economically sensitive company may choose not to tilt to value. A retiree may feel, with ample liquidity, that the business cycle risks of small and value aren't as big a concern to them anymore, so holding them in a reduced equity allocation (from working years) may benefit them. This is no different than the fact that investors without a real income stream in retirement shouldn't buy long-term bonds, while people in the accumulation phase don't have as much inflation risk and maybe better positioned to take TERM risk in bonds. Everyone might believe that, given their future portfolio consumption will take place mostly in the US, that a global equity allocation (40% US, 60% world) should be inverted to preserve diversification but better match assets with liabilities.

It seems we need to be a bit more intentional when we talk about the diversification benefits of small/value portfolios as "diversification across known sources of risk/return" to avoid confusion with the commonly held definition of diversification based on # of securities or broad/distinct asset class diversification (stock/bond/TIP/cash). Personally, I believe small and value stocks ARE asset classes, as an asset class is simply a diversified set of securities with similar underlying risks and returns that cannot be achieved by combining other investments together. In this case, mid caps aren't an asset class (small and large proxy for MID), but small value or small growth is. And TSM isn't so much "the market" as it is a LG oriented stock portfolio given its close association with a LG index or the S&P 500.
The most disciplined investor in the world.

Jerry_lee
Posts: 915
Joined: Tue Mar 06, 2012 1:55 pm

Re: Challenge to the Fama French Three Factor Model

Post by Jerry_lee » Tue Oct 09, 2012 11:56 am

One final thought: I think we get sloppy when we casually dismiss tilted portfolios as "riskier" just as higher stock % portfolios are riskier than lower stock % portfolios. Size and value risks are not additive in the sense that a 100% TSM portfolio has a certain amount of risk, while a levered TSM portfolio is an even riskier portfolio. We cannot manufacture the distribution of returns on (say) a 60/40 tilted balanced portfolio simply by increasing our TSM allocation to 80% in a TSM/bond mix. Since 2000, or from mid 60s to early 80s, for example, no amount of TSM allocation (up to 100%) would have been enough to earn the returns on a balanced 30% Vanguard TSM, 30% Vanguard SV, 40% bond allocation. From 95-99, even 100% small value wasn't enough to match the returns of a 60% TSM, 40% TBM portfolio.

Further, we must be careful to understand the nature of multifactor risks and their diversification benefits. In short-term panics and sell-offs, all risks tend to show up, and your losses maybe exacerbated by the amount of size, value, term, or credit risk you take relative to a traditional "Tobin" market/t-bill allocation. But over intermediate/longer periods, panics give way to risk/return principles, and it is common to see a good period for beta offset but no additional value premium or even a negative size premium or vice-versa.

In all cases, we are discussing an investor with a long time horizon with a desired return necessary to reach real future return objectives while minimizing the risks that are most acute to them. That a TSM/TBM portfolio can be right for every investor (even if it is the average of all investors) simply cannot be true.
The most disciplined investor in the world.

User avatar
ClosetIndexer
Posts: 277
Joined: Mon Mar 19, 2012 11:00 pm
Contact:

Re: Challenge to the Fama French Three Factor Model

Post by ClosetIndexer » Tue Oct 09, 2012 6:10 pm

Great posts Jl. I don't necessarily agree that we can think of TSM as large-growth oriented, but otherwise your points are well taken. I certainly agree that it would be a good thing to be more cognizant of exactly what we're talking about when we say "risk" or "diversification". There are many facets to each, and different ones will matter more to different investors, given their overall financial situations.

Post Reply