WSJ: Why Most Value Investors Will Burn Out

Discuss all general (i.e. non-personal) investing questions and issues, investing news, and theory.
Post Reply
EDN
Posts: 528
Joined: Sat Dec 29, 2012 7:55 pm
Location: Oklahoma City, OK

WSJ: Why Most Value Investors Will Burn Out

Post by EDN » Sat Feb 16, 2013 9:07 am

I enjoyed Zweig's article on value investing in today's WSJ: http://online.wsj.com/article/SB1000142 ... 61978.html

The article notes that active managers aren't particularly good at earning the value premium, underperforming a large value index fund (cheapest 50% of stocks) by about 0.7% per year for the decade thru 12/31 and almost -1.8% relative to a deeper value index fund (cheapest 20% of stocks). So much for the "smart money".

Active investors cost themselves another 1.2% by buying and selling at the wrong time, underperforming (dollar weighted) the non-value oriented S&P 500 index by a wide margin.

Have to be registered to read.

Eric

User avatar
bottomfisher
Posts: 399
Joined: Fri Jan 04, 2013 9:03 am

Re: WSJ: Why Most Value Investors Will Burn Out

Post by bottomfisher » Sat Feb 16, 2013 10:17 am

Have to be registered to read.
Try this; it worked for me. Highlight and copy name of article here: Value Stocks Are Hot—But Most Investors Will Burn Out .
Go to Google and paste article name. Search. First entry on my computer brings you to article. Hope it helps

sdrone
Posts: 108
Joined: Wed Jun 10, 2009 11:17 pm

Re: WSJ: Why Most Value Investors Will Burn Out

Post by sdrone » Sat Feb 16, 2013 12:38 pm

Unfortunately, it's true for me. I drift towards value/smallcap value investing in rising markets (via rebalancing, adding funds), then drift right back to a 3 fund portfolio in falling markets.

Though hmm, as I've done portfolio research I think I've usually used Vanguard value funds/ETFs rather than the iShares Russell 1000 value ETF. Might be something to play with later today for fun. You'd think I'd have more fun hobbies than experimenting with portfolios in spreadsheets.

stlutz
Posts: 4315
Joined: Fri Jan 02, 2009 1:08 am

Re: WSJ: Why Most Value Investors Will Burn Out

Post by stlutz » Sat Feb 16, 2013 1:45 pm

The insights of this article aren't really limited to value stocks. You could substitute any other type of investment in the article (REITs, international stocks, junk bonds etc.) and derive the same lessons.

User avatar
Blue
Posts: 1147
Joined: Sat Jul 12, 2008 10:18 pm

Re: WSJ: Why Most Value Investors Will Burn Out

Post by Blue » Sat Feb 16, 2013 2:01 pm

stlutz wrote:The insights of this article aren't really limited to value stocks. You could substitute any other type of investment in the article (REITs, international stocks, junk bonds etc.) and derive the same lessons.
I was surprised Zweig did not point this out. An abundantly important point that was absent from this article.

cheapskate
Posts: 736
Joined: Thu Apr 26, 2007 1:05 pm

Re: WSJ: Why Most Value Investors Will Burn Out

Post by cheapskate » Sat Feb 16, 2013 2:27 pm

I read the article, which missed out on another important point on why active value investing is doomed to lag Value Indexing.

1) To be adequately diversified, such an active fund must hold at least 100 different issues, at which point it becomes a closet value index fund, and is doomed to lag the index by costs.
2) Active Value Managers who want to not be a closet index fund will be under diversified, at which point the inevitable blow ups of a few of the issues in the portfolio result in larger lags.

Losing proposition all around.

User avatar
nisiprius
Advisory Board
Posts: 35441
Joined: Thu Jul 26, 2007 9:33 am
Location: The terrestrial, globular, planetary hunk of matter, flattened at the poles, is my abode.--O. Henry

Re: WSJ: Why Most Value Investors Will Burn Out

Post by nisiprius » Sat Feb 16, 2013 2:30 pm

"Value" is a euphemism, like "high-yield" in bonds. I learn from Justin Fox's excellent The Myth of the Rational Market that Benjamin Graham called his favorite investments "cigar butts." Well, they are good "value"--they have a meaningful amount of tobacco in them and you can pick them up for free. The reason why they are a good value is that other investors don't want to touch them. Perhaps it would clarify our thinking if we made a habit of doing mental translation: Vanguard Small-Cap Cigar Butt Index Fund, DFA US Targeted Cigar Butt Portfolio, Fama-French Cigar Butt Factor etc. If you cannot say, aloud and unafraid, "I have a cigar butt tilt," perhaps you are in danger of being among the cigar butt investors who burn out.
Annual income twenty pounds, annual expenditure nineteen nineteen and six, result happiness; Annual income twenty pounds, annual expenditure twenty pounds ought and six, result misery.

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

Re: WSJ: Why Most Value Investors Will Burn Out

Post by hafius500 » Sat Feb 16, 2013 2:45 pm

Today's typical value fund might not be able to deliver the superior returns generated by bargain stocks in earlier decades. Nowadays, most value managers search for stocks that are cheap relative to current or expected earnings. The bargain stocks in the best-known Fama-French index were selected on a different measure: book value, a basic yardstick of corporate net worth. Many stock pickers don't pay much attention to book value anymore
I recall Rick Ferri (and others) argued that the book-to-market ratio is not the only reliable value marker:
Value Stocks are in the Eye of the Beholder, February 11, 2013 By Rick Ferri
It’s just a reminder that every stock strategy works sometimes, but none of them work all the time.No one can say which value factor will win the market’s heart this year or over the next ten years. If you love a particular value strategy, then stick with it for the long-term and you’ll be happy that you did. If you’re prone to spread your love around, then a multi-factor strategy may be right for you.
Evanson Asset Management, Updated December 2012, Investing in High Dividend Stocks
DFA's view is that book-to-market is a better strategy for capturing total return in value stocks than dividends but both are value stock measures. However, when small cap high dividend payers are taken out and the small cap premium is removed, DFA's large cap book to market model underperformed high dividend stocks for 1927-2010, ...but US large cap data back to 1927 from Ken French suggest that large cap high dividend stocks are likely to return about the same as high book-to-market stocks and are another avenue by which the value premium can be captured. They indicate ...that differences between the two approaches to capturing value are not large, and that they have about the same volatility and drawdowns.
A Value and Size Tilter's Guide, Style indexes under the microscope, By Samuel Lee | 09-21-11
On the other hand, dividend strategies are a value tilter's dream. They soft-pedal market exposure, have no size exposure, and have value tilts competitive with small-value funds.
linked article wrote: And while small-company returns accounted for much of the historical high returns on the Fama-French index, most managers today favor larger companies.
The small and small value stocks that captured the FF size and (small) value premiums had never been investable (at least, before DFA or other small index/quant funds had been launched)
Value stocks also are less of a bargain than they used to be...So value might fade again before long.
The argument of FF disciples is that value stocks always have a higher expected return because they capture a 'risk' premium. IMO this is not plausible because risk (insurance) premiums can approximate zero so that only a few investors (who have extremely low costs) can earn the insurance premium.
prior username: hafis50

User avatar
baw703916
Posts: 6667
Joined: Sun Apr 01, 2007 1:10 pm
Location: Seattle

Re: WSJ: Why Most Value Investors Will Burn Out

Post by baw703916 » Sat Feb 16, 2013 2:55 pm

nisiprius wrote:"Value" is a euphemism, like "high-yield" in bonds. I learn from Justin Fox's excellent The Myth of the Rational Market that Benjamin Graham called his favorite investments "cigar butts." Well, they are good "value"--they have a meaningful amount of tobacco in them and you can pick them up for free. The reason why they are a good value is that other investors don't want to touch them. Perhaps it would clarify our thinking if we made a habit of doing mental translation: Vanguard Small-Cap Cigar Butt Index Fund, DFA US Targeted Cigar Butt Portfolio, Fama-French Cigar Butt Factor etc. If you cannot say, aloud and unafraid, "I have a cigar butt tilt," perhaps you are in danger of being among the cigar butt investors who burn out.
We can do various translations along these lines:

Deep value: "almost bankrupt"
Core value: "underperforming"
Blend: "run of the mill"
Growth: "overpriced"
Aggressive growth: "bubblishious"
Most of my posts assume no behavioral errors.

EDN
Posts: 528
Joined: Sat Dec 29, 2012 7:55 pm
Location: Oklahoma City, OK

The Value Premium

Post by EDN » Sat Feb 16, 2013 5:38 pm

Hafius,

I disagree with your comments and quoted articles above.

Value Sorts
It is true that book to market isn't the only variable that helps identify higher-returning value stocks, but when it comes to real world implementation, we've yet to find an approach (another variable or a combination of variables) that works better. Why? Price to Book is a very stable measure, so it results in lower real world portfolio turnover than typical cash-flow variables like P/E or P/CF or Sales--therefore higher expected returns after costs. A multiple factor sort is actually the worst in this regard, as only about 30% of stocks that have a combination of low prices to multiple factors stays that way. And most stocks have book values, so value portfolios contain broad diversification. This low turnover and broad diversification make P/B ideal for real world portfolios.

Dividends
Price to Dividend is as stable as Price to Book, it doesn't do a very good job of isolating the source of expected returns. Looking at the period from 1964-2007 from a Jim Davis paper on the subject, high minus low price to book produced a difference in returns of about 4.7% per year with a t-stat of 2.2. High minus low dividend to price only produced differences of 2.2% (less than half) with a t-stat of only 1.0 -- so we cannot even say with significance there is a difference in returns between high and low dividend paying stocks. What about just raw "value" stock portfolios sorted by dividend (highest yielding 30%) and book value (lowest 30% of price)? Over the same period, we see +13.2% vs +15.4% respectively, over 2% in favor of price to book. With that much return difference, you have plenty of room to add a small amount of high-quality bonds to dampen risk more reliably than an all-stock high dividend portfolio does.

So why does this differ from the Evanson article? Well, if you torture the data long enough, it will say anything. Notice he used the FF index for dividends, but the DFA index for price to book value -- one is updated more regularly and includes a drag associated with negative momentum (that the live fund screens for and therefore avoids). If you compared FF BtM value with FF high dividend, you'd see the return spread I mention.

Further, Evanson fails to mention two other drawbacks of a high dividend portfolio:
(a) most small stocks don't pay dividends, so you can't capture the size premium with a dividend sorted approach, which obviously detracts from returns. He says "controlling for size and removing that factor"...but why would you? Of course, if you want to prove dividends have merit, you would play this game, but objectively it makes no sense.
(b) a high dividend portfolio is much more concentrated than a high book to market portfolio. In the Davis study I mention, the average dividend portfolio size was 500 companies, the average book to market value size was over 1200 stocks.

Value Returns
When we read that the value premium has been larger in small than big companies, of course some of that data is during a period where available indexes (DFA or Vanguard) didn't exist. But the out of sample data (non-US stocks) confirmed this, as has live data since 1991. So the concerns over these unrealistic simulations are unrealistic. As a matter of fact, here are the 1928-2012 simulated and last 10 year (through 2/15 in paranthesis) live fund returns for the market, large value, and small value net of fees:

S&P 500 = +9.5% (+8.4%)
Large Value = +10.5% (+10.1%)
Small Value = +13.0% (+13.3%)

So large and small value returns for the last decade are identical to the 85 year history, and only the market portfolio, of all things, has failed to match its historical return.

Risk Premiums
Much research of the 3 Factors finds their existence is linked to different risk premiums available to investors who are wiling to bear the risk and can target it consistently and efficiently (managing costs). Market prices are such that the expected returns are a function of the risk. Those who take the risk get the return, those who don't, won't. If it turns out that the risk was higher than the return payoff, prices will fall to their appropriate equilibrium level and the new risk/return relationship will be in place. This isn't some tenuous or time sensitive phenomena. The market, size, and value premiums are durable, economically and statistically (to a varying degree) significant.

Eric

User avatar
Rick Ferri
Posts: 8291
Joined: Mon Feb 26, 2007 11:40 am
Location: Austin, TX. Twitter: @Rick_Ferri
Contact:

Re: WSJ: Why Most Value Investors Will Burn Out

Post by Rick Ferri » Sat Feb 16, 2013 7:12 pm

Eric,

LOL, you're most definitely a very good "DFA" adviser who has drunk gallons of Kool-Aid. Nothing is going to change your mind that BtM (the DFA way) is not the absolute greatest value stock strategy ever found or will ever be found in the entire Universe - for ever and ever, Amen!

Rick Ferri
The Education of an Index Investor: born in darkness, became enlightened, overdid complexity, settled into simplicity.

EDN
Posts: 528
Joined: Sat Dec 29, 2012 7:55 pm
Location: Oklahoma City, OK

Just How I See It

Post by EDN » Sat Feb 16, 2013 7:20 pm

Rick Ferri wrote:Eric,

LOL, you're most definitely a very good "DFA" adviser who has drunk gallons of Kool-Aid. Nothing is going to change your mind that BtM (the DFA way) is not the absolute greatest value stock strategy ever found or will ever be found in the entire Universe - for ever and ever, Amen!

Rick Ferri
I never said BtM was the end all/be all. There is a lot of robust reasoning behind it as I tried to lay out that makes sense to me.

Snarky comments aside, I'm not a DFA anything. I'm a fee-only advisor who of course tries to find the best vehicles in each asset class for the allocations I develop and manage. I've been very, very pleased with the results, certainly relative to other allocations and fund choices I've seen. We'll see about the future.

Eric

stlutz
Posts: 4315
Joined: Fri Jan 02, 2009 1:08 am

Re: WSJ: Why Most Value Investors Will Burn Out

Post by stlutz » Sat Feb 16, 2013 7:30 pm

Actually, backtests based on metrics like enterprise value to ebitda or price-to-free cash flow have historically outperformed the various P/E, P/B etc. metrics. The former ones are the ones that quant value investors are really keying off of nowadays.

The problem with P/B and P/E is that the denominators of both are really based on accounting constructs. In many industries, most assets are "soft assets" (e.g. patents, trademarks, goodwill, etc.) and might be valued differently at company X vs. company Y. Even in industries with more "real" assets, depreciation is done based on accounting rules, not based on the market value of, say, a steel plant. From a theoretical standpoint, it's hard to argue that P/B is the "best" measure of value. More obviously, "earnings" may or may not relate to whether the company is really ending up with more money in the safe at the end of the year.

The thing about EV/EBITDA or P/FCF is that the denominators are closer to being real numbers--measuring real cash coming into the business and still don't have the biases that result from one time special items and the like.

Is that why they backtest better? I don't know. It may just be related to the fact that for most of the period being backtested, value investors were using the more standard P/B and P/E metrics and thus there was advantage to using the lesser-known ones. Now that the quants are all focusing on the cash flow statement, that may not repeat in the future.

Also, on the dividend models, again numerous backtests show that price/shareholder yield (the sum of dividends and share buybacks) far outperforms simple dividend yield screens.

Don't know that there's really much of anything practical for anyone to take away from this except that something is always going to backtest the best.

User avatar
Rick Ferri
Posts: 8291
Joined: Mon Feb 26, 2007 11:40 am
Location: Austin, TX. Twitter: @Rick_Ferri
Contact:

Re: WSJ: Why Most Value Investors Will Burn Out

Post by Rick Ferri » Sat Feb 16, 2013 7:36 pm

There are many roads to Dublin.

Value factor premiums are period sensitive as I show in my article linked above. The value factor that worked best last decade is not likely to be the winning strategy this decade. That's what makes a multi-factor value approach appealing. Although it's never first, it's also never last and never has to make excuses.

Rick Ferri

PS. The issues with turnover in some value strategies are non-issues when ETFs are used because ETFs are not buying and selling individual securities within a fund. ETFs are also much more tax efficient than open-end value funds.
The Education of an Index Investor: born in darkness, became enlightened, overdid complexity, settled into simplicity.

EDN
Posts: 528
Joined: Sat Dec 29, 2012 7:55 pm
Location: Oklahoma City, OK

Decades Don't Matter & No ETF Tax Advantages

Post by EDN » Sat Feb 16, 2013 8:43 pm

Rick Ferri wrote:There are many roads to Dublin.

Value factor premiums are period sensitive as I show in my article linked above. The value factor that worked best last decade is not likely to be the winning strategy this decade. That's what makes a multi-factor value approach appealing. Although it's never first, it's also never last and never has to make excuses.

Rick Ferri

PS. The issues with turnover in some value strategies are non-issues when ETFs are used because ETFs are not buying and selling individual securities within a fund. ETFs are also much more tax efficient than open-end value funds.
Rick,

You don't base an investment strategy on what worked best in the last decade or on raw index returns alone. You have to consider invest ability. Multifactor sorts are not best, they either aren't as value oriented (they tend to have a wider value sort to allow for the fact that al low P/E might have a high P/B and vice versa), or if they are they require much greater turnover because many value criteria change often and are very volatile (like sales and earns based "flow" variables).

As to value ETF efficiency (relative to TM value funds?), your claims just aren't true. I wish you'd stop the misleading statements. Here are some stats on "% of return lost to taxes since inception" from each provider's website:

iShares Russell 3000 Value Index = -0.45%
iShares Morningstar LV Index = -0.45%
iShares S&P 500 Value = -0.42%
DFA TM Market-wide Value Fund = -0.32%

iShares Morningstar SV Index = -0.52%
iShares Russell 2000 Value Index = -0.51%
iShares S&P 600 Value = -0.31%
DFA TM Targeted Value Fund = -0.49%

So 1 in 6 common value indexes had better tax efficiency than a TM fund. Yes, ETFs can use the arbitrage mechanism to swap out high basis shares, but TM funds can tax loss harvest and use incoming cash flows to their advantage (constantly increasing cost basis). TM funds also have the flexibility to screen out high dividend stocks if taxes on those return to ordinary income levels, which would make them more tax-efficient than ETFs (unless their indexes excluded dividend stocks).

What about simple broad market portfolios? We can compare them to a "tax aware" broad market Core mutual fund portfolios with the same guidelines as above:

iShares Russell 3000 Index = -0.32%
iShares Wilshire TSM Index = -0.31%
iShares S&P 1500 Index = -0.29%
DFA TA Core 2 Index = -0.26%

And with the Core fund, you hold your small cap and value exposure (if that is the right amount for you) inside the fund, so you don't have the additional tax costs associated with rebalancing TSM, value, and small cap funds periodically--your allocation is rebalanced for you inside the fund. Even if you do tilt more than the Core fund, because it "sits" closer to small and value strategies, a "Core + TM" allocation will stay in line more frequently and require less rebalancing as bands are breached. Or you could just use the one-fund Vector option (-0.36% to taxes) which has as strong tilt to small and value as anyone would probably want and would again eliminate the need to rebalance between funds.

As I said, there are no differences in tax efficiency between TM/TA open ended funds and ETFs (funds were more tax efficient in 8 of 9 cases), despite what you say. Both are very tax efficient and render taxation a relative non-issue, allowing one to focus primarily on fund construction and value-added management.

Eric

User avatar
VictoriaF
Posts: 17956
Joined: Tue Feb 27, 2007 7:27 am
Location: Black Swan Lake

Re: WSJ: Why Most Value Investors Will Burn Out

Post by VictoriaF » Sat Feb 16, 2013 11:07 pm

nisiprius wrote:Benjamin Graham called his favorite investments "cigar butts." Well, they are good "value"--they have a meaningful amount of tobacco in them and you can pick them up for free. The reason why they are a good value is that other investors don't want to touch them.
A note in a men's room:
Do not throw butts into urinals. It's difficult to puff at them afterwards.

Victoria
WINNER of the 2015 Boglehead Contest. | Every joke has a bit of a joke. ... The rest is the truth. (Marat F)

SP-diceman
Posts: 3968
Joined: Sun Oct 05, 2008 9:17 am

Re: WSJ: Why Most Value Investors Will Burn Out

Post by SP-diceman » Sat Feb 16, 2013 11:34 pm

VictoriaF wrote:
nisiprius wrote:Benjamin Graham called his favorite investments "cigar butts." Well, they are good "value"--they have a meaningful amount of tobacco in them and you can pick them up for free. The reason why they are a good value is that other investors don't want to touch them.
A note in a men's room:
Do not throw butts into urinals. It's difficult to puff at them afterwards.

Victoria


The Porcelain Special Situations Fund.

:)

User avatar
nedsaid
Posts: 9391
Joined: Fri Nov 23, 2012 12:33 pm

Re: WSJ: Why Most Value Investors Will Burn Out

Post by nedsaid » Sat Feb 16, 2013 11:51 pm

I don't think Value investors burn out. Most "value" investors get bored and ditch the strategy for something more exciting. Good investing is boring. It is as exciting as watching paint dry. Boring is good, boring that pays a dividend is even better.

The cigar butt investing is what others call the value trap. Some companies are "undervalued" because they have a fundamental flaw that probably can't be fixed. Eastman Kodak was a classic example. It was a dying business model and the company could not make the transition from film to digital. There were too many players in the digital imaging business. To me this isn't value investing.

To me Value investing is investing in companies that are out of favor or companies that are having temporary problems. You want to buy good stuff that is out of favor. This is what Warren Buffett does. He just doesn't overpay.

Value investing would also be characterized by long holding periods. A good value investor should have a holding period of at least five years. Doing this properly takes patience.

So how can a boring, patient investor burn out? It seems the Growth and the Aggressive Growth guys are the ones that flame out.
A fool and his money are good for business.

User avatar
baw703916
Posts: 6667
Joined: Sun Apr 01, 2007 1:10 pm
Location: Seattle

Re: WSJ: Why Most Value Investors Will Burn Out

Post by baw703916 » Sat Feb 16, 2013 11:57 pm

nedsaid,

I agree with you, and I really like your summary.

Probably what the article is referring to is that many people call themselves value investors but don't really have the needed patience.

Best wishes,
Brad
Most of my posts assume no behavioral errors.

User avatar
nedsaid
Posts: 9391
Joined: Fri Nov 23, 2012 12:33 pm

Re: WSJ: Why Most Value Investors Will Burn Out

Post by nedsaid » Sun Feb 17, 2013 12:00 am

Wow. I am usually right about something about every six months.

The bummer is that I will have to wait another six months to be right about something again!!

Thanks Brad for your kind comments. Nice to know that I added something.
A fool and his money are good for business.

Random Walker
Posts: 2479
Joined: Fri Feb 23, 2007 8:21 pm

Re: WSJ: Why Most Value Investors Will Burn Out

Post by Random Walker » Sun Feb 17, 2013 12:27 am

Worthwhile to remember that the outperformance of the value asset class is generally due to a few big winners amidst the big collection of dogs. And normal humans can't pick those few in advance, so diversification within the class is critical.

Dave

User avatar
nedsaid
Posts: 9391
Joined: Fri Nov 23, 2012 12:33 pm

Re: WSJ: Why Most Value Investors Will Burn Out

Post by nedsaid » Sun Feb 17, 2013 12:47 am

I can't remember who said this. It might have been Will Rogers, Yogi Berra, or Mark Twain.

The quote went something like this. Pick a portfolio of stocks very carefully. If a stock doesn't go up, don't buy it.
A fool and his money are good for business.

fidelio
Posts: 217
Joined: Sun May 04, 2008 5:28 pm

Re: WSJ: Why Most Value Investors Will Burn Out

Post by fidelio » Sun Feb 17, 2013 4:29 am

nisiprius wrote:"Value" is a euphemism, like "high-yield" in bonds. I learn from Justin Fox's excellent The Myth of the Rational Market that Benjamin Graham called his favorite investments "cigar butts." Well, they are good "value"--they have a meaningful amount of tobacco in them and you can pick them up for free. The reason why they are a good value is that other investors don't want to touch them. Perhaps it would clarify our thinking if we made a habit of doing mental translation: Vanguard Small-Cap Cigar Butt Index Fund, DFA US Targeted Cigar Butt Portfolio, Fama-French Cigar Butt Factor etc. If you cannot say, aloud and unafraid, "I have a cigar butt tilt," perhaps you are in danger of being among the cigar butt investors who burn out.
brilliant! when i was a kid we rolled cigs from old butts. i guess that has something to do w/ book value ....

User avatar
William Million
Posts: 544
Joined: Wed May 05, 2010 4:41 am
Location: A Deep Mountain

Re: WSJ: Why Most Value Investors Will Burn Out

Post by William Million » Sun Feb 17, 2013 5:16 am

Value investors should ask themselves at what stage of subpar returns will they walk. Some can stick it out for 5 years, a few for 10 years. At 15 years, the party gets really small.

Random Walker
Posts: 2479
Joined: Fri Feb 23, 2007 8:21 pm

Re: WSJ: Why Most Value Investors Will Burn Out

Post by Random Walker » Sun Feb 17, 2013 10:05 am

As an investor, all you can do is look forward with a knowledge of the past risk/return relationships. When prices decline all one can really do is look forward to increased future expected returns. Knowing that expected returns increase as prices decrease helps build the tenacity to stick with a plan.

Dave

Stryker
Posts: 373
Joined: Tue Nov 24, 2009 6:13 am
Location: Canada

Re: WSJ: Why Most Value Investors Will Burn Out

Post by Stryker » Sun Feb 17, 2013 10:35 am

When Walter Schloss was alive and running his investment business with his son Edwin, they had around a 100 stocks in the portfolio. In a 1989 interview with Outstanding Investor Digest when he was asked about his feverish pace of turning over his stocks at the torrid rate of once every four years, Walter replied

"That's right, but we have a lot of stocks. If we only had 4 or 5 and we only bought and sold one stock a year - while it may be great for some people, I wouldn't like it. I like a little action in what we do. We like to have a little fun."

texasdiver
Posts: 2420
Joined: Thu Jun 25, 2009 12:50 am
Location: Vancouver WA

Re: WSJ: Why Most Value Investors Will Burn Out

Post by texasdiver » Sun Feb 17, 2013 12:50 pm

Is the WSJ article really unique to value investing?

Seems to me that the same things that cause value investors to "burn out" probably apply equally well to investors who follow any other tilt.

precious metals?
emerging markets?
tech stocks?
energy stocks?
health care stocks?
internet stocks?
REITs?
micro-cap stocks?

etc. etc.

Seems to me that the same arguments about how managers are going to under-perform the indexes and how individual investors are going to botch their attempts at market timing would apply to pretty much any sector tilt, not just value stocks.

User avatar
Rick Ferri
Posts: 8291
Joined: Mon Feb 26, 2007 11:40 am
Location: Austin, TX. Twitter: @Rick_Ferri
Contact:

Re: Decades Don't Matter & No ETF Tax Advantages

Post by Rick Ferri » Sun Feb 17, 2013 1:03 pm

EDN wrote:
Rick Ferri wrote:There are many roads to Dublin.

Value factor premiums are period sensitive as I show in my article linked above. The value factor that worked best last decade is not likely to be the winning strategy this decade. That's what makes a multi-factor value approach appealing. Although it's never first, it's also never last and never has to make excuses.

Rick Ferri

PS. The issues with turnover in some value strategies are non-issues when ETFs are used because ETFs are not buying and selling individual securities within a fund. ETFs are also much more tax efficient than open-end value funds.
Rick,

You don't base an investment strategy on what worked best in the last decade or on raw index returns alone. You have to consider invest ability. Multifactor sorts are not best, they either aren't as value oriented (they tend to have a wider value sort to allow for the fact that al low P/E might have a high P/B and vice versa), or if they are they require much greater turnover because many value criteria change often and are very volatile (like sales and earns based "flow" variables).

Eric
Eric,

Again, your preaching to a choir who has been singing in the church for more than a decade. Multifactor sorts are not "best" or "worst". There will always be a single value factor that outperforms multi-factor over a specific period, but it's not possible to know which factor that will be. You say that multi-factor funds are not as "value intensive" as DFA funds, but that depends entirely on what your starting definition of value is. If you measure value using only Book-to-Market (BtM) as your yardstick (the DFA way), then of course DFA funds will have the most "value" exposure. But if you use Price-to-Cash-Flow (P/Cf) or Price-to-Earning (P/E) or some other yardstick, then those funds are not as "value" intensive because it's not the single factor that DFA is targeting. It gets back to my article, Value is in he eyes of the beholder!

Also, inside an ETF, the multi-factor method captures a smoothed average of several value factors with very low cost and little to no tax consequence. ETFs are an ideal way to invest in value stocks because turnover does not generate taxes (DFA should start offering ETFs for this reason). So, I'm confused why you're insisting that the turnover of an "index" has an effect on the tax efficiency of an ETF that follows that index. That's not true. The securities in an ETF trade in-kind, not in cash, so there are no capital gains distributions. Fama-French understand this very well per the WSJ article, but you seem to have a misunderstanding about it.

Rick Ferri
The Education of an Index Investor: born in darkness, became enlightened, overdid complexity, settled into simplicity.

EDN
Posts: 528
Joined: Sat Dec 29, 2012 7:55 pm
Location: Oklahoma City, OK

Value Sorts, Value ETFs, and DFA ETFs

Post by EDN » Sun Feb 17, 2013 2:30 pm

Rick Ferri wrote:
Eric,

Again, your preaching to a choir who has been singing in the church for more than a decade. Multifactor sorts are not "best" or "worst". There will always be a single value factor that outperforms multi-factor over a specific period, but it's not possible to know which factor that will be. You say that multi-factor funds are not as "value intensive" as DFA funds, but that depends entirely on what your starting definition of value is. If you measure value using only Book-to-Market (BtM) as your yardstick (the DFA way), then of course DFA funds will have the most "value" exposure. But if you use Price-to-Cash-Flow (P/Cf) or Price-to-Earning (P/E) or some other yardstick, then those funds are not as "value" intensive because it's not the single factor that DFA is targeting. It gets back to my article, Value is in he eyes of the beholder!

Also, inside an ETF, the multi-factor method captures a smoothed average of several value factors with very low cost and little to no tax consequence. ETFs are an ideal way to invest in value stocks because turnover does not generate taxes (DFA should start offering ETFs for this reason). So, I'm confused why you're insisting that the turnover of an "index" has an effect on the tax efficiency of an ETF that follows that index. That's not true. The securities in an ETF trade in-kind, not in cash, so there are no capital gains distributions. Fama-French understand this very well per the WSJ article, but you seem to have a misunderstanding about it.

Rick Ferri
Rick,

I assure you there is no misunderstanding on my part--be it the fact there is aren't major tax advantages to ETFs over tax managed funds (the point you tried to make), the value orientation of various investment approaches, or DFA ETFs.

Value Sorts
Who runs muti-factor sorts today? MSCI, S&P, and even Russell. What is their definition of "value"? Cheapest 50%. Using a single factor sort, it is much easier to hold a purer representation of the asset class -- say 25%. If you think a "bottom 50% multi-factor sort" has the same value orientation as a "bottom 25% single factor sort", I'm not sure what else to say.

Now, if you are speaking of hypothetical (that don't exist) multi-factor "deep" value strategies, then this discussion is purely theoretical and lately irrelevant. But even here, you either have:
(a) the option of an "all inclusive" value sort, where stocks must meet the cut for all value metrics (resulting in significant concentration as very few stocks are categorized as value based on multiple considerations), or
(b) "one of the above" where stocks are either low P/B, P/E, P/S, or P/CF -- but the last 3 criteria haven't been shown to produce more dispersion in the distinction of value/growth but require more trading and turnover as E/S/CF are more volatile and stocks migrate more rapidly and violently. So you either have to reconstitute more frequently where negative momentum becomes problematic, or reconstitute only once or twice a year and face being stuck with more stale value stocks with lower expected returns and a less pure value orientation.

ETFs
As for ETFs being the ideal way to invest in value, that is completely wrong. ETFs are fine, but not ideal. To target value stocks most efficiently, you want to:
1. hold just the companies with the deepest exposure to that dimension (ETFs don't do that) while staying diversified and not holding too-few stocks or becoming concentrated by industry
2. update the portfolio continuously to ensure the holdings constantly reflect the underlying asset class (not reconstituting once or twice per year all at once), ideally with new cash in-flows and dividend distributions
3. manage around the harmful effects of momentum and trading in a way so as to not be penalized by spreads/impact or reconstitution pressures.

And, for taxable accounts, you want to be able to:
4. harvest losses continuously to offset gains that result from value stocks that migrate.

Sure, the ETF in-kind redemption loophole helps with tax efficiency, but it doesn't eliminate trading costs as the stale reconstitution process will attest to (that is why they reconstitute so infrequently).

DFA ETFs
Why would DFA want to manage ETFs that have the drawbacks I outline above? Not for tax reasons obviously, given the tax statistics I presented above. Better returns? A quick glance at the 10YR category returns for their 5 core value funds shows the following results:

US Large Value = top 5%
US Small Value = top 11%
Int'l Large Value = top 10%
Int'l Small Value = top 10%
EM Value = top 2%

Surely you aren't saying that those results need to be dramatically improved upon? I can attest to you, those who have held those funds for the last decade aren't complaining or looking for any improvements! I could care less if DFA managed ETFs, I wouldn't use them, I don't believe in trying to fix what ain't broken.

Eric

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

Re: The Value Premium

Post by hafius500 » Sun Feb 17, 2013 2:39 pm

EDN wrote:
Dividends
...High minus low dividend to price only produced differences of 2.2% (less than half) with a t-stat of only 1.0 -- so we cannot even say with significance there is a difference in returns between high and low dividend paying stocks. ... Well, if you torture the data long enough, it will say anything. ...
(a) most small stocks don't pay dividends, so you can't capture the size premium with a dividend sorted approach, which obviously detracts from returns. He says "controlling for size and removing that factor"...but why would you?
The size and yield effects could be related.
I recall some quant (value) managers use different ratios (P/B vs. P/E etc.) for different country markets because no ratio works best in every market.
Data mining and sample bias are dangers.See this comparison of dividend strategies (US vs. UK):
Source
Naranjo et al: Stock returns, Dividend Yields, and Taxes, Journal of Finance, Dec 98:

Abstract:...we document that risk-adjusted NYSE stock returns increase in dividend yield during the period from 1963 to 1994...too large to be explained by a tax penalty...the effect is primarily driven by smaller market capitalization stocks...
p.2031: ...relation betweeen dividends and returns is driven by stocks with below-median NYSE market capitalization. Though the effect is nonlinear due to disproportionately poor returns of zero-yielded stocks, it is not merely a zero-yield effect...
p.2050: Consequently, our dividend effect is not a proxy for previously documented B/M, CF/P, E/P or sales-growth rank effects...Similarly the yield effect appears to be strong and fairly monotonic for the below-median sized NYSE firms, but there is no discernible yield effect for the largest quartile NYSE firms..."
...
A Robust Estimation of the Relation between Stock Returns, Size, Dividend Yield and Payout Ratio, I.D. McManus, O. ap Gwilym and S.H. Thomas (UK market 1958-97)

...if I didn't misunderstand them...they argue smaller dividend stocks had higher (risk-adjusted) returns than large dividend stocks.
In the literature summary (page 3) they write:
"Indeed Keim (1985) and MT (1998) identify a 'U-shaped' relation between the dividend yield rank and returns: the higher the dividend yield, the higher the returns, excepting that the zero-dividend portfolio also exhibits high returns. This is often referred to as a “non-linearity” in the relation between stock returns and dividend yields. Further, both firm size and seasonality also influence portfolio returns in the regression context, but do not dilute the strong, positive relationship between returns and dividend yields".

"Table 3 generally shows the familiar inverse relation between firm size and returns, zero-dividend stocks outperforming all other dividend quintiles excepting the highest yielding stratum; and leading to the familiar 'U-shaped' relation between dividend yield and returns (see the final columns of Table 3)." (page 7)

On risk-adjusted returns(page 8)
"Returns decline monotonically with decreasing dividend yield through to the lowest, non Zero-Dividend paying category, though not including Zero-Dividend stocks; here, positive abnormal returns are once again in evidence, at 0.34% per month (the highest of any dividend-ranked portfolio excepting for the highest-yielding)...Examination, as before, of the 'horizontal' (within dividend category) characteristic reveals a similar pattern; largely constant abnormal returns across the 3 largest size quintiles (2 largest in the case of the Zero-Dividend category), but with rapidly increasing abnormal returns with decreasing size thereafter"
...
prior username: hafis50

User avatar
Rick Ferri
Posts: 8291
Joined: Mon Feb 26, 2007 11:40 am
Location: Austin, TX. Twitter: @Rick_Ferri
Contact:

Re: WSJ: Why Most Value Investors Will Burn Out

Post by Rick Ferri » Sun Feb 17, 2013 6:44 pm

Eric,

Do you honestly think that DFA has a lock on what value stock portfolios should be and how to trade them? Based on your regurgitation of DFA marketing material, it sounds like you believe no other fund company has enough intelligence or skill to capture returns in the value space. Your wrong. No firm has a lock on intellectualism. Just look at Research Affiliates. The papers Jason Hsu and Vitali Kalesnik have put out lately about the value premium are excellent. They exceed anything DFA has written about. And I'm not being naive or biased. I own DFA products, not RA products.

Rick Ferri

BTW, the ETF marketplace is much more robust than you appear to want to know. DFA ETFs in a master portfolio structure would improve the tax-efficiency of existing share class funds and DFA knows this. However, lauching ETFs would bite the adviser's hands that feed them. Exclusivity is cherished by so-called "DFA advisers" who rely on access to DFA funds for new business. I don't know how much longer DFA can resist the reality of the marketplace and start expanding their product offerings.
The Education of an Index Investor: born in darkness, became enlightened, overdid complexity, settled into simplicity.

User avatar
Robert T
Posts: 2496
Joined: Tue Feb 27, 2007 9:40 pm
Location: 1, 0.2, 0.4, 0.5
Contact:

Re: WSJ: Why Most Value Investors Will Burn Out

Post by Robert T » Mon Feb 18, 2013 5:22 pm

.
I like the DFA TM funds. But one of the challenges of TM funds versus ETFs - is sampling risk. Consider the following 10 year returns to December 2011.

DFA Targeted Value fund = 8.20%
DFA TM Targeted Value fund = 5.94%

The 'sampling' of stocks to get better tax-efficiency in the TM fund led to a 2.26% lower annualized return over this ten year period. This is a significant cost which would have likely been much smaller if there was an ETF version of the targeted value fund. And the ETF version would likely have similar tax efficiency to the TM fund. In this respect ('sampling risk') - I see a benefit of ETFs over TM funds.

Robert
.

User avatar
Karl
Posts: 1074
Joined: Sun May 13, 2007 5:52 pm
Location: Milwaukee, WI

Re: Decades Don't Matter & No ETF Tax Advantages

Post by Karl » Mon Feb 18, 2013 6:46 pm

Rick Ferri wrote:ETFs are an ideal way to invest in value stocks because turnover does not generate taxes (DFA should start offering ETFs for this reason).
Seems Vanguard has already beat DFA (and all others) in that area by creating ETF shares that are part of existing open end funds.

User avatar
Rick Ferri
Posts: 8291
Joined: Mon Feb 26, 2007 11:40 am
Location: Austin, TX. Twitter: @Rick_Ferri
Contact:

Re: WSJ: Why Most Value Investors Will Burn Out

Post by Rick Ferri » Mon Feb 18, 2013 6:52 pm

Other companies will soon be offering ETFs as a share class of open-end funds. These master-feeder funds will not operate exactly the same as Vanguard ETFs and mutual funds, but the tax benefits to all share class holders will be the same. Fidelity is one of those companies that has filed for exemptive relief needed to launch such funds.

Rick Ferri
The Education of an Index Investor: born in darkness, became enlightened, overdid complexity, settled into simplicity.

EDN
Posts: 528
Joined: Sat Dec 29, 2012 7:55 pm
Location: Oklahoma City, OK

DFAs Value Process

Post by EDN » Mon Feb 18, 2013 8:58 pm

Rick Ferri wrote:Eric,

Do you honestly think that DFA has a lock on what value stock portfolios should be and how to trade them? Based on your regurgitation of DFA marketing material, it sounds like you believe no other fund company has enough intelligence or skill to capture returns in the value space. Your wrong. No firm has a lock on intellectualism. Just look at Research Affiliates. The papers Jason Hsu and Vitali Kalesnik have put out lately about the value premium are excellent. They exceed anything DFA has written about. And I'm not being naive or biased. I own DFA products, not RA products.

Rick Ferri

BTW, the ETF marketplace is much more robust than you appear to want to know. DFA ETFs in a master portfolio structure would improve the tax-efficiency of existing share class funds and DFA knows this. However, lauching ETFs would bite the adviser's hands that feed them. Exclusivity is cherished by so-called "DFA advisers" who rely on access to DFA funds for new business. I don't know how much longer DFA can resist the reality of the marketplace and start expanding their product offerings.
Rick,

I am amazed that for someone who seems to be well-respected around here, you can't submit a single post without taking some snarky dig at me ("drinking kool-aid", "regurgitating DFA marketing material") instead of offering something that approaches real substance.

YOU said value ETFs were "much more" tax efficient than value mutual fund ETFs, I showed you that was not the case
YOU said that any value management process produces the same results & value exposure, I explained why that wasn't the case
YOU said DFA could vastly improve on their current strategy lineup by using ETF strategies, I showed you the worst performing value fund was in the top 90% for the last decade and the best was +/- #1 in the category, clearly not much room for improvement

Do I think DFA is the only one capable of managing good value portfolios? Of course not. I like the work Vericimetry is doing, Bridgeway seems to be headed in the right direction by shifting focus from their active quant funds to more asset class vehicles. I even think iShares and Vanguard do a good job targeting the mid cap value segment of the market, however limited that maybe. I prefer the use of DFA regular and tax-managed funds based on all the research I have done (some of which I've provided), but I sure don't disparage another advisor if they don't share my opinions. Given that my portfolios are in the 80% to 100% value range (as a % of equities), its something I take great interest in. If I was only putting 10-15% in non-TSM based asset classes, I might not worry as much about the details.

Personally, I could care less whether DFA is made available to every investor or just advisor-clients, so long as DFA can minimize the hot money effects from performance chasers (loading up on DFSVX from 2003-2006, bailing out in 07-08, and jumping back in again after another good 4 year stretch through '12). The clients I have hire an advisor for my holistic approach to their wealth, my expertise in developing asset allocations, and most importantly, the confidence I instill in them to stay with their plan. As for ETFs having better tax-efficiency, we don't know that given that most iShares ETFs are not more tax efficient than existing DFA TM funds (again, see the data above). And even if the TM funds weren't as/more tax efficient than ETFs, I'm happy to leave 0.1% or 0.2% on the table if that means a better ongoing management approach that more consistently targets size and value premiums (with expected returns of 2% to 4%, where small differences in exposure far outweigh temporary tax differences).

Finally, I also don't share your admiration of the RAFI guys. I'm not an active advisor, and I'm not looking for free lunches (which is what Arnott claims his FI strategies are), I don't believe in sector/style rotation, and the fact that his message has changed several times in the last few years (first he said his strategy wasn't a value approach, then it was, etc.), and I try to avoid investing with managers who moonlight as tactical advisors when the price is right.

Eric

EDN
Posts: 528
Joined: Sat Dec 29, 2012 7:55 pm
Location: Oklahoma City, OK

DFA TM Value Funds

Post by EDN » Mon Feb 18, 2013 8:59 pm

Robert T wrote:.
I like the DFA TM funds. But one of the challenges of TM funds versus ETFs - is sampling risk. Consider the following 10 year returns to December 2011.

DFA Targeted Value fund = 8.20%
DFA TM Targeted Value fund = 5.94%

The 'sampling' of stocks to get better tax-efficiency in the TM fund led to a 2.26% lower annualized return over this ten year period. This is a significant cost which would have likely been much smaller if there was an ETF version of the targeted value fund. And the ETF version would likely have similar tax efficiency to the TM fund. In this respect ('sampling risk') - I see a benefit of ETFs over TM funds.

Robert
.
Robert,

That is true, DFAs TM funds took in a ton of assets in 2002 as advisors had an opportunity to TLH out of the taxable funds for the first time, and this was prior to DFA using any momentum filters. It couldn't have been predicted as there was no indication this would be a problem in the years leading up to 2002 (TM funds outperformed non-TM from 1999-2001), and the enhancement they made to minimize these problems in the future were made rather quickly, and I haven't seen the issue reappear meaningfully since.

For the 10 years ending 2012, the return of a US balanced portfolio that I use with their non-TM funds (30% DFA US Large, 30% DFA US Large Value, 40% DFA US Small Value) was +9.28%, using TM funds with the same size and value exposure (20% DFA TM Equity, 40% DFA TM Market-wide Value, 40% DFA TM Targeted Value) earned +9.24%--a dead heat. The non-TM DFA Int'l Value earned +10.20%, the DFA TM Int'l Value fund actually did better, earning +10.29%. And, of course, in both cases the TM mix was more tax efficient than the non-TM version, leading to higher after-tax returns.

And it doesn't appear this similar TM/non-TM fund results are limited to the value funds. Over a shorter period, since 2007, the DFA TA (tax-aware) US Core 2 fund has outperformed DFA US Core 2, the identical taxable version by over 30bps per year. A helpful reminder that "tracking error" can go either way sometimes.

But if I were to try and duplicate the size/value exposure of the US allocation above (30/30/40 or 20/40/40) using ETFs with a similar large/small design, I'd have to go all the way to 50% S&P 500 Value, 50% S&P 600 Value, or 50% Russell 1000 Value, 50% Russell 2000 Value (no growth, so greater tracking error with the ETFs). In each case over the last decade, the raw index returns (no iShares fees, no taxes) were about 1/2% lower than the TM fund mix net of the fund fees and tax distributions (over 1% if we add iShares fees and taxes back in). If I were to try and duplicate the value exposure of DFA TM Int'l Value, iShares EAFE Value is as close as I could get, and the raw index returns (no iShares fees, no taxes) were over 1% lower than DTMIX net of fund fees and taxes.

For the type of allocations I build (usually a mix of large/small and growth/value in the US and in some cases all-value, and always all-value in foreign and EM markets), DFAs TM funds work a lot better for me than ETFs. TM portfolios have tracked and even beat similar non-TM portfolios over time, and have been very tax efficient.

But I don't take DFA at their word, I review TM and non-TM fund attribution analysis frequently to ensure no surprise structural issues.

I know you've had very good results with your approach and seen very good tax efficiency with ETFs and tax-sensitive funds, better I might add, than many advisor-oriented allocations I see.

Eric

EDN
Posts: 528
Joined: Sat Dec 29, 2012 7:55 pm
Location: Oklahoma City, OK

High Dividend Doesn't Necessarily Mean High Return

Post by EDN » Mon Feb 18, 2013 9:17 pm

hafius500 wrote:
The size and yield effects could be related.
I recall some quant (value) managers use different ratios (P/B vs. P/E etc.) for different country markets because no ratio works best in every market.
Data mining and sample bias are dangers.
hafius,

Fama and French followed up their US paper with an International version in 1998 (see here: http://citeseerx.ist.psu.edu/viewdoc/do ... 1&type=pdf). In it they looked at global value portfolios based on different sorts and showed the following returns:

Market = +9.6%
Price to Book = +14.8%
Price to Earnings = +13.7%
Price to Cashflow = +13.5%
Price to Dividend = +12.7%

So while all value approaches "worked" (outperforming the market), P/D underperformed P/B by over 2% per year, similar to the US evidence I submitted in a previous post. Probably more problematic for those assuming high-yield stock approaches will reliably outperform, only 2 of 13 countries produce statistically significant higher returns from High Dividend minus Low Dividend portfolios. In 3 of 13, Low Dividend actually beat High Dividend!

Eric

User avatar
Rick Ferri
Posts: 8291
Joined: Mon Feb 26, 2007 11:40 am
Location: Austin, TX. Twitter: @Rick_Ferri
Contact:

Re: WSJ: Why Most Value Investors Will Burn Out

Post by Rick Ferri » Mon Feb 18, 2013 10:13 pm

Ken French has plenty of factor data on his website. Anyone can slice and dice this data hundreds of ways. I've done it as well as others. It's fairly clear that BtM is a good way toninvest, but is not the highest returning factor. P/E has returned better theoretical performance over the past 5 decades and in most individule decades and P/CF isn't far behind. . This isn't to say that P/E or P/CF or BtM will be the top factor this decade or next, that cannot be known in advance.BTW, Eric is correct about dividends. High dividend yield is the worst value factor by far. It works, but not well.

Rick Ferri
The Education of an Index Investor: born in darkness, became enlightened, overdid complexity, settled into simplicity.

User avatar
docneil88
Posts: 927
Joined: Mon Apr 30, 2007 3:39 pm
Location: Taxable

Re: WSJ: Why Most Value Investors Will Burn Out

Post by docneil88 » Tue Feb 19, 2013 12:01 am

Rick Ferri wrote:Other companies will soon be offering ETFs as a share class of open-end funds. These master-feeder funds will not operate exactly the same as Vanguard ETFs and mutual funds, but the tax benefits to all share class holders will be the same. Fidelity is one of those companies that has filed for exemptive relief needed to launch such funds. Rick Ferri
Hi Rick, As I remember, Vanguard has a patent on combining ETFs with open-end funds. Will Vanguard get a licensing fee when other fund companies use this concept? If so, maybe that'd reduce Vanguard's expense ratios even further given that Vanguard is a mutual co. owned by the shareholders of its funds; that'd be sweet. Best, Neil

User avatar
Rick Ferri
Posts: 8291
Joined: Mon Feb 26, 2007 11:40 am
Location: Austin, TX. Twitter: @Rick_Ferri
Contact:

Re: WSJ: Why Most Value Investors Will Burn Out

Post by Rick Ferri » Tue Feb 19, 2013 11:23 am

docneil88 wrote:Hi Rick, As I remember, Vanguard has a patent on combining ETFs with open-end funds. Will Vanguard get a licensing fee when other fund companies use this concept? If so, maybe that'd reduce Vanguard's expense ratios even further given that Vanguard is a mutual co. owned by the shareholders of its funds; that'd be sweet. Best, Neil
I am not a patent lawyer, but from what I heard, the master-feeder method is far enough removed from the Vanguard patent so that it does not infringe. Vanguard may pursue patent infringement in court, but I've not heard any rumblings.

Rick Ferri
The Education of an Index Investor: born in darkness, became enlightened, overdid complexity, settled into simplicity.

User avatar
Phineas J. Whoopee
Posts: 7046
Joined: Sun Dec 18, 2011 6:18 pm

Re: WSJ: Why Most Value Investors Will Burn Out

Post by Phineas J. Whoopee » Tue Feb 19, 2013 5:48 pm

Rick Ferri wrote:
docneil88 wrote:Hi Rick, As I remember, Vanguard has a patent on combining ETFs with open-end funds. Will Vanguard get a licensing fee when other fund companies use this concept? If so, maybe that'd reduce Vanguard's expense ratios even further given that Vanguard is a mutual co. owned by the shareholders of its funds; that'd be sweet. Best, Neil
I am not a patent lawyer, but from what I heard, the master-feeder method is far enough removed from the Vanguard patent so that it does not infringe. Vanguard may pursue patent infringement in court, but I've not heard any rumblings.

Rick Ferri
Just to add...

Vanguard does not hold a patent on the idea of combining ETFs with open-ended funds, only on a method to accomplish it.

PJW

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

Re: High Dividend Doesn't Necessarily Mean High Return

Post by hafius500 » Fri Feb 22, 2013 12:40 pm

EDN wrote:
hafius500 wrote:
The size and yield effects could be related.
I recall some quant (value) managers use different ratios (P/B vs. P/E etc.) for different country markets because no ratio works best in every market.
Data mining and sample bias are dangers.
hafius,

Fama and French followed up their US paper with an International version in 1998 (see here: http://citeseerx.ist.psu.edu/viewdoc/do ... 1&type=pdf). In it they looked at global value portfolios based on different sorts and showed the following returns:

Market = +9.6%
Price to Book = +14.8%
Price to Earnings = +13.7%
Price to Cashflow = +13.5%
Price to Dividend = +12.7%

So while all value approaches "worked" (outperforming the market), P/D underperformed P/B by over 2% per year, similar to the US evidence I submitted in a previous post. Probably more problematic for those assuming high-yield stock approaches will reliably outperform, only 2 of 13 countries produce statistically significant higher returns from High Dividend minus Low Dividend portfolios. In 3 of 13, Low Dividend actually beat High Dividend!

Eric
I mentioned high-yield stocks to demonstrate data-mining but I didn't recommend them.
A one-factor model may give the purest and highest factor exposure. But we must choose the right factor and unless we can find a theory that explains why a particular factor is superior we cannot be sure that the past predicts the future [*].

More data-mining that does not prove the superiority of book value (here the markets weights are valuation-based):
Indexuniverse March 2013, Mebane Faber, P. Dalmia- Global Value
Samuel Lee has a great article titled “The Hedgehog’s Error”9 on Morningstar’s website that sorts global countries based on value (price/book) using the French/Fama database. Not surprisingly, he finds that sorting on value works well.

We utilize the database to sort the countries (12 in 1975 and rising to 20 by 1991) based on various measure of value. In Figure 12, we demonstrate the results of sorting the countries on a yearly basis and choosing the cheapest x percent of the universe (from 10 to 33 percent). Results are U.S. dollar based, nominal.
Figure 12 - Appendix Data: Cheapest X Percent of Countries, 1975-2011, CAGR in USD:
------- Book Yield-----Earnings Yield----Cash Flow Yield----Dividend Yield

33%--- 3 ------------ 1 ---------------- 4 ------ 2 --
25%--- 1 ------------ 3 ---------------- 4 ------ 2 --
10%--- 1 ------------ 2 ---------------- 3 ------ 4 --

If we consider the 'moderate' portfolios ( 33 + 25) Book Yield, Earnings Yield and Dividend Yield had similar rankings ( 1 = best CAGR ).

Dimson/M/S - Credit Suisse Yearbook 2013 about the 'flaws' of the P/E Ratio:
There was also an evolution in accounting standards and major step changes in the definition of reported earnings, so that early earnings data are not truly comparable with more recent data. Additionally, when comparing different countries’ equity markets, there has been cross-sectional variation in inflationary and economic conditions, and in reporting practices. Consequently, not only is the cyclically adjusted price-dividend ratio PD10 a substitute for the cyclically adjusted price-earnings ratio PE10 in the USA, but the dividend-based series is likely to be a superior metric for making very long-run and crosscountry comparisons
.

FF admit that the book value is not perfect:
Fama/French Forum - Why use Book Value to Sort Stocks?
One fundamental (book value, earnings, or cashflow) is pretty much as good as another for this job, and the average return spreads produced by different ratios are similar to and, in statistical terms, indistinguishable from one another. We like BtM because the book value in the numerator is more stable over time than earnings or cashflow, which is important for keeping turnover down in a value portfolio.

Nevertheless, there are problems in all accounting variables and book value is no exception, so supplementing BtM with other ratios can in principal improve the information about expected returns. We periodically test this proposition, so far without much success.
[*]If the explanatory power of BTM can decline, the power of other valuation ratios could rise:
3-factor regression tests of the FF benchmark portfolios [not shown] indicate the explanatory power of the
book-to-market effect seems to weaken in the period following January 1979. HML coefficients for the FF large cap
portfolios fall monotonically from 1.08 in the pre-1979 sub-period to 0.84 in the most recent sub-period, 1995 to
2006. T-statistics for the coefficients also fall commensurately but all coefficients remain strongly significant. HML
coefficients for the FF small cap portfolios remain fairly stable (0.89, 0.94, 0.94) over the three sub-periods
Page 113 pdf, Scislaw: Three essays on the value premium
Or: page 54: Recent time variation in the power of the 3F Model to explain stock returns
prior username: hafis50

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

Re: WSJ: Why Most Value Investors Will Burn Out

Post by hafius500 » Fri Feb 22, 2013 12:48 pm

With regard to the original article, Scislaw wrote:

Scislaw - Three essays on the value premium : can investors capture the promised rewards? (I don't have the link to the updated version in the Journal of Asset Allocation (?))
Page 34: Persistence of the value premium:
Over the same 80 year period of 966 monthly return observations, value stocks outperform growth stocks only 51% of the time. When value stocks dominate, they do so with greater superiority... This suggests that minor timing errors while using an equity style switching strategy could eliminate any profits compared to those using a simple buy and hold value strategy....Since 1926, a large portion of the premium has been earned during rare (or outlier) monthly return periods. Value stocks dominate growth stocks at return levels greater than 10% in thirteen of the 975 months from July 1926 to September 2007. Interestingly, when the thirteen extreme monthly returns are removed from the sample, the average monthly HML premium is more than halved from 0.41% to 0.19%. Periods of extreme volatility in the value premium illustrate serious problems facing investors who try to tactically capture the value premium.
prior username: hafis50

MarginOfBuffett
Posts: 97
Joined: Fri Nov 12, 2010 11:04 pm

Re: WSJ: Why Most Value Investors Will Burn Out

Post by MarginOfBuffett » Fri Feb 22, 2013 12:58 pm

stlutz wrote:Actually, backtests based on metrics like enterprise value to ebitda or price-to-free cash flow have historically outperformed the various P/E, P/B etc. metrics. The former ones are the ones that quant value investors are really keying off of nowadays.

The problem with P/B and P/E is that the denominators of both are really based on accounting constructs. In many industries, most assets are "soft assets" (e.g. patents, trademarks, goodwill, etc.) and might be valued differently at company X vs. company Y. Even in industries with more "real" assets, depreciation is done based on accounting rules, not based on the market value of, say, a steel plant. From a theoretical standpoint, it's hard to argue that P/B is the "best" measure of value. More obviously, "earnings" may or may not relate to whether the company is really ending up with more money in the safe at the end of the year.

The thing about EV/EBITDA or P/FCF is that the denominators are closer to being real numbers--measuring real cash coming into the business and still don't have the biases that result from one time special items and the like.

Is that why they backtest better? I don't know. It may just be related to the fact that for most of the period being backtested, value investors were using the more standard P/B and P/E metrics and thus there was advantage to using the lesser-known ones. Now that the quants are all focusing on the cash flow statement, that may not repeat in the future.

Also, on the dividend models, again numerous backtests show that price/shareholder yield (the sum of dividends and share buybacks) far outperforms simple dividend yield screens.

Don't know that there's really much of anything practical for anyone to take away from this except that something is always going to backtest the best.
You've been reading the O'Shaugnessy book haven't you?

Burn the Witch!!!!

MarginOfBuffett
Posts: 97
Joined: Fri Nov 12, 2010 11:04 pm

Re: WSJ: Why Most Value Investors Will Burn Out

Post by MarginOfBuffett » Fri Feb 22, 2013 1:02 pm

SP-diceman wrote:
VictoriaF wrote:
nisiprius wrote:Benjamin Graham called his favorite investments "cigar butts." Well, they are good "value"--they have a meaningful amount of tobacco in them and you can pick them up for free. The reason why they are a good value is that other investors don't want to touch them.
A note in a men's room:
Do not throw butts into urinals. It's difficult to puff at them afterwards.

Victoria


The Porcelain Special Situations Fund.

:)
The Porcelain International Special Situations Fund

P.I.S.S.

User avatar
VictoriaF
Posts: 17956
Joined: Tue Feb 27, 2007 7:27 am
Location: Black Swan Lake

Re: WSJ: Why Most Value Investors Will Burn Out

Post by VictoriaF » Fri Feb 22, 2013 1:17 pm

MarginOfBuffett wrote:
SP-diceman wrote:
VictoriaF wrote:
nisiprius wrote:Benjamin Graham called his favorite investments "cigar butts." Well, they are good "value"--they have a meaningful amount of tobacco in them and you can pick them up for free. The reason why they are a good value is that other investors don't want to touch them.
A note in a men's room:
Do not throw butts into urinals. It's difficult to puff at them afterwards.

Victoria
The Porcelain Special Situations Fund.

:)
The Porcelain International Special Situations Fund

P.I.S.S.
Nisi has an uncanny talent for generating and propagating great ideas.

Victoria
WINNER of the 2015 Boglehead Contest. | Every joke has a bit of a joke. ... The rest is the truth. (Marat F)

swaption
Posts: 1182
Joined: Tue Jul 29, 2008 11:48 am

Re: WSJ: Why Most Value Investors Will Burn Out

Post by swaption » Fri Feb 22, 2013 2:06 pm

nisiprius wrote:"Value" is a euphemism, like "high-yield" in bonds. I learn from Justin Fox's excellent The Myth of the Rational Market that Benjamin Graham called his favorite investments "cigar butts." Well, they are good "value"--they have a meaningful amount of tobacco in them and you can pick them up for free. The reason why they are a good value is that other investors don't want to touch them. Perhaps it would clarify our thinking if we made a habit of doing mental translation: Vanguard Small-Cap Cigar Butt Index Fund, DFA US Targeted Cigar Butt Portfolio, Fama-French Cigar Butt Factor etc. If you cannot say, aloud and unafraid, "I have a cigar butt tilt," perhaps you are in danger of being among the cigar butt investors who burn out.
I like the cigar butt analogy. Reminds me of a conversation I was having with a friend that is an international portfolio manager. At the time I was working for a European company and she did not like it as an investment. Did not understand how the company would make money, and I really could not find fault with that perspective. Having said that, at the time my company was in the proverbial penalty box being extremely out of favor. Within a year the stock was up close to 100%.

Does the above make her a bad portfolio manager? No, not really. Essentially, the company may still be a bad company, but just less bad than the market was pricing in. Easy to think of the laundry list of companies that fit this profile (think banks during the financial crisis). You have to be able to systematically invest in "bad" companies (cigar butts). Seems like a really tough way for an active manager to earn a living. Makes sense that a systematic passive approach would be well suited to this.

EDN
Posts: 528
Joined: Sat Dec 29, 2012 7:55 pm
Location: Oklahoma City, OK

Re: High Dividend Doesn't Necessarily Mean High Return

Post by EDN » Fri Feb 22, 2013 2:26 pm

hafius500 wrote:
EDN wrote:
hafius500 wrote:
The size and yield effects could be related.
I recall some quant (value) managers use different ratios (P/B vs. P/E etc.) for different country markets because no ratio works best in every market.
Data mining and sample bias are dangers.
hafius,

Fama and French followed up their US paper with an International version in 1998 (see here: http://citeseerx.ist.psu.edu/viewdoc/do ... 1&type=pdf). In it they looked at global value portfolios based on different sorts and showed the following returns:

Market = +9.6%
Price to Book = +14.8%
Price to Earnings = +13.7%
Price to Cashflow = +13.5%
Price to Dividend = +12.7%

So while all value approaches "worked" (outperforming the market), P/D underperformed P/B by over 2% per year, similar to the US evidence I submitted in a previous post. Probably more problematic for those assuming high-yield stock approaches will reliably outperform, only 2 of 13 countries produce statistically significant higher returns from High Dividend minus Low Dividend portfolios. In 3 of 13, Low Dividend actually beat High Dividend!

Eric
I mentioned high-yield stocks to demonstrate data-mining but I didn't recommend them.
A one-factor model may give the purest and highest factor exposure. But we must choose the right factor and unless we can find a theory that explains why a particular factor is superior we cannot be sure that the past predicts the future [*].

More data-mining that does not prove the superiority of book value (here the markets weights are valuation-based):
Indexuniverse March 2013, Mebane Faber, P. Dalmia- Global Value
Samuel Lee has a great article titled “The Hedgehog’s Error”9 on Morningstar’s website that sorts global countries based on value (price/book) using the French/Fama database. Not surprisingly, he finds that sorting on value works well.

We utilize the database to sort the countries (12 in 1975 and rising to 20 by 1991) based on various measure of value. In Figure 12, we demonstrate the results of sorting the countries on a yearly basis and choosing the cheapest x percent of the universe (from 10 to 33 percent). Results are U.S. dollar based, nominal.
Figure 12 - Appendix Data: Cheapest X Percent of Countries, 1975-2011, CAGR in USD:
------- Book Yield-----Earnings Yield----Cash Flow Yield----Dividend Yield

33%--- 3 ------------ 1 ---------------- 4 ------ 2 --
25%--- 1 ------------ 3 ---------------- 4 ------ 2 --
10%--- 1 ------------ 2 ---------------- 3 ------ 4 --

If we consider the 'moderate' portfolios ( 33 + 25) Book Yield, Earnings Yield and Dividend Yield had similar rankings ( 1 = best CAGR ).

Dimson/M/S - Credit Suisse Yearbook 2013 about the 'flaws' of the P/E Ratio:
There was also an evolution in accounting standards and major step changes in the definition of reported earnings, so that early earnings data are not truly comparable with more recent data. Additionally, when comparing different countries’ equity markets, there has been cross-sectional variation in inflationary and economic conditions, and in reporting practices. Consequently, not only is the cyclically adjusted price-dividend ratio PD10 a substitute for the cyclically adjusted price-earnings ratio PE10 in the USA, but the dividend-based series is likely to be a superior metric for making very long-run and crosscountry comparisons
.

FF admit that the book value is not perfect:
Fama/French Forum - Why use Book Value to Sort Stocks?
One fundamental (book value, earnings, or cashflow) is pretty much as good as another for this job, and the average return spreads produced by different ratios are similar to and, in statistical terms, indistinguishable from one another. We like BtM because the book value in the numerator is more stable over time than earnings or cashflow, which is important for keeping turnover down in a value portfolio.

Nevertheless, there are problems in all accounting variables and book value is no exception, so supplementing BtM with other ratios can in principal improve the information about expected returns. We periodically test this proposition, so far without much success.
[*]If the explanatory power of BTM can decline, the power of other valuation ratios could rise:
3-factor regression tests of the FF benchmark portfolios [not shown] indicate the explanatory power of the
book-to-market effect seems to weaken in the period following January 1979. HML coefficients for the FF large cap
portfolios fall monotonically from 1.08 in the pre-1979 sub-period to 0.84 in the most recent sub-period, 1995 to
2006. T-statistics for the coefficients also fall commensurately but all coefficients remain strongly significant. HML
coefficients for the FF small cap portfolios remain fairly stable (0.89, 0.94, 0.94) over the three sub-periods
Page 113 pdf, Scislaw: Three essays on the value premium
Or: page 54: Recent time variation in the power of the 3F Model to explain stock returns
Haifus,

I never said other factors didn't work. The Fama/French quote you provide says it all -- they are all good (except dividend sorts), (low) price is what matters, BUT when it comes to investability BtM produces lowest turnover therefore lower trading costs and best real world results AND FF periodically test multifactor sorts which would appear to offer an advantage but they've been unable to show with actual results that that is the case.

You do realize that the extremely lengthy articles you quote make it difficult to respond to your posts, your points are never clear, which might be why I don't or cannot address your comments.

Eric

PS -- the issue with multifactor sorts is they are ALL just small variations on the same theme: low prices relative to an accounting variable. The most effective or valuable ADDITIONAL sort or criteria to add would be one that is able to whittle down all of the Low P/X stocks to more reliably isolate which companies in particular tend to drive the premium, as we know that 15% to 20% of the stocks account for all the additional size and value returns but never which ones (so we hold them all).

As the recent research from FF and Robert Novy Marx indicate, this additional sort isn't a Price variable whatsoever, its a Profitability variable. So further refining a P/B sort to exclude the most unprofitable firms and even widening the value sort to hold slightly higher P/B firms with very high profitability is the only thing I've come across that speaks affirmatively about multifactor screening. And, in the spirit of P/B which has low turnover and high stability, highly profitable firms tend to stay that way for several years -- also with very low turnover.

EDN
Posts: 528
Joined: Sat Dec 29, 2012 7:55 pm
Location: Oklahoma City, OK

Re: WSJ: Why Most Value Investors Will Burn Out

Post by EDN » Fri Feb 22, 2013 2:35 pm

hafius500 wrote:With regard to the original article, Scislaw wrote:

Scislaw - Three essays on the value premium : can investors capture the promised rewards? (I don't have the link to the updated version in the Journal of Asset Allocation (?))
Page 34: Persistence of the value premium:
Over the same 80 year period of 966 monthly return observations, value stocks outperform growth stocks only 51% of the time. When value stocks dominate, they do so with greater superiority... This suggests that minor timing errors while using an equity style switching strategy could eliminate any profits compared to those using a simple buy and hold value strategy....Since 1926, a large portion of the premium has been earned during rare (or outlier) monthly return periods. Value stocks dominate growth stocks at return levels greater than 10% in thirteen of the 975 months from July 1926 to September 2007. Interestingly, when the thirteen extreme monthly returns are removed from the sample, the average monthly HML premium is more than halved from 0.41% to 0.19%. Periods of extreme volatility in the value premium illustrate serious problems facing investors who try to tactically capture the value premium.
Well, that one I can make out, and it's wrong. FF HmL over the longest period available (1926-2013) was positive 54% of the time on a one-month basis (market premium = 59%), 63% of the time on rolling 1YR basis (market premium = 68%), 84% of the time on a rolling 5YR basis (market premium = 78% of the time), and 96% of the time on a rolling 10YR basis (market premium = 84% of the time).

So the value premium has been more reliable than the market premium over intermediate/longer periods.

And, no one advocates for a value timing strategy. Quite the contrary, everything I said above is about the best value strategies isolating and targeting value stocks as consistently and reliably as possible, so it looks like we agree in general.

Also, while I'm not going to check the validity of the data-mining exercise of excluding just the very best months of the value premium (a value investor should hold value stocks every month, so this is irrelevant), you could say the same things about the market premium, the size premium, the credit premium, and the term premium: exclude a number of the absolute best returns, and the average result goes down.

Heck, if I eliminate the 5 worst bear markets for US stocks in the last 85 years, you could say stocks don't experience bear markets! :oops:

Eric

Post Reply