Glamour stocks and value tilt

Discuss all general (i.e. non-personal) investing questions and issues, investing news, and theory.
Post Reply
boglerdude
Posts: 634
Joined: Fri Jul 10, 2015 1:28 am
Contact:

Glamour stocks and value tilt

Post by boglerdude » Fri Oct 30, 2015 10:47 pm

Arnott uses glamour stocks (twitter, Facebook, Priceline) to suggest that cap-weighters are throwing money away on overpriced shares.
Eg, starting at 11:36 through 16:06
https://www.youtube.com/watch?v=wK4a-jXMLdk&t=695
(Not necessary to watch just citing source.)

So, his fund sells when a company's price gets beyond (an arbitrary percentage?) of its "actual" value (some mix of book value, earnings, dividends).

He implies that buying growth/glamour stocks is risky because you are paying now for ambitious earnings that might not materialize. So his value tilt is less risky.

On the other hand, Fama says value stocks are companies that have fallen on hard times:
https://www.youtube.com/watch?v=HIKO-t4vU6Q&t=396
(Earlier in the interview he downplays the small cap factor, but that's another story.)

How does he know about these hard times, can you make that determination from looking at publicly available statistics. Or is he saying that if a company has good numbers on paper but is selling low, there must be a problem with it that traders are aware of?

He says the value premium is because you are taking more risk, betting on troubled companies.

Asking because I'm deciding between total stock market or a dividend fund (because I'm in a low tax bracket). The dividend funds are categorized as "large value" and the companies they contain don't seem risky to me. Maybe that's only because I understand their products.

In sum, would like to hear opinions on potential drawbacks of Arnotts RAFI fund, and explanations for the different definition of risk compared to Fama. thx!

small_index
Posts: 631
Joined: Sun Aug 16, 2015 2:47 am

Re: Glamour stocks and value tilt

Post by small_index » Fri Oct 30, 2015 11:05 pm

I think you're taking two definitions of value and mixing them together. Fama/French would define the "value" factor in their model based on price/book. Facebook has a price/book over 7, which is incredibly high - it would qualify as a "growth" stock, not a "value" stock in the Fama/French 3-factor model.

User avatar
JoMoney
Posts: 6120
Joined: Tue Jul 23, 2013 5:31 am

Re: Glamour stocks and value tilt

Post by JoMoney » Fri Oct 30, 2015 11:17 pm

There are two different stories often used by the 'Value' fans, sometimes on this board they confuse their arguments and combine them although they're clearly in conflict with each other, despite both believing 'Value' has higher expected returns.
One story is that the market is efficiently and persistently pricing 'risk premiums', and there is higher returns only for accepting higher risk.
The other is a behavioral story of an in-efficient market where participants frequently mis-price stocks, and that they do so more often to the detriment of "growth" style stocks and the benefit of "value" style.
"To achieve satisfactory investment results is easier than most people realize; to achieve superior results is harder than it looks." - Benjamin Graham

User avatar
in_reality
Posts: 4529
Joined: Fri Jul 12, 2013 6:13 am

Re: Glamour stocks and value tilt

Post by in_reality » Fri Oct 30, 2015 11:22 pm

The drawback on RAFI of course is cost. Will the higher cost actually prove worth it?
Another drawback is that you may in fact take more risk. I don't think anyone can say RAFI has more risk. It has higher costs which is a risk if it doesn't perform well enough to make up for that.

For instance, RAFI had me more heavily into energy stocks before they fell. Given their methodology uses a 5 year average, I suspect my RAFI funds will be rebalancing into energy. I guess I will be well positioned for a recovery. But what if that recovery doesn't happen? Or what if that recovery happens after the 5 year average and then RAFI has moved to reduce exposure. I'm paying the higher ER the whole time...

One reason I like RAFI methodology more than a simple cheap value tilt is that it ends up investing in sectors that are not traditionally value. For instance, their small/mid cap fund has a sizable allocation to mid-cap growth. Why? Because there are "growth" companies that are more valuey than their peers. I am not sure if DFA selects those companies or not.

Another drawback is abandonment risk. After value underperforms for some time, if you sell out low you have really hurt yourself. Putting yourself into a fund with higher costs and that then underperforms is a sure way to put doubt in your head about whether you should have really made that investment in the first place.

Another drawback is the reliance on reported data to compose the index. How reliable do you think the data coming from emerging markets is? I use RAFI in emerging markets but have some doubts there.

I suspect many are drawn to RAFI or DFA in the promise of greater potential returns for taking more risk. Well, how about this, why not just allocate say 5% more to market cap weighted stocks and use high quality treasuries/bonds to rebalance when those glamor stocks crash? It might be easier to stay the course that way. Or if you believe that RAFI has less risk, why not just lower your stock allocation and again rebalance? You won't be guaranteed to be paying the transaction and higher fund costs that way.
Last edited by in_reality on Fri Oct 30, 2015 11:35 pm, edited 2 times in total.

User avatar
baw703916
Posts: 6679
Joined: Sun Apr 01, 2007 1:10 pm
Location: Seattle

Re: Glamour stocks and value tilt

Post by baw703916 » Fri Oct 30, 2015 11:29 pm

JoMoney wrote:There are two different stories often used by the 'Value' fans, sometimes on this board they confuse their arguments and combine them although they're clearly in conflict with each other, despite both believing 'Value' has higher expected returns.
One story is that the market is efficiently and persistently pricing 'risk premiums', and there is higher returns only for accepting higher risk.
The other is a behavioral story of an in-efficient market where participants frequently mis-price stocks, and that they do so more often to the detriment of "growth" style stocks and the benefit of "value" style.
The two don't necessarily contradict each other. Maybe the function that the market optimizes isn't Sharpe ratio as normally defined, but something like return*(1+C*skewness)/(variance)^0.5, where C is a positive constant. So by being willing to give up finding the "next Microsoft/Google/whatever" (positive skewness) you might get a higher expected return.
Most of my posts assume no behavioral errors.

User avatar
Maynard F. Speer
Posts: 2139
Joined: Wed Mar 18, 2015 10:31 am

Re: Glamour stocks and value tilt

Post by Maynard F. Speer » Sat Oct 31, 2015 7:56 am

Ultimately there can be 1,000 different stories for value, but the simple fact is you're getting more (take your pick): earnings, sales, cash-flow, growth or company for your $

Be cynical, but those fundamentals are essentially 'facts' ... Anything else people are pricing/discounting on (such as risk) is based on prediction - which tends to be less reliable

The story I like is that you pay a hefty premium for certainty ... We know Facebook is likely to be there in two years time - when you buy a smaller company, the chance may only be 1 in 3 ... A note on dividend stocks, however ... With bonds yields where they are, there may be a hefty premium on dividend-payers as bond proxies ... It's a bit of a macro bet - if rates stay low, or get lower, the premium could go up ... If rates normalise, or dividends are slashed, these so-called 'stable' assets *could* be hit among the hardest (in some people's opinion)
"Economics is a method rather than a doctrine, an apparatus of the mind, a technique of thinking, which helps its possessor to draw correct conclusions." - John Maynard Keynes

User avatar
nedsaid
Posts: 10500
Joined: Fri Nov 23, 2012 12:33 pm

Re: Glamour stocks and value tilt

Post by nedsaid » Sat Oct 31, 2015 8:53 am

JoMoney wrote:There are two different stories often used by the 'Value' fans, sometimes on this board they confuse their arguments and combine them although they're clearly in conflict with each other, despite both believing 'Value' has higher expected returns.
One story is that the market is efficiently and persistently pricing 'risk premiums', and there is higher returns only for accepting higher risk.
The other is a behavioral story of an in-efficient market where participants frequently mis-price stocks, and that they do so more often to the detriment of "growth" style stocks and the benefit of "value" style.
I am in the behavioral camp. If risk is measured by standard deviation, from what I have seen value stocks are actually less risky than growth stocks. I would be in agreement that Value stocks might have more fundamental risk, tending to have more debt for example. I think it was Packer16 or Backpacker who pointed out fundamental risk and Larry Swedroe has made comments along that line too. In other words, the "bad company" theory. The thing is, I have read that value companies are "riskier" but the arguments are in narrative form and I have not seen value "risk" explained in mathematical terms.

But again, I am in the camp of buying good stuff when it goes on sale. I am not trying to get the last few puffs from a discarded cigar butt. My definition of value would be closer to Warren Buffett. This is a way of saying that value is in the eye of the beholder. There are different ways of measurement. For example, I am more focused on earnings and cash flow. For certain types of companies, book value may not be all that meaningful.
A fool and his money are good for business.

anil686
Posts: 751
Joined: Thu May 08, 2014 12:33 pm

Re: Glamour stocks and value tilt

Post by anil686 » Sat Oct 31, 2015 9:17 am

in_reality wrote:The drawback on RAFI of course is cost. Will the higher cost actually prove worth it?
Another drawback is that you may in fact take more risk. I don't think anyone can say RAFI has more risk. It has higher costs which is a risk if it doesn't perform well enough to make up for that.

For instance, RAFI had me more heavily into energy stocks before they fell. Given their methodology uses a 5 year average, I suspect my RAFI funds will be rebalancing into energy. I guess I will be well positioned for a recovery. But what if that recovery doesn't happen? Or what if that recovery happens after the 5 year average and then RAFI has moved to reduce exposure. I'm paying the higher ER the whole time...

One reason I like RAFI methodology more than a simple cheap value tilt is that it ends up investing in sectors that are not traditionally value. For instance, their small/mid cap fund has a sizable allocation to mid-cap growth. Why? Because there are "growth" companies that are more valuey than their peers. I am not sure if DFA selects those companies or not.

Another drawback is abandonment risk. After value underperforms for some time, if you sell out low you have really hurt yourself. Putting yourself into a fund with higher costs and that then underperforms is a sure way to put doubt in your head about whether you should have really made that investment in the first place.

Another drawback is the reliance on reported data to compose the index. How reliable do you think the data coming from emerging markets is? I use RAFI in emerging markets but have some doubts there.

I suspect many are drawn to RAFI or DFA in the promise of greater potential returns for taking more risk. Well, how about this, why not just allocate say 5% more to market cap weighted stocks and use high quality treasuries/bonds to rebalance when those glamor stocks crash? It might be easier to stay the course that way. Or if you believe that RAFI has less risk, why not just lower your stock allocation and again rebalance? You won't be guaranteed to be paying the transaction and higher fund costs that way.
Thanks for one of the best described (IMO) posts regarding RAFI vs cap weighting - it was really well done IMO

I like the RAFI methodology but I cap weight based on the arguments presented here due to cost. I think the cost savings on a value index approach along with tax efficiency in a taxable account make it hard for me to justify RAFI over a cap weighted value index.
While I agree with Arnott's point that you are overpaying for some growth stocks, there are periods of time where growth does very well despite higher valuations. I think Mr. Bogle looked at the history of 30 of the largest growth stocks from 1969 to the late 1980s (when the 1970s were really a tough time for growth) and it performed very similar to the SP 500 for the entire period(source Coomon Sense on Mutual Funds). For that reason, it is hard to pay so much more when cap weighting a value index is less than 0.1% and cap weighting a total U.S. Market index is now at 0.05% or less ...

stlutz
Posts: 4784
Joined: Fri Jan 02, 2009 1:08 am

Re: Glamour stocks and value tilt

Post by stlutz » Sat Oct 31, 2015 9:52 am

[Fama] says the value premium is because you are taking more risk, betting on troubled companies.
Here is the logic:
a) Value consistently outperforms growth across time/markets in backtests.
b) When something like this happens consistently, it "was expected" to have occurred ahead of time.
c) The market is efficient--i.e. this type of expected outperformance of one type of stock over another can only happen because of risk.
d) Therefore, value is riskier than growth.
He implies that buying growth/glamour stocks is risky because you are paying now for ambitious earnings that might not materialize. So his value tilt is less risky.
Here is the logic on this one:

a) Value stocks tend to go down less during bear markets (though certainly not always).
b) Therefore, value is less risky than growth.

User avatar
nisiprius
Advisory Board
Posts: 36869
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: Glamour stocks and value tilt

Post by nisiprius » Sat Oct 31, 2015 12:34 pm

boglerdude wrote:...Arnott uses glamour stocks (twitter, Facebook, Priceline) to suggest that cap-weighters are throwing money away on overpriced shares...
Well, what do you expect him to say? "A cheap cap-weighted index fund that doesn't make me any money is just as good as a fund that pays me licensing fees?"

When you get past the rhetoric, you have two basic choices:

1) Invest your money the same way the market as a whole invests its money. (That is to say, use a cap-weighted total market index fund).
2) Invest your money in a different, and hopefully better way than the market does.

Probably most of us agree that the market as a whole is often stupid, faddish, and glamor-bedazzled. Unfortunately it's surprisingly hard, even though it looks easy, to identify which are the overpriced stocks. Am I right that he does not include Google as among those "glamor stocks?" Why not? Don't you remember all of the obviously over-the-top hype about Google, and all of the dire warnings about it?

Conrad DeAinlie, writing in the New York Times on October 31st, 2005:
Five years after the tech bubble came and went, the search engine operator has bounded from about $180 a share in April to nearly $300.... So is it different this time, or are investors partying like it's 1999?What separates admirers of Google's stock from skeptics is its price. Based on its close of $282.50 a share, it trades at 88.3 times earnings in the most recent four quarters. But Mr. Jordan estimates that Google will earn $6 this year and $8 to $10 next year. He comes up with a multiple in the 30's, saying this justifies a forecast of $400 for the stock. This contrivance -- make prodigious growth estimates far enough into the future to generate a less yikes!-inducing valuation -- became popular during the bubble. Forecasts are just that, and in the case of Google they have been soaring as analysts outdo one another in unceasingly rosy hues.
Why doesn't Google qualify as a glamor stock? The obvious answer is: it is a glamor stock, it just happens to be a glamor stock that paid off. That often does happen, but when it happens people quit calling them disparaging names like "glamor stocks."

Source: Morningstar
Image

Choice 2) then can be divided into three subcategories:

2a) "We have a magic formula that automatically gives more weight to the undervalued stocks and less weight to the overvalued ones." There are a bunch of them. There's tilting on the Fama-French value factor. Tilting on improved versions of the value factor. Tilting on multiple factors. Overweighting dividend stocks, or investing only in dividend stocks. Equal weighting. "Fundamental" weighting. "Enhanced indexing" (Fidelity). Smart beta. And, of course, Joel Greenblatt's magic formula, which he calls "the Magic Formula."

2b) "There isn't any formula, but we know 'em when we see 'em."

2c) The hybrid: "We use a formula, but we don't apply it blindly--we use it together with our judgement. Even if our formula says to buy a stock, we won't buy it if we know better than the formula."

Shrug. I go with choice #1, personally. If you decide to go with choice #2, then you can't just say "I think the market is inefficient, this guy says the market is inefficient, ergo he must be the one who knows how to beat the market."
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.

User avatar
Maynard F. Speer
Posts: 2139
Joined: Wed Mar 18, 2015 10:31 am

Re: Glamour stocks and value tilt

Post by Maynard F. Speer » Sat Oct 31, 2015 2:32 pm

nisiprius wrote:Shrug. I go with choice #1, personally. If you decide to go with choice #2, then you can't just say "I think the market is inefficient, this guy says the market is inefficient, ergo he must be the one who knows how to beat the market."
I must admit I've not been won over by smart beta yet ... However - looking backwards - it seems almost any method of constructing an index has tended to outperform ... With monkey portfolios returning on average 1.6% higher since 1964 (Re: The Surprising Alpha of Malkiel's Monkey and Upside-down Strategies)

Image

The conclusion for me is it pays to be away from the crowds (I think Ibbotson identified popularity as a sort of negative factor?) ... The obvious problem with smart beta is it signposts where investors should put their money .. A premium could disappear almost as soon as it's investable

It seems to me the best shot an indexer has is to be equal-weighted (the infinite monkey portfolio)
"Economics is a method rather than a doctrine, an apparatus of the mind, a technique of thinking, which helps its possessor to draw correct conclusions." - John Maynard Keynes

User avatar
nisiprius
Advisory Board
Posts: 36869
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: Glamour stocks and value tilt

Post by nisiprius » Sat Oct 31, 2015 3:00 pm

Maynard F. Speer wrote:...It seems to me the best shot an indexer has is to be equal-weighted (the infinite monkey portfolio)...
I think it's been shown that the characteristics of the equal-weighted portfolio are simply the consequences of increased allocation to small-caps. See, for example, Rick Ferri's No free lunch from equal-weighted S&P 500

Does it make sense to you that if a business named AT&T gets split up into seven pieces named Ameritech, Bell Atlantic, etc. that you should immediate invest seven times the original number of dollars in the pieces of the original business? The Baby Bells might be better apart, but can you really believe sum of the parts is seven times as big as the whole?

Another logical problem with equal weight is that if you cap-weight the S&P 500, you effectively have 4/5th of the cap-weighted total market, which is not too bad as a representative sample... and for about twenty years it's been feasible to implement a cap-weighted total market fund and numerous firms have done so.

However, if you equal-weight the S&P 500, you only have about 1/8th of the equal-weighted total market, and thus obviously are nowhere near implementing the desired strategy--unless you carefully snip out all the small-cap stocks from the NYSE list before you start throwing the darts. There is no fund or ETF that equal-weights the whole U.S. stock market. I think it's not feasible--there aren't enough of the smallest stocks to go around and the trading costs of rebalancing would be too large. There are ETFs that follow the I-think-misleadingly-named Russell 1000 and 2000 Equal Weight indexes--but these indexes, despite their names, do not do anything close to giving equal weight to each of the stocks in them.
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.

Levett
Posts: 4177
Joined: Fri Feb 23, 2007 2:10 pm
Location: upper Midwest

Re: Glamour stocks and value tilt

Post by Levett » Sat Oct 31, 2015 3:31 pm

Maynard Speer observed:

"The conclusion for me is it pays to be away from the crowds (I think Ibbotson identified popularity as a sort of negative factor?) ... The obvious problem with smart beta is it signposts where investors should put their money .. A premium could disappear almost as soon as it's investable"

Amen.

Boring is beautiful.

Lev

User avatar
Maynard F. Speer
Posts: 2139
Joined: Wed Mar 18, 2015 10:31 am

Re: Glamour stocks and value tilt

Post by Maynard F. Speer » Sat Oct 31, 2015 5:01 pm

nisiprius wrote:
Maynard F. Speer wrote:...It seems to me the best shot an indexer has is to be equal-weighted (the infinite monkey portfolio)...
I think it's been shown that the characteristics of the equal-weighted portfolio are simply the consequences of increased allocation to small-caps. See, for example, Rick Ferri's No free lunch from equal-weighted S&P 500
I used to assume it was mainly explained by size - however the breakdown in this paper from 2014 (first table), suggested equal weight added 1.8% value .. of which 0.38 could be explained by size; 0.12 value; -0.02 momentum; and a surprising (as per the paper title) 1.63% alpha
http://www.q-group.org/wp-content/uploa ... ll_Hsu.pdf

The same year Roger Ibbotson put out Dimensions of Popularity:
"Liquidity is popular, whereas risk is unpopular ... In general, the less popular a security, the lower the valuation but the higher the expected return"
http://www.iijournals.com/doi/abs/10.39 ... Code=jpm40

Does it make sense to you that if a business named AT&T gets split up into seven pieces named Ameritech, Bell Atlantic, etc. that you should immediate invest seven times the original number of dollars in the pieces of the original business? The Baby Bells might be better apart, but can you really believe sum of the parts is seven times as big as the whole?
Well it doesn't sound very intuitive, but I suppose what you really want is exposure to growth .. I suppose if all seven are still large enough to be make it into your equal-weighted index, you could argue you could have justifiably had seven times the exposure in the first place?

Another logical problem with equal weight is that if you cap-weight the S&P 500, you effectively have 4/5th of the cap-weighted total market, which is not too bad as a representative sample... and for about twenty years it's been feasible to implement a cap-weighted total market fund and numerous firms have done so.

However, if you equal-weight the S&P 500, you only have about 1/8th of the equal-weighted total market, and thus obviously are nowhere near implementing the desired strategy--unless you carefully snip out all the small-cap stocks from the NYSE list before you start throwing the darts. There is no fund or ETF that equal-weights the whole U.S. stock market. I think it's not feasible--there aren't enough of the smallest stocks to go around and the trading costs of rebalancing would be too large. There are ETFs that follow the I-think-misleadingly-named Russell 1000 and 2000 Equal Weight indexes--but these indexes, despite their names, do not do anything close to giving equal weight to each of the stocks in them.
I'd probably add cap-weighted small-cap exposure .. It seems to me small and mid-cap funds overcome much of the problem anyway, as they regularly sell winners

As I was interested to learn the other day, far less than 1% of the 27 million companies registered in the US are listed, so I'm not sure how arbitrary the number you're actually exposed to in a total market fund is anyway? (as far as it's a representation of the economy)
"Economics is a method rather than a doctrine, an apparatus of the mind, a technique of thinking, which helps its possessor to draw correct conclusions." - John Maynard Keynes

stlutz
Posts: 4784
Joined: Fri Jan 02, 2009 1:08 am

Re: Glamour stocks and value tilt

Post by stlutz » Sat Oct 31, 2015 5:04 pm

Another logical problem with equal weight...
I think all weighting schemes have logical problems. Is it really logical to say that because the market as a whole has 3+% in Apple that I must therefore do exactly the same? I don't think so. The "fun" of the Arnott piece is that you can come up with completely illogical approaches and do just fine.

Cap weighting is very useful because is solves a lot of practical problems around turnover costs (both trading and taxes). The ETF structure and the existence of tax-sheltered accounts has done a lot to resolve those. As such, I think it makes most sense to simply focus on costs and risk and then just going with something that is logical enough to me personally that I'll stick with it for the long term.

I personally use a cost cutoff of 20 basis points to filter down to anything that I care about it. Even what that is done, I still have access to value funds, equal-weight funds, momentum funds, low-volatility funds, and of course cap-weight funds. From those choices I've mostly settled on cap weight with some tilt to value and low volatility. That is a "logical" approach to me and I'm happy to stick with it. I don't know that I feel a need to argue that anything else (such as using the equal-weight EUSA ETF) is illogical, however--it's just not a direction I've chosen to go.

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

Re: Glamour stocks and value tilt

Post by hafius500 » Sat Oct 31, 2015 5:37 pm

boglerdude wrote:Arnott uses glamour stocks (twitter, Facebook, Priceline) to suggest that cap-weighters are throwing money away on overpriced shares.
I guess this is the 'Noisy-Markets-Hypothesis' (use the forum search or google) , which was coined by Jeremy Siegel (?).

Several articles published in academic journals concluded this hyothesis is mathematically false unless you make several additional assumptions.

The argument is that portfolios that are not cap-weighted also overweight (small stocks, large stocks, value stocks or growth) securities, when they are overvalued.

If the mispricings (of large, small, value, growth, etc. stocks,) are random, no strategy can save you.
inreality wrote::
One reason I like RAFI methodology more than a simple cheap value tilt is that it ends up investing in sectors that are not traditionally value. For instance, their small/mid cap fund has a sizable allocation to mid-cap growth. Why? Because there are "growth" companies that are more valuey than their peers.
The factor zoo becomes larger. The MSCI Style Factor Indices seem to avoid bets on sectors, too.

Morningstar - This ETF Takes a Different Approach to Value
So, investors who are concerned about taking these types of sector bets might consider a fund like iShares MSCI USA Value Factor ETF (VLUE).

This fund effectively targets the cheapest 30% of stocks in each sector, but it applies sector-neutral weightings, which means that it sets its sector weightings equal to the market-cap-weighted MSCI USA Index.
prior username: hafis50

small_index
Posts: 631
Joined: Sun Aug 16, 2015 2:47 am

Re: Glamour stocks and value tilt

Post by small_index » Sat Oct 31, 2015 7:00 pm

nisiprius wrote:2a) "We have a magic formula that automatically gives more weight to the undervalued stocks and less weight to the overvalued ones." There are a bunch of them. There's tilting on the Fama-French value factor. Tilting on improved versions of the value factor. Tilting on multiple factors. Overweighting dividend stocks, or investing only in dividend stocks. Equal weighting. "Fundamental" weighting. "Enhanced indexing" (Fidelity). Smart beta. And, of course, Joel Greenblatt's magic formula, which he calls "the Magic Formula."
One of these has far more historical data and academic research behind it than the others. Is it really fair to lump Fama-French in with Fidelity's attempt to sell products? It strikes me as disingenuous to write off small/value tilting in this particular manner, lumping it with a long list and not giving it more consideration than the sales teams at various investment shops.

Not saying you in particular should weight investments differently, just saying you're obscuring the very different evidence for some approaches.

User avatar
in_reality
Posts: 4529
Joined: Fri Jul 12, 2013 6:13 am

Re: Glamour stocks and value tilt

Post by in_reality » Sat Oct 31, 2015 11:18 pm

anil686 wrote:For that reason, it is hard to pay so much more when cap weighting a value index is less than 0.1% and cap weighting a total U.S. Market index is now at 0.05% or less ...
It is. RAFI suggests the value effect is mean reverting and their funds will take advantage of that more than a traditional value index. So it's not such a simple comparison to make.

see http://www.iinews.com/site/pdfs/JII_Summer_2014_RA.pdf

Anyway, I think the correct answers are to 1) mind your contribution rate and 2) don't change your AA at an inopportune time.

User avatar
nisiprius
Advisory Board
Posts: 36869
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: Glamour stocks and value tilt

Post by nisiprius » Sun Nov 01, 2015 7:02 am

small_index wrote:
nisiprius wrote:2a) "We have a magic formula that automatically gives more weight to the undervalued stocks and less weight to the overvalued ones." There are a bunch of them. There's tilting on the Fama-French value factor. Tilting on improved versions of the value factor. Tilting on multiple factors. Overweighting dividend stocks, or investing only in dividend stocks. Equal weighting. "Fundamental" weighting. "Enhanced indexing" (Fidelity). Smart beta. And, of course, Joel Greenblatt's magic formula, which he calls "the Magic Formula."
One of these has far more historical data and academic research behind it than the others. Is it really fair to lump Fama-French in with Fidelity's attempt to sell products? It strikes me as disingenuous to write off small/value tilting in this particular manner, lumping it with a long list and not giving it more consideration than the sales teams at various investment shops.

Not saying you in particular should weight investments differently, just saying you're obscuring the very different evidence for some approaches.
(Shrug) Surely Rob Arnott would say there is academic research behind "RAFI indexing," surely Jeremy Siegel would say there is academic research behind WisdomTree's "fundamental indexing."

The general way what I'll call the "factor story" keeps drifting and shifting must give one pause. The very fact that you refer to "small/value" tilting is one case in point. Nobody will ever acknowledge that anything has actually been discredited, but the original grand old factor, the "size effect," was once thought to be powerful and important and now is acknowledged to be relatively weak and unimportant--if indeed it does exist at all. People talk about a "small value tilt" because if the small-company effect does exist, it is more in the nature of a garnish on the value factor than a meal in itself.

As for the value factor, Fama and French are now up to five factors, and while they are careful to explain that "redundant" doesn't mean "isn't," they now call the value factor "redundant." It's no longer really one of the primary factors--it is just magnetic north, and they've now discovered true north in some other factor.

The point, though, is that you have to be careful of the reasoning that says:

a) Cap-weighted indexing is not the best weighting.
b) Therefore, there are departures from cap-weighting that are better than cap-weighting.
c) I am recommending something that is a departure from cap-weighting.
d) Therefore, what I am recommending must be better than cap-weighting
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.

User avatar
in_reality
Posts: 4529
Joined: Fri Jul 12, 2013 6:13 am

Re: Glamour stocks and value tilt

Post by in_reality » Sun Nov 01, 2015 7:20 am

nisiprius wrote: The point, though, is that you have to be careful of the reasoning that says:

a) Cap-weighted indexing is not the best weighting.
b) Therefore, there are departures from cap-weighting that are better than cap-weighting.
c) I am recommending something that is a departure from cap-weighting.
d) Therefore, what I am recommending must be better than cap-weighting
Awesome nishiprius! Simply awesome.

What about the reasoning that says Cap Weighing can beat Cap Weighting?
Rip Van Winkle wakes up and constructs a portfolio reusing the cap weights of the 1,000 largest stocks from when he fell asleep 20 years earlier. He ignores stocks that no longer exist and invests their weight in remaining companies in proportion to their old capitalizations. In subsequent years, he then rebalances back to the stale weights 20 years earlier. For example, since the reliable data start in 1926, the analysis waits until 1946 to use the 20-year-old cap-weighting data. This gives results over the past 67 years (1946–2013) for a portfolio always weighted back to 20 years ago.

The method produced a risk-adjusted outperformance of 1.8% per annum over the normal cap-weighted index
Obviously this is not a viable strategy as if people started using it, it's sure that front runners would be all over it since everyone would know what is coming next!!!

boglerdude
Posts: 634
Joined: Fri Jul 10, 2015 1:28 am
Contact:

Re: Glamour stocks and value tilt

Post by boglerdude » Mon Nov 02, 2015 12:22 am

in_reality wrote:What about the reasoning that says Cap Weighing can beat Cap Weighting? The method produced a risk-adjusted outperformance of 1.8% per annum over the normal cap-weighted index
Why do all these strategies (Van Winkle, monkeys throwing darts, etc) outperform (1-3%) cap-weight in backtesting
(Source: http://www.morningstar.com/cover/videoc ... ?id=613699 )

In reality everyone would copy the winning strategy - so what it is about the backtesting model that causes all the alternative indexes to beat capweight by ~2%

User avatar
in_reality
Posts: 4529
Joined: Fri Jul 12, 2013 6:13 am

Re: Glamour stocks and value tilt

Post by in_reality » Mon Nov 02, 2015 12:34 am

boglerdude wrote:
in_reality wrote:What about the reasoning that says Cap Weighing can beat Cap Weighting? The method produced a risk-adjusted outperformance of 1.8% per annum over the normal cap-weighted index
Why do all these strategies (Van Winkle, monkeys throwing darts, etc) outperform (1-3%) cap-weight in backtesting
(Source: http://www.morningstar.com/cover/videoc ... ?id=613699 )

In reality everyone would copy the winning strategy - so what it is about the backtesting model that causes all the alternative indexes to beat capweight by ~2%
They win for a few reasons:

1) they have more small value which has done better
2) I don't think they include transactions costs and remember smaller stocks are less liquid and more expensive to trade
3) [IMO most important] I don't think backtesting is that accurate. If people did dump more money in small and value, then the market cap weighting of small and value would have been greater and the holdings in the market cap indexes would have changed. Also, active managers would have been looking at a different picture of where money is going too. I mean, by definition, investing will change the market. You aren't changing the market in backtests.

Dulocracy
Posts: 2291
Joined: Wed Feb 27, 2013 1:03 pm
Location: Atlanta, GA

Re: Glamour stocks and value tilt

Post by Dulocracy » Mon Nov 02, 2015 11:43 am

Another argument in favor of Mel's unloved midcaps?
I'm not a financial professional. Post is info only & not legal advice. No attorney-client relationship exists with reader. Scrutinize my ideas as if you spoke with a guy at a bar. I may be wrong.

boglerdude
Posts: 634
Joined: Fri Jul 10, 2015 1:28 am
Contact:

Re: Glamour stocks and value tilt

Post by boglerdude » Tue Nov 10, 2015 3:12 am

What are options for reducing or removing glamor stocks from a portfolio? Wellesley/Wellington? I might prefer not to gamble on high future earnings from Tesla, Facebook, or other trendy growth brand names.

Or do they make up such a small fraction of total stock market index that it's not worth changing? Not looking to beat/outsmart the market, just reduce troughs.

randomguy
Posts: 6499
Joined: Wed Sep 17, 2014 9:00 am

Re: Glamour stocks and value tilt

Post by randomguy » Tue Nov 10, 2015 1:58 pm

boglerdude wrote:What are options for reducing or removing glamor stocks from a portfolio? Wellesley/Wellington? I might prefer not to gamble on high future earnings from Tesla, Facebook, or other trendy growth brand names.

Or do they make up such a small fraction of total stock market index that it's not worth changing? Not looking to beat/outsmart the market, just reduce troughs.
You have a lot of options depending on your definition of glamour stock. There are all sort of value type funds out there. Will you do better? In retrospect sure. At the time who knows. How would your portfolio for the past 10 years done without the following glamour stocks (S&P 500 was 7.7) -
Google 14.8
Appl 30.9
Chipotle (only public about 9 years) ~32%
Amzn ~31%
And so on. Obviously these are the winner glamour stocks. Do they compensate for the loser ones? Depends on your screens.

User avatar
Taylor Larimore
Advisory Board
Posts: 27508
Joined: Tue Feb 27, 2007 8:09 pm
Location: Miami FL

Are Cap-Weighters throwing money away?

Post by Taylor Larimore » Tue Nov 10, 2015 2:17 pm

Arnott uses glamour stocks (twitter, Facebook, Priceline) to suggest that cap-weighters are throwing money away on overpriced shares.
boglerdude:

Morningstar is currently featuring an article which states:

"Just 10% of the large-blend funds that existed 10 years ago went on to beat the (cap-weighted) S&P 500."

How Well Have Large-Blend Funds Performed? It Depends

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

Quark
Posts: 1045
Joined: Sun Nov 01, 2015 5:32 pm

Re: Glamour stocks and value tilt

Post by Quark » Tue Nov 10, 2015 2:42 pm

In one of the recent Bogleheads conference videos, Gus Sauter and Bill Bernstein make good points about value tilts. My paraphrase:

Gus: investment theory tells us we shouldn't expect to get compensated for taking risk we could have diversified away. A value tilt is less diversified than the total equity market, so we shouldn't expect to get compensated for a value tilt.

Bill: everyone knows about the value premium and tries to exploit it. This greatly diminishes if not eliminates it. (Echoes of the Rekenthaler Rule he often cites - if the bozos know about it, it doesn't work any more).

Post Reply