Why not a larger small cap allocation?

[Tyr0ne]

I recently came across this article (note the two charts), and although I completely understand how past performance doesn't predict future returns, the concept of the small cap premium also makes sense to me. My time horizon for retirement is around 45 years. Perhaps I don't understand enough about cap-weighted asset allocations, but with such a long-term horizon, I am tempted to increase my small cap holdings from ~15-20% to ~25-35% (edit: %'s of my equity holdings). Note: this would be strictly for retirement money. My emergency fund and house savings are all CD's and ST bonds.

I am assuming this would be a very risky and potentially grave mistake, and and that there is a perfectly good reason NOT to do this. What do I need to consider when thinking about such a move? Any book suggestions (perhaps Ferri's All About Asset Allocation)?

[mptfan]

I think it is a perfectly reasonable thing to do. It will make your portfolio slightly more volatile, but with such a long time horizon, I think it will result in higher returns. Also, let's keep this in context... the "small cap" companies in the Vanguard small cap index fund have an average value of $1.7 billion dollars each. Small is a relative term.

I picked one of the companies at random from the top 10 holdings of the small cap fund.... Jarden Corporation....

"Jarden Corporation is a leading provider of a diverse range of consumer products with a portfolio of over 100 trusted, quality brands sold globally. Jarden operates in three primary business segments through a number of well recognized brands, including: Outdoor Solutions: Abu Garcia®, Aero®, Berkley®, Campingaz® and Coleman®, ExOfficio®, Fenwick®, Gulp!®, K2®, Marker®, Marmot®, Mitchell®, Penn®, Rawlings®, Shakespeare®, Stearns®, Stren®, Trilene®, Völkl® and Zoot®; Consumer Solutions: Bionaire®, Crock-Pot®, FoodSaver®, Health o meter®, Holmes®, Mr. Coffee®, Oster®, Patton®, Rival®, Seal-a-Meal®, Sunbeam®, VillaWare® and White Mountain®; and Branded Consumables: Ball®, Bee®, Bernardin®, Bicycle®, Billy Boy®, Crawford®, Diamond®, Dicon®, Fiona®, First Alert®, First Essentials®, Hoyle®, Kerr®, Lehigh®, Lillo®, Loew Cornell®, Mapa®, NUK®, Pine Mountain®, Quickie®, Spontex® and Tigex®. Headquartered in Rye, N.Y., Jarden ranks #371 on the Fortune 500 and has over 23,000 employees worldwide."

This is a "small cap" company on the Fortune 500 with 23,000 employees. I have no interest in this company at all, it was a random pick, just to illustrate what small cap means. And the small cap fund has over 1,700 companies like this (although probably most are smaller). You are doing fine.

[Aptenodytes]

I'm assuming these percentages are of your equity holdings, not the total portfolio. Last I looked small-cap was about 9% of the total stock market, so at 15-20% you are already significantly tilted.

Moving to 25% seems perfectly reasonable (that's about what I'm at). But before you start changing things I'd suggest you revisit all your tilts and make sure you are comfortable with them for the long haul -- what's your value tilt, for example?

You don't want to be bouncing around on these things. And there's no way to know the single optimal set of tilts. You make your best guess and stick with it.

I set my tilts about two years ago after a lot of reading, thinking and soul-searching. Now I basically don't think about them.

[gt4715b]

I think you should become familiar with the French Fama model that is the basis for the idea of overweighting small cap stocks (and value stocks).

A really good thread to start with is here:
http://www.bogleheads.org/forum/viewtopic.php?t=7353

This will enable you analyze what your planning to do in a more meaningful way, i.e., analyzing load factors instead of percentages which can yield different results depending on which index family you use.

[rkhusky]

I picked one of the companies at random from the top 10 holdings of the small cap fund.... Jarden Corporation....

Morningstar considers 23% of Vanguard's Small Cap Index to be mid-cap. Jarden probably fits their definition of mid-cap and interestingly, as of 11/30, it is no longer in the top 10. Even more interesting, only two of the top 10 on 9/30 are in the top ten as of 11/30, which shows how much churn there can be in the small category.

[Simplegift]

What do I need to consider when thinking about such a move?

The most important thing, IMO, is the strength of your conviction in the small cap premium. Large and small-cap stocks operate in fairly long-term cycles in which large outperforms small, or vice versa, for years at a time (chart below). The better you understand the behavior of the small premium, the less inclined you'll be to bail out of your small-cap tilt when they underperform for a decade — or some new research catches your eye that questions the whole idea. A long-term, small-cap tilt can be perfectly sensible, but it's best to understand as much as you can before committing yourself. Cheers.

In this chart, large-caps are outperforming when the line is rising and small-caps are outperforming when the line is falling:


Source: Seeking Alpha

[maddyken]

With 45 years to go you are contemplating exactly the opposite approach I would take if I were in your situation.

I would adopt a highly diversified conservative or moderate portfolio to reduce risk and take advantage of compounding over so many years to grow my portfolio, rather than relying too heavily on the small cap premium to grow my portfolio.

[Calm Man]

I have read a whole bunch of posts and scholarly information from Larry, the Fama French papers, etc. There has indeed been a premium for these SCV stocks for many decades in the past. Presumably a premium comes with risks. I am surprised that the computers and smart money haven't arbitraged it away a little. I guess we will know in 40 years if the premium persists for a $ invested by you today (I won't -- I'll be dead). You already have put more of your assets in SCV than market weighting and want to increase that. If you just stay with TOtal Stock Market you guarantee yourself market returns. If SCV outperforms you lose that overperformance bonus. But there is no black swan event likely in which the other 60% of the market crashes and your SCV bails you out. A more likely potential LARGE underperformance would likely be if SCV significantly underperforms TSM. Frankly if I were you I would buy the TSM and forget about it -- no rebalancing, no guessing about percentages and more time to spend on things that can enhance your life. Good luck whatever you do.

[grap0013]

I recently came across this article (note the two charts), and although I completely understand how past performance doesn't predict future returns, the concept of the small cap premium also makes sense to me. My time horizon for retirement is around 45 years. Perhaps I don't understand enough about cap-weighted asset allocations, but with such a long-term horizon, I am tempted to increase my small cap holdings from ~15-20% to ~25-35% (edit: %'s of my equity holdings). Note: this would be strictly for retirement money. My emergency fund and house savings are all CD's and ST bonds.

I am assuming this would be a very risky and potentially grave mistake, and and that there is a perfectly good reason NOT to do this. What do I need to consider when thinking about such a move? Any book suggestions (perhaps Ferri's All About Asset Allocation)?

I agree. Notice how even when small cap value underperforms that it is not by much? Watch out though, if you change strategies often you will quickly trail the overall market by a large margin. Once you go small cap value you never go back.

How disciplined are you in other areas of your life? If you are very diciplined, this strategy may be for you.

[grap0013]

I have read a whole bunch of posts and scholarly information from Larry, the Fama French papers, etc. There has indeed been a premium for these SCV stocks for many decades in the past. Presumably a premium comes with risks. I am surprised that the computers and smart money haven't arbitraged it away a little. I guess we will know in 40 years if the premium persists for a $ invested by you today (I won't -- I'll be dead). You already have put more of your assets in SCV than market weighting and want to increase that. If you just stay with TOtal Stock Market you guarantee yourself market returns. If SCV outperforms you lose that overperformance bonus. But there is no black swan event likely in which the other 60% of the market crashes and your SCV bails you out. A more likely potential LARGE underperformance would likely be if SCV significantly underperforms TSM. Frankly if I were you I would buy the TSM and forget about it -- no rebalancing, no guessing about percentages and more time to spend on things that can enhance your life. Good luck whatever you do.

Why hasn't the equity premium been arbitraged away?

[Calm Man]

I have read a whole bunch of posts and scholarly information from Larry, the Fama French papers, etc. There has indeed been a premium for these SCV stocks for many decades in the past. Presumably a premium comes with risks. I am surprised that the computers and smart money haven't arbitraged it away a little. I guess we will know in 40 years if the premium persists for a $ invested by you today (I won't -- I'll be dead). You already have put more of your assets in SCV than market weighting and want to increase that. If you just stay with TOtal Stock Market you guarantee yourself market returns. If SCV outperforms you lose that overperformance bonus. But there is no black swan event likely in which the other 60% of the market crashes and your SCV bails you out. A more likely potential LARGE underperformance would likely be if SCV significantly underperforms TSM. Frankly if I were you I would buy the TSM and forget about it -- no rebalancing, no guessing about percentages and more time to spend on things that can enhance your life. Good luck whatever you do.

Why hasn't the equity premium been arbitraged away?

Maybe the arbitragers are busy with other matters. Or, maybe it has been. We will only know if we prospectively set some date after which it is followed. I am not much of a betting man as it's a losers game but I would take some nice odds if offered them that 5 years from this second SCV will be statistically indistinguishable or worse than TSM.

[maddyken]

In the long run don't the returns on stocks reflect earnings?

[maddyken]

Last night they showed the episode of Seinfeld in which George convinces Jerry to go in on a stock with him, Cendrex.

After several days of losses Jerry bails and locks in a $1,500 loss, George decides to stick it out.

A few days later, out of curiosity, Jerry and his girlfriend check the stock in the newspaper and it's way up.

Jerry can't believe it, his girlfriend reminds him she told him stocks fluctuate, Jerry responds "I just got fluctuated out of fifteen hundred dollars!"

[nirvines88]

Last night they showed the episode of Seinfeld in which George convinces Jerry to go in on a stock with him, Cendrex.

After several days of losses Jerry bails and locks in a $1,500 loss, George decides to stick it out.

A few days later, out of curiosity, Jerry and his girlfriend check the stock in the newspaper and it's way up.

Jerry can't believe it, his girlfriend reminds him she told him stocks fluctuate, Jerry responds "I just got fluctuated out of fifteen hundred dollars!"

Classic scene! Hilarious how his girlfriend tells Jerry the market goes up and down like Jerry didn't understand that haha. Also, gotta love George at the end of the episode, smoking a cigar and thinking he's a genius while acting like a king.

[maddyken]

The ending was great, especially when George intimates his next hot tip...a business making robot butchers.

[nisiprius]

I think you should read this article: The Small-Company Advantage: Fact or Fiction?

And this one: Rethinking Small Caps

The existence of a small-cap effect is not the universally accepted fact that some people would have you believe.

[Calm Man]

Thank you nisiprius. I am not enough of an expert to analyze and rate the respective methodologies that went into both showing that there is a positive small cap effect and disproving it (and even showing the opposite). I feel like this is comparable to a review of the medical literature for whether vitamins help or not (and then one discovers some studies showing they hurt). I have puzzled over why a premium should persist and not be taken advantage of by the hedge funds, sharpies, etc. So I am very happy to see that there is at least some evidence that a small cap effect may not exist to add to the evidence that is does. For example, I know nothing about DFA funds except that a big part of it is this 3 part Fama French model that somehow gets on to overweighting small cap value I believe and is a big basis for their hauling in assets from advisors who have to be trained, approved and evaluted periodically for the right to be able to invest client funds in DFA. My suspicion is that proponents of the small cap effect will never give it up nor vice versa for opponents. I have no idea how this holds up in real world investing vs a boring Total STock Market index fund. This helps me see that there are the alternate views, meaning that it probably doesn't matter.

[Call_Me_Op]

I have read a whole bunch of posts and scholarly information from Larry, the Fama French papers, etc. There has indeed been a premium for these SCV stocks for many decades in the past. Presumably a premium comes with risks. I am surprised that the computers and smart money haven't arbitraged it away a little.

Computers do not "arbitrage away" risks or returns. What they can do is close the gap when risk and reward are out of balance (by exploiting market inefficiencies). Otherwise, why wouldn't computers "arbitrage away" the returns of stocks compared to bonds? SCV stocks are inherently riskier than large-cap stocks, and thus have higher expected return.

[Simplegift]

My suspicion is that proponents of the small cap effect will never give it up nor vice versa for opponents. I have no idea how this holds up in real world investing vs a boring Total Stock Market index fund. This helps me see that there are the alternate views, meaning that it probably doesn't matter.

Be aware, one can also find papers that question whether the equity risk premium itself even exists — so just the decision to invest in the "boring Total Stock Market Index" requires a certain amount of research and conviction on the part of every do-it-yourself equity investor.

It's not quite as easy as saying, well, there's opposing views, so I don't have to investigate the issue for myself.

[nedsaid]

I have for years overweighted my portfolio towards mid-caps and small-caps. It has worked for me. Mid-Caps give you most of the return of small caps with somewhat less risk.

I believe in the small cap premium. The academics believe in it, John Bogle does not.

Even if there was no premium, the large part of the market and the small part of the market behave differently. Sometimes the small does better than large and sometimes the other way around. Being overweighted in small cap stocks would reduce volatility. There is a chart in the Bogleheads Guide to Investing that shows a 69% correlation between
the Small Cap index and the S&P 500. Small Cap Growth Index correlates at 61% and Small Cap Value Index correlates with the S&P 500 at 59%. Hmm. (I wonder if the 69% is a misprint).

That being said, you will do fine with a Total Stock Market Index for your US Stock Investments. There are a lot of threads on this forum where the Indexers duke it out with the slice and dicers. I slice and dice myself but I think pure indexing is a very sound approach.

Slice and dice is not a hill that I want to die on.

[Default User BR]

I have read a whole bunch of posts and scholarly information from Larry, the Fama French papers, etc. There has indeed been a premium for these SCV stocks for many decades in the past. Presumably a premium comes with risks. I am surprised that the computers and smart money haven't arbitraged it away a little.

Computers do not "arbitrage away" risks or returns. What they can do is close the gap when risk and reward are out of balance (by exploiting market inefficiencies).

Indeed, if small stocks were riskier but provided no extra return, then you could reduce your risk without affecting return through avoiding them altogether.


Brian

[grayfox]

I recently came across this article (note the two charts), and although I completely understand how past performance doesn't predict future returns, the concept of the small cap premium also makes sense to me.

This article has tables that shows historical return and worst year for four corners of market SCV LCV SCG LCG, for 6 periods:

1y 2010
5y 2006-2010
10y 2001-2010
20y 1991-2010
50y 1961-2010
80y 1931-2010

SCV wins in every period from 1-year to 80-years. Well that about covers everything right?
Not so fast. All the end dates are 2010. Someone posts this chart.



So the periods in the paper all end in 2010 after a 10-year run of outperformance by SCV? It also likes like some of the start dates were at the end of periods of underperformance by SCV--at least 2001 and 1991.

So he is just cherry picking (maybe accidently) periods when SCV outperformed.

He should re-title the article, "How to be mislead by historical data" How about making 1991 or 2000 the end date?
After a long run of outperformance, choosing the current date as the end date is going to make it look good. Back in 1999, the same thing was said about growth stocks.