Total Stock Index v. Large/Small/Value/Growth routine

[Rich_in_Tampa]

Is there any good evidence to show that the usual large value/large growth/small value/small growth slice and dice rebalancing routine really outperforms a Total Stock index fund over the long haul?

Wondering if it's worth the hassle. My searches have turned up lots of advice but little data. I do understand that small and value have outperformed the market for a while, but now they say growth will surge, etc. etc. It's beginning to feel that just owning the whole market for a long time is as good a way to go as any.

(this assumes you have your fixed income, cash, and international already in place - just talking about the stock piece)

[wabmester]

The only evidence is historical back-testing, so it depends on what you want to demonstrate. There is pretty good evidence for the value premium. Slightly less convincing evidence for the small premium. Not sure what else you're assuming.

The argument for a cap-weighted total market approach is that the markets are efficient at allocating capital, so let them do it for you without trying to outguess the market about which slice or sector will outperform in the future.

The argument against a cap-weighted approach is that cap-weighting overweights overvalued "growth" stocks and underweights value stocks.

The value/small premium was discovered in the 1990's. The big question about such "premia" is whether they'll continue to exist once discovered. Since everybody is chasing value and small these days, my guess is no.

[audreyh1]

I'm a firm believer that simpler is better. It may initially seem than rebalancing isn't that much work, but it requires taking action, and added up over many years it becomes a hassle. Much better if you can find something that "runs itself".

Sure, theoretically you might be able to "optimize" a teeny amount (i.e. boost your returns very slightly each year), but is it really worth the extra work? And then you might not even be ahead anyway - it might be a lot of trouble for naught or even underperform.

If I had it all to do over again, I would have just put everything in DODBX and never looked back....

So if your instincts are saying maybe it's too much trouble - listen to your instincts!

Audrey

[REWahoo]

Doc, I'm firmly in Audrey's camp...on all counts. Too much complexity spoils the ambiance of retirement.

[Spanky]

Total market, cap-weighted index, is more tax-efficient and simpler than a slice-and-diced approach. One suggestion from Rick Ferri is to implement a 50% total market and 50% small-cap value mix.

[unclemick]

Yep - slowly getting there - went to one Target Retirement Series in 2006 - from Lifestrategy mod, REIT Index, Small Cap value and High Yield Corp.

Alas - still 15% in individual dividend stocks.

heh heh heh - perhaps someday? Somewhat er ah more simpler. Right?

[lswswein]

Below is the graph from DFA saying Small and Value matter. Also pay attention to the Std Deviation numbers in the bottom row which say that in most cases the volatility increases with returns except in the case of smal growth, where returns don't increase. Anyway this is the backtested data and it is your call as to this stuff continues into the future.

-h



Ref: http://www.dfaus.com/philosophy/dimensions/

[lswswein]

Also a couple more notes:

a) Just as your stocks won't return 8% every yr linearly but swing up and down, the small and value premium does keep swinging up and down. So keep in mind that after you move your portfolio towards small and value you might see a period like 1995-2000 when large growth completely obliterated every other asset class around it. You have to be able to hang in with your skew away from total Mkt for a long time, even lasting a couple of decades.

b) It might be a good idea to keep Total Mkt as the core holding of your portfolio. This is important if you compare yourself with your brother in law A lot of people compare themselves against what they hear from people around them and having Total Mkt as your core would help you in these comparisons. A small amount of small and value skew is recommended if the portfolio is all you have funding your retirement. If you have a lot of assets and the risk and volatility does not bother you at all then you should look into a bigger small & value skew.

c) all this is covered in the Intelligent Investor

-h

[samclem]

I'm overweighted in small and value. I'm sure you've probably seen all the significant info I've seen, so I don't have any new info to provide. Here's my reasoning:
-- History: Maybe it won't continue, but small and value have done well in the past. 1% to 4% advantage annually compounded over decades is a LOT. I'm betting on a continuation, even if it's not as pronounced.
-- Growth-mania: We've seen the tech-stock craziness, with valuations well out of line with earnings and depending on the "greater fool" to make money. When that happens again, I want my money in the stocks that will be damaged the least during the fall.
-- Dividends: Each dollar invested in value stocks buys more dividends than the same dollar invested in growth stocks. Though taxed at a higher rate, I like the steadiness of dividends compared to the ups/downs of stock prices.
-- I like bargains in my personal buying decisons, I think buying out-of-favor stocks is much the same thing.
-- If I were buying individual stocks, I'd be in the Ben Graham camp. Buying value index funds is the lowest-cost way to get a rough approximation of this across a wide stock diversification.

Anyway, that's the reasoning. I do have some Total Market, and it will be a do okay when growth is doing better than value.

YMMV

[laurencewill]

Ditto on Samclem and the others re: value - for me as a young dreamer, I just stick my value/dividend slice in my retirement accounts to avoid tax issues. Even as a retiree, couldn't you just have money in the IRA be in those accounts since they are going to get taxed as regular income as you withdraw anyway?

[Rich_in_Tampa]

Originally Posted by wab

The argument against a cap-weighted approach is that cap-weighting overweights overvalued "growth" stocks and underweights value stocks.

That's what I thought, but here's how Morningstar x-rays VTSMX:

Value Core Growth
25 27 21 Large
6 7 6 Medium
3 3 3 Small



So much for total stock being grossly growth weighted, or is Morningstar wrong?

[lswswein]

Portfolio1 vs Portfolio3 demostrates the excess returns and decrease in volatility that you gain by adding value and small caps. Note that this is rebalanced monthly which is not practical for the DIY individual investors. Also do note that Portfolio1 is infact Total Mkt not Dow Jones 30. So you gain 1.5% in annual returns and a decrease in SD of 0.5% It is left as an excercise to the reader as to what that adds to the portfolio in actual returns since all this is a fictional index with no expense ratios added. Small value might cost you so much that it might take away all the excess return. Think about it and do the math before taking a decision.

-h



Ref: http://www.dfaus.com/philosophy/structure/

[runchman]

I'm a slice-and-dice DFA guy, but I can see the wisdom of the total market approach from a slightly different perspective:

It will probably make you less likely to be tweaking with (and potentially screwing up) your portfolio. You avoid the thinking of "Well maybe I should add commodities to my portfolio so they amount to 3.5% of the total, and decrease international value", "REIT's are only 10% of my allocation, maybe I should bump that to 15%......"

If keeping it simple tends to keep your hands off it, probably a good idea.

- John

[wabmester]

Originally Posted by Rich_in_Tampa

That's what I thought, but here's how Morningstar x-rays VTSMX:

Value Core Growth
25 27 21 Large
6 7 6 Medium
3 3 3 Small



So much for total stock being grossly growth weighted, or is Morningstar wrong?

The issue isn't the percentage weighting of various classes. The argument against cap-weighting is that *by it's nature* cap-weighting will put more GOOG in your portfolio than, say, F. GOOG is arguably overvalued, and thus overrepresented by cap-weighting. F is a value stock, so cap-weighting will underweight by design.

But these are the weightings that the market says is right. Who are we to argue?

[Rich_in_Tampa]

Originally Posted by wab

The issue isn't the percentage weighting of various classes.

Wab, I don't follow.

From the style box, I take it that 25% of VTSMX value is comprised of value stocks. In that sense, it is weighted against growth and toward value. I could not find the answer on their site, but maybe the style box simply lists the percent of companies in each cell, regardless of market cap?

[lswswein]

Originally Posted by Rich_in_Tampa

That's what I thought, but here's how Morningstar x-rays VTSMX:

Value Core Growth
25 27 21 Large
6 7 6 Medium
3 3 3 Small



So much for total stock being grossly growth weighted, or is Morningstar wrong?

Hi Rich
What you are seeing is what happens when there is a sustained period of value overperforming growth. As value stocks outperforms growth, the cap sizes of the value stocks increases and also their weighting in the total index also increases.

-h
p.s: more numbers and data here:
http://diehards.org/forum/viewtopic....ighlight=#6213

[wabmester]

Originally Posted by Rich_in_Tampa

From the style box, I take it that 25% of VTSMX value is comprised of value stocks. In that sense, it is weighted against growth and toward value. I could not find the answer on their site, but maybe the style box simply lists the percent of companies in each cell, regardless of market cap?

All that tells you is the current percentage breakdown. The question is whether or not it's the "right" weighting. The argument against it being "right" is that cap-weighting by it's nature will give you too much overvalued stock and too little value stock. It doesn't tell you anything about the relative market percentage breakdown of value vs growth.

Anyway, I couldn't find a more recent chart, but here's a chart of the relative outperformance of value over the years.



It's not shown here, but recent outperformance is the highest since WWII. That's why I think small/value may be due for a period of underperformance, but who knows.

[Rich_in_Tampa]

Excerpt from the Diehard forum:

This reminds me of a John Bogle interview where he said something to the effect "Half the stocks in the market are 'value' and half are 'growth', but the problem is nobody for sure can tell which half is which going forward".

[lswswein]

Originally Posted by Rich_in_Tampa

Wab, I don't follow.

From the style box, I take it that 25% of VTSMX value is comprised of value stocks. In that sense, it is weighted against growth and toward value. I could not find the answer on their site, but maybe the style box simply lists the percent of companies in each cell, regardless of market cap?

Value is defined as the higher 50% of the Price/Book ratio and Growth is defined as the lower 50%. (This is a crude definition and varies between how it is defined by Morningstar, DFA and Vanguard).

There is no rule saying that Value should be 25% of Total Mkt. Total Mkt just holds all the stocks based on their cap weighting. Around 1999, all the growth stocks had higher capitalization and ended up being a larger % of total mkt. If you looked at the top 10% in 1999, I think 7 of them were growth stocks. Right now 7 of the top ten total mkt stocks are value stocks.

-h
A spreadsheet I generated for the top 10 holdings of s&P500


Ref:
http://tinyurl.com/yuoefk

[wabmester]

Good recent interview with Fama and French in which they fight about whether value is a risk premium or a consistent mispricing:

link and discussion at raddr's