How much diversification is over-diversification?

[Olav23]

I'm about to put about 100k into a few etfs. I feel its time ot make the leap. I'm trying to decide how much diversification is "over-diversification". I guess everyone has a different sweet spot, but I'm thinking about this:

In %s:

Large:
15 Value
10 SP500

Small:
20 Value

Intl:
25 MSCI EAFE
10 Emerging Markets

10 REIT
10 Oil/Natural Resources (as the commodity portion)

Does anyone see any asset classes that are either out of whack in %, or completely forgotten that could put a bit more juice in my port? Keep in mind its 100k so I can't diversify it much more before the transaction fees become prohibitive.

BTW, I'm 28 so I have quite a few years ahead as well as a relatively high tolerance for risk to achieve higher returns.

Thanks!

[brewer12345]

Two comments:

1) I would not be willing to tie myself to value that much. Growth and value go in and out of favor over time. Value has been in favor for several years and growth has been on the skids. You really want to pile on the ship that has already sailed so far, so fast? I would just buy index funds.

2) I would NOT be interested in just buying oil. I think a diversified commodity index delivers the hedge/inflation protection you are looking for but with much less volatility than oil alone. Something like PCRIX looks much better to me than QRAAX.

[wildcat]

Just my comments and I differ from Brew (although what he said is correct)

I stick with value index funds because A) it reflects my investing approach - I would rather be at the lower end of the valuation spectrum than higher end; growth index P/E 23+ and value index around 14-15 plus better 2% or so yield but heavy in financials and B) it, meaning a value approach, will outperform over time and my time horizon is very long.

Agree stick with index funds although I have one active

Like I said what Brew said is right but to me valuation determines returns and the lower the better, i.e. there really should be no distinction between growth and value and it should all boil down to how much you pay for future ***. I would rather see some of the growth index companies fall into the value index rather than buy into the growth index.

Agree diversified commodities would be better and oil play would be more suitable for play money

[Bookm]

Good points made thus far. What you have listed is a great start towards holding a diversified portfolio. It's close to what William Bernstein suggested in his initial online journal entry on portfolio composition - minus 3 or 4 sub-classes. Here's what Bernstein suggested if one wanted to hold a highly complex and diversified port:

Quote:

A fixed mix of U.S. stocks, bonds, and foreign stocks (one third each Vanguard European, Pacific, and Emerging Markets Index funds), plus foreign bonds, foreign small stocks, US small value stocks, REITs, precious metals, natural resources, utilities, junk bonds, etc.

As long as you aren't adding funds that overlap your other funds, and your adding ones that vary in style and/or size, an argument can be made for their inclusion, IMHO. Just ask ben.*

Which energy/nat'l resource ETF were you looking towards? The most diversified would be VDE, with approx. 120 companies, while XLE holds a more concentrated group, with about 30 companies. In the middle is iShare's IYE. VDE is one of Vanguard's VIPERs, is rather new and doesn't have a long track record (a year maybe?). On the mutual fund side, PCRIX is a little broader, in terms of various other commodity derivitaves. It doesn't just play with oil and gas, from what I know about it.

Bookm

[preben]

He,he - bookm; you know me so well!

I think it is a super portfolio but if we HAVE to tinker :
I would want SAME foreign exposure as US (but I AM a crazy foreigner....)
I would give US Large blend and US (large) value same weight
I would want to add some precious metals/mining: VGPMX (sweet 13% return YTD...nice...)

In case you look at fixed income at later stage; do not forget some foreign bonds (PFUIX/PEBIX or GIM).

And I agree that PCRIX is the way to go for commodities in general.

Cheers!

[lazyday]

What Brewer said.

Also, I'd consider BRLIX instead of s&p500etf for large us. Perhaps lower total costs (your commissions+fund er+hidden costs the fund incurrs), less of a buy high sell low fund, and might get an internal rebalancing yield in the fund itself.

For EAFE, if you plan to use a very cheap broker and rebalance infrequently, could consider more diversification: VUG and VPL. They might not always zig/zag together, and total fund ER is lower than the eafe etf, tho your commissions will be higher of course.

If you are US based, I'm not sure you need more foreign than you list. You have a high exposure to risk of a rising dollar, if your commodity portion is PCRIX. (25+10+apx10=apx45%)
I have a similar issue--my expenses are in $, but my portfolio has over 50% exposure to currency risk.

[unclemick]

Rats!

I keep thinking - I should get single fund(Target or similar) to bookend Ben - pun intended.

BTY - if I were sitting overseas somewhere - knowing myself - I would morph toward Ben's direction - for currency/various worldwide inflation differences, etc. I remember the Terhorst's early experiences during Argentine inflation - but I understand they're back - in Argentina that is - building a house/homebase.
.

[Olav23]

First of all, thanks everyone for all the great pointers and help. I've looked into BRLIX, but unfortunately its a "transaction fee" fund through my broker. Which means something like 30-40$ to buy. Same with PCRIX, but since that had high diversification potential, and this is a lump sum, I thought I might go in on some PCRIX just to hold onto it.

My point was to hold as many ETFs as possible, and steer clear from the funds, as my fund "supermarket" has very few low cost funds, the vanguards of the world, etc. Closest I can get are Dreyfus for no transaction fee, which are around .2-.5% for the bigger funds.

I've also read lots of information regarding the "Value Premium" so thats why I wanted to gear it a bit towards value. I know what you are saying, assets zig/zag and Value has been hot lately over Growth, but I have read studies that go wayyy back, before the 90s. I think the research showed that you get a few % more out of Value for the past hundred years, if I am reading the report correctly. I can dig it up if anyone is interested. And I don't think the argument that it's "sailed too far". The Value index is a passive index and not like an actively managed entity. It does not act like a single stock, where as it becomes in favor, the price goes up and you are paying more for the same earnings. As soon as an equity outperforms its breatheren in the index, it would most likely transition to "growth" and thus be removed from the index being that its a simply mechanical system of P/E or P/B or whatever. So, you're left with nothing but the laggards. I would intuitively make that argument, anyway.


After your opinions, perhaps this is a bit better::

ETF Portfolio 2.0

Large: 25%
Large Blend: 15 SPY or IVV or IWB or VV (S&P 500 or Russell 1000 or Vanguard LC)
Large Value: 10 IVE or VTV (S&P 500 Value or Vanguard Value)

Small: 20
Small Blend: 10 IWM (Russell 2000)
Small Value: 10 IWN or VBR (Russell 2000 Value or Vanguard SC Value)

Intl: 35
Emerging: 10 VWO
EAFE: 25 EFA or VGK+VPL (EAFE or Europe + Pacific)

Commodity: 20
Hard comm: 10 PCRIX
Real estate: 10 RWR or ICF or VNQ (Wilshire REIT or Cohen Steers Realty or Vanguard Reit)

I think adding too many more ETFs with only 100k will get a bit commission expensive. I think its about 12$ or so per trade.

I know I'm never going to get complete buy-in from EVERYONE here, but hopefully I can get a lowest common denominator

Thanks again!

[wildcat]

Quote:

I think the research showed that you get a few % more out of Value for the past hundred years

"Yesh, you are correct sir"

[preben]

Instead of BLRIX you can consider the ETF RSP which is equal weighted SP500 which I find very interesting. Will give it more value and mid+small twist. Cheers!

[Olav23]

Wow.. Did some research. That's certainly an interesting twist on the SP500 methodology. How exactly would you place RSP in a portfolio? Replace the Large Cap value and growth with it? It seems like a good replacement (almost) for LC + SC since its equal weighted among all constituents, which can include smaller companies.

I'm intrigued but not sure exactly where it "fits". Would you recommend replace the SPY with this and still adding the LargeCap Value? or replace the whole LC level?

It also looks relatively tax inefficient with a 50% turnover ratio, compared to the SPY turnover of 2%. Maybe this should go into the retirement account?

[preben]

I don't pay US (capital gains taxes) but naturally the re-balancing will mean more turn over so non-taxed account sounds wise.

I would not look down on a RSP replacing both the blend+value etf. But it depends on how much value twist you want. Then add some small value via IJS or similar.

Cheers!

[wildcat]

Interesting Ben....

I read an article some months back in a journal that basically said there are probably better ways than the current methods used to determine the S&P 500 index. The author's research was based more on financial measures than market cap and it outperformed over the long run. It did not suggest equal weight index per se but your recommendation is one potential answer to the "flaws" of current index measurements.

[preben]

RSP follows my train of thought being; I have no clue which stock will do best - so why should I overweigh the ones that was yesterdays BIGGEST winners? Some argue that the market is showing its choice of winners (of the future) through the given market cap, and that the "others" therefore are "riskier" - I don't see that with RSP as all have been through stringent screening, and have reached a certain size in order to even be included in the SP500 at all.

I see it as"riskier" to let 10-15 stocks control most of my index funds returns....

Another twist; combine RSP with VXF (vanguard extended market I.e. EXCLUDES the SP500) as there is not NEARLY the same topweight in the VXF, and you will have a resonably equal share of almost ALL the stocks in the US market...(VXF in NOT equal weighted - but the top 10-15 are much less influential market cap wise than the top 10-15 in the SP500).

Just throwing ideas around here! Cheers!

[Bookm]

Quote:

Originally Posted by wildcat Interesting Ben.... I read an article some months back in a journal that basically said there are probably better ways than the current methods used to determine the S&P 500 index.* The author's research was based more on financial measures than market cap and it outperformed over the long run.* It did not suggest equal weight index per se but your recommendation is one potential answer to the "flaws" of current index measurements.

Greetings wildcat, I think you're referring to Robert Arnott's recent study which was published in the Financial Analyst Journal. Yesterday in the "peal oil thread" I mentioned Arnott and PIMCO's attempt at recreating a equal weighted index mutual fund thru the use of derivatives. His study is discussed here a few months back at Mauldin's site.

Arnott and Sauter also discussed their indexing philosophies, which was summarized in this M* article.

Bookm

[wildcat]

Ahh there it is and thanks for the links