A reader asked what I thought about mutual funds from Vanguard or Fido as substitutes for iShares MSCI EAFE (EFA).
Well, I am not a fan of EFA or any broad-based, total (or close to total) foreign market fund. A big reason to own foreign stocks is so they might offer diversification to a portfolio; zigging when the US is zagging is how I like to think of it.
I used a similar chart in a recent TSCM article. The time captured is mid-2001 to mid-2003 and it compares EFA to SPY.While I was not concerned about increased correlations during the recent dip I would be concerned about increased correlation the next time we have another market that looks like the one in this chart.
During the time period charted the Australia All Ordinaries (ticker ^AORD on Yahoo) declined by only 15%. The couple percentage points from EFA is not significant to me but the 15 percentage points from Australia is.
Canada, as measured by iShares Canada (EWC) correlated closely to the US until November 2002 before outperforming big time so I am not sure what to expect from Canada during the next bear market.
Each country offers different things to a portfolio but when all (or most) countries get lumped into one product a lot of the things offered by the countries individually get blended away as evidenced by the above chart.
To answer the reader's question I prefer the flexibility of the ETF format so I'll say EFA but you now know how little I think EFA brings to the table in terms of diversification. It does offer a chance for outperformance over the S&P 500 but that is a different issue.
The typical account I manage probably has exposure to ten or eleven foreign countries via individual stocks. In terms of priority I would say Australia is first followed by Ireland and then Norway. There are many ways to capture Australia in an ETF, while there is no ETF for Ireland there is a CEF (I use an individual stock, not the CEF) and for Norway there are only individual stocks unless you have direct access to the Oslo Exchange.
7 comments:
On the other hand, there are several well know analysts who'd disagree with this.
I can't find a link; Jeremy Siegel writes occasionally for either Kiplinger's or Smart Money magazine and noted that foreign market indexes do correlate with U.S. markets now, perhaps due to a worldwide growth in market liquidity. However, they have not necessarily correlated in the past. He does not expect close correlation for the future. Of course, it's Siegel, and he was making his usual effort to write only in very general terms.
Paul Merriman writes that he believes foreign indexes (via his preferred DFA funds or Vanguard funds) offer not only diversification via the fact that they're foreign indexes, but from the fact that they invest in hundreds or thousands of stocks (http://www.fundadvice.com/fehtml/bhstrategies/0108/0108a.html). His stats show that this lowers the risk of a portfolio.
As for myself, I did tons of research and still had performance problems with individual stocks. So indexes it is.
Roger makes a good point - the broader international indexes are increasing becoming more correlated to the U.S. market, but who knows how long that will continue. And how long will the countries Roger cites continue to be non-correlated to the big indexes?
Merriman's portfolio concepts are based on historical data - international and U.S. markets used to be a lot less correlated. After you eliminate firm, industry, and sector risk, you don't get much benefit from diversification if the indexes are closely correlated.
I really like the concept Dalio/Bridgewater presented but I haven't seen anyone put forward a feasible example of how an individual investor could put together such a portfolio in a tax advantaged retirement account. Here's the link:
http://www.bwater.com/PDFs/engineering_targeted_returns_and_risks_pmpt_060215.pdf
Also, finding non-correlated asset classes that provide long term returns similar to equities is a challenge. Commodities for example are non-correlated but their long term real returns are negative.
A tiny url to the Bridgewater paper:
http://tinyurl.com/334r43
This is a very intersting article. Like you, I wish it had a bit more detail.
To start to answer the reader's question: Vanguard recently started a FTSE All World ex-US fund (VFWIX). It's new enought not to have too many details, and I'm unclear how it differs from the Total International Stock Index (VGTSX).
FYI - Morningstar just did an article on the 10 funds that least resemble the S & P (looking at R-squared). Not all are good bets simply because of this...
http://news.morningstar.com/article/article.asp?id=190892
Paul
The precision of your portfolio allocation depends on its size and your time to manage it.
For example, I recently opened a Roth IRA and the $4,000 annual limit makes it rather silly and/or costly to split it 18 ways. Something like Vanguard VTI (total US ETF) and VEU (total world ex-US ETF) seem like the best way to diversify while keeping the costs down. As the balance grows to $25K or $100K it will make sense to split it more finely.
Similarly, for those who choose to do other things in life than study investing a lazy portfolio of VTI, VEU and bonds (AGG or BND) makes a lot of sense.
(On top of that, with so many investors seeking low correlations for diversity there are bound to be groupthink effects that undermine the original purpose--such as gold dropping with stocks a couple months ago.)
I think all stock investments' correlations are moving closer to 1.00. If/when the markets move down, they will all go down. Too much globalization. We are all connected in some way. The days of really diversifying by buying diff. correlated items are over. Mainly because EVERYBODY does it. You want to diversify for safety? Just buy bonds.
Post a Comment