Tuesday, January 23, 2007
How To Think About Emerging Markets?
I stumbled across this post about emerging markets. Most of it is an interview with an investment advisor named Jeffrey Troutner who is a DFA guy.
Troutner is questioning the value of investing in emerging markets due to increased correlation to US markets and returns over the last 12 years that have not justified the volatility taken on to get those returns.
The benchmark for these conclusions was a DFA emerging market index that, according Troutner in the interview, has outperformed the MSCI Emerging market index.
A tipping point for Troutner's concern was the correction of both the emerging markets and the S&P 500 last spring. Emerging markets did not zig when the S&P 500 zagged, he is right. Also supporting his case, but he did not mention it, was the Asian contagion back in 1997. The magnitudes may have been different in each instance but during these times of panic (or if you prefer, stress tests) emerging markets did not offer a low correlation.
There are several issues here that cause me to view this much differently.
In no particular order; making this decision based on six week or 2 month stress tests that pop up every now and then seems to contradict the long term approach that DFA people believe in. I am a believer in blending different things together but in times of crisis I would not expect another stock market to provide something positive and in fact they don't. Gold does this, inverse funds can do it, certain parts of the bond market can do this as can defense (not defensive) stocks, depending on the nature of the crisis. I conceded this on May 30, 2006 for TSCM.
During normal rises and declines, emerging markets can add diversification but it depends how you invest in them. The broader an emerging market index is the more likely that index will correlate with the US market. A few days ago I wrote a post comparing Turkey and Malaysia that pointed out the low correlation between the two due to vast differences between them. There is potentially some offset between the two which during a crisis could be a good thing.
An entire broad-based index with all sorts of countries zigging and zagging is going to dilute some of portion of the effect of investing in emerging markets. This makes the case of investing in countries or regions as opposed to something like EEM. I realize that many people may not be comfortable with a single country but the argument that broad-based emerging market funds is not a great way to go has merit.
If you look at the correlation between SPY and just about any single country product you will see a lower correlation than the number for EEM; I realize I am stating the obvious.
However as a investment advisor I don't think I could justify ignoring the space to a client because the broadest vehicles out there may not deliver. Top down analysis of a country combined with a bottom up pick of a country fund or, heaven forbid, an individual stock seems like it should be within the wheelhouse of someone with my job title.
Troutner is questioning the value of investing in emerging markets due to increased correlation to US markets and returns over the last 12 years that have not justified the volatility taken on to get those returns.
The benchmark for these conclusions was a DFA emerging market index that, according Troutner in the interview, has outperformed the MSCI Emerging market index.
A tipping point for Troutner's concern was the correction of both the emerging markets and the S&P 500 last spring. Emerging markets did not zig when the S&P 500 zagged, he is right. Also supporting his case, but he did not mention it, was the Asian contagion back in 1997. The magnitudes may have been different in each instance but during these times of panic (or if you prefer, stress tests) emerging markets did not offer a low correlation.
There are several issues here that cause me to view this much differently.
In no particular order; making this decision based on six week or 2 month stress tests that pop up every now and then seems to contradict the long term approach that DFA people believe in. I am a believer in blending different things together but in times of crisis I would not expect another stock market to provide something positive and in fact they don't. Gold does this, inverse funds can do it, certain parts of the bond market can do this as can defense (not defensive) stocks, depending on the nature of the crisis. I conceded this on May 30, 2006 for TSCM.
During normal rises and declines, emerging markets can add diversification but it depends how you invest in them. The broader an emerging market index is the more likely that index will correlate with the US market. A few days ago I wrote a post comparing Turkey and Malaysia that pointed out the low correlation between the two due to vast differences between them. There is potentially some offset between the two which during a crisis could be a good thing.
An entire broad-based index with all sorts of countries zigging and zagging is going to dilute some of portion of the effect of investing in emerging markets. This makes the case of investing in countries or regions as opposed to something like EEM. I realize that many people may not be comfortable with a single country but the argument that broad-based emerging market funds is not a great way to go has merit.
If you look at the correlation between SPY and just about any single country product you will see a lower correlation than the number for EEM; I realize I am stating the obvious.
However as a investment advisor I don't think I could justify ignoring the space to a client because the broadest vehicles out there may not deliver. Top down analysis of a country combined with a bottom up pick of a country fund or, heaven forbid, an individual stock seems like it should be within the wheelhouse of someone with my job title.
Labels:
emerging market,
ETF,
portfolio strategy
Subscribe to:
Post Comments (Atom)





6 comments:
The time period in question seems too short to make such a radical decision (putting an asset class "on probation"). That said, he raises an interesting issue for Fama/French to ponder. Do they need to add another factor to their 3-factor model?
As someone who has spent a great deal of time examining correlations, covariances and such I can say that all markets (emerging, developed, etc.) indeed have higher correlations during market panics and downturns than at other times. This would seem to dilute the diversification benefit for many people however one must remember that as long as the correlation between 2 assets is not exactly +1 then having it in the portfolio will decrease the variance (volatility) of that portfolio.
I will agree with Roger wholeheartedly that the best way to take advantage of foreign zagging is to choose narrower themes, individual countries, etc. Having a broad market foreign exposure will ultimately dilute diversification more than owning several narrower themes. It may seem as if having narrower themes would increase volatility but if their correlations are lower than say that of EEM then the portfolio variance will be smaller even though the individual volatility may be higher. Just a thought.
While we're on the topic I must also agree with Roger that choosing the COUNTRY FIRST is more key than choosing a sector first. Several studies have shown that securities within the same country but different industries have higher correlations than securities in the same industry but in different countries. This means the macro aspect of choosing the country is more important than the actual sector/security you wish to use to capture that effect.
This is something you don't hear a lot about in the mainstream financial press. I've only read about this topic on sites like seekingalpha and Roger's blog.
The conventional wisdom is diversification = uncorrelated assets/asset classes = reduced risk/volatility. Now we know broad diversification doesn't work like it used to.
Question: Is there a stat that summarizes how correlated/uncorrelated the holdings of a portfolio are? It seems like there should be.
TomK, there are services/sites that do this, if I understand your question correctly.
I use portfolioScience, it allow you to imput a portfolio and then compare it for several things, including correlation, to various indices including SPX.
The last I looked I believe the generic client portfolio had a correlation about 0.70 to SPX. The tool allows you to tweak it some to see what would increas or decrease the correlation
Roger,
Please say a word or two regarding uncorrelated vs. anticorrelated. The term low correlation confuses me - low as in negative? or low as in near zero? If two items have near zero correlation then we can expect that they do not move together or simultaneously away from on another. Is that what you look for? Thanks.
A quick update on my UCPIX hedge. I dumped the position yesterday after a 4.4% gain. This market doesn't want to sell-off even when there are many reasons why it should. I might reload at some point down the road.
Post a Comment