Seeking High-Alpha, Low-Beta Countries (Part II)
This piece is part of an ongoing project to identify countries that have a potential to provide high investment returns while exposing investors to less market risk. It constitutes part two of my "Seeking High-Alpha, Low-Beta Countries" series - part one was published by Seeking Alpha last week .
Ever since I've been involved with investing, I've been on a quest to find exceptions to the dictum that high returns are associated with higher risks.
I believe that this is possible because one of the things missing from the "high-risk, high-return" puzzle is the level of perceived risk of a particular investment in the marketplace. Sometimes a specific investment is perceived to have significantly more risk than it actually does. Other times the opposite is true.
Is Emerging Market Investing Risky?
My opinion is that the "average" investor exaggerates risks that he or she perceives in the emerging markets, mainly because of fear of the unknown. Mainstream media tends to promote this viewpoint as well.
Having had the perspective of an "outsider" to the mainstream U.S. opinion, I've come to the conclusion that many are more comfortable losing money to the "familiar" than making money from the "unknown". If this is indeed the prevailing bias, then there is money to be made. I realize the phrase "prevailing bias" is too Soros-esque but I'm really talking behavioral economics here.
Of course, there are periods of euphoria when all investments get discounted with less risk premiums than warranted, and this includes emerging markets as well. Also emerging markets are vulnerable to larger price moves merely because their smaller trading "pools" are not always equipped to properly "digest" trading decisions made by global players. However, if a U.S. investor has ever dealt with a midcap U.S. stock, they know exactly how that kind of volatility works.
This piece is about analyzing global data and I'm here to provide it to you.
So, Is Emerging Market Investing Riskier Or What?
This is precisely what I've been trying to figure out. For that, I've looked at betas and various other metrics along with performance measurements during different time periods. If you're up to analyzing more data including correlations as well as standard deviations of different markets, I highly suggest that you take a look at the piece published a year ago entitled, "Calculating Country Risk Observed by Betas".
If you look carefully at the two-year and ten-year ranking tables in the piece, you will notice that each emerging country has a unique "risk" characteristic. Some emerging markets are "riskier" to invest in, but some are not. In fact, some emerging markets are equally or even less risky than certain "developed" economies, specifically economies of the European Union.
The article comes to the conclusion that certain emerging countries, specifically those in emerging Asia as well as Middle East and North Africa, are perhaps less risky than we may have initially thought. They rank better than many "developed" nations. On the other hand, some other emerging nations are indeed riskier than global averages.
The U.S. market constitutes a significant portion of the global marketplace, and if we exclude U.S. data from the "independent" world index variable in our calculations, we also get a lower risk metric for the U.S. market. In fact, among the "developed" countries, only U.S. and Japan are associated with the "lesser" risk that our average investor may be interested in.
Country Performance Attribution Analysis
Last week's article lists and ranks aforementioned countries' subsequent yearly investment returns as a follow-up on the preceding year's article, while observing that the remarkable performance of the "lower-risk" countries benefited tremendously from market timing.
Standard & Poor's Index Services makes its country specific index data available in a few currencies, and I chose data provided in U.S. dollars. So all results implicitly assume the perspective of the U.S. investor and that the investor chose not to hedge the currency risks associated with these markets.
The same table from last week is reproduced below:Country Returnsfrom December 2, 2008 to December 2, 2009 Country "Raw" Return (( USD )) Country "Raw" Return (( USD )) Indonesia Canada Brazil Malaysia Peru Netherlands India Portugal Norway Poland Turkey Denmark South Korea Greece Australia Israel Singapore New Zealand Thailand Egypt Russia Spain China Germany South Africa Ireland Sweden United Kingdom Taiwan France Chile Switzerland Austria Czech Republic Hungary Italy Belgium United States Hong Kong Finland Philippines Japan Mexico Morocco
I'm going to attempt to undertake an informal performance attribution analysis on these returns.
As I mentioned in the article last week, there are two things I'm concerned about: One is the fact that these returns are measured from a rather low base. In fact, the "base" data for many of these indices are not very different from the global market lows of March 2009.
The other is the fact that U.S. dollar has lost significant value to many of the currencies of these markets during the measurement period. We should keep that in mind especially if we want to look deeper into the relative underperformance of U.S. equities, or the success of emerging markets vis-à-vis U.S. equities, during this time period.
So I had to get a hold of these markets' currency data for the periods concerning these results. I fetched all the necessary currency data from oanda.com.
Return Diagnostics (1): Currency Performance
The next table breaks down the return metric for all markets to see how much is attributable to respective currency movements. The columns break down the return into three components:
First one is the local return, which is the return associated with the home currency, second is the currency return, which is the return of the currency with respect to the U.S. dollar. The third component is the cross component which accounts for the effect of the change in the local currency on the change in the local market. In other words, it's a mathematical cross term that should be added to the two components to come up with the U.S. return.
If you follow my stuff, you not only like looking at data, but also like looking at how data get sliced and diced. I hear you. So embrace for more. This time, country index returns are ranked from top to bottom according local returns produced in their home currencies:Country Returns from December 2, 2008 to December 2, 2009 Country "Raw" Return (( USD )) = Local Return + Currency Return + Cross Return Peru Indonesia Russia India Turkey China Brazil Hong Kong Thailand Taiwan Singapore Philippines Norway Mexico South Korea Hungary Sweden Austria Belgium Egypt Malaysia Canada Israel Poland Australia Chile South Africa Netherlands United Kingdom United States Portugal Denmark Greece Spain Germany Ireland France Czech Republic Italy Switzerland New Zealand Finland Japan Morocco
There is somewhat of a shift in countries' places when ranked with respect to local returns. Some may argue that changes in ranking should not be important because stock markets adjust for currency devaluations or revaluations and the law of one price should make changes in currency levels irrelevant. However, that premise would hold only if we believed that purchasing power parity functioned accordingly.
As a side point, original regressions I ran a year ago on U.S. returns implicitly assumed that currency fluctuations as such were irrelevant and that PPP held. Otherwise, the model would have had to account for currency risk as another control variable, which would have created a different story. So the reader will have to make do with a simple attribution analysis like this one for now.
The metrics in the table above tell a few narratives, and I'll briefly summarize some I find worth mentioning. First off, some countries' returns have been influenced by significant revaluation of their currencies vis-à-vis the U.S. dollar, among them Indonesia, Brazil, Australia, South Africa, Chile, and New Zealand (34% - 42%). Secondly, South Korea, Norway, Canada, Sweden and Switzerland show a revaluation in high teens and low twenties, and the euro displays a revaluation of 19%.
On the other hand, India, Russia, Turkey, Mexico, Peru, Thailand, Taiwan, Philippines, Malaysia, and Israel show remarkably lower rates of revaluation (there is even a small devaluation in Ruble's case) considering the ensuing buying pressure on the emerging markets' currencies. Elsewhere, China and Hong Kong speak for themselves.
From a fundamental vantage point, I have spoken out and written about dangers of currency overvaluation in smaller emerging markets due to global funds' rush to shallower waters creating bubbles of sorts, specifically in relation to the Turkish economy. For instance, I know that quite a few business players in Turkey would react very negatively had the Turkish currency shown more revaluation than the mere 5.7% shown in the table above during this time period.
In order to be consistent with my country-specific evaluations, I cannot applaud a strong currency valuation when a much longer investment horizon is concerned (2 to 5 years and more). However, when the local market return constitutes a greater portion of the total return, as in the case of Brazil and Indonesia, I will be more forgiving because it is clear that the investment remains a bet on the country rather than on its currency. Whether I view a currency revaluation as good or bad would also depend on whether that specific country is experiencing a trade surplus or deficit, but this type of analysis is beyond the scope of this piece.
However, in the cases of Australia, South Africa, and specifically New Zealand, the currency component constitutes more than half of the distribution of returns. Therefore it's not clear whether the investor is making money because of a bet on that specific country or because of a bet on that country's currency.
Yet, a positive note for U.S. equities is due after this breakdown. The U.S. equity performance has moved up in rank a bit after taking care of the miserable performance of the U.S. dollar vis-à-vis other currencies.
Return Diagnostics (2): Market Timing
Determining how much of a return is due to market timing is not an easy job. If you are interested in my previous efforts to determine this, I suggest that you take a look at last week's article. That piece multiplies daily betas with the yearly global market return to come up with such a timing metric, and subtracts this metric from the yearly asset return to come up with an "excess" return, but the methodology is quite suspect because of the mingling of daily metrics with yearly returns. On top of that, not only are betas subject to change, but alphas also tend to be positive during good performance periods and negative during bad performance periods. So any alpha produced may be partly or wholly due to the right timing.
One of the more convenient ways to look at the "timing" effect is to analyze returns from peak to peak to see whether a specific country is outperforming or underperforming the global index during such periods. In other words, if I lengthen my measurement period to two years, my "base" data for country indices will capture numbers that are very close to the world highs established in October of 2007.
I would like to state that the initial spirit of this "country risk" study was to capture lower-risk countries while benefiting from the higher than warranted risk premiums that investors required from emerging markets a year ago. I figured, while capturing high-risk premiums (hence low valuations), why not try to accomplish it with less risk?
So I will now look at return data from a base period when global markets were priced closer to "perfection". This way we can have a better idea about which country is outperforming the global index when measured from a place close to the global peak.
For that I will lengthen the measurement period to two years as opposed to selecting the world highs established in October 2007, although December 2007 is close enough. (I am doing enough data mining and slicing as it is, and don't want to add an ex-post "selection bias" to what I'm doing.)
Below are the results. The global index return during this time frame remains -25.6 percent:
Country Returnsfrom December 3, 2007 to December 2, 2009 Country "Raw" Return (( USD )) Country "Raw" Return (( USD )) Peru 6.64% Denmark -28.77% Israel 0.70% Netherlands -28.82% Chile 0.49% France -29.02% Brazil 0.35% Japan -29.51% Taiwan -13.64% Czech Republic -29.64% Mexico -14.34% Turkey -29.91% Malaysia -14.47% South Korea -30.14% Switzerland -15.05% Hungary -30.25% Indonesia -15.42% Germany -32.11% South Africa -15.66% United Kingdom -32.17% Thailand -16.41% Norway -33.59% Canada -17.00% New Zealand -35.38% Morocco -17.09% Egypt -35.59% Singapore -18.04% Portugal -36.75% Sweden -20.45% Poland -39.15% Philippines -20.65% Italy -41.19% Australia -22.89% Austria -41.35% Hong Kong -23.34% Russia -44.00% United States -23.79% Belgium -44.57% China -23.91% Finland -50.33% Spain -25.68% Greece -50.55% India -27.49% Ireland -56.88%
As we can see, the majority of these markets have not reached levels observed during the highs of 2007. If we divide the 44 countries in this table into four quartiles for performance evaluation purposes, we see that the only "developed" market in the first quartile is Switzerland. However, Canada, U.S. and Australia remain relatively healthy in the second quartile. This exercise bodes badly for European countries, including those within and around the European Union.
Nevertheless, part two of these series has established that U.S. equities are doing better than most of their European counterparts.
Below is a currency attribution analysis on this period's returns. This time, countries are ranked according to their respective local currency returns: Country Returns from December 3, 2007 to December 2, 2009 Country "Raw" Return (( USD )) = Local Return + Currency Return + Cross Return Mexico -14.34%1.04%-15.22%-0.16% Peru 6.64%0.96%5.62%0.05% Chile 0.49%-0.41%0.90%0.00% Israel 0.70%-0.79%1.50%-0.01% Brazil 0.35%-2.11%2.51%-0.05% South Africa -15.66%-9.23%-7.09%0.65% Thailand -16.41%-10.40%-6.71%0.70% Turkey -29.91%-10.94%-21.31%2.33% Philippines -20.65%-12.29%-9.53%1.17% South Korea -30.14%-12.31%-20.33%2.50% Canada -17.00%-12.80%-4.81%0.62% Sweden -20.45%-13.69%-7.84%1.07% Taiwan -13.64%-13.84%0.23%-0.03% Malaysia -14.47%-14.20%-0.31%0.04% India -27.49%-15.09%-14.60%2.20% United Kingdom -32.17%-15.58%-19.66%3.06% Indonesia -15.42%-16.10%0.81%-0.13% Morocco -17.09%-19.03%2.40%-0.46% Singapore -18.04%-21.81%4.82%-1.05% Hong Kong -23.34%-23.72%0.49%-0.12% United States -23.79%-23.79%0.00%0.00% Switzerland -15.05%-24.78%12.94%-3.21% Australia -22.89%-25.87%4.02%-1.04% Hungary -30.25%-27.19%-4.21%1.15% Spain -25.68%-27.76%2.89%-0.80% China -23.91%-29.76%8.33%-2.48% Netherlands -28.82%-30.82%2.89%-0.89% Denmark -28.77%-30.95%3.16%-0.98% France -29.02%-31.01%2.89%-0.90% New Zealand -35.38%-31.66%-5.45%1.73% Czech Republic -29.64%-32.31%3.95%-1.28% Norway -33.59%-32.36%-1.82%0.59% Poland -39.15%-32.40%-9.98%3.23% Russia -44.00%-33.48%-15.81%5.29% Germany -32.11%-34.02%2.89%-0.98% Egypt -35.59%-36.67%1.71%-0.63% Portugal -36.75%-38.53%2.89%-1.11% Italy -41.19%-42.84%2.89%-1.24% Austria -41.35%-43.00%2.89%-1.24% Japan -29.51%-45.03%28.24%-12.72% Belgium -44.57%-46.13%2.89%-1.33% Finland -50.33%-51.73%2.89%-1.49% Greece -50.55%-51.94%2.89%-1.50% Ireland -56.88%-58.09%2.89%-1.68%
When global markets were discounting a financial Armageddon almost a year ago, U.S. dollar registered significant gains. Most of the disappointing dollar performance since then has been the undoing of that value gain.
However, when we look closer into currency performances registered beginning with global highs at the end of 2007, we see that a couple of countries' currencies devalued against the U.S. dollar and remained that way, most notable among them Turkey, South Korea, United Kingdom, Russia, Mexico, and India (-22% to -15%). What stroke me most about this table is that while most countries' currencies have recovered against the U.S. dollar, the currencies of the above-mentioned countries did not.
In fact, when country performance is measured and ranked according to local currency returns, both Turkey and South Korea move from the third quadrant into the first one.
As a side point, this devaluation has been an opportunity for Turkey for improving its current account deficit standing, but whether that alone will help the country to move towards an export-oriented growth economy remains to be seen.
I'm no Japan expert here, but having registered a currency gain of 28 percent against the dollar and even more against some other currencies, there appears to be a significant overvaluation problem in their hands. In fact, when ranked according to local returns, the Japanese market is doing as badly as European markets.
This exercise started as a risk-ranking project and morphed into something bigger. It's a big world out there, and it's a bigger challenge to fit the world's supreme diversity into some type of a quantitative model. However, calculating basic risk metrics with available index data has been a good starting point for a top-down analysis of the world markets.
A real-time observation of results required a further analysis into the whys and hows of individual country performances. So I prepared performance attribution tables to provide a snapshot of what is going on in many global markets as well as these markets' respective currencies.
I must admit that performing a currency break-down analysis raised more questions in my mind than answers I was preparing to provide. These types of questions belong to fundamental analyses of respective countries' economies, which is beyond the scope of this write-up.
Everyone is telling a story about the markets - that the recent rally has been too fast, or that a correction is due. I always like to look at numbers and hope that they will manifest a less "biased" narrative. Via strictly looking at available data without anyone clouding our judgment, perhaps we can better arrive at our own version of the tale that is about to unfold.
Disclosure: No positions
See also Top 10 Emerging Markets Metals and Mining Stocks on seekingalpha.com