Seeking High-Alpha, Low-Beta Countries (Part I)

by: Suna Reyent, CFA

A year ago, Seeking Alpha published a piece of mine entitled “Calculating Country Risk Observed By Betas”.

Since equity markets tend to go up and down in a synchronous fashion with the world markets, owning a country risk itself means owning a piece of the general “market risk”. The article intended to unearth systematic risks that an investor undertook merely by the action of going long a certain country via an ETF or any such investable index.

The article talked about methodology related to capital asset pricing model and displayed betas of various investable countries with respect to a chosen world index, with an eye toward distinguishing those that exposed investors to more risk and those that exposed them to less risk.

Regressions: Finding Beta and Alpha Coefficients

A short review of capital asset pricing model is in order here. Very briefly put, the model displays a security’s return as a function of the general market return. This relationship is manifested by the security market line, which relates expected returns to their non-diversifiable risks, in other words, their respective betas.

A beta greater than one signifies high systematic risk, whereas a beta less than one signifies low systematic risk.

Regressions were run for each country index return (dependent variable) against the world index (independent variable), and slope coefficients were found to be the countries’ respective betas. After ranking the results of countries according to betas manifested during two-year and ten-year periods, the article made the following conclusion, word for word:

An inference of this study (or so I reckon) is that if all else is equal, it is more advisable to invest in nations with higher than average growth potential that also show a low degree of risk with respect to the world markets. The diversification effects are also valuable. Of course, all else is never equal and such quantitative studies should always be complemented by fundamental analysis as well as due diligence.

For ease of reference, I’m including those countries that returned a beta lower than one, or the market itself, directly from the article published a year ago:

Country Betas : Two-Year Returns (as of December 2, 2008)
Country Beta = Correlation w/ Global Portfolio x Standard Deviation / Global St Deviation
Greece 0.95 0.69 2.06% 1.50%
Chile 0.95 0.71 1.99% 1.50%
South Korea 0.92 0.50 2.77% 1.50%
Portugal 0.88 0.73 1.80% 1.50%
India 0.86 0.53 2.43% 1.50%
Switzerland 0.83 0.80 1.55% 1.50%
Singapore 0.76 0.64 1.79% 1.50%
Indonesia 0.76 0.47 2.42% 1.50%
Hong Kong 0.70 0.55 1.90% 1.50%
Thailand 0.63 0.44 2.14% 1.50%
New Zealand 0.60 0.52 1.73% 1.50%
Philippines 0.51 0.38 2.01% 1.50%
Taiwan 0.48 0.39 1.83% 1.50%
Israel 0.42 0.48 1.32% 1.50%
Japan 0.41 0.34 1.81% 1.50%
Malaysia 0.38 0.40 1.43% 1.50%
Egypt 0.34 0.27 1.92% 1.50%
Morocco 0.28 0.33 1.28% 1.50%

An interesting result I had also mentioned in the article was that the bottom of the list (i.e., countries exhibiting less “systematic” risk) included those with higher than average growth potential with respect to the rest of the world.

I must add that among the BRIC countries, only India made it into this list. China barely missed the list with a somewhat low beta. On the other hand, both Brazil and Russia missed this list with a wide margin because they consistently keep showing betas greater than one.

Identifying “Low Risk” Countries with High Potential for Growth

As an investor, if I were looking for “high-growth potential” countries in this list, I would directly exclude any nation I find that happens to be in the European region, perhaps with the exception of those that are classified as Eastern or Emerging Europe. No such luck here, so Greece, Portugal, and Switzerland would be eliminated right off the bat from MY list of “high-growth potential” countries. Japanese market clearly shows low systematic risk, but it is just not feasible to expect aggressive growth from Japan like China or India.

When Japan is thrown off this list, the investor is faced with mostly emerging Asia exposure (South Korea, India, Singapore, Indonesia, Hong Kong, Thailand, New Zealand, Philippines, Taiwan) as well as some Middle Eastern and North African market exposure (Israel, Egypt, Morocco), and finally Chile from South America.

The other caveat of this exercise is that it “suggests” buying in baskets without paying any specific attention to any of the countries mentioned in the list. It does not offer any due diligence on the specific firms that may dominate these markets nor does it provide fundamental analysis on the macro environments governing any of these countries. Yet this is precisely what such ranking methodologies and quantitatively oriented studies generally offer, so the reader will have to go along with me here.

Jensen’s Alpha: Computation and Interpretation

Jensen’s Alpha is the intercept term remaining after a regression is run, where the dependent variable is the country return and the independent variable is the index that the country return is compared against (i.e., world market return). Calculations are performed on daily data in this exercise. Strict calculations require subtracting the daily risk-free rate from both the dependent and the independent variables before performing regressions.

In finance, Jensen’s Alpha signifies whether the security (or the country) has outperformed or underperformed the market (or the world) index after accounting for the security’s (or country’s) systematic risk, i.e., the beta.

However, while an in-sample alpha gives a good idea on how a security or country has performed after getting rid of its beta effect, it is an ex-post measure, meaning that it does not tell the practitioner about how the specific security or country is going to perform in the future with respect to its level of systematic risk. So while alphas of each country regressions were informative, I did not see the point of publishing the results because I was not interested in the over or underperformance of countries up until that time. I was specifically after trying to capture future alphas while forming a basket of countries that manifested the least amount of systematic risk.

Also, I ran the regressions without subtracting the risk-free rates from independent and dependent variables, which has been done in practice before. However, this action will affect the intercept term, which will make the computed alphas less reliable.

Country Performance One Year Later: Results

Standard & Poors Index Services has continued maintaining and publishing daily trading data for most of the countries that I had calculated betas for a year ago. The world performance since the time I wrote this piece has in general been stunning, so this specific buy-and-hold study has definitely had considerable luck in terms of timing.

Country Returns from December 2, 2008 to December 2, 2009
Country "Raw" Return Country "Raw" Return
Indonesia 188.66% Canada 73.60%
Brazil 154.66% Malaysia 61.16%
Peru 133.59% Netherlands 61.11%
India 119.45% Portugal 57.94%
Norway 107.21% Poland 56.80%
Turkey 103.77% Denmark 56.01%
South Korea 103.70% Greece 54.66%
Australia 100.31% Israel 54.20%
Singapore 97.94% New Zealand 53.61%
Thailand 95.72% Egypt 53.15%
Russia 95.41% Spain 52.71%
China 92.19% Germany 47.82%
South Africa 89.47% Ireland 47.56%
Sweden 89.34% United Kingdom 46.99%
Taiwan 87.02% France 44.81%
Chile 86.67% Switzerland 42.30%
Austria 86.59% Czech Republic 41.59%
Hungary 86.26% Italy 40.25%
Belgium 86.25% United States 33.93%
Hong Kong 83.49% Finland 30.61%
Philippines 82.25% Japan 17.90%
Mexico 73.87% Morocco 7.45%

Returns have indeed been quite spectacular for the buy-and-hold folks out there. It was definitely the right time to be optimistic, specifically with regard to emerging market investing. What is more surprising is that certain emerging nations that now occupy the top of the list have been shown to display a consistently lower beta in my earlier calculations that were published in the article.

In short, the performance of low-beta countries such as Indonesia and India is even more spectacular than the performance of countries like Brazil and Turkey, two countries that have been known to display higher than average betas.

One of the great things about such “basket” portfolios is the relative ease with which one is able to hold some “losers” among the “winners” without knowing (or caring) a lot about why these investments are losing money. Although there are no “losers” here in absolute terms, Morocco, which was one of the countries within the initial basket of 13, has been the worst “absolute” performer among all the countries for which I could find continuous trading data.

Of course, the investor should not complain much given the fact that Indonesia, India, South Korea, Singapore, Thailand, and many other successful performers were also in the same basket.

For disclosure and clarification purposes, I did not personally execute this strategy nor do I hold any of the ETFs associated with the countries mentioned above. However, when I do buy stocks, I tend to buy ten or twenty of them at the same time precisely because I know that I cannot have a good grasp on everything that has to do with one particular stock.

What About Future Returns? Didn’t This Study Benefit from Market Timing?

Answer #1: Dude, why would I bother trying to have such a buy-and-hold strategy published with an unalterable time stamp if I felt like it was the wrong time to go public with it?

Answer #2: As much as I hate to admit it, the answer to this question is certainly “yes”. Even though my timing to write about such a strategy had to do with self-selection, the article “recommends” a very specific strategy. And this strategy has benefited tremendously from the fact that world markets have registered a return of 47 percent according to global index data provided by Standard & Poors Index Services.

Such a strategy would face its true test if undertaken during a period of falling global markets.

However, to shed some further light on the topic, I tried to look at these returns after stripping away the general “market” effect. I did that via subtracting the world index return multiplied by the country’s beta from the specific return of the country. The excess return stripped from the general market influences could give me more information on a specific country’s return.

To clarify the point above, let’s look at China’s return:

Country Returns from December 2, 2008 to December 2, 2009
Country "Raw" Return - ( Beta x World Index Return ) = "Excess" Return
China 92.19% 1.02 46.87% 44.39%

Standard & Poors Index Services measures China return as 92.19%, some of which is attributable to the world index. Multiplying China’s registered beta with the world index return gives 47.79%, and subtracting this number from the “raw return” leaves us with an “excess” return of 44.39%.

This method looks like another round-about way to come up with Jensen’s alpha, but mathematically it’s not equivalent because I’m multiplying a yearly return number with a daily beta. Also the world risk-free rate needs to be subtracted from both the country and the world index return in order to comply with the theory, which is not done here.

I know the methodology would face due academic criticism, but I’ll have to counter that via saying that the yearly “excess returns” for all the countries are more or less in line with the daily alphas that came out of the regressions I performed for the past year’s data.

The reason I decided to provide the yearly “beta-adjusted” excess return estimate instead of the daily alpha is because it looks more intuitive and tells more. Also, this metric falls when you increase the country beta (undesirable) and rises when you decrease the country beta (desirable).

For instance, if the Chinese market were more “risky”, then the beta-adjusted “excess” return would be less. If the Chinese beta were 1.97 or above, this metric would turn negative. That would mean that all of Chinese market’s positive returns could be attributable to the fact that the world markets had a successful year.

Which is precisely why investors should care about betas, or the non-diversifiable risks that they are holding.

Let’s look at the same list of returns after adjusting for each country’s respective beta:

Country Returns from December 2, 2008 to December 2, 2009
Country "Excess" Return = "Raw" Return - ( Beta x World Index Return )
Indonesia 153.19% 188.66% 0.76 46.87%
Peru 81.04% 133.59% 1.12 46.87%
India 79.19% 119.45% 0.86 46.87%
Brazil 70.27% 154.66% 1.80 46.87%
Thailand 66.05% 95.72% 0.63 46.87%
Taiwan 64.58% 87.02% 0.48 46.87%
Singapore 62.10% 97.94% 0.76 46.87%
South Korea 60.37% 103.70% 0.92 46.87%
Philippines 58.30% 82.25% 0.51 46.87%
Australia 53.07% 100.31% 1.01 46.87%
Hong Kong 50.76% 83.49% 0.70 46.87%
China 44.39% 92.19% 1.02 46.87%
Malaysia 43.36% 61.16% 0.38 46.87%
Norway 42.89% 107.21% 1.37 46.87%
Chile 42.18% 86.67% 0.95 46.87%
Belgium 38.15% 86.25% 1.03 46.87%
Turkey 37.65% 103.77% 1.41 46.87%
Egypt 37.04% 53.15% 0.34 46.87%
South Africa 35.76% 89.47% 1.15 46.87%
Russia 35.16% 95.41% 1.29 46.87%
Israel 34.29% 54.20% 0.42 46.87%
Sweden 31.06% 89.34% 1.24 46.87%
Austria 28.29% 86.59% 1.24 46.87%
New Zealand 25.59% 53.61% 0.60 46.87%
Hungary 18.26% 86.26% 1.45 46.87%
Canada 16.66% 73.60% 1.21 46.87%
Portugal 16.52% 57.94% 0.88 46.87%
Mexico 12.42% 73.87% 1.31 46.87%
Netherlands 10.05% 61.11% 1.09 46.87%
Greece 9.99% 54.66% 0.95 46.87%
Poland 6.40% 56.80% 1.08 46.87%
Denmark 5.43% 56.01% 1.08 46.87%
Switzerland 3.42% 42.30% 0.83 46.87%
Spain 1.79% 52.71% 1.09 46.87%
Japan -1.14% 17.90% 0.41 46.87%
Germany -4.72% 47.82% 1.12 46.87%
Morocco -5.65% 7.45% 0.28 46.87%
Ireland -6.78% 47.56% 1.16 46.87%
United Kingdom -7.91% 46.99% 1.17 46.87%
Italy -8.53% 40.25% 1.04 46.87%
France -8.60% 44.81% 1.14 46.87%
Czech Republic -12.06% 41.59% 1.14 46.87%
United States -15.60% 33.93% 1.06 46.87%
Finland -20.89% 30.61% 1.10 46.87%

The list changes somewhat, but Indonesia still manages to occupy the first place. Brazil drops from second to fourth place due to the high beta associated with the country, which is still indicative of a stellar performance and high alpha. High beta countries can register positive alphas, but it should be worth mentioning that both Turkey and Russia drop in rank after taking their higher than average respective betas into performance consideration.

The updated list of “excess returns” is even more biased towards Eastern and Emerging Asia countries. We see an improvement in their standing according to this metric because of their low betas combined with high absolute returns.

On the other hand, the performance of G-10 countries, perhaps with the exception of Belgium, Sweden, and Canada, has been specifically disappointing.

Indeed, most of the so-called “developed” world has suffered from disappointing performance, with Australia clearly bucking the trend. It is unfortunate that Japan does not benefit in the same fashion from the dynamism of Eastern and Emerging Asia as much as Australia does.

Yet the performance of the United States as the fourth worst absolute performer and the second worst “relative” performer is also worth mentioning.

These performance metrics are telling a story that is quite discomforting for the U.S. equities. While dollar depreciation during the measurement period may provide for an explanation, the mathematical magnitude of the depreciation does not account for the relative underperformance of U.S. equities. I am not even going to go into a discussion about how currency depreciation could work in favor of the equities of the country whose currency depreciates in the long run.

The second part will try to look deeper into the currency movements of most of these markets and mention some notable trends. I will also provide an updated list of countries this strategy “suggests” with a calculation of their last year’s betas.

Disclaimer: This portfolio basket strategy is not a recommendation to buy specific stocks or ETFs. This is not a beta-neutral or portable alpha strategy and does not provide for downside protection for general market risk.

Disclosure: No positions

About this article:

Want to share your opinion on this article? Add a comment.
Disagree with this article? .
To report a factual error in this article, click here