Using Valuation And Volatility To Develop Optimal International Stock Allocations
My discovery of Crestmont Research and their work on secular investing led me to develop a dynamic asset allocation approach using valuation. I have also discovered that valuation can be used effectively for allocating to international stocks. The following paragraphs illustrate an optimal approach for international stock investing at the country level.
For US stocks we have the cyclically adjusted price to earnings ratio (CAPE) to measure valuation.
CAPE data is available for international countries through Global Financial Data and some including Meb Faber have started utilizing CAPE to manage his own international stock ETF, Cambria Global Value (GVAL).
Here is his paper on using CAPE globally:
Global Value Building Trading Models with the 10 Year CAPE
We can also measure stock valuation using Market Cap/GDP.
This chart gives an idea as to the correlation between Market Cap/GDP and CAPE for US stocks.
Source: The World Bank, Shiller
This begs the question: which set of data do you trust more when it comes to international stocks, earnings or GDP?
I prefer GDP data simply because I am not as skeptical of GDP data (China is an exception) internationally as I am the accounting standards in emerging market countries.
The following paragraphs illustrate a Market Cap/GDP/Volatility approach.
At the end we’ll develop an example of an optimal allocation across developed and emerging countries.
iShares MSCI ETFs were utilized for monthly country returns.
Only ETFs with the greatest history were utilized.
These ETFs cover a majority of countries which make-up the MSCI EAFE index.
EAFE countries missing: Denmark, Finland, Ireland, Israel, New Zealand, Norway, Portugal.
The source of the annual Market Cap/GDP data was the World Bank. Period: 1996-2012.
Benjamin Graham first proposed the “margin of safety” investing principle in his book “Security Analysis.” The idea is to focus on buying cheap high quality stocks. This same concept can be applied at the country level by first focusing on countries with low Market Cap/GDP ratios.
When comparing the iShares MSCI EAFE ETF against an equal weight strategy and a Market Cap/GDP strategy you can clearly see the advantage of using low valuation allocations.
1997-2013
Additionally, I found that on a valuation per volatility (trailing 1 year standard deviation) basis, allocating to countries with low Market Cap/GDP/Volatility improved performance.
Here is the back-test including formulas, allocations and returns.
So how could we apply the Market Cap/GDP/Volatility ratio today to arrive at an international allocation?
First let’s see where a majority of the world’s GDP is generated from.
Here are the top countries by GDP.
Next let’s rank them on the basis of Market Cap/GDP/Volatility from smallest to largest.
Couple of observations.
Generally, we can see a majority of inexpensive countries have moderate volatility.
Russia’s high volatility translates into poor quality.
Therefore we’ll omit Russia.
China is another case. The reliability of China GDP numbers is questionable.
Therefore we’ll omit China.
Here are the global allocations with Russia and China omitted and ranked based on Market Cap/GDP/Volatility.
To extend our quality requirement, one consideration is to cut off allocating to countries with high volatility.
Since 15 has been the average historical volatility for US large cap stocks one may consider countries with volatility well above 15 for omission. This means Brazil is omitted.
Finally, to ensure that the overall portfolio contains inexpensive stocks, countries with a Market Cap/GDP Ratio above 80 should be omitted.
Disclosure: None.