Meb Faber’s Sovereign Bonds Study – What Would Buffett Do?
Sovereign Bonds – Finding Yield in a 2% World
Do you know what the largest asset class in the world is?
Many investors are surprised to learn that the answer to this question is foreign debt. The below chart is from a Vanguard article on sovereign bonds.
How much of your global allocation do you invest in foreign bonds?
Likely very little.
Following the global market capitalization weighted portfolio, investors should have about 30% of their portfolio in foreign government bonds, but very few do. (Likewise US investors should have about half of their global stock allocation in foreign stocks but most only have about 30%.)
Why does this matter right now?
As of January 2016, US 10-year government bonds yield 2.25%, and 30-year bonds yield 3.00%. Many investors that rely on income, particularly retirees, struggle with such paltry yields.
Thankfully, US investors are not limited to investing within our borders.
Would adding foreign bonds help diversify a US-centric portfolio?
Global bonds have seen similar real returns as US bonds all the way back to 1900. The Dimson, Marsh, and Staunton team examined investing in 16 countries stock and bond markets in their outstanding book The Triumph of the Optimists. They demonstrated that US bonds had real returns of 2.0% from 1900-2014, and the median country had returns of 1.7%. (Note: real returns are returns after inflation.)
The best performing sovereign bond market experienced real returns of about 3.3% (Denmark), and the worst, well, there are some unfortunate examples of hyperinflation that destroyed investor’s capital. But in general a diversified portfolio of sovereign government bonds did an admirable job of protecting purchasing power overtime. Businesslnsider.com has a good article on the causes of the nine worst episodes of hyperinflation in the past 100 years.
However, even in global developed and emerging markets there is wide disparity between yields. On one hand you have many European countries that are yielding less than 0.5% (and in some cases negative yields!), and on the other, many countries, particularly in the emerging markets, have yields above 5%. The median and average yields for the combined developed and emerging country universe are 2.2% and 3.7%, respectively. So, even in the global bond space you’re not getting much more yield than in the US.
However, most bond indexes are market capitalization weighted, which means you invest more in the countries that have the most debt outstanding. Does that make much sense? Would you lend more to your neighbors or family members based only on how much debt outstanding they have?
Global bond indexes are dominated by the five biggest issuers: the United States, Japan, Germany, France, and the United Kingdom. Those five countries alone account for about 70% of total debt outstanding, but less than half of global GDP and about 10% of global population. (For a wonderful overview of the sovereign bond space, with discussion of index construction you can view the Research Affiliates piece, Debt be not Prowl)
Is there a better way to invest in global bonds? We know that moving away from market cap weighting in stocks is a smart move, and in particular a value approach has performed well over time. Does applying the same logic to global bonds lead to higher returns?
Value Investing With Sovereign Bonds
Value investors have long focused on such metrics as dividend yield in stock selection, and the historical results confirm this has been a valid approach to outperformance. How does one define value in bonds? One approach is simply sorting bonds based on their yields. There is ample research that demonstrates that sorting government bonds based on this measure of value has historically produced strong returns. We are not going to go into an exhaustive literature review, but you can find a thorough summary in the book Expected Returns by Ilmanen, as well as some papers in the appendix. Below we run our own test to confirm the results in the literature.
We decided to examine a global value approach to sovereign bonds back to 1950 with 30 countries from the Global Financial Data database. We sort the universe by yield and invest in the top 33% of countries by nominal yield. We compare this strategy to a few different benchmarks. First, we compare the returns to an ”Equal Weighting” of all the countries in the universe. Second, we compare the returns to the ”Foreign 10 YR” label which uses a GDP weighted index of 10-year bonds from the countries of Australia, Austria, Belgium, Canada Denmark, Finland, France, Germany, Ireland, Italy, Japan, Netherlands, New Zealand, South Africa, Spain, Sweden, and the United Kingdom.
The results: an approximate 2% outperformance for the high yield strategy. More importantly the strategy outperformance is consistent across decades, including both rising and falling interest rate environments.
Note that the high yielding portfolio outperformed both the equal weight and GDP weight bond portfolios in five out of six decades. And in each case where it trailed the other indexes, the underperformance was fairly negligible. These are USD based returns, but local real returns should be very similar. The currency exposure should not matter much over time since real currency returns are fairly stable, although in the short term currency gyrations can have major impact on returns. Vanguard’s has a nice whitepaper on the subject ”The Impact of Currency Hedging in Foreign Bonds”.
See full PDF below.