December 19, 2013

Warning: Bond Indexes Are Broken

Fixed income indexes present distinct challenges in comparison with their equity index counterparts. The fixed income investment universe is considerably larger, with included securities issued by government issuers, corporations and other entities.  Index turnover is necessarily higher than for stock indexes, as outstanding debt matures naturally and new debt is issued continually to meet a particular issuer’s financing needs. Fixed income investments are traded over the counter, typically with much less trading activity than equities, making their value more difficult to determine.  

The theoretical underpinnings used for equity indexes don’t work as well for fixed income indexes.  Cap-weighted equity indexes have a beneficial form of success bias, in that companies viewed positively by the market will grow in price and value, increasing their weight in the index automatically. Fixed income indexes arguably offer a perverse reward system, in that the companies or countries that issue the most debt “earn” higher weights in traditional indexes.

The shortcomings of fixed income indexes matter, as investors gravitate to bond ETFs that are tied to flawed benchmarks.  Some of the most popular bond ETFs are tied to benchmarks that may not provide the safety or performance profile expected. For example, one of the most popular bond ETFs, iShares Total Bond Market ETF (AGG), is tied to an index dominated by Treasuries and is far from a total bond market proxy. Popular ETFs used as proxies for the U.S. high-yield market— iShares iBoxx High Yield Corporate Bond ETF (HYG) and SPDR Barclays High Yield Bond ETF (JNK)—are deeply diversified but are vulnerable to the perverse rewards system tied to issuance volumes. Outside the U.S., ETFs such as SPDR Barclays Short Term International Treasury ETF (BWZ) and iShares International Treasury Bond ETF (IGOV) were overly concentrated in the debt of troubled issuers during the European financial crisis.  

The Troubling Role of Ratinsg Agencies

Ratings agencies offer a troubling aspect to bond indexes, in the prominent influence they have on index composition. Ratings agencies—much maligned for their slow response to the financial crisis—classify bonds as either investment grade or below investment grade, influencing whether a bond is included in an investment-grade index or a high-yield index. It’s fair to say that the market could or should impose discipline lacking from the ratings agencies, but we often see exuberance in the bond new-issuance market that has disproportionate influence on bond indexes when compared to equity indexes.

The flaws in current approaches can be illustrated by examining historical index weights. For example, in September 2009 Italy, Spain and Greece represented 22% of the Barclays Global Treasury ex-U.S. Index.  Greek debt was weighted more than Australia in the index, not a reasonably intuitive outcome given the relative size, economic output and population of the two countries. Australia, for example, has three times the GDP of Greece. These index weights were driven by high debt issuance, something that could have been seen as a warning sign instead of a buy signal.

Click to enlarge 

A couple of years later, Italy, Greece and Spain were among the world’s most troubled borrowers.  Adding Japan to the list, which in September 2009 represented approximately 24% of the index, these four countries represented nearly half of the global bond index.  Critics of traditional approaches to fixed income indexing say that upon any measure other than debt, it would be hard to argue that these countries should comprise nearly half of a global bond index. Each of these countries was over-represented in many respects—viewed by population, GDP and economic strength.

Critiques of the Barclays Aggregate Index

Domestically, critics including John Bogle argue that Treasury securities are over-represented in the Barclays U.S. Aggregate Bond Index, otherwise known as “the Agg.”  Bogle argues that many purchasers buy Treasury securities for policy purposes rather than as investments.  After adjusting for Treasury securities owned for policy-driven purposes, Bogle says that U. S. government securities should represent about one-third of the investable U.S. universe for bonds rather than the more than two-thirds weighting in the Barclays U.S. Aggregate Index Bond Index.

Click to enlarge

Although the Barclays Aggregate is the most widely used bond index, critics also observe that the index isn’t fully representative as it excludes significant and popular bond classes such as high-yield and municipal bonds.

Alternative Bond Indexes

Research Affiliates published research in 2010 into alternative fixed income indexes, drawing from the methodology used to create its popular series of fundamental equity indexes. Their published research is the foundation for a series of bond indexes offered in partnership with Citigroup.  Research Affiliates’ objective was to “sever the link between price and portfolio weight.”  For corporate bonds, Research Affiliates devised and back-tested indexes with weighting schemes based on assets, dividends, cash flow and sales. The Research Affiliates methodology excluded privately owned or employee-owned companies, as well as companies based in foreign countries. Back-tested results for the corporate bond indexes (investment grade and high yield) showed that each index outperformed the corresponding traditional index.   

For emerging market bonds, Research Affiliates created indexes based on multiple factors, using actual economic data or proxies as necessary for such attributes as GDP, population, land mass and energy consumption. The emerging market bond index also outperformed the corresponding traditional index in both bull markets and bear markets.

ETFs have been developed to track the fundamental corporate bond indexes, with mixed results to date.  ETFs based on the high-yield and investment-grade bond indexes have lagged the indexes they are meant to replicate since inception, as well as their respective traditional Barclays index counterparts.  Some of the tracking error for the high-yield ETF may be attributable to transition issues in moving from a legacy index to the RAFI Index, but liquidity and transaction costs may also be a factor. Although the intellectual capital behind the fundamental indexes is compelling, we are carefully monitoring performance of the ETFs to see whether performance over a market cycle delivers compared to the back-tested results. 

What We Do in the Meantime

With the choice of investment vehicles based on broken indexes or ones based on unproven alternative indexes, how are thoughtful investors building fixed income portfolios? Many use a two-pronged approach. The first prong is diversification. Start with a core fund that tracks the Barclays U.S. Aggregate Index Bond Index; then add funds with exposure to high yield, international debt and other areas that are absent or under-represented in the Agg. The second prong is active management. Identify funds managed by skillful portfolio managers who are mandated to find opportunities in the bond market and are not tied to an index.

In short, until the indexes are fixed, we have to make our own.

Page 1 of 2
Single page view Reprints Discuss this story
This is where the comments go.