The number of indexes available to investors is forever growing. While equity indexes such as the ones provided by Standard & Poor’s, Dow Jones, MSCI and Russell remain the most widely used, indexes covering other asset classes are receiving an increasing share of activity. Over the last few years, a number of indexes covering commodities (S&P GSCI, DJ-AIG), credit derivatives (Markit CDX and Markit iTraxx), bonds (Barclays, Markit iBoxx), structured finance (Markit ABX.HE), real estate (S&P/Case-Shiller) and others have generated significant interest. This broadening of the index market and the growing recognition that active asset management does not necessarily warrant the high fees it demands have resulted in a huge increase in index-based investment vehicles—such as index tracking funds, exchange-traded notes (ETNs) and exchange-traded funds (ETFs)—that track nonequity markets. The benefits to investors of index-tracking investment vehicles are many: low fees, reduced risk and independence from individual managers’ capabilities.
Overall, the expansion of index-based investment products into new asset classes is a positive development. New asset classes, however, require new types of indexes and present new index challenges. Index-based products will only benefit investors if the indexes that these products track are reliable, efficient and functional as investable benchmarks. This is a topic of immediate concern in the fixed-income arena.
Bond Indexes In The U.S. And Europe
Although equity indexes have been in evidence since the early 1800s, it was not until the early 1970s that efforts were initiated to establish bond performance indexes. Two of the main protagonists in index development for bonds were Kuhn, Loeb and Salomon Brothers. Salomon Brothers introduced its Long-Term High-Grade Corporate Bond Index in 1973, at the same time that Kuhn, Loeb (ultimately acquired by Lehman Brothers) set up three U.S. bond indexes.
For many years, most bond index providers were investment banks. The reason is simple: There is no central price-reporting facility in the bond market the way there is with equities. As a result, compared to stocks, the bond market is opaque: There is no one “price” for IBM’s bonds the way there is for IBM’s stock. Investment banks with large bond trading operations are able to draw on their internal trading to price securities, gaining firsthand visibility into a challenging market.
Leveraging this process, by the early 1990s, U.S. indexes were being published by J.P. Morgan, Lehman Brothers, Citibank and Merrill Lynch. The Lehman indexes, now rebranded under their new owner Barclays Capital, became the dominant index family in the U.S., and today the majority of bond portfolios are benchmarked against these indexes.
European bond indexes, by contrast, were quite fragmented. Different currencies meant that each country had its own dominant index for sovereign issues. The corporate bond market was largely still underdeveloped until the introduction of the euro on Jan. 1, 1999, when the advent of a single market helped investment portfolios become cross-national.
Recognizing the limitations that come with an individual approach, especially in such a fragmented market, a group of leading European and U.S. banks active in Europe established a consortium and launched the Euro and Sterling iBoxx indexes in 2002. The iBoxx franchise uses a multidealer pricing model, aggregating pricing from multiple bond dealers into a single consolidated price. The iBoxx indexes have since established themselves as the market-leading indexes in Europe and the U.K. both for benchmarking and as the basis for ETFs and structured products.
The two contrasting models—single- versus multidealer indexes—have become a major topic of debate in the fixed-income indexing arena.
Before exploring the dynamics and differences between the single-dealer index model and the independent index provider in more detail, it is worth briefly summarizing the consensus view of what constitutes a good index. Most of these views apply whether an index is primarily used for benchmarking or product structuring, and whether it tracks a market or an investment strategy.
A well-designed index should be clearly defined with transparent rules. The components should ideally be investable and have pricing available. The index should also reflect current market dynamics and have daily performance data available. For an index to gain acceptance and become an established benchmark, other aspects of index provision should also be considered, such as independence in the calculation and management of the indexes. Lastly, an additional issue when assessing an index is the systems and technology competence available to support the index effort.
A good index for product structuring contains similar characteristics to those of a broad benchmark, but with more focus on investability. Replicating a broad benchmark for an index-tracking product would require purchasing the few thousand issues in the index to ensure minimum tracking error, which can be difficult. There are three approaches to solve this. The asset manager can invest in a sample portfolio that has a low tracking error to the broad index. Alternatively, the index provider can create a smaller index that has a small tracking error to the broad index using bonds and other instruments such as swaps. The last option is to use a smaller index (“liquid index”) that represents the asset class but with a much smaller set of bonds.
Sponsors of index-tracking products use some or all of the above options, which gives the investors a good choice. Investors requiring that their investments reflect the performance of the broad benchmark may prefer the first option, and many asset managers have built an excellent track record in replicating broad benchmarks using sophisticated quantitative sampling techniques. Other investors may prefer to be less dependent on an asset manager’s sampling skills and may prefer a fully replicating product that tracks a smaller index.