Bond illiquidity poses a significant challenge to fixed income ETF providers as a bond ETF has to track its index closely and efficiently. The lack of liquidity in the bond market, highlighted by the quasi-absence of a secondary market as most bonds are held until maturity, makes it even more difficult for a bond ETF price to converge to its stated NAV and fulfil its advertised promise of instant liquidity. This challenge is bigger for corporate bonds than for government bonds. Bond ETF providers will reduce this risk by limiting the number of bonds held to a representative number of the index’s constituents, although it’s often unclear how the choice of bonds is arrived at. For example, the Barclays Aggregate Bond Index contains more than 6,000 bonds, but the Barclays iShares Aggregate Bond Fund, which tracks the index, contains just over 100 of those bonds.
Another solution is to track a bond index made up of a limited number of bonds. Several corporate bond index providers (namely Barclays, Citigroup, Bank of America Merrill Lynch, J.P. Morgan, Dow Jones, iBoxx and Deutsche Bank), offer “liquid” versions of flagship indices, consisting of a limited number of bonds, typically those with the largest outstanding market value.13 However Campani and Goltz (2011) have shown that reducing the number of bonds could lead to higher instability of interest rate risk exposure of the index, and the smaller the number of bonds the greater the instability. As discussed previously, replacing a constituent with one with different characteristics will impact the overall risk exposure of the index. This impact will clearly be more important for an index with a small number of securities (say 50) than for an index with thousands of securities.
Overall, although liquidity risk is rightly one of the main concerns of index investors, the natural lack of liquidity in the corporate bond market makes it a problematic issue for anyone investing in this asset class. Index providers’ initiatives to provide liquid versions of their indices offer a potential solution but it should be noted that such indices also come with other pitfalls.
Despite their increased popularity, corporate bond indices and ETFs still rely on traditional market-value-weighted indices. Newer index versions, using equal or fundamental weighting, may partly address the well-known concentration problem of debt-weighted indices. However, alternative weighting schemes are often straightforward extensions of equity index methodologies and were never designed to address important challenges in bond index construction, such as duration instability, credit risk classification and liquidity.
When managing credit risk, most existing indices offer little granularity and rely on backward-looking information. Different approaches have been developed by bond index providers to manage indices’ interest rate exposures. Such approaches are often based on targeting specific market segments within single-issuer government bond indices.
For corporate bond indices, where one has to address questions of issuer diversification and interest rate risk control, such an approach may not be straightforward. Finally, many index providers offer corporate bond indices which control liquidity through security selection. Such indices, however, face a natural trade-off: it has been shown that higher investability will lead to higher instability of interest risk exposure.
Overall, it is clear that standard corporate bond indices serve a useful purpose as market benchmarks. They represent the performance of the average investor and thus allow for comparison with a peer group. However, relatively few indices allow investors’ specific requirements in managing credit, interest rate and liquidity risk to be integrated. At best, existing methodologies address only some of these questions. Index providers face a challenge in devising new methods of bond index construction without taking the easy route of simply transposing techniques used in building equity indices.