Robeco: Bond ETFs hampered by low liquidity

In a recent report, Patrick Houweling, head of quantitative credit research at the Dutch firm, found the tracking error for high-yield ETFs ranged from 7.3% to 12%, while the highest average tracking error for investment grade ETFs reached 1.63%.

There are only three high yield ETFs in the market, and all significantly underperformed their respective benchmarks, Houweling found.

From December 2007 through March 2010, the SPDR Barclays Capital High Yield Bond fund managed by State Street Global Advisors (SSgA) fell short its benchmark, the Barclays Capital US High Yield Very Liquid index, by an annualised six percentage points.

The PowerShares High Yield Corporate Bond fund underperformed its benchmark, the Wachovia US High Yield Bond Index by an annualised 5.9 percentage points. The iShares iBoxx USD High Yield Corporate Bond fund, run by BlackRock, was down by an annualised 2.5 percentage points against its benchmark, the iBoxx USD Liquid High Yield index.

"The research is showing that it's quite hard to generate index-like returns, particularly in corporate bonds. It all has to do with the illiquidity of the market," said Houweling. "You can't turn an illiquid asset class into a liquid investment by magic," he said.

At least one ETF provider agreed that high tracking errors and lower returns were greatly impacted by illiquidity in the markets, particularly in early 2009, but said the same market issues plagued active managers.

SSgA senior managing director Jim Ross said: "The challenge broadly in the bond universe is an issue of illiquidity."

"Can ETFs track bonds on a laser point like they can in equities? No. But that doesn't mean it doesn't work," he added.

He described indices as "theoretical" because they don't include costs, where as ETF returns are ultimately impacted by the cost of the product and the cost of trading. "The cost of trading is significantly higher in fixed income than in equities," he said.

He said about three-quarters of the tracking error can be attributed to the first quarter of 2009.

In an e-mailed statement to GP, BlackRock officials said: "Our dedicated BlackRock portfolio management team controls that all the risk features of the fund are in line with the index on an ongoing basis. However, usage of the optimisation technique means that the benchmark tracking might not be perfect. Market volatility, low transparency of the bond prices and trading costs are additional factors that can influence tracking error."

Optimisation is when the fund owns some of the securities within a benchmark with a goal of replicating the broader benchmark, a tougher task when there are more constituents in the index.

BlackRock added: "It is important to emphasise however, that for a strategically focused ETF investor the tracking difference between the fund and the respective benchmark would be a more valuable and applicable parameter to assess. For the majority of the Fixed income iShares ETFs tracking difference stays very close to the annual TER (Total Expense Ratio) of the fund."

iShares is by far the largest provider of investment grade credit ETFs. According to the Robeco reports, iShares offers 11 different funds, versus three each for SPDR and Vanguard, two for Lyxor and one for Amundi.

iShare's average tracking error for its combined investment grade funds was 1.61%, and the average performance versus their benchmarks were -0.67%.

The SPDR investment grade ETFs had an average tracking error of 1.63%, but an average underperformance of 2.56%.

Officials at PowerShares could not comment before press time.