[Editor’s note: The following is the final installment of a two-part series analyzing best practices in using fixed-income exchange-traded funds in the most cost-effective manner as part of a broadly diversified portfolio.]

In the first part of this study, we examined the controversy surrounding the premiums and discounts in fixed-income ETFs.

The piece reviewed how and why those premiums developed. It also looked at the oft-cited case of the iShares High Yield Bond ETF (NYSEArca: HYG) during the credit crunch of 2008. During that period, HYG traded at wild premiums and discounts to its underlying NAV, causing many to wonder if fixed-income ETFs make sense for any category of bond. (See Part 1 here.)

In this second and concluding portion of our analysis, we take an in-depth look at that question by providing a breakdown of premium/discount data in various corners of the fixed-income market.

HYG’s performance during the market meltdown is interesting, but ultimately, doesn’t tell us much about the typical experience of an investor in a bond ETF. HYG captures the most illiquid segment of the bond market, and October 2008 was a uniquely illiquid period in the history of the bond market.

To find out how ETFs perform during more normal periods, we examined the premiums/discounts for all relevant domestic fixed-income ETFs during the second quarter of 2009.

The study compared the closing bid/ask midpoint to the net asset value for the funds on each day of the quarter. It excluded Vanguard ETFs, which operate as share classes of Vanguard’s fixed-income mutual funds and thus have different trading characteristics than other bond ETFs. It also excluded money market funds, as well as actively managed ETFs, where premiums and discounts would naturally be more variable.

Finally, it excluded international fixed-income funds, since the premium/discount information here is irrelevant: Premiums and discounts are calculated by comparing the closing price of the ETF with the net asset value of the funds’ underlying holdings. For international funds, the NAV freezes when the domestic markets close. For a fund holding European Treasuries, for instance, the NAV would be based on prices at 10 a.m. ET (when European markets close), while the ETF will trade until 4 p.m. It’s not apples-to-apples.

The study shows one thing very clearly: Asset class matters when trading fixed-income ETFs.

The more liquid the underlying market, the lower the average premiums and discounts. The study also highlights a technical glitch in the bond market: ETFs tended to close almost exclusively at a premium to NAV, rarely dipping into discount mode. This is nearly baked in by design: Bond prices are marked on the “bid” when calculating net asset values, while premiums and discounts are measured based on the closing midpoint of the bid and the offer for the ETF itself.

That difference explains much of the positive bias in premium/discount distribution.

Treasuries

The place to start is with Treasuries, the most liquid corner of the ETF market. The message here is clear: Investors should not worry about premiums and discounts in traditional Treasury ETFs.

Of the 10 Treasury ETFs examined, the average premium or discount during the second quarter ranged from 0.00% to 0.12%. The absolute maximum premiums and discount were small as well, ranging from .64% to -0.61%, and staying in much tighter realms for most of the funds.

Importantly, Treasury ETFs traded at both premiums and discounts, which is what you would expect to see in a random walk when both sides of the creation/redemption mechanism are functioning. Sometimes the market closed when the ETF happened to be trading slightly above its true value, and sometimes not. There was no discernible pattern.

Interestingly, premiums and discounts got larger as you moved out the yield curve, which is exactly what you would expect given the higher liquidity of short-term T-bills.

Previous Page

1  |  2  |  3  |  4  |  5

| Full Article  |Next Page