Why Bond ETFs Often Do Not Track their NAVJune 8th, 2012 by David Waring
Net Asset Value (NAV) is the value of all the the holdings, bonds, stocks and all other assets, held by an ETF, divided by the number of shares. In plain english, NAV is what an ETF should be worth. However, ETFs which trade just like stocks, have the potential to have a market price that is very different than their their NAV. To understand why the market value and NAV might be different, you must first understand how ETF shares are created and eliminated.
ETF companies allow “Authorized Participants” (APs) to create new ETF shares by buying a specific basket of shares or bonds, and exchanging that basket of securities for shares of the ETF. Conversely, an AP can exchange shares they own of an ETF for a basket of securities. These exchanges are based on the NAV of the ETF and not the market price. In theory, an AP can make “free money” by creating new shares when the market price of the ETF is above the NAV and eliminating shares when the market price is below the NAV. This system works very well for stocks and keeps most stock ETFs trading within a couple pennies of the NAV.
The system works differently for bond ETFs.
- In most cases, the price of a bond ETF will be slightly higher the ETF’s NAV.
- Frequently, a bond ETF will trade at one or even two percent away (at a premium or discount) to an ETFs NAV.
Why do Bond ETFs behave differently than stock ETFs?
With stocks the prices where you can buy (the “ask”) and and sell (the “bid”) are usually very close. With bonds, the spread between the bid and ask are farther apart. This creates issues around “deciding” the value of a bond. Does the ETF company use the bid, ask or midpoint value of the the two? Normally, the value which is assigned to a bond for the purpose of calculating the NAV is below the bond’s “ask” price. Thus the cost of buying a basket of bonds under normal market conditions is greater than the NAV of that basket calculated by the ETF company. As a result, APs don’t tend to create new shares of bond ETFs unless the shares are trading at a small premium to NAV.
Many bonds, such as junk bonds don’t trade actively, and therefore don’t have published prices. When trying to establish the NAV for ETFs like high yield bond ETFs, ETF companies therefore have to guess the value of the bonds in the portfolio (they normally use third parties to help them establish the bond’s value). If you actually tried to sell some of the bonds in the ETFs portfolio however, there is not likely to be a ready buyer, since there is not an active market for the bonds. As there is not anyone actually “asking” to buy the bonds, a seller must be willing to give a buyer (normally a bond dealing desk) a price attractive enough to compensate them for potentially being “stuck” with the bonds. In other words when you sell bonds which do not have an active market, you are likely to get a price which is less than the “fair value” of the bonds. Because of this fact, bond ETFs in categories such as high yield can often sell at a couple percent discount to their NAV.
Once in a while, the reverse happens. Demand for a particular type of bond, for example municipal bonds, overwhelms supply. In this case, the cost of acquiring new bonds may be much higher than the estimated NAV. This means that the Bond ETF can trade at a significant premium to the ETFs NAV.