New to bond ETFs, or just looking to freshen up your knowledge? We tackle several of investors' most frequently asked questions about how bond ETFs work.
How Are Bond ETF Yields Calculated?
Yield is the relationship between a bond's coupon and its current market price. A bond's coupon is fixed. Its market price is not. The math behind yield can get complicated, but the basic gist is: When a bond's price goes down, its yield goes up, and vice versa.
Simple enough, right? It is—until you consider that bond ETFs hold hundreds—even thousands—of individual bonds, each of which have different individual yields. So how do you calculate the yield of the ETF as a whole?
There are four main statistics that track the yield of a bond ETF portfolio. They include:
- Yield to maturity. Yield to maturity is the weighted average yield of all the bonds in an ETF's portfolio, assuming they were held until maturity. As such, it reflects behavior of the ETF's underlying bonds, versus the ETF itself. Note that yield to maturity doesn't factor in the fund's expenses, which means it may overestimate the yield that investors actually experience. Plus, ETFs don't often hold the bonds in their portfolio until maturity, making this estimate somewhat hypothetical.
- 30-day SEC yield. 30-day SEC yield is calculated by annualizing the ETF's last 30 days of income, then subtracting fund expenses. By using the ETF's actual distributions and expenses, it gives investors a real-world sense of how the ETF will actually perform. However, because it's so short term, it's susceptible to being pushed around by large, spiky distributions.
- Distribution yield. Distribution yield is calculated by annualizing the most recent distribution—income, capital gains, etc.—then dividing by the ETF's current net asset value (NAV). It's useful as a shorthand, given that it uses the fund's up-to-date NAV and cash distributions that it's actually made. But distribution yield can also fluctuate wildly should the size of an ETF's distributions spike up or down between payments.
- 12-month yield. 12-month yield is calculated by adding all the interest paid over the past 12 months, then dividing it by the sum of the ETF's most recent NAV and any capital gains distributions made over the past year. As such, 12-month yield is rooted in real-world observations about an ETF's behavior. However, it will always lag any big changes in yield. Plus, it's impossible to calculate for ETFs less than one year old.
These yield calculations can often differ significantly for the same fund, often by tens or even hundreds of basis points. At ETF.com, we supply average yield to maturity figures for all bond ETFs, as it is the most accurate look at the current real-world holdings yield of a bond, in our opinion. But no one measure is perfect: To get a complete picture, take a look at all four statistics:
Why Do Bond ETF Values Drop When Interest Rates Rise?
In a rising-rate environment, prices of older bonds must drop to stay competitive. To understand why, let's look at an example (See "Fixed-Income ETFs: Understanding Duration"):
- A hypothetical $100 bond has a 5 percent coupon—meaning, every year, the bond will pay out $5 to investors until it matures. Then interest rates rise 2 percent. The bond issuer decides to issue a new bond that's identical to the first, except that it now carries a 7 percent coupon.
- Given the choice between the 5 percent bond and the 7 percent one, who would choose to own the one that pays less money? Nobody, if the price remained the same $100. It doesn't, of course. The bond's price drops to make it comparable to the yield of the 7 percent bond. So if you own the 5 percent bond, you'll still receive your $5 every year, but the bond's market price—or the price you'd get if you sold the bond today—would decrease.
- That's vastly simplifying the matter, though. Different bonds react very differently to the same movement in interest rates.
The further out a bond's maturity date, the more sensitive it is to rate increases. All else being equal, a 10-year bond carries more interest-rate risk than a five-year one, because your money runs the risk of rising interest rates for a longer period of time.
Duration is a time-weighted measure of interest-rate risk. Duration estimates how a bond's price will change in response to changes in interest rates. Higher durations mean more interest-rate risk. A duration of 3.5, for example, means a bond's value will decrease 3.5 percent, given a 1 percent rise in interest rates.
There are a few things to keep in mind about duration:
- Duration is an estimate, not a certainty. Though bond values go up when interest rates go down, it isn't a one-to-one relationship. Duration tends to underestimate price increases from falling yields, while overestimating price decreases from rising yields.
- Duration assumes a simplified interest-rate environment. Duration is calculated assuming that interest rates change by 1 percent across all maturities; in other words, when rates change, the entire yield curve will shift up or down by 1 percent. Reality is never so precise.
Do Bond ETFs Pay Interest?
Bond ETFs usually make monthly income payments. One of bonds' biggest benefits is that they pay out interest to investors on a regular schedule, usually every six months. But bond ETFs hold many different issues at once, and at any given time, some bonds in the portfolio may be paying their coupon. As a result, bond ETFs usually make coupon payments monthly, rather than semiannually. The value of this payment can vary from month to month.
How Do Bond ETFs Track Their Indexes?
Traditional bond indexes make great benchmarks, but terrible portfolios. Most equity ETFs hold every security in their index. But that's usually not possible with bonds. Bond indexes often include hundreds, even thousands, of individual securities. Buying up all those bonds for an ETF's portfolio is not only difficult, but expensive. The cost to purchase thousands of bonds in illiquid markets, even those with minimal impact on the index, can quickly erode returns.
Bond ETF managers often optimize their indexes. To avoid high costs, fund managers often must pick and choose which securities in the bond index to include in the ETF. They'll opt for bonds that provide the most representative sample of the index, based on credit quality, exposure, correlations, duration and risk. This process is known as optimization, or sampling.
Optimizing saves on costs but has its risks. Depending on how aggressively an ETF's portfolio was optimized, its returns may drift away from those of its index over time. Most ETFs track their indexes very well; however, a few have missed their benchmark by a few percentage points or more per year. (See "How To Run An Index Fund: Full Replication Vs. Optimization")
Why Do Bond ETFs Sometimes Show Big Premiums And Discounts To NAV?
It comes back to how the individual bonds making up the ETF are priced.
Individual bond values are hard to calculate. Without an official exchange, there's no single agreed-upon price for the value of any particular bond. In fact, many bonds don't even trade daily; certain types of municipal bonds, for example, can go weeks or even months between trades.
Fund managers need accurate bond prices to calculate NAV. Both mutual fund and ETF managers rely on bond pricing services, which estimate the value of individual bonds based on reported trades, trading desk surveys, matrix models and so on. It's not a sure thing, of course. But it's a good guess.
An ETF's share price isn't exactly its NAV. A bond mutual fund's share price is always exactly its net asset value, or the value of the underlying securities in its portfolio. A bond ETF's share price, however, can drift, depending on market supply and demand. Premiums develop when share prices rise above NAV, and discounts develop when prices fall below NAV. There is, however, a natural mechanism in place to keep a bond ETF's share price and NAV aligned: arbitrage.
APs use arbitrage to keep ETF share prices and NAV in line. A special class of institutional investors known as "authorized participants" (APs) have the ability to create or destroy shares of the ETF at any time. Should an ETF's share price dip below its NAV, APs can make money on the difference by buying up shares of the ETF on the open market and trading them into the issuer for an "in kind" exchange of the underlying bonds. To profit, the AP simply needs to liquidate those bonds. Likewise, if an ETF's share price rises above NAV, then APs can buy up the individual bonds and trade them in for ETF shares. Arbitrage creates a natural buying or selling pressure that tends to keep an ETF's share price and NAV from drifting too far from each other.
In stressed or illiquid markets, an ETF's price may be below its reported NAV by a lot, or for a long period of time. When that happens, it essentially means the ETF industry thinks the bond pricing service is wrong, and that they're overestimating prices for the fund's underlying bonds. In other words, the APs don't believe they can actually liquidate the underlying bonds for their reported values. For ETF investors, this means the ETF price falls to a discount to its NAV. (The reverse process is true for any premiums that may arise.)
Large premiums and discounts don't necessarily signal mispricing in a bond ETF. An ETF's market price can actually be a better approximation of the aggregate value of the underlying bonds than its own NAV, and highly liquid bond ETFs can perform price discovery for the bonds they hold.
Other Articles Of Interest