A new ETF of strip bonds

Back in February I asked readers to share their ideas for new ETFs that might fill gaps in the current marketplace. The idea for that blog post came from a contest run by First Asset, who invited advisers to submit their suggestions and promised to launch a new product based on the best idea.

I offered a couple of my own: an international equity ETF that doesn’t use currency hedging, and an international bond ETF. As it happens, iShares answered the first call in April with the launch of the MSCI EAFE IMI Index ETF (XEF), which I wrote about here. Vanguard listed an international bond fund in the US last week, though this isn’t suitable for Canadian investors who want to avoid currency risk with their fixed income.

Justin Bender strips for charity

The winning suggestion turned out to be more innovative than either of these ideas, and it came from none other than my colleague Justin Bender at PWL Capital. The new ETF, which will launch in June 11, is called the First Asset DEX 1-5 Year Laddered Government Strip Bond Index ETF (BXF). As the winner of the contest, Justin secured a $10,000 donation for his favourite charity, the Centre for Addiction and Mental Health.

As the name suggests, the ETF will hold a five-year ladder of strip bonds. These fixed income products (also called zero-coupon bonds) do not make semi-annual interest payments like traditional bonds. Instead, you buy them at a discount and they mature at face value. Their yield is calculated by comparing the bond’s purchase price and its future value. For example, if a five-year strip bond with a face value of $10,000 is purchased for $9,057, it has a yield of 2%— because $9,057 invested for five years at 2% and compounded semi-annually would grow to exactly $10,000.

Investors like strip bonds because they remove what’s called reinvestment risk. With traditional bonds, the yield to maturity is only an estimate, because it assumes all interest payments will be reinvested at current rates—and that’s not likely to be the case, since rates fluctuate constantly. With strip bonds, however, there are no interest payments to reinvest, so this risk disappears and the investor knows exactly how much she will receive on the maturity date.

Building a ladder of strip bonds is a popular strategy for fixed-income investors, and it’s one advocated by Hank Cunningham in his excellent book, In Your Best Interest: The Ultimate Guide to the Canadian Bond Market. However, Cunningham points out that such a ladder is only appropriate if you don’t need income (since there are no interest payments). He also argues that strip bonds are unsuitable for taxable investors, and that opinion is widely held. The most common place for a ladder of strip bonds is in a long-term investor’s RRSP.

How strip bonds are taxed

Every investor knows that fixed-income investments are best held in registered accounts, because interest is fully taxable at your marginal rate. But with strip bonds, there’s another wrinkle: you pay that tax on income you don’t actually receive. That’s because the difference between your purchase price and the bond’s face value is amortized over the life of the bond and taxed annually as though it were interest.

To return to the example above, let’s look at the tax consequences of that five-year strip bond with a yield of 2%. The bond is purchased for $9,057 and matures at $10,000, and this discount is amortized over five years and taxed annually. Meanwhile the adjusted cost base of the bond increases each year by the amount of this notional interest, until it reaches $10,000 at maturity:

The holder of this bond would therefore pay tax on about $181 to $196 of income annually, despite receiving no cash payments. It’s easy to see why the conventional wisdom says strip bonds should not be held in non-registered accounts.

It turns out, however, that when Justin came up with the idea for an ETF of laddered strip bonds, he was looking to solve a problem in taxable accounts, not RRSPs. And the new First Asset ETF is designed with a couple of surprising features that offer a possible solution. I’ll explain more later in the week.