Back in September, my colleague Justin Bender and I published a white paper entitled After-Tax Returns: How to estimate the impact of taxes on ETF performance. Justin has now updated his Excel calculator and made it available for free download on his blog.
Recently Justin put his own methodology to work by measuring the 2014 after-tax returns of 10 short-term bond ETFs. As it happens, 2014 was a relatively good year for short-term bonds: as interest rates fell again, ETFs in this asset class delivered returns between 2.3% and 3.5%. But as we have written about before, traditional fixed income ETFs tend to be full of premium bonds, which are notoriously tax-inefficient because of their high coupons. If you held one of these ETFs in a non-registered account, your after-tax return would have been lower.
Launched in 2013, the First Asset 1-5 Year Laddered Government Strip Bond (BXF) was designed to be a tax-friendly alternative. Strip bonds do not make interest payments like traditional bonds do: rather, they are sold at a discount and mature at par value. This means a strip bond is more tax-efficient than a high-coupon premium bond, because you pay tax only on an amount of interest equal to the strip bond’s yield to maturity. Even if a traditional bond ETF and a strip bond ETF enjoyed the same pre-tax return, one should expect the latter to outperform on after-tax basis.
Putting bond ETFs to the test
That’s the theory, anyway. But how would all of this play out in the real world? As Justin describes in his recent blog, BXF delivered on its promise in 2014. Not only did it produce the highest return before tax, it was more tax-efficient than all of its peers—and by a wide margin. Here are the performance numbers for 2014:
| Short-Term Bond ETF | Before-Tax Return | After-Tax Return | Tax Cost Ratio |
| First Asset 1-5 Year Laddered Government Strip Bond (BXF) | 3.49% | 2.76% | 0.70% |
| Vanguard Canadian Short-Term Corporate Bond (VSC) | 3.39% | 1.84% | 1.51% |
| iShares Core Canadian Short Term Corporate + Maple Bond (XSH) | 3.39% | 1.83% | 1.51% |
| BMO Short Corporate Bond (ZCS) | 3.25% | 1.63% | 1.57% |
| Vanguard Canadian Short-Term Bond (VSB) | 2.82% | 1.58% | 1.21% |
| BMO Short Provincial Bond (ZPS) | 3.09% | 1.54% | 1.50% |
| iShares Canadian Short Term Bond (XSB) | 2.80% | 1.48% | 1.28% |
| BMO Short Federal Bond (ZFS) | 2.31% | 1.25% | 1.03% |
| iShares 1-5 Year Laddered Corporate Bond (CBO) | 3.32% | 1.22% | 2.03% |
| iShares 1-5 Year Laddered Government Bond (CLF) | 2.94% | 1.20% | 1.70% |
Sources: www.canadianportfoliomanagerblog.com, CDS Innovations, BlackRock Canada, BMO Asset Management, Vanguard Investments Canada, First Asset
Justin’s methodology is explained in detail in our white paper, but it’s complicated. If you’re not prepared to wade through the details, here’s what you need to know. The before-tax returns in the table above are those published by the ETF providers themselves. These are total returns that include any increase in price during the year, as well as all distributions, which are assumed to be reinvested immediately. The after-tax return assumes that each distribution is first taxed at the highest marginal rate, and only the remaining amount is reinvested. The tax cost ratio in the far-right column offers a simple way of comparing different funds: the lower the number, the more tax-efficient the ETF.
What jumps out from this table is not only BXF’s low tax cost ratio, but also the extreme tax-inefficiency of iShares’ popular laddered bond funds, CBO and CLF. These two funds have obvious appeal for income-oriented investors: CBO has an average coupon of almost 4.8%, while CLF comes in at 3.8%. But in this era of low rates, the yield to maturity on short-term corporate and government bonds is nowhere close to that—in fact, for CLF it’s now well under 1%. That means unless rates keep falling, investors in these ETFs should expect hefty payouts but also a significant decline in price. If interest remain unchanged, both funds could easily deliver negative after-tax returns despite their high yields.
My guess is that many investors (and their advisors) are unaware of this danger. CBO has more than $2.2 billion in assets, making it the fourth-largest ETF in Canada, and CLF is in the top 20 with almost a billion. Surely not all of those shares are being held in RRSPs and TFSAs. Meanwhile, BXF has attracted just $18 million despite its inherently more tax-friendly structure. Another example of how juicy yields can seduce investors into making poor decisions.
How’s your brokerage?                  Â
This year MoneySense and the research firm Surviscor are again teaming up to review Canada’s top discount brokerages. (I wrote the first two editions of this feature, the most recent of which is available here.) The magazine is looking for DIY investors who are willing to share their online brokerage experiences—positive or negative—for this article.
If you’re interested in participating, please contact MoneySense through its Facebook page or email features editor Romana King directly.
Is there any reason why the most frequently used bonds ETF (ZAG, VAB, XBB, XQB, etc.) were not included. It would also have been interesting to include HBB since it has a different structure.
@Linda: All of the ETFs you listed track the broad bond market (average term about 10 years). Those in the table above are short-term bonds only (average term about 3 years). The comparison would be unfair. In any case, HBB (as well ZDB, another potentially tax-efficient fund) do not have a full calendar year of returns yet. As soon as they do, we’ll do a comparison.
But for the record, the before-tax and after-tax return of HBB would be exactly the same, because Justin’s methodology only measures the effect of taxes on distributions, and HBB makes no distributions. The only tax consequences of HBB would show up when the shares are sold, and that is not considered in the methodology.
Great post! Would you have any additional information on how you and Justin arrived at the after tax returns?
I took a brief dive into the white paper, but was hoping you might have a simple comparison of BXF vs CLF (or CBO).
Thanks,
Ty
Great Post Dan ! thank you.
You are right: the tax inefficiency of even short term high grade corporate and gov bond ETFs is a big headache.
1- The only drawback is that fact that the fund is so small; less than 20M in assets. Otherwise looks great (no highly taxed coupons, short duration, good credit quality, low fees). As a general principle I tend to stay away from funds/ETF with less than 100M in assets as recommended at a seminar I attended. A detailed study last year presented by Ernst & Young (where I managed to slip into a seminar directed specifically at investment bankers) showed that in the current cutthroat ETF environment, the chances of survival of small (less than 100M, new ETFs are rather dismal. It was not a sales pitch: the panel included top ETF managers from Blackrock, Vanguard and others, and they were basically discussing where the ETF industry is going. In other words, the idea of a strip laddered ETF is great, but I doubt that BXF can survive.
2- for all the reasons above, I would tend to favor TSE:XFR (iShares Floating Rate). It has all the qualities mentioned above, including favorable after tax returns, and in addition AUM of 247 M.
I would be very happy to know your perspective on these 2 points,
thanks again very much for a great post,
Peter
Sorry for posting a second comment.
Yes, in retrospect, the after tax performance is great, but the current, as of today, pre-tax weighted average YTM is only 0.85%. Do you still think that it’s a worthy investment?
Just had a question; Seeing that BXF by far performed the best, are there any particular reasons why the BXF ETF is no where to be found in the recommended ETF listing or the Model Portfolio sections of the site?
Cheers!
Thanks Dan. I try to hold all of my bonds inside my RRSP. I’ve been using a high interests saving account (TDB8150) for short term funds in my non-registered account. But this looks like a good alternative.
@Peter: ETF closures are a pretty minor risk. If an ETF does fail, investors simply have their money returned to them: the worst that can happen is that there is a forced capital gain. Remember, too, that BlackRock and Vanguard have a vested interest in telling investors to avoid smaller ETFs, since they are the biggest players in the industry. A smaller firm like First Asset isn’t likely to attract $100 million in an ETF in a year or two, and I don’t think it’s fair to simply cut them out of the picture for that reason, especially when they are innovating. (Where are the tax-efficient bond funds from iShares and Vanguard?)
Floating-rate bonds are a different asset class and not likely to behave in the same way as traditional short-term bonds, so it’s an apples to oranges comparison. Notably, the price of a traditional bond goes up when rates fall, while a floating-rate fund would benefit when rates rise. I’m not sure I really see the benefit in floating rate bonds when you can simply use a high-interest savings account that also has a floating rate.
@Justin: BXF is a specialized product: a short-term bond ETF designed specifically for taxable accounts. My model portfolios are far more generic, and I assume they will mostly be used in registered accounts. That said, I probably should add it to the recommend ETF list, with the caveat about how it should be used.
Dan, thanks very much for your reply.
What do you make of the current average pre tax weighted YTM of only 0.8%.
Ah of course, I didn’t even stop to think that. Thanks as always Dan.
@Peter: Unfortunately that is the low-interest rate environment we’re in. CLF has an even lower yield-to-maturity. You can certainly make an argument for just using a high-interest savings account instead, though the advantage of a bond fund is that bonds tend to go up when stocks fall (or interest rates decline) so they can be a better diversifier than cash.
I am enjoying – and learning from – your posts and now read each one thoroughly. Unfortunately I used to skim through some and “cherry pick” the ones to read more closely. Thanks for your efforts.
Dan, any insight on the new iShare bond ETF: XSI?
@JFCCP: Seems like an expensive actively managed fund to me.
@Peter – your question about the size of an ETF is a common one. Size/less trading volume is often highlighted by some market participants in order to dissuade some investors. It is ironic, because every ETF provider launches new ETFs consistently, and they all start small! I would say the key is to know the ETF company and assess whether the company is dedicated to the business. I can state definitively as President and CEO that First Asset is dedicated to ETFs and to being a consistent provider of smart solutions like BXF.
Since it passed it’s one year anniversary, we are getting more and more inquiries and interest in BXF. We traded a $10MM block and a $2MM block this year already. I think it is just a matter of time before it’s inherent benefits catch on, and I hope we can make a convert of you!
Thanks
I fail to see how they got the 3.49% before tax return figure for BFX. Can someone enlighten me?
In 2014 BFX.TO trading price went up from $10.06 to $10.11. That is about 0.5% in capital gain. Combine that with approximately 1.75% from the distributions from 2014. Wouldn’t that be closer to 2.25% before tax?
Even factor in the small compounding effect of reinvesting the distributions. That is still quite far away from the 3.49% figure. What am I missing here?
@Wali: The price appreciation is measured using NAV values from December 31, 2013 ($9.97) to December 31, 2014 ($10.14). That’s 1.71%. Add another 1.75% in distributions with some compounding and you get 3.49%.
I still have trouble figuring out all these taxes. Do you think you could post recommended funds for a taxable account? Including all US, Can and Int? I have my stuff mostly in a taxable account and I keep flip flopping on what to have everything in. My financial guy is against corporate class funds (for some reason). But I am starting to invest myself and have used your funds as a guide. But as I learn more I constantly doubt what I am doing and think I should buy all these tax favorable items. I do save in a corp account so that could change the tax ramifications?
In your opinion should I sell my VSB and get a more tax favorable bond?
Thx in advance.
Fascinating, thank you for the explanation. ETF and stocks in general are new to me and your website is very helpful and informative. Please keep up the good work.
I am a bit shocked at the discrepancy between the before tax return derived from the market price and NAV. Would it be correct to say that the market price for BFX was trading at a premium at the end of 2013 and at a discount at the end of 2014?
I can see why NAV is used for the comparison but as an investor isn’t the market price more relevant because ultimately that’s what I pay and what I will ultimately get when I sale?
At last, is it generally a good time to buy when NAV is higher than the trading price?
Thanks in advance.
Which is better in what?
Let’s say bonds get 4%/year, while stocks get 8%/year.
Capital gains are taxed half so I’d get taxed on 4%, which is same as the bonds portion.
So should I put bonds in Reg accounts (TFSA/RRSP) or stocks instead?
The original plan was to put bonds in TSFA/RRSP, but if the returns are 4 and 8 percent, I’d get taxed the same.
Or is the answer: Taxed now versus taxed later?
@Wali Le: Back in late 2013 BXF was not trading frequently, so the last trade in December occurred on December 16, 2013. The daily posted market by the market-makers would have always reflected fair value, so if it had traded on Dec 31, you would get the result you expected. To be clear, you can always buy or sell it at fair value on any given day, but especially in cases where an ETF didn’t trade on the last business day of the year, you get an accurate picture of the ETF’s performance from the NAV calculation. All performance calculations are done on the basis of NAV for this reason, regardless of the ETF company. Hope that helps
Barry
Will you or Justin be able to do a similar comparison between ZAG, VAB, XBB and XQB? It would be really helpful for DIY investor. Thanks!
Thank Barry for the explanation.
So, for a less popular or less frequently traded fund, what happens when I want to sell and cash out?
Would I have to wait days for a matching bid?
Will the ETF fund issuer sell some underlying asset and buy the back the ETF like a mutual fund?
Thanks again.
@Wali Le: Barry may want to add something, but in general you should understand that ETFs have designated “market makers” whose job it is to fill orders close to the NAV at any time. So you will never have to wait to see your order filled, even if there is no other individual investor on the other side of the trade. These may help:
https://canadiancouchpotato.com/2011/12/28/an-etf-creation-story/
https://canadiancouchpotato.com/2012/09/10/etf-liquidity-and-trading-volume/
@Wali Le: Sorry, I should have been clearer. At all times throughout a trading day there are market makers who post a bid/ask based off underlying NAV, so you can buy or sell anytime, any day the markets are open. BXF typically is quoted 2 cents wide ie: 10.32 bid and 10.34 offered. So if you want to trade, place a limit order – ALWAYS trade with limit orders, no matter what ETF you are buying or selling – and you will transact at the posted price.
The liquidity of an ETF is driven off the liquidity of its holdings, which is reflected in the bid/ask spread. We’re trained to think of liquidity in terms of historical volume traded when we’re dealing with individual stocks or bonds, but that is of far less importance for an ETF. An ETF may trade nothing one day, then $10MM in one trade the next.
Thanks
Thank you guys. I learned a lot.
For the bond portion of my portfolio, how much emphasis should I place on the credit rating?
For instance, XSB has over 50% in AAA but also 8.5% in BBB. First Asset BXF only has 19.6% in AAA, but nothing below A.
If we know the capital loss that the bond fund is accumulating, surely they should publish this so that unitholders could claim this against capital gains from our stock holdings and null out this loss.
Sounds like a reason to buy bonds directly to have access to the potential capital losses (although I know the spreads on bonds look huge in today’s low rate environment).
@rob Information on a fund’s accumulated capital losses is, I believe, included in their annual reports. However, there’s no way that you could use this information for the purpose you propose. There’s no mechanism to flow a loss out to unit-holders. How do you distribute a loss? Ask your unit-holders to pay you?
@WS: We have tried to argue that you should never hold traditional bond ETFs in a taxable account, so I am not sure what benefit there would be in comparing the tax-efficiency of the various bond funds you listed. They will all look bad. Comparing the after-tax performance of ZDB with traditional bond ETFs would be useful, but we don’t yet have a full calendar year of data.
@rob: Not sure if I am misunderstanding your question, but as YYC81 mentions, a mutual fund or ETF cannot pass along capital losses to unitholders. It can only use those losses to reduce the gains it passes along. As for buying bonds directly to take advantage of capital losses, this isn’t wise either. With a premium bond, the capital loss will be offset by higher interest payments, which are taxed at twice the rate of the capital gains you’re trying to offset. It’s like paying a dollar to save 50 cents.
@Wali Le: There is a big difference between high-yield (junk) bonds and investment-grade bonds, but I’m not sure I would worry too much about the credit risk of bonds at the low end of investment grade (that is, BBB as opposed to A). This is especially true of short-term bonds, since the fortunes of the company issuing them would have to change pretty dramatically and swiftly to put these bonds at risk of default.
If the greater after tax return of the BXF is largely due to having a relatively lower yield/distribution payout and higher capital gain compared to its competitors,
can we always expect BXF to behave similarly? That is having lower yield and higher CG relative to the traditional short term bond ETF?
Is the higher CG lower yield an inherent property of using strip bonds?
Thanks
@Wali Le: Strip bonds, by definition, always trade at a discount, so they will always be inherently more tax-efficient than premium bonds.
https://canadiancouchpotato.com/2013/06/07/why-use-a-strip-bond-etf/
However, traditional bond ETFs won’t always hold so many premium bonds. If interest rates trend upwards over the next several years, older bonds will see their prices decline, perhaps even below par. At some point, that will make traditional ETFs much more tax-efficient than they are now.
Thanks again. The bond sector of ETF investing is really complicated and can be confusing at times.
It is easy to compare ETF tracking a particular stock index e.g. S&P 500. Compared the MER, structure, currency exposure, tax implications and that’s about it. With bond ETF, it is not nearly as straight forward.
As a novice investor to ETF, it is easy to be drawn to the established products with x billions in asset. But I failed to consider that a large portion of that could be hold in registered accounts. Thank you for pointing that out.
I think for the non-registered account, products like BXF and to some extend HBB are worth considering. Especially for someone younger like myself, the difference compounded over 20-30 years can be very significant.
Very good info as usual. I have been catching up on my reading. I have 2 questions. I presently hold xef, xec and vun. But looking at the new ishares fund does it not makes sense to just sell these and buy one fund to keep trading costs low? That way only need to keep allocation right with one find instead of 3?
Reading about about bonds. Does it make sense to buy the bmo bond etf in a taxable account or wait cause if the rates go up the advantages will be marginalized? I have vsb and xbb. But it seems they are virtually worthless in tax accounts. Is it better to sell and buy these more tax efficient funds?
Thanks
@Bruce: If you hold all three asset classes (US, international, emerging markets) in the same account, then a fund like VXC or XAW is certainly convenient. That’s what we often recommend for our DIY clients.
My next blog post should answer your question about ZDB, so stay tuned!
CCP, the article mentioned that if interest rates go up, high distribution Bond ETFs such as CBO will see a price decrease and significant payout. So if interest rate does go up, then it is not sustainable for CBO to continue the high distributions right?
@Alan: While the yield on a bond changes when interest rates move up or down, the coupon does not change. (The coupon is the interest rate established when the bond was first issued.) So the distributions will continue at the same level as long as CBO holds the same bonds. But if rates rise and prices fall, then the gap between the average coupon and the yield to maturity will narrow.
This is all pretty confusing, I realize. Bottom line, don’t hold CBO in taxable account.
@CCP. Since CBO holds short term bonds, I assume they need to purchase new bonds every year. And if interest rate rises, then I assume coupons would be higher for these new bonds. I plan on holding this ETF in a RRSP/TFSA to generate income for retirement, is there anything I should consider/worry besides price falling? Thanks.
Would it still not make sense to choose bxf even in a sheltered account over the other funds in the list?
Phil
@phil: In hindsight, sure. But there is no reason to expect BXF to continue its performance on a before-tax basis. Its tax-efficiency, however, should be expected to continue.
@CCP:
Do you think that BXF could be used for very short term investment purpusose (6-12 months) or you would recommend a plain saving account for short term investments? For example, while saving for the cash down for a new house, or saving for a new car purchase, etc.
@Jas: I would not recommend any bond fund for such short-term savings. There is always the risk such a fund could fall in value. A savings account is the only safe bet for short-term money.
I suggest updating the article title to “Which Short-term Bond ETF Is Most Tax-Efficient” so it’s easier to understand what this is about for future reference. I’d love to see another post about tax-efficiency of broad bond market ETFs.
This article is now more than 4 years old. Any significant development have occured that might change its conclusion i.e. for unregistered account BXF is the favorite short term bond tax effcient ETF out there?
@NRH: ZDB is also a good choice for bonds in a taxable account, but it has a significantly higher duration than BXF, so it will be more volatile. There have not been any other tax-efficient short-term bond index ETFs launched.