If you’re investing outside of tax-sheltered accounts like RRSPs and TFSAs, you need to choose your investments carefully—otherwise you risk giving a big slice of your returns to the good people at the Canada Revenue Agency. Today we’ll look at ways to create a tax-efficient index portfolio using some innovative ETFs.
Swapping dividends for capital gains
In Monday’s post, I explained that Canadian dividends are not always as tax-advantaged as people believe. Capital gains are not only taxed at a lower rate in the highest tax brackets, but investors can also control when to take them—dividends, on the other hand, are taxable in the year they’re paid, even if you reinvest them.
Horizons’ swap-based ETFs—which I wrote about here—were designed to address this issue. They use a type of derivative that allows investors to earn the same return as the index, without collecting any distributions. Dividends paid by the companies in the index are reflected in the fund’s return, but all of the growth is characterized as capital gains and deferred until the fund is sold. There are currently just two funds in the family: the Horizons S&P/TSX 60 (HXT) for Canadian large-cap stocks, and the Horizons S&P 500 (HXS) for US large-caps.
The tax advantage is especially large for HXS, because dividends from US companies are fully taxable, while capital gains are taxed at half that rate. Consider this: if you held the iShares S&P 500 Index Fund (XSP) in 2011, you received $0.24 per share in cash dividends, a yield of about 1.6%. At the highest tax bracket, you would have lost almost half of that to taxes, reducing your return by about 80 basis points. If you held HXS instead, you would have received a similar 1.6% price appreciation instead, and you would have paid no tax. If you eventually sell the fund at a profit, you’ll pay tax on only half the gain.
Forward thinking
Claymore’s Advantaged ETFs—read a detailed description here—use forward contracts that “recharacterize” bond interest or foreign dividends as capital gains or return of capital (ROC). Unlike swap-based ETFs, which pay no distributions, the Advantaged ETFs are designed for investors who want current income.
Return of capital is the most tax-efficient of all distributions, though it’s not a free lunch. ROC is not taxed in the year it’s received: instead, it lowers your adjusted cost base, and if you sell your shares at a profit in the future, you’ll incur a capital gain. So you’re not getting truly tax-free income—you’re really just getting your own money back—but you are generating tax-deferred cash flow. This document from Claymore explains the idea.
You can see why ROC is preferable to bond interest, which is fully taxable. In 2011, the Claymore Advantaged Canadian Bond (CAB) returned 6.8%. Roughly half of that came from price appreciation, while the other half came from distributions. However, unlike every other bond ETF, those distributions were return of capital, not interest. So you would have collected that entire 6.8% return without a tax bill.
Before you get too excited, there are downsides. All of the 2011 distributions from Claymore’s Advantaged ETFs were return of capital, but that won’t always be the case. In 2010, for example, CAB’s distributions were all capital gains. These would have been taxable—albeit at only half the rate of bond interest.
More important, the Advantaged ETFs have considerably higher costs than plain-vanilla index funds, which will lower their pre-tax returns. Those added costs offset some of the tax savings, and may even wipe out the advantage altogether. For example, both the iShares DEX Universe Bond (XBB) and BMO Aggregate Bond (ZAG) returned well over 9% last year—dramatically outperforming CAB. Even if you lost half your interest income to taxes, you might still have been better off with XBB or ZAG.
If you’re out of RRSP and TFSA room, you could use these ETFs to build a reasonably well diversified and tax-efficient portfolio of Canadian stocks (HXT), US stocks (HXS) and bonds (CAB). In some years—including 2011—you’ll have no tax payable at all. Just spend some time researching these complex products first, and don’t invest in anything you don’t understand simply because you think you’ll save some tax.
H&R Block software giveaway
Speaking of tax, the folks at H&R Block have offered to give away copies of their DIY tax software to five lucky Canadian Couch Potato readers. To enter the draw, leave a comment below or tweet this post to your followers before midnight EST on Sunday, January 29. I’ll announce the winner next Monday.
Hi Dan
First, congratulation to you for being the (acting) editor on the new February/March 2012 issue of MoneySense.
Well written article to address the tax efficiency issue of the couch potato portfolio. I hope you can add a new tax efficiency section in your model portfolio section to include the after-tax return of the potato with different tax scenario.
Please enter me to the draw. Thanks.
I’m in for the draw too – thx!
@Wallace H: Many thanks for congrats.
I may add a tax-friendly model portfolio at some point, though it would be impossible to calculate the after-tax return, because this would be different for each individual.
Great article. Please add me to the draw, thanks.
Interesting post. In Europe I understand there is a move to better regulate derivative based ETFs that use techniques such as swaps as described above. Many people bought ETFs in Europe without understanding the nature of the ETF. It is more of a concern there because of the current fragility, potential instability and undercapitalization of the banks which might put the swap agreements into question should anything bad start happening.
I do not consider it a problem here from reading the prospectuses of the swap ETFs given the capitalization levels of our banks and other indicators of stability.
@Andrew: I agree with you that the swap structure seems to be very low risk in Canada, where the collateral requirements are quite strict. In Europe, not so much.
Dan,
I look forward the tax-friendly couch potato model portfolio. Although tax situation is different for each individual, still, it helps your readers to get an overall picture of how to assemble tax-efficient portfolios outside RRSP and TFSA. For many people, the size of registered portfolio is small because of lack of capital, or deferred contribution to a later time for a better result. The February/March 2012 issue of MoneySense features the article by Sarah Efron included my story of deferred RRSP contribution to minimize tax and keep most money.
Speaking of tax, you gave an interesting example of “re-characterizing” tax-unfriendly income to capital gain/loss by way of forward contracts. By the no-arbitrage forward pricing formula, the value of expected interest/dividend income is reflected in the forward price, which determines the capital gain/loss against the spot price at disposition or at maturity.
I am interested to know how the income replication by forward contracts is affected in response to interest rate change, which is not mentioned in their articles. Stock (spot) price and dividend is affected by many factors, but the forward price is directly affected by (risk-free) interest rate per pricing formula.
@Wallace: Thanks for the comment—I thought I recognized your name. :)
I confess I don’t understand your question about forward contracts. I’d suggest you contact Claymore directly and ask them about how this works in the Advantaged products. They are quite good at responding to investors’ questions.
I too really look forward to the tax-friendly couch potato model portfolio!!!!
@Yves: A simple portfolio of HXT, HXS and CAB would be a good start. If you wan’t to draw some income, have a look at my Yield-Hungry Couch Potato, but take out the REITs and the iShares high-yield bond fund. The other asset classes and ETF choices in that portfolio are all quite tax-efficient.
@CCP:
Very good explanation of the tax efficiencies of Horizon’s Canadian and US Total Return Swaps. But the explanation for the CAB Return of Capital raises some questions in my mind, likely due to my incomplete understanding of Bond Index Fund accounting. I initially thought that interest in ordinary Bond funds was captured tax-free somehow and reflected as an increase in the price of the Fund. I see I may have got that wrong; interest is interest and is taxable.
Say in a Non RRSP, Non-TFSA couch-potato portfolio you hold 40% in CAB as your Income or Stable portion.
In a year, at rebalancing time, say you have received X dollars in cash from ROC (which you can use to buy more CAB or HXT or HXS depending on how the preceding 12 months has affected your original 40:30:30 mix). There is no Canadian Income Tax due on that ROC, right? There may be Capital Gains Tax due on any portion of CAB sold (for cash if needed for rebalancing), and that Capital Gain is increased due to the ROC manoeuvre, but is only triggered in the portion of CAB sold, right?
Meanwhile, in a parallel portfolio, identical except for the 40% Income/Stable Portion which is represented by XBB (iShares DEX Universe Bond). Bond Interest from a fund is fully taxable in the year it is paid out? It is not kept within the fund and tax protected? So let’s say your gain is represented by X dollars in Interest, and Y in increase in value of XBB due to drop in prevailing interest rates. Then X is fully taxable at marginal rate, and Y is capital gains taxed on the portion that got sold. Correct?
You say that in 2010, CAB paid no ROC, and all its distributions as Capital Gains. Why would all that be taxable any different than the Y portion of XBB in the prior paragraph if you didn’t sell any or all of it at that time or subsequently?
I understand that CAB might be too poorly performing to be useful in an RRSP or TFSA. But the benefits of a TRS are more than tax advantage — don’t you get to skip transactional fees as well as hassle in getting more equity automatically during the swap? So if the MER’s were comparable, would not HXT and HXS be reasonable components in an RRSP and a TFSA?
@Oldie: Lots of questions here, but I will do my best.
First, and most important, interest (and dividends, for that matter) is always taxable in the year it is paid out. It doesn’t matter whether your fund reinvests the distributions: you cannot defer the taxes.
With that in mind, your description of CAB and XBB sound correct: you pay no tax on any return of capital, and you pay capital gains taxes if you sell CAB at a profit. Previous ROC payments will lower your adjusted cost base, which will raise your capital gains. However, this is preferable to receiving the income as interest, since cap gains are taxed at only half the rate of interest income. If you hold XBB, you will also pay capital gains on any portion sold at a profit, and any interest payments are fully taxable.
There’s one other issue here. Sometimes funds incur capital gains from selling securities, and they must pass these gains along to you. So even if you buy a fund and never sell any shares, that doesn’t mean you won’t incur capital gains taxes. The fund has a choice of either passing along the capital gain to you in cash, or reinvesting the proceeds and simply sending you the bill at tax time: you get a capital gain recorded on your T3 slip.
CAB makes absolutely no sense in a tax-sheltered account: the additional costs are never worth it unless you can get a significant tax savings. In fact, with yields so low these days, you pretty much have to be in the highest tax bracket for it to be obviously superior.
With HXT and HXS, it’s a bit different: they use total return swaps, which are more cost-effective than forward agreements. HXT has no additional fee for the swap, so the 8 bps MER is all you pay, which makes it a great choice in a tax-sheltered account. HXS adds an additional 30 bps for the swap for a total cost of 45 bps, so it’s more expensive than XSP, but still less than most index mutual funds. It might still be a good choice in a tax-sheltered account because of the constant reinvestment of dividends.
I have done a lot of posts about these products, for example:
https://canadiancouchpotato.com/2011/06/06/understanding-swap-based-etfs/
https://canadiancouchpotato.com/2011/06/20/understanding-claymores-advantaged-etfs/
https://canadiancouchpotato.com/2012/02/06/how-claymores-advantaged-etfs-pay-investors/
@CCP
Thanks for the prompt reply, which was densely packed with thoughtful answers to all my questions. I think you meant HXS rather than the last mention of HXT in the last paragraph (HXT adds ann additional 30bps…). Assuming so, I wonder if HXS would be such a good deal after all. The 0.45% MER compared to 0.08% for HXT is a huge jump. The expense might be associated with the hedging of the Canadian Dollar, which I believe you have mentioned before is not necessarily a good strategy.
I guess evaluating the benefit of HXS in a taxable account has to take into consideration how much foreign dividend would be expected to be paid annually in a conventional US index fund; I have no idea — I was only focussed on the expected long-term capital appreciation. What is the usual amount? (I understand that for conventional US Index Funds, Canadians would be taxed on dividends as if it were 100% income, and there would be a with-holding tax). Depending on the amount of US dividend expected annually, HXS may be a reasonable choice in a large taxable account of a high income Canadian, until a cheaper MER, non-hedged version came around. But the advantages of using HXS in an RRSP seem less stellar now, considering the rather high MER, and that a (cheaper) conventional US Index fund wouldn’t be attracting with-holding tax on dividends, which then are tax deferred in the Canadian RRSP.
@Oldie: Thanks for pointing out the typo, which is now fixed.
I would agree that HXS is not necessarily a great choice for a tax-sheltered account: its structure is primarily designed for its tax benefit. The one exception is if you want to trade in Canadian dollars, you want hedging, and you happen to use a brokerage that offers HXS for zero commissions. :)
Horizons produced a document showing the tax advantage of HXS would be about 90 bps if the yield on the S&P 500 was 2% and you are in the highest tax bracket (46%). That would more than offset its higher cost. Of course, if you are in a lower tax bracket, the savings are smaller. I am trying to track down this document, and I will send it to you.
Are there any US-listed S&P 500 total return swap ETFs? I imagine there will be some soon, once dividend tax rates jump for US investors in January.
That would make a good option for Canadians looking for an unhedged alternative to HXS. Of course, I would worry more about who the counterparties are as I have less faith in US banks.
@Andrew F: I couldn’t find one in the US, but they are common in Europe. Here’s one:
http://www.pensionsource.ie/pmo/funds/SOU023%281%29.pdf
I think you’re right: if dividends are fully taxable, then the swap structure can be a big benefit for long-term investors. I imagine it will be common to spread the counterparty risk around, as the above ETF seems to do.
@Andrew F: For a hypothetical Total Return Swap ETF in the US, regarding ” I would worry more about who the counterparties are as I have less faith in US banks.”, while, admittedly US banks are as a group more sketchy than Canadian banks, the failure of the counterparty bank will only result in the loss of the gain in equity. Your principal as well the interest gained in the initial step prior to the swap is still retained by you. At least that is my understanding of the risk (see CCP post June 8, 2011).
@CCP: OK, one scenario triggers another fantasy– I don’t understand how this works well enough to predict the outcome of the failure of the counterparty bank at the same time that the underlying equity value drops. If the bank is responsible for paying the increased value of the dividend-enriched and value-enriched ETF, and defaults on this responsibility when it fails, then, by the same token, doesn’t a bank failure also carry with its assets (claimed by the creditors) the diminished value of the ETF? Seems to me the individual investor would still hold on to his original investment principal plus accrued interest, as in the scenario of the failed bank when the ETF underlying value increased.
@Oldie: Actually, you’re right on with that observation. It seems unlikely that the counterparty bank would fail during a time of a rip-roaring positive equity returns. If the counterparty failed during a widespread market meltdown like 2008, it seems likely that the index itself would also be way down. And, therefore, the cash collateral held by the ETF sponsor could actually exceed the value of the index. Horizons actually explained this in my interview with them. Not sure if you have seen these posts, but they probably will answer most of your questions:
https://canadiancouchpotato.com/2011/06/08/swap-based-etfs-what-are-the-risks/
https://canadiancouchpotato.com/2011/06/10/more-swap-talk-with-horizons/
@CCP:
“@Oldie: Actually, you’re right on with that observation.”
Which observation are you agreeing with? Actually I made one observation to Andrew F and a conjecture to you. In the conjecture about the failure of the counterparty bank at the same time that the value of the underlying Equities represented in the TRS etf DECLINED in value, I wondered (and still don’t know) if the individual investor could escape a capital loss. You agreed with me that in that scenario…
“the cash collateral held by the ETF sponsor could actually exceed the value of the index”, but I didn’t and still don’t understand the ultimate consequences of that imbalance. I checked again with the June 8 and 10, 2011 posts you referred to, and indeed the Horizons front man referred to that hypothetical happening, but left it at that, and didn’t say what would happen next.
So then what? If the counterparty bank (which has failed) owns the underlying ETF fund value (which has decreased) but has no claim on the Principal, nor the interest on the original investment before the swap, could it be true that the original investor gets to KEEP HIS ORIGINAL PRINCIPAL AND INTEREST AND THE COUNTERPARTY ENDS UP WITH THE LOSS ON THE INDEX SHARES?? This would seem too good to be true, so common sense tells me that it can’t be so, but the logic trail seems to lead there.
@Oldie: Sorry if I misunderstood. As I understand it, the collateral held by the ETF sponsor must always be equal at at least 90% of the value of the index. So if the counterparty fails, the absolute worst case scenario is that you could lose 10% of your capital. However, in practice, it seems likely that the counterparty would only fail during a period when equities crashed. In that case, it’s possible that the collateral could be worth more than the index, which means that investors in the ETF would lose nothing. (I doubt they would have a claim to any excess return, however.)
If you still have concerns, you can always email Horizons directly. They will answer any specific question you have about these products.
@Andrew F:
Further to my comment about the safety of HXS despite the uncertainty of US banks in general, I e-mailed Horizon and got a reply from Marc Burroni, Business Development Manager, and he informed me that the counter party bank for HXS is the same Canadian bank as for HXT, that is, National Bank. His actual words were:
“HXS – the S&P500 tracking ETF does not use a US counterparty, but rather a Canadian one so the index does not contain the common stock of the counterparty.”
which does not make complete sense to me, because, if the intent was for the index not to contain the common stock of the counter party, well they violated that rule in having National Bank as the counter party for HXT; surely the TSE index includes National Bank. But that’s what he said.
I think I knew from earlier discussions that NB was the counterparty for these two funds. You make an interesting point about why the counterparty holding an index of which they are a component should matter, but not in the Canadian example.
@Andrew F:
OK, I got it sorted out — apparently a misunderstanding from an unfortunate turn of phrase.
“Morning,
I think you misunderstood my previous email. I was simply trying to point out that the only difference between HXT and HXS is the underlying index. TSX60 vs the S&P500 – And you are correct, National Bank (our counterparty for both ETFs) is in the TSX60, it is not in the S&P500.
Regards,
Marc Burroni
Business Development Manager”
So there is, after all, no prohibition of the index holding stock of the counter party.
I have funds with 2.5% MER with my actual advisor but would like to transfer it to a Canadian Couch Potato portfolio. My RRSP and TFSA are full. Here is the model I would like to follow for the amount I have to invest:
TFSA
Fixed Income
RRSP
Fixed income
Canadian real estate
Emerging markets equity
NON-REGISTERED
Canadian equity
US equity
International equity
Assuming I put half of my money in a taxable account, what are your recommandation for tax efficiency and long term returns. Does TD e-series index fund would be a good choice, should I choose specific ETFs or should I keep a 2.5% MER corporate class mutual fund with my actual advisor for the non-registered part?
@Sebastien: Corporate class funds seem appealing on the surface, but once you charge 2.5% the tax advantage is probably going to be wiped out by the high fees. You are likely to be better off using proper asset location and low-cost ETFs or index funds.
What about someone who holds between 100k$ and 200k$ in a non-registered account? Would that be better to switch from mutual funds @ 2.5%MER to index funds @ 0.45% if we take into account the taxes to pay on capital gains and penalty? Lets say capital gain and penalty is about 15% of the value of the funds.
“If you’re out of RRSP and TFSA room, you could use these ETFs to build a reasonably well diversified and tax-efficient portfolio of Canadian stocks (HXT), US stocks (HXS) and bonds (CAB). ”
@CCP:
Do you think excluding international markets (outside North America), would be reasonable for portfolio 100% inside taxable accounts in order to keep a tax efficient portfolio?
I ask this because I am considering a portfolio similar to what to suggested in your post, a mix of : HXT (CAN), HXS (US) and HBB (Bonds).
John Boogle doesn’t believe investing outside North America is essential for the average investor…do you agree with him?
http://www.gurufocus.com/news/110008/john-bogle-on-bond-and-international-investing
@Jas:
I am not asking this question for myself, but because I am building tools to help peoples make the right decisions. I already posted an Excel workbook on CCP that does lot of the hard work for asset allocation and the best way to optimize the amount in each type of account (RRSP, TSFA and non-registered). All what I learned from CCP and MoneySense are included in this workbook. I would like to add features to that workbook to have all the important informations at the same place.
My question was not about how to reinvest the money. I wanted to know if a portfolio with 100k$-200k$ value @ 2.5% MER should be sold out and reinvested in ETF @ 0.45% if the fees to sell are about 15% of the value of this portfolio (15000$-30000$ in fees).
@Jas: I think the decision to use HXT, HXS and HBB while ignoring international equities may letting tax issues determine your investing strategy. I would agree it’s not worth holding REITs in a taxable portfolio, for example, but international equities are too big of an asset class to ignore. Is international (non-US) equity diversification absolutely necessary for Canadians? I suppose not. But I would not exclude half of the world’s markets from a portfolio simply because the dividends are taxable.
As for Bogle’s comment, remember that he is looking at the issue from an American perspective. Although I disagree with him, I will at least acknowledge that investing internationally is less important when you live in a country with the world’s largest and most broadly diversified stock market.
How do you go about rebalancing swap etfs without triggering a taxable event?
@Steve: In a non-registered account, you often can’t avoid triggering a taxable event when rebalancing. You can use periodic tax-loss harvesting to reduce the impact, however. You can also use new contributions to rebalance as much as possible: this involves buying asset classes that are below their targets and not selling anything.
If I were to understand this correctly, does that mean holding HXT or HXS will not generate a T3 even upon selling it?
@Billy: That’s correct: if the fund does not make any distributions you would not receive a T3 slip since there is nothing to report.