It’s been almost three years since Dan Bortolotti and I released our popular Tax Loss Selling white paper. Throughout the paper, we showed investors how to implement a disciplined tax loss selling strategy, using Canadian-listed ETFs. Near the end of the paper, we even provided suggested ‘tax loss selling pairs’ (which consist of a primary ETF and a secondary ETF that is very similar to the initial holding). Our process proved to be very useful after the recent Brexit announcement, as we were busy realizing losses on the international equity ETFs in our taxable client accounts.
Although most of the advice contained in the paper is still relevant, new ETFs have been released over the years, triggering us to update our recommended tax loss selling pairs. We have also updated the monthly tracking errors of the underlying indices over the past ten years (from 2006 to 2015). The ‘tracking error’ in this context is the standard deviation of the differences in their monthly returns – just remember that a lower tracking error is generally preferred for tax loss selling purposes).
Table 1: Recommended primary (initial) ETFs
| ASSET CLASS | PRIMARY ETF | VENDOR | UNDERLYING INDEX | MANAGEMENT EXPENSE RATIO |
| Canadian Stocks | VCN | Vanguard | FTSE Canada All Cap Index | 0.06% |
| US Stocks | VUN | Vanguard | CRSP US Total Market Index | 0.16% |
| International Stocks | XEF | iShares | MSCI EAFE IMI Index | 0.22% |
| Emerging Markets Stocks | XEC | iShares | MSCI Emerging Markets IMI Index | 0.26% |
| Global Stocks | XAW | iShares | MSCI ACWI ex Canada IMI Index | 0.21% |
Sources: BlackRock Canada, Vanguard Investments Canada Inc., MSCI Indices, Morningstar Direct, Dimensional Returns 2.0
Table 2: Recommended secondary ETFs
| ASSET CLASS | SECONDARY ETF(S) | VENDOR | UNDERLYING INDEX | MANAGEMENT EXPENSE RATIO | TRACKING ERROR TO PRIMARY ETF |
| Canadian Stocks | XIC/ZCN | iShares/BMO | S&P/TSX Capped Composite Index | 0.06% | 0.33 |
| US Stocks | XUU | iShares | S&P Total Market Index | 0.10% | 0.09 |
| International Stocks | VIU | Vanguard | FTSE Developed All Cap ex North America Index | 0.22% | 0.46 |
| Emerging Markets Stocks | VEE | Vanguard | FTSE Emerging Markets All Cap China A Inclusion Index | 0.24% | 0.64 |
| Global Stocks | VXC | Vanguard | FTSE Global All Cap ex Canada China A Inclusion Index | 0.27% | 0.08 |
Sources: BlackRock Canada, Vanguard Investments Canada Inc., MSCI Indices, Morningstar Direct, Dimensional Returns 2.0
Canadian Stocks: We’ve stuck to our guns on this one. The Vanguard FTSE Canada All Cap Index ETF (VCN) remains our primary holding in client accounts. When markets get rocky, we’ll sell this holding and purchase either the iShares Core S&P/TSX Capped Composite Index ETF (XIC) or the BMO S&P/TSX Capped Composite Index ETF (ZCN) (you’ll notice that the replacement ETFs track the same index). The tracking error of the underlying indices has been 0.33%, so the ETFs are expected to have similar risk and return characteristics going forward. Just be careful not to use XIC and ZCN as your tax loss selling pairs. The CRA has made it very clear that switching from one ETF to another that follows the same index will result in a denied capital loss.
Most tax loss selling advice recommends switching back to the primary ETF after holding the secondary ETF for at least 30 days (this avoids CRA’s superficial loss rules). We do not feel that this is necessary in most cases, as our secondary holdings are still great long term substitutes for our primary holdings. However, if after 30 days your replacement ETF is showing a significant loss, it may make sense to trigger a second loss by switching back to the original holding.
US Stocks: After the release of our paper, we didn’t come across many tax loss selling opportunities for US stocks (which was a good thing). We continue to use the Vanguard U.S. Total Stock Market Index ETF (VUN) as our primary ETF, but have been using the iShares Core S&P U.S. Total Market Index ETF (XUU) as our secondary ETF. At the end of 2015, XUU switched its underlying index from the S&P Composite 1500 to the S&P Total Market Index (which includes a whopping 3,843 stocks), so it is now even more similar to VUN (which tracks 3,624 stocks). The tracking error between the pairs has been extremely low, at 0.09%.
International Stocks: We’ve left the primary ETF as the iShares Core MSCI EAFE IMI Index ETF (XEF), but changed our secondary ETF to the Vanguard FTSE Developed All Cap ex North America Index ETF (VIU). VIU was released at the end of 2015, and holds the underlying stocks directly (making the ETF more tax-efficient than the past recommendation from our white paper).
You may notice the tracking error of 0.46% is higher than most of the other asset classes. This is expected, as the underlying MSCI and FTSE indices have different construction methodologies. The biggest difference is that FTSE considers South Korea to be a developed market, and includes it in their developed markets indices. MSCI is not as convinced, and still includes the country in their emerging markets indices. Once South Korea is added to the MSCI developed markets indices, the tracking error would be expected to decrease.
Emerging Markets: We have left our recommendations the same for this asset class. Our primary ETF is still the iShares Core MSCI Emerging Markets IMI Index ETF (XEC), and our secondary ETF is the Vanguard FTSE Emerging Markets All Cap Index ETF (VEE).
Similar to the international indices, the tracking error for the emerging markets indices has been noticeably higher, at 0.64%. It has the same issue as our international index pairs – MSCI considers South Korea to be an emerging market, while FTSE does not (and excludes it from its emerging markets indices). FTSE also includes a portion of China A shares within its emerging markets indices, which may be causing some of the increased tracking error.
Global Stocks: Many DIY investors are now using ETFs that track a combination of US, international and emerging stock markets. Even though the returns from the various regions may offset each other at times (resulting in less tax loss selling opportunities), there will still be moments when global stocks as a group plummet. For our primary global stock ETF, we would recommend using the iShares Core MSCI All Country World ex Canada Index ETF (XAW). When this ETF is down in the dumps, consider switching it to the Vanguard FTSE Global All Cap ex Canada Index ETF (VXC) in order to realize the capital loss. The tracking error has also been the lowest of the bunch, at 0.08%.
Justin
Would like to take some tax loss on CPD
What would be a good switch
Regards
ETF
@Andres: There is not much alternatives for ETFs in “Preferred shares fixed income” category:
iShares S&P/TSX Cdn Preferred Share Idx ETF(CPD.A)
iShares S&P/TSX Can Preferred Share Idx ETF (CPD)
Horizons Active Preferred Share ETF (HPR)
iShares S&P/TSX N.A. Preferred Stock C$ Hgd (XPF)
PowerShares Canadian Pref Share Index ETF (PPS)
Horizons Active Preferred Share ETF (HPR.A)
BMO Laddered Preferred Share Idx ETF (ZPR)
Horizons Active Floating Rate Pfd Share ETF(HFP.A)
Horizons Active Floating Rate Pfd Share ETF (HFP)
First Asset Preferred Share ETF (FPR)
RBC Canadian Preferred Share ETF (RPF)
If we exclude all ETFs that does not invest primarily in Canada or are not passively managed, here are the remaining ETFs:
iShares S&P/TSX Cdn Preferred Share Idx ETF(CPD.A)
iShares S&P/TSX Can Preferred Share Idx ETF (CPD)
PowerShares Canadian Pref Share Index ETF (PPS)
BMO Laddered Preferred Share Idx ETF (ZPR)
So, what does each ETF tracks:
CPD follows the the S&P/TSX Preferred Share Index
PPS follows the NASDAQ Select Canadian Preferred Share Index
ZPR follows the Solactive Laddered Canadian Preferred Share Index and is a different type of preferred shares
In my opinion, PPS would be the best alternative to CPD and you could claim the tax-loss as you’re not following the same index.
I am not a professional, I just analysed the datas available online. Here are my sources of information for this post:
List of ETFs: http://www.fundata.com/
List of ETFs and fund datas for each ETF: http://www2.morningstar.ca/
CPD: https://www.blackrock.com/ca/individual/en/products/239836/ishares-sptsx-canadian-preferred-share-index-fund
PPS: https://www.invesco.ca/ivz/powershares/exchange-traded-fund/powershares-canadian-preferred-share-index-etf/at-a-glance/5?lang=en
ZPR: https://www.bmo.com/gam/ca/advisor/products/etfs#fundUrl=%2FfundProfile%2FZPR
Types of preferred shares: https://www.bmo.com/pdf/Understanding%20Preferred%20Shares_E_FINAL.pdf
Hope this will help you make your choice.
Thanks so much.
It’s nice to have a forum to clarify these things. Its really appreciated.
Hello Justin,
I am a bit tardy on this one. I have your white paper on this topic in the past. I continue to have difficulty with this strategy.
For someone who does not sell to rebalance – which has the advantage of not triggering a capital gain in the first place – the strategy appears essentially useless.
In a CCPC, for someone that does trigger capital gains – for good reasons – it seems to me that the benefit of deferring taxes would have to be weighed against the significant cost of keeping the CDA “clear” at all times for a serendipitous event like Brexit.
Kindly assist.
@Kulvir: Tax-loss selling is a good strategy when the cost of doing it is low. If you pay 20$ for a loss of 100$, it is too expansive as the ratio is 20$/100$ = 20%. If the cost is still 20$, but for a loss of 1000$ (ratio of 2%), then it could be a good strategy. 20$ @ 7% x 20 years = 77$, that’s the money you could have done with your 20$. Your loss of 1000$ will allow you to save taxes on capital gains up to 500$. If you pay 20% in taxes at retirement, that means you’ll save 100$ in taxes. You made about 23$ with this strategy. If your sell it before retirement and you have a marginal tax rate of 38%, you saved 190$, so you made about 113$ with this strategy.
@Kulvir: If you never intend on rebalancing your portfolio, it may not have as much of a benefit. But even for this small group of investors, it could still make sense to have capital losses booked if they decide to exit a security that has become relatively too expensive or tax-inefficient. For example, many investors wanted to get out of VDU and into XEF or VIU when our work on foreign withholding taxes was released, but they were stuck, as they had no capital losses carrying forward.
For a CCPC, tax loss selling does not always make sense, depending on what income your accountant is planning to withdraw. It is recommended that you work with your accountant on this strategy.
I understand that if I hold a fund for a long period, chances that I could claim a tax-loss in future is less probable.
If I add 100$ to one of my funds every month and hold it for 20 years, is there a situation where I could still be able to claim tax-loss in 20 years based on past datas using market averages ?
Just curious. Any specific reason not to include ZEA with VIU or ZEM wit VEE in table 2 ? Looks very similar to me or am I missing something? ZEM also holds some stocks directly which may offer a slightly lower wotholding tax on dividends? Thanks
@Anne: ZEA and ZEM exclude small cap stocks, while VIU and VEE include them. You can certainly use them as tax loss selling pairs, but if you wanted broad-market exposure in the first place, you may want to switch back to VIU or VEE after 30 days. The tax loss selling ETF pairs that I have chosen allow an investor to continue holding the replacement ETF going forward without switching back (this can be extremely useful if markets recover after placing the tax loss selling trade, and switching back would realize a gain, potentially offsetting some or all of the deferral benefit).
Justin,
Your post got me wondering about tax loss selling for ETF pairs where one is CAD hedged and the other is not. For example ZEA and ZDM, or XSP and XUS, (and I’m sure there are lots of other pairs like these).
Hedging introduces a significant additional factor, and as we saw last year when the loonie took a nose-dive, it can produce significantly different results. Does the CRA consider them to be identical, even though they clearly are not?
@Don M – I haven’t heard of any comments from CRA on this particular scenario, although there are plenty of products that you could use as pairs and avoid this grey area entirely. For instance, if you sell ZEA (and wanted a hedged product instead), consider buying VI instead of ZDM. If you sell XSP (and wanted an un-hedged product instead), consider buying VUN instead of XUS.
So I guess this means there is less a benefit for tax loss selling strategy in a corporate CCPC account since passive investments in corporate accounts are always automatically taxed at the highest marginal tax rate.
Would you recommend more a buy and hold (almost forever) strategy for corporate CCPC accounts?
@D: Since passive investments are taxed at the highest rate in a CCPC account, tax loss selling would be expected to have a significant benefit (not less of one). In terms of buy and hold, most investors would be rebalancing their portfolios at times, so having some losses banked that could offset these gains would be useful.
This strategy brings down the tax bill in the present (or near future), in exchange for a higher tax bill in the future?
Remember, the ACB will be lowered after this move, so future gains will be that much higher.
Isn’t this just a wash? Or is there real benefit?
@slacker: The tax bill is deferred until a future date – if you are in a lower tax bracket when you eventually realize the gain, there is an additional benefit. If you are in a higher tax bracket when the gain is eventually realized, then you need to weigh this potential cost against the benefit of additional income/growth.
Would you rather pay capital gains taxes of $10,000, or invest that $10,000 and earn income on it? This deferred tax benefit is what investors are usually aiming for with a tax loss selling strategy.
Excellent
Odds on the government cracking down on this, as more and more people start to employ this strategy?
@slacker – it would likely be very difficult for the government to put an end to tax loss selling (investors are going to realize capital losses, even if they are not specifically implementing a tax loss selling plan. If they put an end to it, they would also have to propose a plan for what is to be done with any losses that are realized). They could always argue that these pairs are not dissimilar enough, but I think they would have difficulty doing so – all of the chosen pairs track different indices (i.e. FTSE/S&P/CRSP/MSCI), with different construction methodologies, holdings and security weights.
Hi Justin,
I’m curious as to why you use VUN as the primary US ETF and XUU as secondary, when XUU is half the cost of VUN and are tracking similarly sized indexes? Is it just the age and longer history of VUN (to evaluate tracking error)?
Thanks!
@Tyler: I tend not to make any product changes based solely on a competitor’s recent reduction in fees. Most ETF providers will follow suit with their own fee reduction – I expect this will be the case for VUN. However, if an investor prefers to use XUU as their primary and VUN as their secondary, I don’t see any issue with that at all. The same is true for (XIC or ZCN)/VCN, VIU/XEF or VEE/XEC. This is one of the reasons why implementing a tax loss selling strategy using broad-market ETFs is superior to tax loss selling using active mutual funds or individual stocks – the differences between the replacement securities is expected to be negligible for the broad-market ETFs.