HXEM vs XEC or ZEM ETF in a Personal Taxable Account

Emerging market exposure might add another layer of diversification to your portfolio. If you use an all-in-one ETF, then it is already in there. Those investing within a personal taxable account and concerned about tax efficiency may be attracted to Global X (formerly Horizons) HXEM ETF to track emerging markets. It has a swap-based corporate class ETF structure that functionally avoids taxable foreign dividend income. Instead, the ETF accrues that amount as a capital gain. Conventional ETFs like XEC or ZEM held on their own or within an asset allocation ETF cover similar markets for a lower cost but with more annual taxable income.

How does HXEM compare to its conventional competition? There are potential advantages for those with high incomes and a long time frame. However, there are other risks to consider as well. Learn more and consider how the potential risks and benefits may apply to your portfolio.


Mutual Fund Corporation vs Trust Investment Income

A corporate-class ETF has a structure different from that of conventional ETFs. The corp class ETF is one share class in a mutual fund corporation. So, the income from that ETF can be offset by expenses in other ETFs in the corporate family. If the net income across the MF corporate family of funds is zero, there is no tax. Instead, the value of each ETF unit value rises by the amount of its net income. So, for the HXEM ETF, dividend income effectively becomes a capital gain rather than a taxable dividend. Capital gains are taxed at half the usual income tax rate and only when the ETF units are sold.

In contrast, conventional ETFs, like XEC/VEE/ZEM, are a trust structure. So, the income received must be paid out and taxed in the hands of the unit holders each year. That income is taxed at the usual personal marginal rates. Money lost to tax each year is money that is not invested and growing.


Foreign Dividends & Withholding Tax

Foreign dividends are taxed as regular income in Canada. There is also a foreign withholding tax (FWT) on dividends collected in their country of origin. It varies by country, but in the emerging markets covered by the MSCI Emerging Markets Index, FWT recently averaged ~13% overall. So, for a 2.6% dividend, there would be 0.34% FWT. Fortunately, many of these countries have tax treaties with Canada, and you recoup most of that FWT when you file Canadian income taxes.

The HXEM ETF uses a swap contract instead of holding stocks directly and receiving dividends. So, it avoids FWT altogether.

In contrast, VEE has an extra non-recoverable layer of FWT. It is a wrapper ETF for holding Vanguard’s US-listed VWO. While the US FWT is recoverable as a tax credit when filing Canadian income tax, the emerging market FWT is non-recoverable. So, it has a ~0.34%/yr tax drag relative to XEC and ZEM, which hold stocks directly.


Mutual Fund Corporation Tax Risk

The other feature of an MF corp is that losses from previous years are carried forward and used to offset future income. In the case of Global X, they booked a large loss pool during the 2020 market swoon. Since then, they have had more income than expenses each year, which has caused the loss pool to shrink, but they have still not had net income.

The trouble occurs if there is net income inside the MF corp. If that happens, it is taxed at the business tax rate, which is 39.5% in the case of Global X. That tax paid within the corporation would be assigned to the ETFs that contributed to the net income and reduce their share value. When you sell, the total of that hidden tax plus your personal tax on a capital gain is higher than interest or foreign dividend income. It is much higher than dividends would be.


Is HXEM at risk of destabilization?

Emerging markets have less stable regulatory and geopolitical environments. It is part of why they are historically classified as emerging rather than developed. However, the HXEM ETF has other stability risks due to its special structure. So far, there seems to be a significant loss pool before net income becomes a problem. However, you can only see the loss pool update for a year when Global X issues its year-end report for the ETFs each spring.

That said, Global X has made some changes this year that may help keep the lights on longer. Global X changed up its all-in-one ETFs (HEQT and friends) to use more conventional ETFs instead of the corporate-class ones. They have also increased their swap fees as of Jan 1st, 2025, which increases expenses. That may decrease income growth (but also decrease ETF return).

I suspect that Global X would initiate a transition into a conventional structure if they saw trouble coming. However, there is always a risk of error. There would also be the management question of how to handle all of the unsettled swap income.

zea xef viu hxdm

While North American markets dominate the world currently and are likely to dominate Canadian investor portfolios, investing outside of North America adds diversification. The developed markets dominate those international markets. Whether adding emerging market exposure on top of that is beneficial is debatable.

Some would argue it adds diversification. There is also the idea that rapidly growing economies translate to stock market profits for investors. That is not necessarily true. Here is an explanation of why. Actually, emerging markets have historically underperformed when you consider very long-term data.

Several ETF options from the major providers make investing in emerging markets easy. HXEM is a corporate class option, and the major Canadian-listed conventional ETF options are ZEM, VEE, and XEC. For the modeling in this post, I will use ZEM. It uses a similar index to HXEM.


Global X’s Emerging Market Index

Global X’s HXEM ETF uses a swap contract to track its own emerging market total return index. It basically mirrors the MSCI EM index, which is comprised of ~1,400 large-cap companies. BMO ETFs’ ZEM directly holds stocks that track the MSCI EM index. So, it is the most tax-efficient, directly comparable conventional ETF to HXEM.

An important nuance is that Global X’s index tracks the futures roll. So, the dividend is not added in. Instead, the collateral held for the swap earns the overnight US Fed rate. When the Fed rate is lower than the index yield, the ETF lags the conventional index. Conversely, it is a bonus when the Fed rate is higher.

For example, the Fed rate is currently 4.33% – higher than the current MSCI EM yield of 2.61%. So, HXEM has had a beneficial tracking error recently. It was detrimental when the Fed rate was extremely low when HXEM was started in 2020. Since inception, HXEM’s tracking error vs ZEM is -0.76/yr (actual) vs -0.29/yr (expected based on higher fees).* There is limited data because it has been under five years, but that would be expected to narrow further if interest rates stay higher than the index yield. For this analysis, I will ignore this wrinkle since it is so unpredictable. However, the impact could be significant, and it certainly adds volatility.


Other Emerging Market Indices

Blackrock’s XEC also holds stocks directly but tracks the MSCI EMM IMI index. That index is slightly broader and includes some smaller companies.

Vanguard’s VEE holds the US-listed VWO ETF and tracks the FTSE Emerging Markets Index and holds almost 6000 companies. The FTSE index treats some countries, like South Korea, differently from the MSCI indices. So, for the modeling in this post, I will focus on XEF and ZEM. VEE would be a reasonable alternative. However, it still has tax inefficiency due to a second layer of FWT.

The slight differences in index holdings can translate into periods of one leading another. For example, over the last few years, MSCI EM (ZEM) has trailed MSCI EM IMI (XEC). However, over the last 12 years, there has only been a 0.10%/yr difference.* That is not surprising since the IMI version includes smaller companies that would be expected to have a higher risk/return.


Comparative Fees & Costs

HXEM has an MER of 0.28%/yr. Trading costs due to the swap contracts are also added to that. Previously, they were up to 0.3%/yr but were recently raised to 0.5%/yr. For the analysis, I will assume trading costs of 0.5%/yr for a total fee drag of 0.78%/yr.

The other ETFs that cover the comparable indices just have an MER and some trading costs. BMO ETF’s ZEM has an MER of 0.27%/yr and most recently had a Trading Expense Ratio (TER) of 0.07%/yr for a total fee drag of 0.34%/yr. Blackrock iShares’ XEC ETF has an MER of 0.28%/yr and had a TER of 0.07%/yr. If the TER from each ETF’s 2023 financial report is used, Vanguard’s VEE ETF had an MER of 0.25%/yr, but also embedded FWT drag of ~0.34%/yr on top of that for a total drag of 0.59%/yr.


Dividend Yield & Taxes

Even though the FWT for XEC or ZEM would be recoverable in a personal taxable account, size also matters. The MSCI EM index tracks companies that pay moderately large dividends. The yield varies, but an average of ~2.6%/yr is a reasonable estimate. That is the gross yield before fees and FWT.


Net Distributions or Growth

The gross yield differs from the distributions you see as an investor because distributions are made after fees and costs of the ETF are deducted. For example, with dividends of 2.6%/yr upon which ~13% of FWT is applied, it leaves 2.26%/yr. Further, ZEM has MER & TER of 0.34%/yr. So, the net dividend received is 1.92%/yr. When the FWT credit is recovered at tax time, then it is effectively still 2.26%. That 2.26% is taxed at Canadian income tax rates.

For HXEM, the dividend yield would be captured as an increase in the ETF’s share price. The 2.6%/yr would be blunted by the 0.28%/yr MER plus an up to 0.50%/yr swap fee, which would add 1.82%/yr to the share price.

Future returns are unpredictable, but I have arbitrarily set the capital gains at 4.8%/yr for modeling. So, for ZEM, that would be a 4.80% capital gain plus a 2.26%/yr dividend for a total return of 7.06%/yr. For HXEM, it would be a total return of 6.62%/yr, all as a capital gain.


Passing Through All Gains Annually

Generally, a capital gain is attractive because it is taxed at half the usual rate compared to a foreign dividend. That is the simplified appeal of corporate-class ETFs.

If the ETFs were sold every year and tax paid on all dividends and capital gains, that lower capital gains rate would result in less tax for HXEM vs. ZEM. However, at low personal tax rates, the higher cost of HXEM overpowers that benefit. Conversely, at higher marginal tax rates, the tax owed on those fat dividends is worse than the additional 0.5%/yr swap fee. This is shown using Ontario tax rates below.

xec etf vee zem

Tax Deferred Growth

A lower taxable income is only part of the advantage of capital gains. Since that tax is not due until the ETF is sold and the capital gain is realized, there can also be tax deferral. Tax deferral is generally advantageous because the money that would have been paid as tax stays invested and compounds. This is why it doesn’t make sense to sell the ETFs and realize gains before you have to.

The growth boost from having a larger unrealized capital gain instead of a taxable dividend is shown below. Ontario tax rates are used, but the situation would be similar for other provinces. Using corporate-class ETFs in Alberta, Saskatchewan, and Manitoba would have slightly less of a tax deferral advantage since their income tax rates are lower. HXEM has a tax deferral advantage that grows at higher income levels. It has a 0.77%/yr advantage at the highest personal tax rate.


After-Tax Value When You Do Sell

The taxes on tax-deferred growth eventually come due. When sold, the extra capital gains that accrued (instead of dividends) will be partially taxable. One of the risks with ETFs is that you may be forced to realize the capital gain earlier than you’d planned if the ETF gets liquidated. Usually, that is a low risk if you stick with popular broad ETFs from major providers. Global X is a major provider, and these are broad ETFs. However, it is not inconceivable that there could be a problem with corporate-class income management,

To quantify the advantages and risks, I modeled how much more money you could have after-tax if you sold at different points in the future. The higher your tax bracket now, the more of an advantage HXEM could have (as mentioned). Deferral of the extra capital gains tax further into the future also helps.


HXEM vs ZEM

The chart below models the amount more after-tax due to the HXEM vs ZEM ETF over time. The illustration below uses a taxable income of $125K in Ontario, both now and in the future. An income level where HXEM could make sense. Tax on dividends is removed annually for the conventional ETFs. The value at each time point is after the capital gains tax from liquidating the ETF. As you can see, there is a very small increase in the amount of after-tax money that grows over time. About 1% more after-tax money at 10 years. Double that at 15 years (gotta love compounding). Less if the capital gain is taxed at a higher inclusion rate, like 66.67%.

emerging market etfs canada

The HXEM advantage is slightly better at higher tax brackets. At the highest Ontario tax bracket, the after-tax advantage is 1.9% at 10 years and 3.1% at 15 years (data not shown).


Impact of Different Future vs Current Income Tax Rate

The benefit of tax deferral is also impacted by the current vs. future tax rate. If you defer from a high current to a lower future tax rate, the benefit grows. The reverse also applies. For example, deferring from a low tax bracket now and realizing a large gain that bumps your tax brackets later could be detrimental.

With the high fee drag, HXEM investors could be worse off if they defer taxes from a low current rate and face a much higher future tax rate. In contrast, someone currently in the highest marginal tax bracket who is able to realize the gains at a lower tax rate in the future could have 3.5% more after-tax money at 15 years in this model.

zem vs xec etf vee

Whether to use HXEM or a conventional ETF is a personal decision. I cannot give specific investing advice. Hopefully, this post will help you to make a more informed one as part of your own due diligence. It is a balance of potential risk and reward. In a tax-sheltered account, like an RRSP or TFSA, it is easy. The cost of a corporate-class ETF is not worth it. In a taxable personal account, it depends.


Potential Reward

The potential reward of using a corporate-class ETF is moderate at mid-high income levels. That could add up over longer time periods. However, a lot can happen over time. The recent increase in the swap fee is a good example. It really cut into the potential tax benefit of this ETF. If it were to increase further, that would make it really close or even detrimental.

The potential advantage is greater in the highest tax brackets, and the risk of being bumped into higher tax rates is also lessened. Taxes can only go so high. Supposedly. They have been higher at different times and in different parts of the world. Using HXEM for tax-deferred investing from a high current to a lower future tax rate is when it shines.


Potential Risks

The unknowable future underlies risk. We expect that there is investment risk with equities and to be compensated for taking it. However, the corporate class structure adds other risks. This analysis assumes that HXEM has no net corporate income assigned to it from the corporate class structure. If there is net income, then it quickly becomes inefficient. Global X might be able to manage a conversion to a conventional ETF if they see that coming, but I am not sure how that would play out.

HXEM has swap-based income. So, the risk of income is greater if they must settle swaps to manage their counterparty risk, which could happen with massive market rises. The counter-party exposure for HXDM is only moderate (28.7%). In this case, the swap is against a futures roll. That introduces another source of tracking error. If the Fed rate is higher than the conventional index dividend yield, it is beneficial, and the reverse is also true. That is tough to predict or model, but could have a meaningful impact.

I stuck with the analysis of HXEM vs. ZEM. Using a slightly broader index, like XEC, could narrow HXEM’s advantage slightly—historically, by about 0.10%/yr. That doesn’t change the results substantially, but it could nudge the scales if you’re in a moderate tax bracket now and facing a higher future tax rate. Because HXEM uses a futures roll index, tracking errors also increase. That may be beneficial when Fed rates are higher but detrimental when they are lower than the index’s dividend yield.

The potential good news is that while the recent swap fee increase has made HXEM less attractive, the increased costs may also make the corporate structure more sustainable. I’ve done my best to model what the benefits could look like and discuss the potential risks. However, the biggest risks are always the ones you don’t see coming.

Disclosures: We don’t currently use any emerging market ETFs in our portfolio.

*Thank you to Andrew Jones, CFA of Verified Beta, for helping with the tracking error analysis.

8 comments

  1. Thank you for this very educational and thorough analysis! This has helped me to determine the best asset location for foreign equity in my portfolio.

    However, I did notice a minor error regarding the calculation of fees for ZEM. After reviewing the annual reports, I verified that ZEM has mgmt fees and trading fees but the MER incorporates both so no need to add the TER to the MER – that is double counting the trading fees on this fund.

    On the other hand, it was a little surprising to note that in the case of HXEM, while the swap fee is incorporated within the trading fee, they do in fact report the TER independently of the MER for that fund so they do need to be added in that case. Quite confusing for most investors I think.

    1. Thanks Scott for bringing this up.

      It is confusing & hard to really nail down. It was a struggle for me too. Still, I took a stab at it.

      Here is a link to BMO ETFs most fact sheet for ZEM. It is dated Jan 17, 2024 and shows a TER of 0.04% at that time on top of MER of 0.27% for a total cost of 0.31%/yr at that time. It is down at the bottom. It doesn’t appear on all of their literature.

      In their 2023 financial statement on page 33 (Statement of Comprehensive Income), for 2023, the management fee listed is only ~23.75% of NAV and the trading costs another 0.07%. So, I see what you are saying. However, for 2022, it was 21% for management fees & another 15.5% for transaction and other portfolio costs. Note 6 on page 24 also describes the management fee as separate from brokerage and a variety of other portfolio costs. So, it is hard to really know the trading cost & MER from their financial statements as it swings around so much. Their ETF Facts sheet shows them as additive. So, I took a stab at it and added a small TER.

      The TER on Global X’s site is exactly what their swap fee used to be. So, I think it is not additive to the swap fees. The 2024 ETF Fact sheet shows a total cost of 0.58% (MER plus TER) at bottom – not an extra swap fee on top of that. They just started reporting them separately when they raised the max swap fees. So, I would expect it to go up to 0.5%/yr as that transition happens.
      Mark

      1. Thanks for the additional explanation Mark. I was incorrect about some of those fees. Appreciate your thorough response!
        Scott

        1. No worries. Got me looking at it again. It is really hard to figure out! I don’t think my guess is perfect either. Fortunately, I don’t think a small difference either way changes the big messages.
          Mark

  2. Very illuminating post! I remember using the Roboswat calculator back a few years ago, and it showed that the tax+fee drag for HXEM in the TFSA (0.71%) was lower than the tax+fee drag for XEC in the RRSP (0.82%). In this post you stated that the gross returns for HXEM and ZEM were assumed to be the same, so the tax+fee drag for XEC in RRSP has to be lower than HXEM in TFSA for it to be preferred as described in the post. I just can’t wrap my head around the discrepancies. Am I missing something? Is it due to the recently increased swap fees?

    1. Hey Mark,

      Thanks! I just dug into my Robocorp SWAT archive. I’ll need to revitalize it at some point. It was a useful tool. Two things have happened since then. First, XEC moved from being a wrapper ETF to holding stocks directly. So, it removed a layer of FWT – the fee and tax drag is now closer to 0.6%/yr in a TFSA/RRSP. ZEM has the same single layer of FWT. Also, the raise in the HXEM fee will make it higher now (0.78%/yr). With those improved efficiencies by the conventional ETFs, they’d expected to be a better and reliable option for a TFSA/RRSP (in my opinion).
      Mark

  3. Hi Mark,

    Thanks for another great post. Regarding these swap ETFs from Horizon’s, when they are held in an informal trust (not RESP), must you always do an income tax return for the child annually? I saw you comment that you have an informal trust with such funds in your other post about the American ETFs. I appreciated the article you wrote on the informal trusts too: It was very thorough.

    I started my kids off with another investment (which had dividends) in their trust, but now have moved to HXS only. So last year (2024) and the year before (2023) there were no dividends (which I know would have been attributable to me) and no capital gains realized either year either. In 2023, my accountant advised that because I had previously filed returns for my kids (because of dividends and the capital gains realized when I switched to HXS in 2022) that I should file a return for 2023. Since I started with some tax returns for my kids, do you know if I would have to file them every single year until they start filing their own? I know I need to track ACB, which is very easy with HXS as the only holding so far.

    1. Hey Diego,

      If you have an informal trust and there is income, then the adult account holder will get a T5 for the interest/dividends. If there is none (like a properly functioning corp class ETF), then there is no T5 issued. If there are capital gains, those are attributed to the minor. We have realized capital gains when our kids were ~16+ because they had income from student jobs and even with the gain, they paid no tax.

      If the informal trust (a kind of bare trust) has over $50K in it, there was going to be a requirement to do an annual filing that basically says who the trustees are and who the beneficiary is. That was put on hold because they realized it had a larger impact on voters than intended.

      Sounds like you are doing it right. I don’t see why you would need to file a return for them every year if there is no income or capital gain (but I am not an accountant). I could see if there were capital gains maybe, but if nothing at all, I don’t see what would be filed. If you find out another reason from your accountant, please let me know!
      Mark

Leave a Reply

Your email address will not be published. Required fields are marked *