Trouble on the Global Ex-Horizon? Management Risks of Corporate Class Funds

Corporate class funds have been around for a long time. The main advantage of corporate class funds is to avoid taxable income and generate capital gains instead. However, their popularity has waned and many funds have been winding down. That is often due to fee compression and the generation of net income to distribute. In contrast, Horizons ETFs switched to a corporate class structure a few years ago in response to a legislative attack on their swap ETFs. They recently changed their name to Global X for rebranding.

The Global X corporate class ETF family structure is well-designed. They have low enough fees to be attractive for their potential tax benefits. They are now less susceptible to legislative attacks, but the risk of managing income still exists. Learn more about the corporate class structure and how Global is doing with that after several years of real-world experience. Will they buck the trend through their innovative design? Or is there trouble on the Global ex-Horizon?

Updated: May 16, 2024

Disclosure: I use the Global X total return index tracking corporate class ETFs in my various tax-exposed accounts. For me, the tax benefits (net of fees ) are worth the risks. I plan to continue using them, but I am also aware of the potential risks and monitor them. You must judge for yourself and make your own decisions. Further, I have no business relationship with Global X at the time of writing this. I will present publicly available data, but also my opinions and understanding as a finance nerd. Not a financial professional with knowledge of the inner workings of Global X. I can make no specific recommendations.

Mutual Fund Trust Income

Conventional ETFs have a mutual fund trust structure. That means that each individual ETF has to distribute whatever investment income it receives, net of expenses. The management expenses and trading costs are used to offset the less efficient income, like interest or dividends, and the leftovers are paid out. Net capital gains from rebalancing or holdings turnover are also distributed in the year incurred.

Mutual Fund Corporation Income

With corporate class ETFs, each ETF is a class of shares in the larger mutual fund corporation. Different share classes allow each ETF to be assigned a different value or make different distributions. However, the income for the corporation is net across all of the different ETFs. So, the income from one ETF can be offset by the management expenses or losses of another ETF. Losses can also be carried forward to be applied against future income. If the net income is zero, then there is no corporate tax. That also means no dividend distribution. So, no taxable income for the investor either.

As the holdings allocated to each ETF grow, the net asset value (NAV) of the corresponding share class grows also. The ETFs that have more management costs would have that deducted from the value of their shares. No one has a disadvantage worse than the costs they would normally incur. When the shares are sold, the growth in their value is taxable at the favorable rate of a capital gain.

fund trust vs fund corporation

Global X was pretty clever in designing their corporate class ETF suite to take advantage of how mutual fund corporations work. The underlying indexes are still tracked using swap contracts. A swap is a derivative where Global X holds cash for collateral and they have a contract with a bank that invests in the index.

The bank holds the stocks or bonds and collects the dividends or interest. Banks are able to handle that tax efficiently. When the swap contract is settled, the bank pays Global X the increase in value of the underlying stocks plus their income. If the total return (capital gains plus income) of the underlying index has been negative, then Global X pays the bank the difference. It is a nebulous structure, but there are three practical effects.

Strategically settling contracts.

A major advantage is that Horizon can strategically decide when to settle contracts to book losses or delay taking income. That helps with managing income for the mutual fund corporation. For example, they settled contracts and booked massive losses during the Covid-Crash of 2020. Those can be carried forward to offset future income.

Settling swaps to manage counter-party risk.

If the unsettled swap has a lot of money owing to Global X, then there is some counter-party risk. That is because if the bank were to go bankrupt, it could default on payment. This shows up as a positive number in the products section of their site.

Previously, counter-party risk was limited by law to 10%. However, that limitation no longer applies to the mutual fund corporation. For example, HXE currently has 59% counterparty risk as I write this. In contrast, slower-moving funds (like the bonds HBB fund) only have a counter-party exposure of 0.35%. Still, Horizon has had to settle some swap contracts and take income to be prudent at managing the potential risk. For example, that happened when the markets took off like a rocket post-Covid and the counter-party exposure became huge.

Derivative income is business income.

The third practical implication is that the income generated from settling swap contracts is considered regular income. Not dividends, capital gains, or interest – even though some of the total return of the underlying indexes was due to a mix of those different income types. That is an advantage for foreign dividend income because it avoids foreign withholding taxes. However, what would normally be tax-favored capital gains and eligible dividends are taxed more as regular income.

In summary, the swap structure offers the advantage of better control over realizing income in the mutual fund corporation. However, if there is income – it is more highly taxed. More on this later.

The ETF Mix In Horizon’s Mutual Fund Corporation

horizon etfs

Another way that Horizon was clever in designing its corporate class ETF family is the mix of funds.

They have a mix of passive total return index (TRI) funds and expensive boutique funds. The passive TRI funds have low costs (0.03-0.5%/yr) and would be expected to grow and likely generate income over the long run.

Some of the ETFs, like HXQ and HULC, hold underlying securities that will pay dividends regularly.

The potential derivative income from swap settlements and growing regular dividend income could present a problem since they want to keep the net corporate income zero. They need growing expenses to offset the income.

Fortunately, the boutique BetaPro funds come with lots of expenses and are geared towards active traders. They use daily contracts for leverage, short-selling, and management in niche sectors. Those expenses make those funds like money pits, with management and trading costs in the 1-3.5%/yr range. Bad for the speculative investors that use them, but a normal cost of trading using those strategies. Good to offset the income of the TRI investors.

That money sump is an advantage that other corporate class fund families have not had. Well, other than more expensive active management costs for all of their funds. However, if the growth of index investing rapidly outpaces the usage of the more costly boutique funds, then the income could eventually overwhelm the expenses. The passive index-tracking ETFs are already about 90% of the assets in the corporation. Horizon seems to recognize this threat and has been adapting.


I think not. However, this section is just me speculating and management-splaining.

Global X’s asset allocation ETFs all currently use the efficient TRI ETFs. Asset allocation ETFs are being recognized as a great way to invest. Their popularity would also increase the pace of potential income growth as they buy more of the underlying corporate class TRI ETFs. Recently, Global X changed their AAETFs to allow them to hold conventional ETFs that pay distributions. That could defray some of the problem. Or maybe not. If new money flows into the swap ETFs are faster than market growth, then purchasing new swap contracts could lower the counter-party risk and the potential need to settle old ones.

Global X also suspended new subscriptions to their HSAV ETF which has a high-income yield. Presumably, to help limit the income generation in their mutual fund corporation from this type of investing. They now have CASH as a high-interest ETF outside of their corporate class structure instead. Notably, HSAV is not swap-based. So, it receives interest payments at a current yield of 5.5%. The interest income has already jumped from ~$47 million in all of 2022 to $108 million in 2023. Removing it from the corporate class to a conventional ETF would not be popular with its unit holders, but might help the corporation as a whole if interest rates stay elevated.

The Sum Of The Parts

If or when income could outstrip the expenses of the mutual fund corp are not predictable. This is because most of the actual income is only realized when swaps are settled, and swaps dominate the holdings. There are also losses that can be strategically harvested and carried forward. Risk management fits into that decision and is also impacted by market performance. Global X does receive some regular dividends and interest income. While they cannot control that, it is a small amount overall.

Those are a lot of unpredictable moving parts. However, we can see how that is playing out overall by looking at Global X’s mutual fund corporation’s net income or loss pool over time.

At the end of the annual report for funds in the corporate class family is a section called “Notes To Consolidated Financial Statements”. In section 12, there is a listing of non-capital losses carried forward. That is the loss pool that Global X has as a buffer against income. The sum is shown for each year in the chart below.

So, you can see that Horizon got off to a great start by building a $4 billion dollar buffer. They booked massive losses by settling swap contracts during the Covid-19 market bottom in 2020. Since then, there has been income exceeding expenses each year. You can see that by the loss pool shrinking. Probably due to growth in the TRI ETFs and settling swaps to manage counter-exposure risks prudently as markets went on a tear to the upside. Even though markets pulled back again in 2022, they were not able to grow their loss pool that year. However, despite the recovery of 2023 they did keep control and there loss pool only shrank a little.

The “burn rate” of that buffer is another way to look at it. In the chart below, I show the same data as the annual change in the loss pool. At the post-pandemic burn rate, the loss pool would be exhausted and taxable income generated in about five years. If the markets take off like a rocket, that burn rate could increase to keep counter-party risk in check. If Global X finds opportunities to harvest some losses, opens popular expensive funds, or can manage swap settlement really well – then the burn rate could slow.

It is also important to put the burn rate into perspective. For the last three years, it has been about $500 million dollars per year. The assets in Global X’s mutual fund corporation were ~$25 billion dollars at the time of writing this. So, if Global X used up its loss pool and continued to generate $500 million/yr in net income. That would represent ~2%/yr net income yield.

What happens if there is net income in Global X’s ETF corporation?

Net income would be subject to taxation within the corporation as a whole. How that burden would be assigned to each ETF share class is uncertain. There are also features of corporate taxation for mutual fund corporations and derivatives that make it potentially nasty. With the exception of having high-interest savings account ETFs in their corporate class structure, I think that Global X has built it pretty well. I plan to stay on board. The current ride is too luxurious from a tax efficiency standpoint and there looks to be open seas ahead for now. However, I am going to keep my eye on the horizon and know where the lifeboats are.

horizon etf management risk

There looks to be some time away, but the current loss pool burn rate shows that the income management risks are real. Global X will surely be monitoring that. Hopefully, they will also have a plan to shift to a conventional structure or make other changes before the corporate taxes kick in.

They may also come up with other innovative management solutions along the way that I haven’t even considered. They have before.

Still, it will be useful to model what hitting a tax-berg in a mutual fund corporation looks like. Just in case of foggy markets or captain-error. We’ll do that in the next post. Like an iceberg, the taxes on an eligible dividend distribution on the surface are easy to spot. However, there is much more lurking beneath the surface.


This series of posts has complex material. I would like to thank Benjamin Felix (Head of Research, PWL Capital) for helping me to identify issues, understand nuances, and find data. I am also grateful to the folks at Global X ETFs for answering our tough questions as honestly and completely as possible. This post is not endorsed by anyone and any errors are my own. I try to be accurate and complete, but this is complex material. If I get something wrong, please let me know (with references if applicable) so that I can fix it.


  1. Mark,
    What a brilliant piece of analysis, thank you very much. I am also a unitholder of many TRis and when I look back I think I took too long to jump on board.The term ” swaps ” were given a bad name by Warren Bufet after the GFC in 2008-09 as he called derivatives-” weapons of mass destruction”. In retrospect I could have had 5 years of not paying taxes on distributions.It seems to me after reading the post that ” swaps management is key in the arrangement. I thought a swap was a firm committment between the 2 parties with an end date, but your article implies that Horizon can redeem swaps as they see fit ?Why would the bank agree( counter party agree) ? Also I recall National Bank was the main counterparty but I believe now they use CIBC also. A few years ago I could not understand why the bank ( counterparty) would recive a lousy money market return and give Horizons the market ( index) return. I think this had something to do with the tax advantage of the bank receiving the dividends from the portolio.I think the recent Federal budget was tinkering with dividends banks receive…and closing some advantages. Perhaps a subsequent post can educate me/us on this as well clarify the ” allocation to redeemeers methodology that they use”. Thank you again for taking the time to explain the intracies of the ” footnotes” and ” burn rate. Very interesing and worth keeping an eye on. Hopefully we have many more years of the free ride….

    1. Hey Lyndon,

      I asked about that back when I did an article on the corp class ETFs when the made the switch. Either party can choose to settle the swap in their arrangement. The bank(s), I didn’t know about CIBC, but I do know they actually have had multiple options lined up. That could possibly be another way to manage counter-party risk. The swap arrangement actually works out well for the banks. They make money on the swap fees for the non-Canadian index ETFs. The Canadian equity ETFs don’t have swap fees because the bank can actually profit from eligible dividends received and have no net tax because they pay out eligible dividends to shareholders anyway. There may be some other nuance in there that I can’t remember, but basically they benefit from the arrangement too and have no real incentive to end it.

      The allocation to redeemers methodology issue is no longer an issue because they side-stepped it with their change to corporate class. I wrote about that here. It is a bit mind-bending, but the issue was in their old structure they were recharacterizing income. In the corporate class, they are not. So, I have filed it under “complicated stuff I don’t remember anymore”.

      1. Thanks Mark. I went back to your posts from 4 years ago, great illustrations of the mechanics of the redeemers methodology and T2s attempts to squash the swap structure !
        Thanks for all the great work, this is a very complex topic, but you make it so much more digestable !
        Please give us regular updates from what you hear at Horizons on the burn rate etc, Felix would be in the know.
        I wasnt aware they eleimiated the 10% counterparty restriction.

        1. Thanks Lyndon,

          I think that the switch to corporate class addressed the previous legislative risk issues pretty well. This management risk is the main issue remaining at this point. Now that I know how to find and look at it, I plan to update it annually. I use some of these funds in parts of my portfolio. So, I am keeping an eye anyway. Currently, the potential tax-berg is still over the horizon, but we’ll have to see how Horizons navigates the waters.

  2. Hi LD,

    > Recently, Horizon changed their AAETFs to allow them to hold conventional ETFs that pay distributions. That could defray some of the problem. Or maybe not. If new money flows into the swap ETFs are faster than market growth, then purchasing new swap contracts could lower the counter-party risk and the potential need to settle old ones.

    I’m trying to understand what’s going on with these AA ETFs with the new changes this year.

    The investment objectives of these asset allocation ETFs has been changed from “long-term capital growth” to providing “moderate long-term capital growth and a moderate level of income”.

    However these asset allocation ETFs are mutual fund trusts whose holdings are still all TRI ETFs as of this moment. Horizons has started paying out distributions from these ETFs. That would mean that these distributions are composed of realized capital gains/return of capital from selling shares of the underlying TRI ETFs, or perhaps paying out cash from new subscriptions.

    I’ve been seeing complaints that these asset allocation ETFs are no longer tax-efficient due to passing out distributions. However, it seems to me that if they’re still utilizing the TRI ETFs, they should still be expected to pass through the tax efficiency of the underlying ETFs to the unit holders, at least until Horizons actually decides to reallocate away from TRI ETFs. The distributions not sourced from eligible dividends should still be more tax-efficient than receiving non-eligible dividends/foreign dividends/income.

    For now, my understanding of the current state of these asset allocation ETFs would be that unit holders are still getting tax efficient exposure to the underlying asset classes. The main change now is that unit holders are losing a bit of control over when they realize the capital gains in the portfolio due to the new distribution policy. Unless the asset allocation shifts away from TRI ETFs to plain vanilla ETFs, taxable investors should still want to hold these ETFs, assuming that they can accept the asset allocation choices made by Horizons, i.e. the focus on large caps and overweight in the NASDAQ 100.

    Does your understanding match with mine?

    Forgive me if this discussion is somewhat off-topic from the blog post, it just seems that you have spent the most time online analyzing, dissecting and presenting information about Horizons TRI ETFs.

    – Bill

    1. Hey Bill,

      I have been speculating on what the asset allocation ETF changes will mean. I think I originally put something about it in this article in a pre-publication draft, but shied away from guessing. From their literature, there are two points that I honed in on. One is that they particularly reference HBAL and HCON for producing income and hint that they are often held in registered accounts where the tax doesn’t matter. So, it may be an attempt for them to expand the appeal of those ETFs in that setting. For HEQT, the big benefit would be to improve diversification and they claim that as the reason. They do have an underexposure to small and mid-caps that some conventional ETFs could remedy. I don’t know their plans, but if I were them, I would hold a conventional Canadian market ETF and add a small cap ETF or two. I think they’ll still mostly hold their on TRI corp class ETFs. Regardless, I suspect that the distributions for HEQT will be small and tax-efficient as capital gains, return of capital, and maybe some dividends. Who knows. I am just watching. The have made some tiny distributions in the past. Capital gains from rebalancing from what I can tell.

      In the back of mind, I also wonder if this is part of their plan to manage the growth of the underlying corp class ETFs. That could go either way. New money flowing in could mean new swap contracts and lower their need to settle and take income. On the other hand, growth in the TRI ETFs and not the expensive Beta Pro ETFs isn’t great for managing income and expenses. Not sure if it is a strategy or not, but it could be a factor.

      1. Hey LD,

        While we’re speculating here anyway, I did some closer reading of that notice and circular for the special meeting.

        > If each Proposed Change is approved and implemented, while each ETF may continue to invest in TRI ETFs, at the discretion of the Manager, an ETF may also invest in non-TRI ETFs, consistent with its Proposed Investment Objectives and Proposed Investment Strategies (as defined herein). Accordingly, each ETF is not anticipated to offer the same tax-efficiency to Unitholders, particularly when Units are held in a non-registered account.

        HEQT still indicates that its primary objective is long term capital growth with no mention of income. This could be taken to indicate that they intend to continue to use TRIs in there, especially since HGRO (now HEQT) is excluded from this clause:

        > Currently, each ETF’s investment objective seeks to provide long-term capital growth using a portfolio of exchanged traded funds. If each Proposed Change is approved and implemented, each ETF will change its investment objective to instead seek to provide a combination of long-term capital growth and income, in the case of HCON and HBAL.

        I wonder if the ETFs with non-zero fixed income allocations will be the ones to start straying from the TRI holdings. This could provide a bit of a release valve, although if HEQT’s AUM shoots up due to it being the only one that still targets long term capital growth as its objective, that may undo any income pressure released.

        Another thought I do have now is that although these asset allocation ETFs still provide a tax-efficient alternative to the other asset allocation ETFs by the inertia of the existing holdings, there appears to be no more forward-going promises of tax efficiency in these ETFs. Given this lack of any future tax efficiency promises, and their growth and large cap tilts (which result in presumably negative loadings to the size and value factors), I find it hard to justify using these ETFs as a long term holding in a taxable account over the alternatives that just hold vanilla ETFs.

        – Bill

        1. Hey Bill,

          I read that the same way. I think that the main income will be on the fixed income side. I see a couple of easy solutions.

          1) If wanting some fixed income and a tax exposed account. Just use HEQT and HBB (or ZDB) in your desired mix. This also gives the advantage is using more bonds peri-retirement if concerned about sequence risk and then going back to normal for longer-term growth when out of the danger window. The downside is that you have to have the guts to rebalance yourself when there is a big market swing.

          2) If worried about small cap value under-exposure, add some AVUV and/or AVDV as appropriate. That is how I decided to address some of the large cap growth overweighting in parts of my own portfolio. Not as simple as an all-in-one, but still pretty simple. I honestly think that the factors-piece is tweaking. Probably does more to satisfy my need to tweak than anything, with little harm done. I also hold a conventional Canadian ETF in my corporation – HEQT is a bit less Canadian than the plain vanilla all-in-ones. Optimal home country bias is probably in the 20-30% range and eligible dividend income is very efficient in a corporation anyway.

          Of course, the further you move from an all-in-one strategy, the more potential you introduce for behavioral errors.

  3. Hi, Mark,
    The whole aim of investing in the Horizon asset allocation ETFs was to have no income distributions and have only capital gains when you sell them. With these ETFs now giving distributions from 2-2.6%, there seems to be no advantage in holding them, compared to the regular asset allocation ETFs from Vanguard or iShares, which also distribute income in that range, with lower MERs and no swap fees. The other alternative would be BRK.B, which is almost like an index fund due to it’s diverse holdings and pays no dividends, so you have only capital gains when you sell it.

    1. Hey Victor,

      The corporate class ETFs themselves (HXT, HXCN, HXS, HXDM etc) are still not planning on making distributions.

      The all-in-one ETFs that bundle them (HEQT/HBAL/HCON) will. It remains to be seen what those distributions will be comprised of. If it is capital gains, probably not a big deal. If it is all income (interest/dividends) then I agree that their advantage would be lost. I am not sure how that will play out. If it is simply from rebalancing etc, then it would be capital gains. However, they mentioned improving diversification which to me implies they may also add some ETFs that do make income distributions. Will be interesting to see how it actually unfolds. I suspect it will remain mostly capital gains for the HEQT version, but I am just guessing. The HCON may pay more interest/dividends.


        1. That’s correct. They are not corporate class – they are a convention trust structure that hold corporate class ETFs inside. They are also classified as being actively managed. They tweak their asset allocation. Most recently, they dropped HXX for example. I am curious what they plan/tweak to add that makes income distributions.

  4. Hi Mark.

    Do we have an idea where is the Horizons Non-Capital Income Loss Pool at the end off 2023? What was the burn rate for 2023?

    Thanks for everything on Loonie Doctor and for the Money Scope (awesome!).


    1. Just updated. Their 2023 report was finally out. The loss pool burn rate dropped substantially for 2023 despite a market up-year (a good thing). They are still using up runway, but it was at a slower pace last year and still quite a bit of room.

  5. Hello Mark,

    Thanks for this update on this. I really appreciate it, as many of us do.

    Sorry in advance of this question was answered elsewhere. I could not find it with my search.

    Within my CCPC, I did set myself up with the global-x funds initially. I do not think that I will have the long-term issue of bumping up against passive income limits, noting also that I am non-Ontario and New Brunswick physician, so your previous BC physician analysis would apply to me.

    Based on the analysis that you have done and with the recent changes in the capital gains inclusion rate, I was planning to sell HXT, which I probably would not have needed anyways based on your later analysis.

    I would imagine that you are updating your posts, or generating a new ones, with respect to how the capital gains inclusion rate would impact the other global-X funds. My overall take on this as a nonexpert would be to possibly get rid of HXS as well as this fund had the least benefit compared to a conventional fund.

    It would appear that HXDM and HXEM would have the most benefit of having the dividends realized as capital gains based on your previous analyses. I do not hold much prolonged so HBB is not as relevant.

    Hopefully in answering this question, this may help others with the same question.

    Again, thanks so much for your work as well as Benjamin Felix’s work and all of the education that you have provided over the years.

    1. Thanks Erik. I am planning to redo them all once the new capital gains changes are more clear. I have the same take as you and I don’t think the capital gains stuff will change it. HXDM, HXEM, and HBB are really good. Canadian ones I just use conventional (in corp) – we use them in personal.

      The US ones are kind of either way. If you are really trying to have full control and just use capital gains harvesting here and there – then maybe.

Leave a Reply

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