Complexities & Assumptions of Asset Location Strategies

Tax optimization using asset location in an attempt to reduce the drag of taxes on investment returns is optional. For most DIY investors, it is of minimal potential benefit and increases the risk of no should be ignored. However, for high-income investors with large portfolios, particularly if incorporated, there could be worthwhile benefits. Scenarios to illustrate that will be in the next page. However, before considering the potential benefit, you must consider the factors that make asset location complex, different approaches, and possible solutions. That is the focus of this page.


In The Wrong Spot?


You Must Use More Than One ETF & Rebalance

In order to match investments to where they are most tax efficient, you must use multiple ETFs that bundle holdings with different income and growth characteristics. That could still be simple, like using an all-equity globally diversified ETF (like XEQT, VEQT, or ZEQT) for your stocks and a bond ETF for your bonds. That is minimal hassle once you get used to DIY investing. It is also easy to rebalance and rebalancing does not need to be precise or often. However, it is not automatic and effortless like with an all-in-one asset allocation ETF.

To get really serious about tax optimization, you’d need to separate out your Canadian stocks, US stocks, Non-North American stocks, and bonds. You may even buy your US stocks/ETFs using US-dollars for lower management fees and foreign withholding taxes. That usually means 5-7 ETFs. You could even consider using corporate class ETFs in your tax-exposed accounts if you face a high tax rate. That could easily add up to 10 ETFs.

The cost and time for buying/selling ETFs is negligible when you become experienced and have a large portfolio. Probably not so, for a small portfolio or a beginner. While the complexity of using more ETFs is higher, it does offer opportunities to lower fees and strategically use corporate class ETFs for extra tax efficiency.

Minimize Tax Drag While Preserving Asset Allocation

Minimizing the overall tax drag while preserving the overall asset allocation makes for a dynamic multi-dimensional jig-saw puzzle. Mind-blowing without a computer aid. I have made portfolio building calculators to do this, but it is important to understand what they are trying to do.

You can try to shelter your most tax-inefficient investments in your RRSP and TFSA. However, those accounts are limited in size, and it may not all fit. If the only investments you have fit into these tax-sheltered accounts, asset location is of minimal benefit since your tax burden is low.

If you must spill investments over into tax-exposed accounts, that is where asset location could matter more. Due to limited registered account size, you may not be able to simply do something like put all your International Equity and bonds in your RRSP and all your Canadian Equity in your TFSA. Some needs to go to the taxable accounts. Which one? Generally, the asset class that has the least worsening of tax efficiency by spilling over. It is the relative tax drag worsening of a holding between account types that matters rather than simply within an account. You have to fit the pieces together across the whole portfolio and not just within accounts.

You still need to preserve your asset allocation as the primary objective. Getting your asset allocation right to maximize potential returns and avoid behavioural sabotage will have a much larger impact on your outcome. To complicate this juggling of relative tax drags of an asset class between account types, their contribution to asset allocation also changes between accounts if you are using post-tax asset allocation.

What is Pre-Tax vs Post-Tax Asset Allocation?

An RRSP is 100% tax-deferred and a corporate account is about 85% tax-deferred. You have a bigger lump to invest but must still pay the tax when the money comes out. A TFSA and personal account is post-tax. You paid the tax up front, but capital gains are preferentially taxed at half the rate of regular income in a cash account (or not at all for a TFSA). Essentially, different accounts have different tax-liabilities baked into them.


Pre-tax asset allocation is what you see on your investment statements. Post-tax allocation is what you would have to spend after the investments are sold, taxes paid, and the cash put into your hands.


Whether you choose to use a pre-tax or a post-tax approach to asset location changes where the best places to put assets are. The next two sections explore these two approaches further.

Pre-Tax Asset Location & Tax Optimization

When optimizing pre-tax, the total return (capital gains plus income) of each asset class is important. The assets with the lowest expected total return, after tax-drag on the income (interest or dividends), get located in the tax-deferred accounts. This usually translates into the bonds going into the RRSP. The types of bonds that you want to use to stabilize your portfolio are low risk, but that also means low expected return.

Pre-tax optimization may trick the behavioural beast, but it also increases risk volatility poses when you need the money.

The government basically owns the part of your RRSP that is the tax liability. They are like your investing partner and get their share when you access the money. With pre-tax asset location, they get to share your low return on the bonds, and you keep most of the higher return from the stocks. Seems like a great way to subsidize your behavioral insurance. You may trick the behavioral beast that just looks at your investment statements and doesn’t do hard stuff, like math and taxes. However, the argument against using pre-tax asset allocation and location optimization is that you are increasing the volatility in after-tax terms. Important, since you spend after-tax money.

For example, if you have $100K of stocks in your non-registered personal taxable (cash) account and $100K of bonds in your RRSP, your pre-tax asset allocation is 50:50. The post-tax asset allocation is not 50:50 because the RRSP has a tax-liability. Let’s say your average tax-rate as you access the RRSP in the future is 50%. That makes your portfolio$100K stocks in the non-registered account and $50K bonds in the RRSP.

Post-Tax Asset Location & Optimization

You can’t spend the money until it is in your hands, after the tax has been paid. Spending the money is what matters in the end. So, you need to account for the taxes-owing when considering asset allocation and how much risk you are taking. This is the argument for using post-tax optimization.

The main driver for optimal location using post-tax allocation is the tax drag on the investment income. Investments that get the biggest reduction in tax drag by being put in an account type get targeted there first. More on how complicated that is later, but recently that has meant putting low-yielding bonds into the taxable accounts and sheltering high-dividend paying foreign stocks.

This highlights the contrast to pre-tax optimization where both the drag on growth and the total return influence location. That is good because it means that you don’t need to guess at future capital return rates. Just tax drag on income. The capital growth is neutralized when you discount for the baked in tax liabilities.

Since you need to account for the tax liability of a tax-deferred account, the value of the RRSP holdings are discounted. Some stocks are also held in the taxable account to make up for that discount. This is simplistically illustrated below.

You will note that the volatility for how much money you have in your hands after-tax, in the post-tax allocation example above is what you were targeting. However, the volatility that you would see on your investment statements is actually higher. Potentially more provocative to your behavioral beast, depending on how smart they are.

Pre-Tax or Post-Tax Volatility: Which Matters More To You?

The argument made by the most knowledgeable people who look at this stuff is that post-tax asset allocation and volatility is what matters. The money you have after-tax. Accounting for the tax liability to access the money, a post-tax optimized approach may yield a slightly higher after-tax return. That is the math. However, how much you would have after-tax not only depends on the math, but on what we actually do along the way. So, I think it is actually debatable which is better.


In favor of pre-tax optimization.

If you look at your investment statements, see the number, and don’t automatically apply higher thought processes like thinking about the tax liabilities. That favors pre-tax allocation which has less volatility on your statements. The function of asset allocation is to have a volatility that suits your risk. Your emotional beast is simple and provoked by volatility – they don’t do math well. They are happy. Lower behavioral risk.

The downside is that the after-tax volatility is higher, and that is important for when you need the money. For, example, it increases sequence of return risk if you need to take out money during a major drawdown. Some people will have other ways of dealing with sequence risk, like a flexible budget, timeline, part-time work around retirement, or a big cash reserve. For them, the behavioral risk may be the main limiter to asset allocation. That could favor using pre-tax optimization. Plus, it is much simpler and has less execution risk than post-tax optimization.


In favor of post-tax optimization.

If you spend so much time thinking about this stuff that when you see your account statements, you mentally start discounting the value of your tax-deferred accounts, then you have a smart beast. You get more upset when you see your TFSA or personal account shrinks than when your RRSP takes a hit because you know that is going to impact you more. The emotional beast is not fooled.

Further, if your risk capacity is the main limiter for asset allocation, then post-tax money and risk to that is what matters. Again, that means post-tax optimization.

You must make assumptions about the future to optimize for it.

You will have probably noticed by this point that you need to take some guesses about the future. What will your tax rate be like when you draw the money out? Will tax rules change making one type of income better or worse than now? What will bond yields do to impact their tax drag?

The future is unknowable. However, the practical question is can we either guess close enough or find a way to remove the unknown variable?


What will your future tax rate look like?

The simplest thing to do is to apply taxes as if you were liquidating the portfolio today. That would be an accurate reflection of the after-tax risk for today. The present-day approach makes sense if you have low risk capacity and a significant risk of a forced liquidation of your portfolio. However, if you are unlikely to draw from your portfolio until retirement, then tax rates during a gradual decumulation phase is more appropriate.

Another approach is to take a guess at your taxable income during retirement and apply the average tax rate for that. That is a guess. However, if you guess that it will be 40% and it is 30% or 50%, then that is still closer than assuming it is zero or all at the highest tax rate.

The reality is more complicated when you have multiple account types to optimize a drawdown plan. For example, if you have a personal taxable account, corporation, RRSP, and TFSA, then you may draw from these in different mixes and sequences during retirement. You may not even touch some of them during your lifetime, if you have more than enough. Each account would have a different tax rate applied to it over your lifespan.

The only way to get that type of information is to do a simulated tax-efficient drawdown. I built The Decumulator to do that and it runs this in the background of Robocorp SWAT to estimate tax-rates to apply to accounts for after-tax allocation. Even using a simulator, the guess is likely not what will actually happen. The question is, is it close enough? It is probably close enough for me, but you would need to answer that for yourself.


Future tax rates or rules will change. Will it be meaningful?

The future may be a long way off and what if the tax laws change? It is almost a certainty that there will be tweaks. We have already seen that with corporations recently. However, the important question when trying to tax-optimize is whether it would be a major relative change. If tax rates go up for all forms of investment income, then eligible dividends will still be more tax efficient than foreign dividends or interest. In contrast, if eligible dividends became taxed as regular income, that would change optimization.

Eligible dividends and capital gains receive lower personal taxation rates for a reason. Foreign governments charge withholding taxes for a reason. The tax system was designed intentionally. While governments may not always act intelligently and can do just about anything, a radical change that breaks everything seems unlikely.


Dividend yields will fluctuate. Does it matter?

The dividend yield is the dividends paid/price of the shares. So, fluctuations of dividend amount or price makes it change. They do this all of the time. The tax drag on investment income is the tax rate multiplied against the yield. Therefore, the tax drag for a given security, or bundled as an ETF, is in constant flux. That would change tax optimization if there is a major change of one relative to another. Fortunately, there are a couple of ways to deal with this.

First, major world markets move independently, but are still very strongly correlated. So, yields generally move in the same direction and important relative changes less likely. Particularly over longer timeframes. There could still be changes on a year-to-year basis though.

Second, you could smooth fluctuations and likely eliminate the impact on asset location optimization. This could be accomplished using the average dividend yield from multiple years and keep that constant for your algorithm. For example, I used the average yield from the underlying ETFs/markets over 10 years in comparing the tax efficiency of asset allocation ETFs. The Robocorp portfolio builders, that automate the asset allocation and tax optimization, also use those long-term yields to compare tax drag.


Interest rates fluctuate. It matters, but there is a solution.

The yield on bond ETFs fluctuates with the interest paid and price per share. This often does not move in sync with dividend yields. So, if bond ETF yields fluctuate above and below the yield of ETFs covering different equity markets, then their optimal location would bounce all over the place. Reshuffling your portfolio every time that happens would be both a hassle and a capital gains tax trigger.

One solution to the impact of interest rate fluctuations on asset location is to use HBB.TO. It is a corporate class ETF that tracks the Canadian Bond Universe. However, with its corporate class structure, the income should be zero (or close to zero) and its total return be made up of only capital gains/losses. That would result in it always being directed towards tax-exposed accounts. Even when you add for the relative fee-drag increase. This is because capital gains are so much more tax-efficient compared to other bond ETFs in a personal or corporate account.

Holding bonds in your tax-exposed accounts has another advantage beyond tax efficiency. When you need some cashflow, selling some bonds from a tax-exposed account is likely to have minimal tax consequences compared to other options.

Of course, using HBB also means that you comfortable with the risks. Their swap structure means that if National Bank defaults, there is potential for some exposure. If that happens, it implies financial system issues and I think nowhere is safe. There is a risk that Horizon can’t offset the interest. That would result in an eligible dividend. However, that still is likely to be smaller and more tax efficient than interest. Same with a forced closure of the fund triggering a capital gain. That is better than having paid full tax on interest along the way instead.

Optimal Could Change As Your Income Changes

Your income during your accumulation years could look very different from the years that you draw down your portfolio. It could even change during your accumulation years, if you have a major career change. You don’t want to have to suddenly change everything at retirement or for a temporary income fluctuation. So, when considering how you tax optimize, you really need to consider income over your lifespan. Fortunately, what is good during accumulation is also good during drawdown. The possible exception is how you treat Canadian dividend-paying ETFs/stocks. This may change slightly with income fluctuations or if an exceptionally large corporation.

The Canadian market is dominated by stocks that pay relatively high dividend yields.

Those eligible dividends are taxed at a lower rate in personal hands. The tax drag is the dividend yield times the tax rate, and in some provinces, the top tax rate is still so high that eligible dividends are better off sheltered. In lower tax brackets, the eligible dividends are very tax efficient. A similar phenomenon can happen with corporations.

Eligible dividends are extremely tax efficient when flowing through a corporation. Usually.

They may even lower your tax bill if you would otherwise be paying ineligible dividends to live on. However, if your passive income gets high or the dividends that you pay out of your corp are low, then they can become inefficient. That does not usually become an issue until later in your accumulation years. During retirement, eligible dividends flow out very tax efficiently.

So, when considering optimal location for Canadian dividend payers, you must consider efficiency changes across the projected lifespan.

I have not seen any blogger or planner comment on this issue. However, I uncovered its importance when I was building Robocorp SWAT and that is why that deluxe portfolio builder requires data entry to run a retirement simulation as part of how it determines after-tax allocation and optimization over the lifespan. Eligible dividends in a corporation are usually tax efficient until late career and the efficiency during a long retirement period almost always makes up for any issues. However, that could be different in rare cases.

Summary of Major Considerations in Tax Optimizing

The most important consideration is whether the potential benefit is worth the execution risk. This page was focused on the factors that make tax optimization complex and could hamper implementation. These are summarized in the table below. For the average person, the potential benefit is miniscule. However, for an incorporated high-income professional, the potential benefit could be substantial but is highly variable. I will illustrate that with some case studies on the next page.

portfolio tax optimization