Minimizing investment friction

Over the years, I have started to pay more attention to friction in my portfolio, which I define as any charges, fees, or penalties, which ultimately deter from my earning potential. When I am making decisions to buy/hold/sell investments, there are three primary types of friction I pay attention to, and try to avoid: Tax Friction, Rounding Friction, and Commission Friction.

1. Tax Friction

Taxes are a reality, and ultimately the tax man (or woman!) always gets his (her) due. In Canada, there are three key types of taxes to pay attention to with your investments. The first, is the capital gains tax, which is applied on any capital gains (i.e. profits) from your investments when you sell them. Following this, is taxes on dividend income; and finally, there are taxes on interest income.

There are a number of ways to reduce tax friction with your investments. The most obvious one, is to keep your investments in a tax free account; in Canada, this would be your TFSA, otherwise known as a tax free savings account. Any capital gains, dividends, or interest, you receive in the TFSA are received tax free. The reason for this is that any contributions you make to a TFSA (i.e. money or investments transferred into the TFSA) are made from after tax dollars, so you have already been taxed on the inflows to the account. The one downside to using a TFSA is that you cannot use capital losses incurred in the TFSA to offset capital gains outside of your TFSA.

The second vehicle at your disposal is to keep your investments in a tax deferred account, e.g. your RRSP. Similar to your TFSA, any gains, dividends, or interest, or not taxed in the account (well, not immediately; see the second key difference below). Moreover, any losses cannot be used to offset gains outside of the RRSP. That being said, there are two key differences between a TFSA and an RRSP. The first difference is that contributions to your RRSP lower your taxable income in the year in which you make the contribution1. As an example, say your salary in 2016 is $45,000, and you contribute $5,000 to your RRSP. This lowers your taxable income to $40,000, which means that your income tax for the year is on $40,000, not $45,000. Taking this even further, if you review the marginal tax rates for your province, you may actually move yourself into a lower tax bracket. In the example we just cited, in the province of Ontario, at $45,000 your marginal tax rate is 9.15%, but at $40,000 your marginal tax rate is 4.05%!

The second difference is that you are taxed when you take money out of the RRSP. The theory is that when you take the money out however, you will already be in a lower tax bracket. So while you may be in a $45,000 tax bracket today, when you take the money out when you retire, you will likely be in a lower tax bracket. Again, by forcing yourself into a lower bracket, you are ultimately paying less tax (and keeping more money in your pocket!).

The third way to reduce taxes is to leverage your capital losses against your capital gains. This option is only available to you in a non-registered account (e.g. not a TFSA and not an RRSP). With this reduction, you reduce the amount of tax you pay on your gains by your losses in that year. For example, if you sell a stock for a profit of $10,000, and you sell another stock at a loss of $4,000, you will only pay tax on $10,000 -$4,000 = $6,000. In general this applies provided you claim the gain and loss in the same reporting period (or carry forward any losses and/or gains to future years); seek advise from a tax professional for details.

2. Rounding friction

Rounding friction is exactly what it sounds like: losses due to rounding. As an example, assume we are able to trade stocks with no commissions, and with no tax friction (e.g. in our TFSA). For our example, say there are two companies, A and B, and we wish to sell company A and purchase company B, because the yield on B is higher:

Line # Company A Company B Notes
A # shares 100 20
B share price $12.34 $59.80
C total value $1,234.00 $1,196.00
D dividend yield 5.00% 5.10%
E dividend $ $61.70 $61.00

After we have completed all of our trades, our absolute dollar return is less even though Company B is the higher yielding stock. The reason for this is that we cannot trade fractional shares. When we sold Company A and took the proceeds of $1,234.00, the proceeds divided by the price of Company B would have had fractional shares: $1,234.00 ÷ $59.80 = 20.635 shares. But, since we can only trade in whole shares, we lost out on 0.635 shares. Even if we take into account the residual cash from selling Company A (i.e. the cash leftover from the trade), our net value is still less. Note that this is only in Year 1, however in subsequent years your net dividend income would still be lower as well with Company B due to the loss of 0.635 shares.

Line # Company A Company B Notes
A # shares 100 20
B share price $12.34 $59.80
C total value $1,234.00 $1,196.00
D dividend yield 5.00% 5.10%
E dividend $ $61.70 $61.00
F Cash in Lieu n/a $38.00
G Net Value $1,295.70 $1,295.00

The only way to get around this is to either luck out and fund companies where the net proceeds of the first will exactly pay for the net cost of the second, or to purchase fractional shares. Luckily, there are ways to perform the latter. If one uses Optional Cash Purchases for companies that allow it, you can purchase any number of shares, and the the total shares purchased will be exactly equal to the amount of capital divided by the going price for the shares.

3. Commission Friction

The most common type of friction, and often one of the hardest to avoid, is friction caused by commissions on your trades. As investors we are all familiar with commissions, and they are a cost of doing business when investing.

Other than choosing a (discount) brokerage which has very low commissions (Personally I use BMO InvestorLine, which charges $9.95/trade), to my knowledge there are really only two ways to get around commission friction.

The first, is to use Optional Cash Purchases for those companies that allow it, and that do not charge commissions on OCPs. Not all companies that offer OCPs do so commission free. For example, the McDonald’s OCP program charges $6.00 per share, whereas the Emera OCP program does not.

The only other way to reduce commission friction is to trade in larger quantities of stock, thereby reducing the average commission. For example, if you purchases 100 shares of a stock at $9.95 commission, your average commission per share is only $0.0995. But if you were to purchase only 50 shares, your average commission would be $0.1990. While you are not completely eliminating the commission, you are reducing it on an average basis.

Onward and Upward!

1 This isn’t exactly true. See a tax professional, but you can may be able to defer your contributions to a later year, hence reducing taxes in a later year.

Homemade Dividends

When looking for companies to invest in, as a dividend investor, the initial screen is to weed out any companies that do not pay a dividend. This has weeded out powerhouses such as Apple, Inc., which have experienced a considerable amount of capital appreciation over the years, but which are otherwise overlooked because they horde cash. (Up until recently of course, when Apple officially announced they would start paying a quarterly dividend). However, in the world of finance, if an investor wishes to invest in a company that does not pay a dividend outright, they can use homemade dividends to replicate the income stream of a dividend paying firm. A good definition of homemade dividends can be found at Investopedia:

A form of investment income that comes from the sale of a portion of shares held by a shareholder. This differs from dividends that shareholders receive from a company according to the number of shares the shareholder has.


So, how does this strategy work? The theory is, that if you require an income stream, you can sell off shares of the company that you own, replicating the dividend. This is a relatively simply operation, and easy to illustrate in Excel. For example’s sake, say we have the following information:

  • The current stock price is $10.00
  • We require a starting dividend payout of $0.15/year (yielding 1.5%, which not uncommon)
  • We would like to see annualized dividend growth of 5%; this is in line with many dividend aristocrats

The above illustrates a common example: when doing research we usually know what the current price of a company is, and what our expectations of the dividend — and growth in that dividend — are. A homemade dividend would sell a partial amount of the share to replicate the dividend payout. So at the end of year one, if you require a dividend of $0.15/share, you would sell off 0.015 shares of the company, leaving you with 0.985 shares at the end of the year. With your required dividend growth of 5%, in year 2 you would require a dividend of $0.1575/share, and you would sell off an appropriate partial share to generate that cash flow. The following table summarizes 20 years of homemade dividends, using our example above:

Year Starting Shares Share Price Value of Holdings Dividend Effective Dividend Yield Shares to Sell Closing # of Shares
1 1.0000 $10.0000 $10.0000 $0.1500 1.500% 0.0150 0.9850
2 0.9850 $10.0000 $9.8500 $0.1575 1.575% 0.0155 0.9695
3 0.9695 $10.0000 $9.6949 $0.1654 1.654% 0.0160 0.9535
4 0.9535 $10.0000 $9.5345 $0.1736 1.736% 0.0166 0.9369
5 0.9369 $10.0000 $9.3690 $0.1823 1.823% 0.0171 0.9198
6 0.9198 $10.0000 $9.1982 $0.1914 1.914% 0.0176 0.9022
7 0.9022 $10.0000 $9.0221 $0.2010 2.010% 0.0181 0.8841
8 0.8841 $10.0000 $8.8407 $0.2111 2.111% 0.0187 0.8654
9 0.8654 $10.0000 $8.6541 $0.2216 2.216% 0.0192 0.8462
10 0.8462 $10.0000 $8.4623 $0.2327 2.327% 0.0197 0.8265
11 0.8265 $10.0000 $8.2654 $0.2443 2.443% 0.0202 0.8063
12 0.8063 $10.0000 $8.0634 $0.2566 2.566% 0.0207 0.7857
13 0.7857 $10.0000 $7.8566 $0.2694 2.694% 0.0212 0.7645
14 0.7645 $10.0000 $7.6449 $0.2828 2.828% 0.0216 0.7429
15 0.7429 $10.0000 $7.4287 $0.2970 2.970% 0.0221 0.7208
16 0.7208 $10.0000 $7.2081 $0.3118 3.118% 0.0225 0.6983
17 0.6983 $10.0000 $6.9833 $0.3274 3.274% 0.0229 0.6755
18 0.6755 $10.0000 $6.7546 $0.3438 3.438% 0.0232 0.6522
19 0.6522 $10.0000 $6.5224 $0.3610 3.610% 0.0235 0.6287
20 0.6287 $10.0000 $6.2870 $0.3790 3.790% 0.0238 0.6049

Observing the above, we can see that as time goes on, the dollar amount of your investment is decreasing over time, and this is because we are assuming that there is no appreciation of the share price over time. Realistically we would want to see some share price growth. With that in mind, we can add in one more column which adjusts for share price growth. I will save you the derivation of the formula, but g was calculated from the following:


Where the i subscript are growth, dividend, and price, for year i. With this additional information, our example then becomes:

Year Starting Shares Share Price Value of Holdings Dividend Effective Dividend Yield Shares to Sell Closing # of Shares Required Price Growth (g)
1 1.0000 $10.0000 $10.0000 $0.1500 1.500% 0.0150 0.9850 1.523%
2 0.9850 $10.1523 $10.0000 $0.1575 1.551% 0.0153 0.9697 1.576%
3 0.9697 $10.3123 $10.0000 $0.1654 1.604% 0.0156 0.9542 1.630%
4 0.9542 $10.4803 $10.0000 $0.1736 1.657% 0.0158 0.9384 1.685%
5 0.9384 $10.6569 $10.0000 $0.1823 1.711% 0.0161 0.9223 1.741%
6 0.9223 $10.8424 $10.0000 $0.1914 1.766% 0.0163 0.9060 1.797%
7 0.9060 $11.0373 $10.0000 $0.2010 1.821% 0.0165 0.8895 1.855%
8 0.8895 $11.2420 $10.0000 $0.2111 1.877% 0.0167 0.8728 1.913%
9 0.8728 $11.4571 $10.0000 $0.2216 1.934% 0.0169 0.8559 1.972%
10 0.8559 $11.6831 $10.0000 $0.2327 1.992% 0.0170 0.8389 2.032%
11 0.8389 $11.9206 $10.0000 $0.2443 2.050% 0.0172 0.8217 2.093%
12 0.8217 $12.1700 $10.0000 $0.2566 2.108% 0.0173 0.8044 2.153%
13 0.8044 $12.4321 $10.0000 $0.2694 2.167% 0.0174 0.7869 2.215%
14 0.7869 $12.7074 $10.0000 $0.2828 2.226% 0.0175 0.7694 2.277%
15 0.7694 $12.9967 $10.0000 $0.2970 2.285% 0.0176 0.7518 2.339%
16 0.7518 $13.3006 $10.0000 $0.3118 2.345% 0.0176 0.7342 2.401%
17 0.7342 $13.6200 $10.0000 $0.3274 2.404% 0.0177 0.7166 2.463%
18 0.7166 $13.9555 $10.0000 $0.3438 2.464% 0.0177 0.6989 2.526%
19 0.6989 $14.3080 $10.0000 $0.3610 2.523% 0.0176 0.6813 2.588%
20 0.6813 $14.6783 $10.0000 $0.3790 2.582% 0.0176 0.6637 2.651%

The logic in the above is that we must maintain a share price which allows us to keep paying a dividend forever. Obviously, over time if we are selling partial shares, this number of shares that we have available to create our homemade dividend, thus we the price of the share must appreciate to compensate for the lower number of shares. Observing the table, growth rates of 1.52% in Year 1 up to 2.65% in Year 20 are actually pretty realistic; over a 20 year time period, the share price has actually gone up less than 50%, and has a CAGR of only 1.94%! So this annual required growth of the share price is not out of line with what we would expect to see for any well managed firm; in fact, we would expect to see the same for a regular dividend company anyways.

So, what are the challenges with using homemade dividends?

The obvious first challenge is that selling partial shares is not realistic; most brokerage houses will not allow you to buy or sell partial shares. In our example above, in year 1 we would have to sell 0.015 shares, which likely isn’t possible for the majority of retail investors.

The second challenge is related to cost. In a perfect world, there would be no transaction costs for a homemade dividend. Alas, we do not live in a perfect world. So the activities involved in selling the shares to create the homemade dividend would likely incur brokerage fees which far exceed the dividend itself. This compares to a traditional dividend where the firm pays us the dividend, so the transaction is virtually free for us: all we have to do is sit back, and wait for the dividend to arrive.

Third, we haven’t taken taxes into account. Dividends are taxed at a more favourable rate than capital gains, but by selling partial shares, we are triggering a capital gains tax. This results in more taxes being taken out over time.

Finally, the required growth price of the company, while realistic in terms of value (less than 2% CAGR in share price gives us an annualized dividend growth rate of 5%), having a constantly increasing g may not be realistic.

In theory, homemade dividends allow an investor to replicate the income stream from a conventional dividend paying firm. However, when you take into account brokerage fees, taxes, and the difficulty in selling partial shares, it becomes evident that for the average retail investor, homemade dividends are not as effective in practice as in theory.

Borrowing to Invest in Dividend Stocks

In Canada, the CRA allows one to deduct interest expenses (Line 221 in the 2011 tax return) used for income producing investments from their net income, which results in lower taxes paid by an individual in any given tax year. I was curious as to whether or not borrowing to invest in dividend paying stocks would result in more income after taxes. To investigate this, I created a spreadsheet model to test a number of scenarios. The results I found were pretty interesting, and are summarized below:

Scenario 1 Scenario 2 Scenario 3
Employment Income $100,000.00 $100,000.00 $100,000.00
Dividend Income $0.00 $425.00 $500.00
Interest Paid $0.00 $425.00 $425.00
CPP $2217.60 $2217.60 $2217.60
EI $786.76 $786.76 $786.76
Employment Credit Amount $1065.00 $1065.00 $1065.00
Ontario Health Premium $750.00 $750.00 $750.00
Loan for Investing $0.00 $10,000.00 $10,000.00
Loan Annual Interest Rate n/a 4.25% 4.25%
Dividend Yield n/a 4.25% 5.00%
Net Income $100,000.00 $100,000.00 $100,075.00
Federal Taxes $17,578.41 $17,525.22 $17,535.34
Provincial Taxes $9726.15 $9,696.66 $9,704.51
Total Taxes $27,304.56 $27,221.88 $27,239.85
Income After Taxes $72,695.44 $72,952.37 $73,040.15

Some notes about the above. First, to keep things simple I used employment income of $100,000. I wish I made that much, but it makes things simpler for the model. $100,000 net income simplified the Ontario Health Premium, which was $750.00. I also assumed the maximum for CPP and EI. All of these numbers are based on the 2011 tax year for the Province of Ontario. And as one final check, I punched the same numbers into the Canadian Income Tax calculator found at Tax Tips and came up with identical results.

So, what do those results tell us? Scenario 1 is the control scenario, where we don’t borrow and don’t receive dividends. In that scenario, an individual pays out $27,304.56 in income taxes for the given year. Scenario 3 was a more realistic scenario: borrowing to invest in a company whose dividends were higher than the borrowing interest rate. Thanks to the tax credits one receives on dividend income, and the fact that interest income is deducted from your net employment income, the total taxes paid were less than in Scenario 1 by $64.71. What was surprising Scenario 2: borrowing to invest, and then purchasing shares in a dividend paying corporation with a dividend yield identical to the interest rate on the loan, yielded less overall taxes than Scenario 1 by $82.68! This means that borrowing to receive dividends where the net gain between interest payments and dividend income is $0.00 still yields lower taxes.

Of course, there are a couple of caveats to this approach. First off, when you borrow, you have to pay off the principal as well. However, if you are borrowing to invest in quality companies, then in theory the company’s share price should increase, so over time you could unload your position to pay back the principal of the loan. Also, most loans require you to pay both interest and principal. CRA rules only allow you to deduct the interest payments, not the principal payments; this means that your net income before taxes would be lower than in the model here, which assumes that you never pay down the principal. Finally, when you do go to sell the position, you will be hit by a capital gains tax.

Given the caveats, one, the extra cash in hand is less than $100 (which is less than 0.1% of $100,000 net income), two, the extra work involved, and three, the risks of the company’s share price dropping, this strategy may not be worth it. However, it was still an interesting exercise; no matter which way you cut it, you are still technically better off since you end up with more cash in pocket at the end of the day.