Dollar-Cost Averaging With ETFs: Part 1

One of the most common questions I hear from budding Couch Potatoes is, “Should I use index mutual funds or ETFs?” While index funds generally have higher annual costs (MERs), they do allow investors to add and withdraw money with no fees. ETFs have the opposite problem: they can have much lower management fees, but they incur commissions (typically $10 to $29) every time you buy or sell shares. That usually makes them unsuitable for people who contribute to their accounts every month. As a result, dollar-cost averaging with ETFs can be impractical.

There’s no perfect solution to this problem, but a reader in Regina explained that he’s come up with a compromise, and I think other ETF investors might consider adapting it to their own situation.

Donald has $65,000 in his RRSP, and he’s using iShares and Vanguard ETFs to achieve the following asset allocation:

Asset class ETF Allocation Amount
Canadian equity XIU 30% $19,500
US equity VTI 20% $13,000
International equity VEA 20% $13,000
Emerging markets equity VWO 10% $6,500
Canadian bonds XBB 20% $13,000
$65,000

Donald and his wife recently paid off their mortgage and he’s now planning to supercharge his savings by contributing $2,000 a month to his RRSP. The problem is that Donald’s brokerage, TD Waterhouse, charges $29 for each ETF purchase in accounts under $100,000. (Beyond that threshold, the commission drops to $9.95.) That  would add up to hundreds of dollars a year in commissions if he were to add new shares each month.

Then Donald hit on an idea that would help him minimize trading costs and keep him fully invested, which would allow him to take advantage of dollar-cost averaging. His strategy involves making his contributions to index mutual funds in the same asset classes and the same proportions as the ETFs in his portfolio. Once a year, Donald plans to withdraw the money from the mutual funds and move it to the corresponding ETFs.

Donald’s goal here is to have the best of both worlds. By mirroring his ETF portfolio with index mutual funds, he can set up preauthorized payment plans with each fund, adding money automatically each month with no fees. (While he could have simply made the contributions to a money market fund, that would leave several thousand dollars languishing in the account for up to year, earning next to nothing.) At the same time, he can take advantage of the lower ETF fees, an important consideration in a long-term portfolio.

I think Donald’s strategy has a lot of promise, though he will have to manage things carefully to make it worthwhile. In tomorrow’s post, I’ll look at the math details, and I’ll ask readers to share their own suggestions.

13 Responses to Dollar-Cost Averaging With ETFs: Part 1

  1. Paul T September 14, 2010 at 7:53 am #

    One thing to watch out for is early redemption charges on the mutual funds. Most no load mutual funds have a minimum holding period (generally 90 days). You’ll end up paying a fee if you sell within the min holding period.

  2. Word of caution September 14, 2010 at 8:13 am #

    And then Donald will be surprised (just like I was) by a TD representative telling him that that this approach is considered by TD Waterhouse to be “market timing”, which is a no-no for mutual funds. He will also be told that this is a warning for now, and now that he is warned he will be asked to take his business somewhere else should he be caught doing this again (i.e. using mutual funds for market timing).
    The 2% TD fee if sold in less than 30 days and 2% eFunds fee is sold in less than 90 days are something else entirely and TD will not care that Donald waited for over 90 days to sell his units. “Market timing” definition is somewhat different.
    So, proceed with caution…

  3. Paul T September 14, 2010 at 9:42 am #

    And if you’re already with TD, the e-series ETF funds are a good option for dollar cost averaging people.

    I have my kids RESP with TD for that reason alone. And the MER differential is so minimal that it would require a massive portfolio for the MER difference to outweigh the trading fees of the ETFs.

  4. Greg September 14, 2010 at 10:15 am #

    I like “set it and forget it” investing. 1) Choose some index funds, 2) start a monthly auto contribution plan for dollar cost averaging, 3) review holdings once per year to see if rebalancing is needed. Trying to merge this “set it and forget it” with an annual purchase of ETFs takes away from the simplicity of the set it and forget it approach especially if you need to manage it to align with the holding period of the discount broker (usually 90 days).

    The real problem is the lack of low MER/cost mutual funds in Canada. The US is worlds above Canada in terms of having a low cost mutual fund options in addition to low cost ETF options. Vanguard, Fidelity, etc. all offer mutual funds in the US where the cost of ownership is not that much different than the ETFs. While ETFs are propular these days, the mutual fund format still does some things well that ETFs generally do not do as well – namely – dollar cost avereging and dividend reinvestment (especially with respect to fractional units of a fund).

  5. Alex C September 14, 2010 at 12:17 pm #

    How can TD call that market timing? Do they simply watch if you do it each year at the same time? Maybe if you switched up the redemption times by a month or so each year they wouldn’t notice.

  6. Canadian Couch Potato September 14, 2010 at 12:33 pm #

    @Word of caution: There must have been some other factors at work. The strategy Donald proposes can’t possibly get him into trouble, except by incurring early redemption fees. In any case, market timing with mutual funds may be foolish, but as far as I know, it doesn’t violate any securities law, unless it is done to try to take advantage of “time zone arbitrage”:
    http://www.investopedia.com/terms/m/mutualfundtiming.asp

  7. ABC September 14, 2010 at 12:53 pm #

    There is one more reason to maintain both an ETF and its similar index fund cousin: the imperfect ETF DRIP.

    When the ETF distribution is set to DRIP, the reinvestment plan will only purchase whole shares. Mutual funds give access to micro units, so the entire distribution can be part of the reinvestment plan, the DRIP. For a small investor the ETF drip program will not work, leaving 100% of the payment in cash in the account. This is the case if the monthly distribution is too small to buy even one share in the drip.

    However the minimum purchase is very low for the TD index e-funds, only $100. An investor could transfer there any cash distribution that remains from the imperfect ETF drip. It would be easy to do on a quarterly or semi annual basis to make sure all cash is re-invested and compounded.

    Compounding can be very important, particularly for a bond portfolio. For this reason it is possible the higher fee charged by a mutual fund is justified to make the entire distribution payment part of the drip.

    Here’s a comparison using a bond fund and its ETF equivalent:

    The TD Canadian e Index bond fund, TDB909, has a MER of 0.48% compared with its ETF cousin XBB’s 0.3%. Is the better drip of the e fund worth the additional cost of 0.18% or $18 a year for each $10,000 invested?

    Say an investor puts $10,000 in XBB. In this case the monthly distribution will not be enough to buy a single share and so 100% of the drip, about $25 monthly, around $336 a year, will not be compounding. Is the value of compounding $336/y worth the additional $18/y in fees?

    With the current low yield and taking into account the commission to purchase the ETF, the bond fund is the better answer in this example. But if the investor could increase her XBB holding to $15k, the ETF could be the winner, depending on the size of the purchasing commission. In this example the distribution would be high enough for one additional share every month, leaving only about 30% of the distribution uncompounded. The cash left in the account after the DRIP could then be pooled and invested into an e-fund, but at a frequency compatible with… true couch potato style.

  8. Mark September 14, 2010 at 1:13 pm #

    Does this plan require someone to use TD Waterhouse to buy ETFs or will one have to transfer the money to Questrade (or a broker with cheaper fees)? And transferring money takes time and effort so it is getting away from the couch potato mentality!

    If Donald plans on using TD Waterhouse, then he is incurring an additional $25 commission fee over using Questrade. The $24k he puts into the TD eFunds would have to generate an extra $125 to make this profitable. Which means the $2k/month he puts in would have to generate about 0.96% on top of the MER (~0.35% so ~1.3% in total) returns to break even (assuming a constant 0.08% return each money). Considering the average yearly returns have been higher, this seems like a solid strategy.

    My suggestion is to borrow money to invest at the beginning of each year and instead of making contribution monthly, you use that money to repay the loan. The benefits of this plan are four-fold, one, you enjoy the compound interest throughout the year, two, you only have to make 1 trade per ETF/year so you save on commision fees, third, interest on loan to invest are tax deductible and lastly, you only have to do it once a year so this goes along with the couch potato theme!

    There is a limit to this plan however, if the loan is too large, then the interest on the loan becomes much greater than the commission fee savings thus becoming less profitable.

    Another simple and consistent route would be to throw money into a HISA until it is time to buy ETFs, no need to calculate the breakeven cost or worry about commission fees or transferring money between institutions.

  9. ABC September 14, 2010 at 3:38 pm #

    Borrowing the yearly savings in advance could very well increase returns. It’s a sophisticated approach and it may help set your savings rate in stone. But doesn’t it come with increased leveraging risk? And what about margin rules?

    Also: it does away with the concept of dollar-cost averaging.

    Also: given the CCP strategy of simplicity, maybe we’re trying to pile too much stuff on that proverbial couch? I like the previous comment about “set it and forget it” investing.

  10. 2 Cents September 14, 2010 at 4:02 pm #

    I may be mistaken, but I thought that I read something about Claymore offering pre-authorized contribution plans for their ETFs. It seems to me that some brokerages were trying to opt out of these programs as they allow customers to contribute to ETFs regularly without paying commissions.

  11. Greg September 14, 2010 at 4:40 pm #

    I presume part 2 of this topic will focus on the Claymore offerings. BMO Investoreline does not support Claymore’s pre-authorized plans so that option is not available to me.

  12. Dejan September 14, 2010 at 5:49 pm #

    Last time I checked, CIBC Inverstor’s Edge did support Claymore’s PACC (Pre Authorized Cash Contribution plan). However, Couch Potato’s comment was that when a PACC was used in the context of an RRSP in Scotia McLeod brokerage, PACC portion was invested but not counted as part of RRSP. It went on like that for 4 months until it was detected and sorted out.

    The moral of the story is that dollar cost averaging may be possible with Claymore PACC but it must be carefully set up and monitored to avoid nasty surprises.

    The second disadvantage (no fractional shares) is real – I have DRIP set up with Claymore and while my dividends are more than 1 share, unused fractions are piling up as cash in my acount. The situation was made worse by making dividends paid out monthly instead of quarterly (which makes them similar to PACC when it comes to fractional share issue).

  13. James September 14, 2010 at 10:06 pm #

    Credential brokerage supports no fee DRIP and PACC with Claymore. If the objective is to finally get the funds into an ETF, why not just do it in one transaction. Thanks for the comments about fractional shares. That of course would have to be taken into consideration.

Leave a Reply