ETF choices are less important than you think

The comments following my recent post about new ETFs from Vanguard were another reminder of just how often investors get overly focused on products. I accept part of the blame here, since I write a lot about the relative merits of specific funds. But I want to be clear that the choice between similar ETFs or index funds is not nearly as important as many people believe. In many cases, it’s trivial. Successful investing is about saving regularly, keeping costs and taxes low, diversifying broadly, and sticking to a plan. For index investors, selecting appropriate ETFs is important, but it should only come at the end of the planning process. These important steps should come first: 1. Determine the asset allocation appropriate to your goals. This is the most important decision of all, as it will determine your portfolio’s expected risk level and expected rate of return. My model portfolios are all based on a mix of 60% equity and 40% bonds, which isn’t appropriate for everyone. The right mix of stocks and bonds for you depends on your current savings rate and your ability, willingness and need to take risk. Do this part properly you are at least 75% of the way there. 2. Decide which account types are best for your situation. In most cases, long-term investors should only use non-registered accounts after their RRSPs and TFSAs are maxed out. Most people can’t max out both registered accounts, so they should choose a priority. 3. Determine the most efficient asset location. If you’re investing across registered and non-registered accounts, you should pay attention to where you hold each asset class. The least tax-efficient asset classes should go in your RRSP or TFSA and the most tax-efficient in your non-registered accounts. (If you have a spouse, it typically makes sense to think of all family accounts as one large portfolio.) I can’t stress enough that asset location decisions should be based primarily on income tax, not foreign withholding tax. Justin Bender explains why in this example. 4. Make a shortlist of ETFs or index funds for each asset class. At this stage your focus should be on the ETF’s strategy. For now you’re asking questions like this: Do I want to hold a total-market US equity fund, or one that includes only large caps? For international equities, do I want developed countries only, or emerging markets, too? Do I want a broad-based bond fund, or does my plan call for short-term bonds only? Cap-weighted indexes or some alternative? A good place to start your search is my list of recommended ETFs and index funds. 5. Choose the best product from your shortlist. Only now, after you’ve narrowed your choices to a handful of appropriate ETFs, does it make sense to dig deeper to see if one offers a genuinely meaningful advantage. You should consider, in order of importance:

The potential for US estate taxes
Your cost of converting currency (if the shortlist includes US-listed ETFs)
The impact of foreign withholding taxes
The difference in MER (a few basis points is meaningless)
Whether your brokerage offers one or more of the ETFs commission-free

Let’s say you’ve narrowed your US equity choices to either the Vanguard Total Stock Market (VTI) or its new Canadian equivalent, the Vanguard U.S. Total Market (VUN). How do you know which ETF is the better option?

VTI may be subject to US estate taxes, whereas VUN is not.
VTI trades in US dollars, which is extremely costly unless you’re able to reduce the cost of currency exchange.
In an RRSP, you will pay a 15% withholding tax on dividends if you use VUN (this works out to a cost of 0.30%, assuming a 2% yield). With VTI you would be exempt from this tax. In a non-registered account or a TFSA, the two funds are equivalent.
VTI has a lower management fee (0.05% versus 0.15%). That difference works out to $10 a year on a $10,000 investment.

Clearly the right decision for you may not be appropriate for another investor. And some of the criteria are subjective: for example, VTI might be cheaper overall, but you might be willing to pay a bit for more for the convenience of trading in Canadian dollars. But here’s the larger point: if you’ve made good decisions during the first four steps in this process, this final step is really not that important. We’re often talking about a difference of 10 or 20 basis points a year on one asset class in a diversified portfolio. In a five-figure portfolio the impact is almost certainly trivial, and it should never delay you from implementing your plan. In a larger portfolio it’s worth considering carefully, but only after you’re sure you got the first four steps right.