I read an article recently about a formula to calculate the true cost of owning an ETF. The idea behind this formula was to capture all the costs of an ETF, not just the annual expense ratio or MER.
Related: What are ETFs? – Exchange Traded Funds
If you are trying to add up your costs for something, it makes sense to try to include every single cost involved. However, from a practical point of view, it might also make sense to just include the significant costs and ignore the insignificant ones if it is too much work and the final result will be similar.
This second approach definitely applies to the cost of ETFs. For most investors, the MER or expense ration of an ETF is the dominant cost and the main one they have to worry about.
Bid-ask spread costs [explained below] and to a lesser extent, trading fees are mainly relevant for frequent traders.
Related: How to buy an ETF or Stock at a discount brokerage – Step by step directions
The reason this article caught my eye was mainly because of the example given – the “true” annual cost for an ETF that had an expense ratio of 0.10% ended up being 0.76%, which is almost 8 times higher than the original expense ratio or MER. That seems pretty suspicious to me, so I took a closer look.
In the example, the investor has $10,000 which they use to buy an ETF. After six months they sell the ETF and the author calculates the “expense ratio” which includes the trading fee and the bid-ask spread. While their math isn’t wrong, I’m not sure how common that scenario is.
Do most people buy ETFs if they only have $10,000? I would think that low-cost index funds would be a far better choice. The author is American and they have access to a lot of very low cost index funds, so in most cases – someone with a smaller portfolio wouldn’t likely be investing in ETFs.
Related: ETFs or Index funds – the best strategies
The other problem is that someone who changes their entire portfolio every six months is a reasonably active trader in my opinion. Yes, I realize that investors make changes, but do they sell everything every six months? Not likely.
Most investors are somewhat passive with a good part of their portfolio. Even active traders often have a good portion of their money in static investments and just “play” with a smaller amount of money.
The trading fee and bid-ask costs are only relevant for buys and sells, so the annual cost for ETFs that are invested long term will be very close to the MER of the ETF. If an investor does more frequent trades on a portion of their portfolio, that will increase their expenses, but not by much since the extra costs are spread over the entire portfolio – not just the actively traded portion.
Related: How to sell an ETF or Stock at a discount brokerage – Step by step directions
Trading fees
ETFs are securities traded on public stock exchanges and require a commission to be paid to a broker in order to be bought or sold. There are some commission-free ETFs, but they are in the minority.
Common sense is important when it comes to trading fees. If you are paying $25+ for your fees, then you need to find a cheaper broker – see the Canadian Discount Brokerage comparison for more information.
If you are paying $10 or less per trade, you still need to make sure that it makes sense to be buying ETFs. If you are buying $100 of an ETF each month and paying a $10 fee, that is not a good idea since your fee is 10% of the transaction. Way too high!
One rule of thumb which I like, is to not do any stock/ETF transactions where your trading fees are higher than 1% of the trade amount. So if you are paying $10 per trade, your minimum trade amount should be $1,000.
Here are some other suggestions on how to lower your trading costs when buying ETFs or stocks.
Bid-ask spread
The bid-ask spread is an investment cost which is difficult to understand (for me at least). The easiest way to think about it is pretend you buy some ETF shares right now for $15 per share. If you are a day trader, you might want to sell those shares in one hour. Let’s say the share price doesn’t change in that hour.
When you buy, you are paying the ‘ask’ price and when you sell, you receive the “bid” price (in theory at least). Therefore, if you buy and sell an ETF (or any individual stock) and the share price doesn’t change, you won’t get the same price you paid.
This may seem odd – if you buy a stock at $15.00 and sell at $14.97, didn’t the price go down and you lost three cents per share as a result. In reality, that’s possible, but in my example where the price didn’t change – your loss was a result of the bid-ask spread.
XIU.TO (my favourite Canadian ETF), for example has a bid of $17.57 and ask of $17.58. If you put in a market order, it will get filled at the ask price of $17.58.
The ‘ask’ price is the lowest price that sellers are offering their shares for on orders that are unfilled.
If at that same moment of time, sold the shares at market – you would likely get $17.57 per share for a loss of one cent per share. This loss is due to the difference or “spread” between the bid and ask offers for outstanding shares.
In our example, a one cent “commission’ represents 0.057% of our trade.
Related: Stock Purchase – Bid/Ask Prices
Let’s look at another ETF. VTI (all American stock ETF from Vanguard) which happens to be my favourite American ETF, right now also has a spread of one cent. The difference with VTI is that the price of each share is $72.53, so a one cent commission is only 0.014% which is about one quarter of the spread ‘commission’ we paid on the XIU share.
Obviously the spread commission as a percentage is not only a function of the difference between the bid and ask price, it’s also a function of the share price. The higher the share price is, the lower the commission as a percentage. One cent is a smaller fraction of a $75 stock vs a $18 stock.
The “ETF cost” author’s bid/spread example could be an ETF that has a three cent spread and a sub-$20 share price. While this isn’t unreasonable, if you stick to the larger ETFs, this shouldn’t happen.
I decided to check the bid/ask for all the ETFs I have in my portfolio and most of them were one penny:
- XIU – one cent
- VTI – one cent
- VEA – one cent
- XRB – two cents
- XSB – one cent
- BSV – two cents
This isn’t to say that these ETFs never have larger spreads. It can and does happen, but it’s not likely to get more than a couple of cents.
Both XRB and BSV are not heavily traded compared to the other ETFs which helps explain the larger spread. Even my most ‘expensive’ spread which occurs on XRB still only results in 0.07% commission which is half of the example in the article.
Let’s do a fun example
Since the author did a fairly arbitrary example which increased the ETF costs dramatically, I will do the same thing to show that the trading fees/spreads are insignificant – at least in my example.
Assumptions:
- Portfolio is $100,000 at the beginning of the year.
- Portfolio is made up of one ETF.
- ETF share price is $28 and never changes or pays out dividends.
- ETF annual expense is 0.17%.
- Investor purchases $1500 of one ETF at the end of each quarter with a trading cost of $10 and a two cent spread which is 0.07% of the share price. The trading fee comes out of the account.
- They can buy fractional shares (this makes the example cleaner).
- The time period is five years.
My theory is that the annual expense of 0.17% is the only one that matters.
After five years, the investor has $129,400 in their account and has paid:
- $954.61 in MER
- $200 in trading fees
- $5.36 due to the spread commission
The total cost over the five years is $1,160.
Now I will admit that in this case, trading fees are more significant than I originally thought, however the spread commission is not.
So depending on what kind of investor you are, the annual MER and to a lesser extent your trading fees will determine your total costs.
Moral of the story
Bottom line is that as long as your transaction fees are low and you don’t trade a lot, the annual ETF expense ratio or MER will be the largest part of your ETF cost. If you want to be a cost-conscious passive investor, buy broad-based, commonly traded ETFs with low annual expense ratios (or MERs as we like to say in Canada) and make sure you aren’t paying too much for trading fees.
Even if the combination of bid-ask spread and trading fees is high, assuming your transaction amounts are small portion of your portfolio – the MER is the main cost. Work on minimizing the MER before worrying about the other stuff.
If you don’t have a large portfolio ($100,000+), then I would take a good look to see if ETFs or low cost index funds are more appropriate.
5 replies on “The True Cost Of Owning An ETF – The MER Is The Main Thing”
Really great example here Mike, it’s great to see the numbers crunched like that. Even with extremely low MERs, it is still the biggest expense by far.
Mike: I agree that ETF bid-ask spreads are generally not a major expense, especially if you’re investing in large caps stocks or government bonds. But I do think trading commissions are too often overlooked by investors. Even in your example they represent 17% of the overall costs, and your example includes just one ETF purchased quarterly. Some people have a half a dozen ETFs in a small portfolio and trade more often, and I would not be surprised if trading commissions well exceeded MER in their overall costs. One more reason to consider index mutual funds for small accounts.
@SM – Thanks.
@CCP – Yes, even in my arbitrary example the trading commissions were more significant than I expected. It’s really hard to justify using ETFs for a smaller portfolio.
One justification might be Selection. I have a smaller portfolio and like to choose high performing ETFs like ITB. I can’t find a similar vehicle in mutual funds.
Would you agree that the MER (expense ratio) is a function of time aka the longer it is held the more it is? e.g. an ETF with an annual expense ratio of 0.001 (0.10%) on a $10000 investment would be approx. $10 for 1 year (annual) and $2.5 for 1 quarter (4 months) and for a day trader let’s say holding the investment for 4 hours the MER would be approx. $.005 ($10 / 365 / 24 * 4 =)???