Recently, Cinthia Murphy posed the question: Does the presence of a trading commission (or lack thereof) make a difference in whether or not advisors decide to use a given ETF?
What she found was, generally speaking, no: "For most advisors who manage 'fairly sizable ETF portfolios,’" she writes, "having a fund be offered commission-free 'rarely' sways their ETF pick."
Which begs the question: Should advisors care? What impact could a single flat fee have on how an ETF behaves in the wild?
Turns out, more than you might think.
The general agreement among the ETF industry is that, all else held equal, no fundamental behavioral difference exists between an ETF with a commission and one without. Why should there be? A trading commission introduces no structural differences to an ETF, nothing that should help or hamper a fund's ability to track its benchmark. It functions exactly the same as an ETF without a commission. Commission and no-commission funds aren't horses of a different color—just the same horse, at different prices.
That said, common sense suggests that no-commission ETFs might attract more trading activity than funds you have to pay for to buy and sell. After all, nobody likes paying fees they don't have to, and everyone likes a price tag of zero.
The facts on the ground seem to support this. Investors have been pouring their money into no-fee ETFs over the past few years they've been available. According to Murphy, assets in the funds offered through Charles Schwab's commission-free platform, Schwab ETF OneSource, are growing at twice the pace as ETFs not on the platform. And that's just one platform—Fidelity, E-Trade, TD Ameritrade and Interactive Brokers all offer their own commission-free platforms, with rapidly growing asset bases.
But are investors pouring money into these funds because of their lack of trading commissions, or because they find these funds highly desirable in and of themselves? Or does one inform the other?
We decided to dig into the numbers and find out.
The first hypothesis was this: ETFs available without trading commissions should show higher average volumes than ETFs only available with trading commissions.
And since higher volume should naturally result in tighter spreads, we added a secondary hypothesis: No-commission ETFs should show tighter average spreads than commission-only ETFs.
The focus here was specifically on ETFs in illiquid markets, such as emerging market equities, where funds inherently trade more thinly and with wider spreads. These markets can be higher-risk and/or their behavior harder to understand, meaning there's less chance of investors choosing funds in this space simply because they come without a fee.
Furthermore, if the presence of commissions does affect volumes and spreads, then this influence could make a substantive difference on how an illiquid ETF trades. After all, a difference of $100,000 in trading volume doesn't mean as much for a large-cap equity ETF as it does for, say, a municipal bond ETF.
We examined ETFs in five illiquid markets: emerging market bonds, emerging market equities, U.S. small-cap equities, municipal bonds and U.S. real estate.
The funds in the study are all traded commission-free on at least one platform, but in many cases, are commission-free on multiple platforms.
The sample set was intended to be as inclusive as possible: Wherever relevant, we opted for funds with a total market approach over single-country or region-specific ETFs. Factor funds were also included, such as those implementing dividend or low-volatility weighting, as well as size plays; a few narrow sector funds also found their way into the batch. However, leveraged or inverse ETFs were not considered, due to their unique structure and trading behavior.
That left a sample set as follows:
- 15 emerging market bond ETFs, including sovereign debt and corporate debt, and bonds denominated in U.S. dollars and local currency. (Only total market funds were considered; no Chinese-specific ETFs, for example, were included.) Eight ETFs had commissions, seven did not.
- 56 emerging market equity funds. Only broad emerging markets funds were considered, meaning single-country/region-specific funds or frontier market blends were excluded. Thirty-four ETFs had commissions, 22 did not.
- 32 U.S. small-cap equity ETFs. Microcap funds and small/midcap blends were excluded. Twenty-two ETFs had commissions, 10 did not.
- 35 U.S. muni bond ETFs were included. No state-specific plays were included (namely, California and New York). Twenty-nine ETFs had commissions, six did not.
- 17 U.S. real estate equity ETFs. Mostly, the funds in this space hold REITs and real estate holding companies. Twelve ETFs had commissions, five did not.
Then average volumes for each ETF were examined, as reported in the ETF.com database. These values are calculated by averaging the daily dollar value of shares traded over the past 45 trading days. (Dollar volume versus share volume was chosen to get a better sense of how much investor money was really moving in and out of these ETFs.)
Average spreads values were also taken from the ETF.com database. As with average volume, average spreads are calculated over a 45-day period.
All data was taken as of Dec. 3, 2016.
The results are listed in Table 1. Rarely do numbers tell a story as loudly as this.
Table 1: Spreads and Volume for Commission vs. No-Commission ETFs
|Sector||Mean Spread, Has Commission||Standard Deviation||Mean Spread, No Commission||Standard Deviation||Mean Volume, Has Commission ($M)||Mean Volume, No Commission ($M)|
|U.S. Small Caps||0.27%||0.14%||0.12%||0.10%||3.59||31.60|
|U.S. Real Estate||0.26%||0.25%||0.07%||0.09%||10.93||240.97|
Across the board, for all five market segments, average volumes for no-commission ETFs were higher than those for commission ETFs. Not just by a little bit, either.1 The smallest difference was for municipal bonds, where volume was eight times higher for no-fee ETFs than for ETFs with commissions.
The biggest? Emerging market bond ETFs, where no-fee funds had an average trading volume that was a whopping 250 times higher than that of commission ETFs. Differences in spreads are equally as dramatic.
For all five market segments, spreads for ETFs with commissions are higher than for those without. Much higher. Spreads for emerging market equities ETFs, for example, are more than 50% higher for ETFs with commissions than for ETFs without. They're more than twice as high for U.S. small-cap ETFs, and almost four times as high for U.S. real estate ETFs. Municipal bond ETFs with commissions have spreads that are five times as high as funds without.
The widest difference, however, is again in emerging market bonds, where the presence of a trading commission can mean a spread that is almost six times higher.
What's more, spreads for no-fee ETFs show much less deviation from the mean than spreads for commission ETFs. The greatest difference is, once again, in emerging market bonds, where standard deviation is 7.6 times higher for commission ETFs than for no-fee funds.
The chart below plots average spreads versus volume for all the funds in our sample set. (Volume is plotted on a logarithmic scale.) As to be expected, ETFs with higher volume exhibit smaller spreads, while ETFs with lower volume exhibit wider spreads. But the 102 ETFs with trading commissions exhibit higher variance within the sample set than do the 51 no-fee funds.
All in all, these results fall in line with the above-stated hypotheses: At least in illiquid markets, ETFs without trading commissions do indeed exhibit higher volume and tighter spreads than those with trading commissions.
Furthermore, it also appears that no-fee funds tend to accrue greater assets than ETFs carrying commissions. The average assets under management (AUM) for the 51 no-fee ETFs in our sample set was $3.87 billion, while the average AUM for the 102 commission-only ETFs was just $477 million.
There are many possible explanations for the results, but one is most obvious: People hate paying money when they feel they don't have to.
All things held equal, the lack of a commission on a given ETF entices more traders to trade it more actively, because investors won't have to worry about getting nickel-and-dimed on every single purchase and sale. In illiquid markets, where ownership costs may already be high (expense ratios are often much higher, for example), one less fee can make a big impact on an investor's bottom line.
Plus, when it comes to fund liquidity, herd mentality matters. It's not just about whether you personally choose to pay a $9.95 flat trading fee; it's whether everybody else chooses to pay it, too. If lots of people opt out, deciding that a trading commission is just too much to pay for a given ETF, then the fund won't trade well. Volumes will drop, spreads will widen and the ETF becomes less liquid as a result.
This means that funds with commissions and funds without commissions do not fundamentally behave the same. Apparently, no-fee ETFs are functionally more liquid than ETFs with commissions, at least for illiquid markets.
This has significant consequences for those illiquid markets, where investors struggle with less-than-ideal trading conditions. Lower trading volumes mean you might get stuck holding shares you can't sell, or unable to purchase shares when opportunity calls; wide spreads mean expensive roundtrip trades and higher costs. If a no-fee ETF in this space is more liquid than its competitors, it can reduce investors' market impact and offer a real cost advantage over the alternatives.
It's possible that this connection is correlation, not causation. Perhaps it isn't the lack of trading commissions that has made these funds more liquid. Maybe they were just more liquid already, for other independent reasons.
Some might suggest, for example, that only ETFs that were already very liquid and popular with investors were chosen for inclusion on commission-free platforms like Schwab's OneSource, introducing a selection effect that explains much of the results.
To an extent, that's true, in that Fidelity's initial run of 25 no-fee iShares funds, launched in February 2010, contained mostly big names like the iShares Barclays Aggregate Bond Fund (AGG), the iShares MSCI EAFE Index Fund (EFA) and the iShares MSCI Emerging Markets Index Fund (EEM).
However, the existence of commission-free ETFs actually predates Fidelity's launch. Schwab pioneered the concept in 2009, when it brought to market its own house-brand ETFs that carried no trading commissions for Schwab investors. (Vanguard followed suit in May 2010 with a similar setup for Vanguard investors.)
Commission-free trading remained mostly a novelty at Fidelity until March 2013, when it doubled the number of funds offered through the program. Not coincidentally, this was just one month after Schwab introduced OneSource, with 105 ETFs offered no-fee. From there, it became an arms race to see who could offer more (and more attractive) no-fee ETFs to investors.
Essentially, what history tells us is that the initial motivation for forgoing ETF trading commissions wasn't necessarily about improving access to ETFs that investors had already made popular. It was a marketing tool for brokerage houses to entice investors toward their specific brands.
Secondly, while some commission-free ETFs are large, liquid funds, many more are not. Hundreds of no-fee ETFs of modest assets exist, far outnumbering the more well-known names, and many of these funds were introduced commission-free at inception. Meanwhile, some of the biggest ETFs—like EFA and EEM—no longer trade commission-free, or they never went commission-free in the first place, such as the SPDR S&P 500 ETF (SPY).
What's interesting is that, as our results show, the average AUM of no-fee ETFs in illiquid spaces is higher—roughly eight times higher, in fact—than that of comparable ETFs carrying commissions. That further suggests investors do, in fact, tend to favor commission-free ETFs when given the choice between two similar funds in the same space.
All that said, there's one obvious conclusion to draw from our results: Even if advisors plan to buy and hold ETFs, they ought to care more about their trading commissions, especially in illiquid markets.
- Values for average trading volumes do not include EEM (emerging market equities) and IWM (U.S. small-caps), both of which possess trading commissions. The reason is that these two ETFs trade in volumes that are so much higher than their peers that including them in the sample dramatically skews the results. Average volume for emerging market stock ETFs with commissions shoots from $710,000 to $72 million, just by including EEM in the calculation. Likewise, including IWM in the sample alone increases volume for small-cap ETFs with commission from $3.59 million to $188 million. Since these two funds are clearly outliers, they are therefore excluded from the average volume analysis.