Here's a statement of the obvious: The opinions expressed here are those of the participants, not those of the Mutual Fund Observer. We cannot vouch for the accuracy or appropriateness of any of it, though we do encourage civility and good humor.
I know that David has expressed his distaste for ETF's, but I either don't remember his reasons or maybe he didn't mention them. Here though are some concerns that can impact the selection of ETF's.
I'm also not a fan of ETFs. Though I can't speak for David, I suspect some of my objections overlap with his.
- bid/ask spread. With an ETF you're almost guaranteed to lose. (Even if you put in a limit order, that's primarily going to protect you against momentary market jiggles, not spread.) An OEF gives you 100c on the NAV dollar.
- tracking error. Not vs. the index (OEF index funds have the same propensity), but vs. the NAV (even after allowing for bid/ask spread). Authorized participants act to keep the price close to NAV, but only if their cost (trading the underlying securities plus the fee to convert creation units) is less than the deviation of the market price from NAV. They are less inclined to act when the market is in turmoil, i.e. when you're more likely to want to get out of a position.
There is a counter argument for foreign funds, viz. the market can price foreign index funds better than an OEF fund board using fair value pricing. Color me skeptical here. Even ETFs use fair value pricing when generating intraday indicative values.
- often no cost advantage (e.g. Vanguard Admiral class shares vs. Vanguard ETF class shares of the same portfolio).
On the plus side for ETFs, in theory ETFs are more tax-efficient. I say in theory because many OEF broad based funds are well managed and distribute little if any cap gains as well. Nor does this matter in tax-advantaged accounts.
Another plus is that you can buy ETFs with lower commissions than you can buy TF funds.
@MFO Members: Nonsense ! The truth is some thinly traded ETFs have very wide bid/ask spreads even if they hold very liquid stocks. The spread is the difference between the lowest price a trader is willing to sell the ETF and the highest price a buyer is willing to pay. The wider the spread, the bigger the immediate loss upon buying the ETF. Here is the current spread on SPY the worlds largest fund with 230.5 billion in assets. Bid/Ask/Spread 234.08/234.09. QQQ with 46.1 billion in assets 131.96/ 131.97. Regards, Ted
If I want to buy a $10K interest the Vanguard 500 portfolio, I could pay exactly $10K for VFIAX. Or I could pay a bit more for VOO shares representing the same $10K interest in the portfolio because of the added cost of the spread. I'd prefer not to pay more than $10K for $10K worth of securities.
Regardless of how large or small that added cost is, it is a drawback inherent in the ETF design. Though perhaps it is one that you may not personally care about.
Regarding the size of that spread, Ted's 1 basis point spread is more the exception than the rule. For example, CAPE has a typical spread of 15 basis points. Here's Vanguard's table of spreads on its ETF share class.
With SPY and VFINX Ted is comparing oranges and tangerines (close but still different). The question was what downsides there were to ETFs, not whether ETF 1 was better or worse than OEF 2.
The only way I know to do an apples to apples comparison using concrete funds rather than discussing different attributes of ETFs and OEF is to compare ETF and OEF shares of the same underlying portfolio. Say VFIAX vs. VOO. Since these are shares of the same portfolio, and since they have the same ER, the only factors affecting performance should be due to the nature of the shares and not of the particular fund.
As of April 18, 2017, VFIAX'sVOO's cumulative return was 97.23%, while VOO's was 96.59%. That doesn't include dinging VOO for the cost of the spread.
If you add SPY to the M* comparison chart, you'll find its performance was even worse, at 95.66%. But that's because of the design of the fund (cash drag due to UIT structure plus higher ER). Those additional variables confound the data.
A tip for the linkster - to link to a M* page comparing funds, one needs to link to the chart page (there's a "share this chart" link there). The only fund that shows up when one links to a M* performance page is the original fund; the compared funds don't get passed through the link.
>> I'd prefer not to pay more than $10K for $10K worth of securities.
haha, there goes the system. I'd prefer not to pay any markup ever for anything, cars, groceries, furniture. Actually no, I don't mind; I know that's how the service gets provided. Everything in life should be like VFIAX.
CAPE is an etn. But to switch fruit, since it outperforms VFIAX so consistently, from inception and every interval since, again, who cares?
>> I'd prefer not to pay more than $10K for $10K worth of securities.
haha, there goes the system. I'd prefer not to pay any markup ever for anything, cars, groceries, furniture. Actually no, I don't mind; I know that's how the service gets provided. Everything in life should be like VFIAX.
We should be clear on what service we're talking about. The portfolio management service is provided via management fees (part of the ER). It is the service of selling the product (here, fund shares) where the markup in question is added.
You have a choice. We all do. One can go directly to the "manufacturer" (fund company) and buy the product directly without paying a middleman a markup, or one can pay that third party for the service. In the case of ETFs, the third party is the broker who gets paid via spreads and possibly commissions.
Even if one wants to avoid middleman costs by using an OEF, one may not have that choice. Some funds already have that middleman cost priced in (higher ERs). Thus those funds charge a markup whether one uses a middleman (broker) or not.
Other funds like VFIAX offer the choice of paying a markup or not. One can go directly to Vanguard, or one can pay a middleman (broker) a commission (transaction fee) for the service.
You wrote that everything should be like VFIAX (i.e. available without a markup), and that using middlemen services entails markups. So we may reasonably conclude that you'd prefer to have the choice that VFIAX offers, viz. buying directly from the fund sponsor or paying the middleman (broker) a transparent fee for each transaction. None of these hidden fees like spreads on ETFs or shelf space kickbacks for NTFs.
Transparency and lower costs sound good to me too.
Roger all. Also, I was misdirecting your point, about fees for the exact same thing.
Still, what did this mean?
>> prefer not to pay more than $10K for $10K worth of securities.
Does it intend to say that paying the least amount above $10k for $10k of securities is the most frugal course of action? True, and it goes without saying, almost.
Not to make too fine a point of it, but…we are talking about buying a portfolio or a basket of securities. To replicate the holdings myself, I'd pay a fortune in commissions. The convenience of having the OEF or ETF buy and sell the securities is surely worth something, maybe even more than the derisory low fees Vanguard charges. I'm glad to pay for that, but I'm careful with the spread on an ETN or CEF.
I think we're all in agreement that funds (whether OEF or ETF) provide a valuable service for which they deserve fair compensation (management fees) and fees to cover reasonable costs (legal fees, servicing your account, handling proxy material, etc.).
Open end shares and exchange traded shares are two ways to gain access to this service. ISTM that the more middlemen that are involved, the more costs are layered on, even if they're not called out explicitly. These might include spreads with ETFs, and shelf space fees at brokerages for OEFs (especially ones sold NTF).
One may feel that these additional fees are also worth it for value received. But if one does not want to pay the middleman (explicitly or implicitly), then one needs to go directly to the fund sponsor. That's something one can't do with an ETF. (One also can't do that with CEFs or with some open end fund companies like Janus, that sell to retail customers only via brokers.)
That's what I meant by getting $10K of securities for $10K. I don't want to pay the middleman. (Though I confess that I find Fidelity's $5 fee worth the price to gain the convenience of not having to open yet another account elsewhere.)
There can be oddball cases. Vanguard Emerging Market Index fund used to charge a fee (I think it was 1%) when you bought open end shares of the fund. This was to cover the trading costs incurred by the fund in procuring stocks with your investment dollars.
That fee made it possible for the ETF share class VWO to be cheaper to buy, at least assuming the market price was close to the NAV. All you had to pay for the ETF was the spread and any commissions. So you might get $9950 worth of securities by buying the ETF (after all trading costs were considered), as opposed to only $9900 worth of securities by buying the OEF (after paying the 1% entry fee).
Roger all, again, ... but why do you keep putting it like this?
Only in a world of no trading costs, and no other costs either. Not only does it never happen, it ought never to happen. (And it never does happen.) That was my point. I remain perplexed about why you put it as you do. It costs something to get you your $10k of securities, does it not?
Mutual fund X has $200 worth of assets, and has 10 shares outstanding. I go to the distributor of fund X, hand it $20. I am issued one share. Fund X now has $220 worth of assets, and with 11 shares outstanding, each share represents a 1/11 interest in those assets, i.e. $20.
You may choose to view this as my getting a single share of the fund, a security worth $20 based on current market price. Or you may choose to regard this transaction as providing me a 1/11 interest (worth $20) in a $220 portfolio of securities and cash. Either way, I've received 100c on the dollar. No overhead.
When one purchases shares through a DRIP plan, one may even get a discount. See, e.g. WTR's plan (5% discount on dividend reinvestments).
Not many companies offer DRIPs at discounts with no set up fees, no maintenance fees, no purchase fees, but a few like WTR do. (WTR does charge an annual fee if you want the shares in an IRA.) Can't say it never happens.
Comments
Regards,
Ted
http://www.mutualfundobserver.com/discuss/discussion/32520/ben-carlson-when-an-etf-changes-its-stripes#latest
- bid/ask spread. With an ETF you're almost guaranteed to lose. (Even if you put in a limit order, that's primarily going to protect you against momentary market jiggles, not spread.) An OEF gives you 100c on the NAV dollar.
- tracking error. Not vs. the index (OEF index funds have the same propensity), but vs. the NAV (even after allowing for bid/ask spread). Authorized participants act to keep the price close to NAV, but only if their cost (trading the underlying securities plus the fee to convert creation units) is less than the deviation of the market price from NAV. They are less inclined to act when the market is in turmoil, i.e. when you're more likely to want to get out of a position.
There is a counter argument for foreign funds, viz. the market can price foreign index funds better than an OEF fund board using fair value pricing. Color me skeptical here. Even ETFs use fair value pricing when generating intraday indicative values.
- petty SEC Section 31 fees on ETF sales.
- often no cost advantage (e.g. Vanguard Admiral class shares vs. Vanguard ETF class shares of the same portfolio).
On the plus side for ETFs, in theory ETFs are more tax-efficient. I say in theory because many OEF broad based funds are well managed and distribute little if any cap gains as well. Nor does this matter in tax-advantaged accounts.
Another plus is that you can buy ETFs with lower commissions than you can buy TF funds.
Regards,
Ted
>> bid/ask spread. With an ETF you're almost guaranteed to lose.
>> tracking error.
What do these mean in consequence? That is, even if true, so what ?
Regards,
Ted
http://performance.morningstar.com/fund/performance-return.action?t=VFINX®ion=usa&culture=en_US
Regardless of how large or small that added cost is, it is a drawback inherent in the ETF design. Though perhaps it is one that you may not personally care about.
Regarding the size of that spread, Ted's 1 basis point spread is more the exception than the rule. For example, CAPE has a typical spread of 15 basis points. Here's Vanguard's table of spreads on its ETF share class.
With SPY and VFINX Ted is comparing oranges and tangerines (close but still different). The question was what downsides there were to ETFs, not whether ETF 1 was better or worse than OEF 2.
The only way I know to do an apples to apples comparison using concrete funds rather than discussing different attributes of ETFs and OEF is to compare ETF and OEF shares of the same underlying portfolio. Say VFIAX vs. VOO. Since these are shares of the same portfolio, and since they have the same ER, the only factors affecting performance should be due to the nature of the shares and not of the particular fund.
Here's the M* comparison over the past ten years.
As of April 18, 2017, VFIAX'sVOO's cumulative return was 97.23%, while VOO's was 96.59%. That doesn't include dinging VOO for the cost of the spread.
If you add SPY to the M* comparison chart, you'll find its performance was even worse, at 95.66%. But that's because of the design of the fund (cash drag due to UIT structure plus higher ER). Those additional variables confound the data.
A tip for the linkster - to link to a M* page comparing funds, one needs to link to the chart page (there's a "share this chart" link there). The only fund that shows up when one links to a M* performance page is the original fund; the compared funds don't get passed through the link.
haha, there goes the system. I'd prefer not to pay any markup ever for anything, cars, groceries, furniture. Actually no, I don't mind; I know that's how the service gets provided. Everything in life should be like VFIAX.
CAPE is an etn. But to switch fruit, since it outperforms VFIAX so consistently, from inception and every interval since, again, who cares?
You have a choice. We all do. One can go directly to the "manufacturer" (fund company) and buy the product directly without paying a middleman a markup, or one can pay that third party for the service. In the case of ETFs, the third party is the broker who gets paid via spreads and possibly commissions.
Even if one wants to avoid middleman costs by using an OEF, one may not have that choice. Some funds already have that middleman cost priced in (higher ERs). Thus those funds charge a markup whether one uses a middleman (broker) or not.
Other funds like VFIAX offer the choice of paying a markup or not. One can go directly to Vanguard, or one can pay a middleman (broker) a commission (transaction fee) for the service.
You wrote that everything should be like VFIAX (i.e. available without a markup), and that using middlemen services entails markups. So we may reasonably conclude that you'd prefer to have the choice that VFIAX offers, viz. buying directly from the fund sponsor or paying the middleman (broker) a transparent fee for each transaction. None of these hidden fees like spreads on ETFs or shelf space kickbacks for NTFs.
Transparency and lower costs sound good to me too.
Still, what did this mean?
>> prefer not to pay more than $10K for $10K worth of securities.
Does it intend to say that paying the least amount above $10k for $10k of securities is the most frugal course of action? True, and it goes without saying, almost.
Open end shares and exchange traded shares are two ways to gain access to this service. ISTM that the more middlemen that are involved, the more costs are layered on, even if they're not called out explicitly. These might include spreads with ETFs, and shelf space fees at brokerages for OEFs (especially ones sold NTF).
One may feel that these additional fees are also worth it for value received. But if one does not want to pay the middleman (explicitly or implicitly), then one needs to go directly to the fund sponsor. That's something one can't do with an ETF. (One also can't do that with CEFs or with some open end fund companies like Janus, that sell to retail customers only via brokers.)
That's what I meant by getting $10K of securities for $10K. I don't want to pay the middleman. (Though I confess that I find Fidelity's $5 fee worth the price to gain the convenience of not having to open yet another account elsewhere.)
There can be oddball cases. Vanguard Emerging Market Index fund used to charge a fee (I think it was 1%) when you bought open end shares of the fund. This was to cover the trading costs incurred by the fund in procuring stocks with your investment dollars.
That fee made it possible for the ETF share class VWO to be cheaper to buy, at least assuming the market price was close to the NAV. All you had to pay for the ETF was the spread and any commissions. So you might get $9950 worth of securities by buying the ETF (after all trading costs were considered), as opposed to only $9900 worth of securities by buying the OEF (after paying the 1% entry fee).
Roger all, again, ... but why do you keep putting it like this?
Only in a world of no trading costs, and no other costs either. Not only does it never happen, it ought never to happen. (And it never does happen.) That was my point. I remain perplexed about why you put it as you do. It costs something to get you your $10k of securities, does it not?
You may choose to view this as my getting a single share of the fund, a security worth $20 based on current market price. Or you may choose to regard this transaction as providing me a 1/11 interest (worth $20) in a $220 portfolio of securities and cash. Either way, I've received 100c on the dollar. No overhead.
When one purchases shares through a DRIP plan, one may even get a discount. See, e.g. WTR's plan (5% discount on dividend reinvestments).
Not many companies offer DRIPs at discounts with no set up fees, no maintenance fees, no purchase fees, but a few like WTR do. (WTR does charge an annual fee if you want the shares in an IRA.) Can't say it never happens.