Total returns

Discuss all general (i.e. non-personal) investing questions and issues, investing news, and theory.
Post Reply
coffeecup333
Posts: 32
Joined: Wed May 17, 2017 9:00 pm

Total returns

Post by coffeecup333 » Wed Jul 11, 2018 3:26 pm

A few questions I hope can be clarified:

1) I am invested in total international index admiral and investor shares. Looking at the total returns for 2017, the investor shares have returned 27.40% vs the benchmark of 27.41% and the admiral shares have a total return of 27.55% vs the 27.41% benchmark. If they are measuring the same index, why is there a difference in return in share classes? I called the Vanguard representatives to clarify this but it seems no one seems to know (with the explanation difference in expense ratios also don't account for this). Anyone know why this is?

2) What are capital returns, income returns, and total returns? Clearly they equal the total return but how are they calculated? What are they used for?

3) How can I figure out the expected rate of return of an asset allocation? What is reasonable to expect as a return for each asset class investing in the 3 fund Portfolio vs the overall return of a particular AA? In other words how does the expected return of each investment class affect or not affect the expected return of the AA? Looking at 75/25 more or less.

bloom2708
Posts: 4285
Joined: Wed Apr 02, 2014 2:08 pm
Location: Fargo, ND

Re: Total returns

Post by bloom2708 » Wed Jul 11, 2018 3:33 pm

Investor shares have a lower entry point but a higher expense ratio.

The difference is the higher cost of the Investor shares compared to Admiral shares.

Income is the dividends received. The other part of Total Return is the increase (or decrease = net change) in stock price.
"We are here not to please but to provoke thoughtfulness" Unknown Boglehead

alex_686
Posts: 3604
Joined: Mon Feb 09, 2015 2:39 pm

Re: Total returns

Post by alex_686 » Wed Jul 11, 2018 3:49 pm

1. I would point out that you have the same benchmark return for both, so I am a bit confused by the "If they are measuring the same index". Having done this for a living, different returns for different share classes must come down to the expense ratios. Or something very very wrong is happening.

2. There have been really thick and technical books written on this subject. Generally:
Dividends = Qualified / Ordinary Income
Gains from sale of securities held for less than 1 year = short term
Gains from sale of securities held for more than 1 year = long term

The above items are used to calculate your tax return. Uncle Sam needs to be paid.
Total Return assume the above are reinvested back into the fund plus unrealized capital gains.

3. Ah, the million dollar question. Here is a recent ongoing discussion. Probably lower than historical averages.
viewtopic.php?t=253256

User avatar
vineviz
Posts: 524
Joined: Tue May 15, 2018 1:55 pm

Re: Total returns

Post by vineviz » Wed Jul 11, 2018 4:13 pm

The difference in published expense ratio accounts for a good amount of the difference, but also note that the fund has a fiscal year that ends in October so the published expense ratio at any given moment may not be the actual expense for any particular calendar period.

Also, the published expense ratio does NOT include all the expenses a fund incurred. It's basically just the management expenses. Trading costs, for instance, are not included and its possible that they are lower (on a percentage basis) for Admiral shares. Also revenue from securities lending gets netted against expenses, and that can vary by share class as well.

FactualFran
Posts: 668
Joined: Sat Feb 21, 2015 2:29 pm

Re: Total returns

Post by FactualFran » Wed Jul 11, 2018 4:24 pm

A likely reason that the difference in the total returns is larger than the difference between the expense ratios is that the total returns are calculated with the NAVs rounded to two digits after the decimal point. With the investors shares of the fund, if the more precise NAV of the fund had been 0.0049 less at the end of 2016 than the reported rounded value and the more precise NAV of the fund had been 0.0049 more at the end of 2017, then the total return would have been 27.47%, not the reported 27.40%.

alex_686
Posts: 3604
Joined: Mon Feb 09, 2015 2:39 pm

Re: Total returns

Post by alex_686 » Wed Jul 11, 2018 4:25 pm

vineviz wrote:
Wed Jul 11, 2018 4:13 pm
Also, the published expense ratio does NOT include all the expenses a fund incurred. It's basically just the management expenses. Trading costs, for instance, are not included and its possible that they are lower (on a percentage basis) for Admiral shares. Also revenue from securities lending gets netted against expenses, and that can vary by share class as well.
I get where you are going with this but the 2 examples you gave are wrong. You can only charge expenses to a class where that expense can be specially associated with that class. Publishing, postage, and marketing are the classic examples. Trading costs - you are correct here that it is not counted - can't be specifically associated with a specific share class.

Further, the portfolio manager's fee is specifically broken out. Operation's expenses tend to be the same amount as the manager's fee.

User avatar
vineviz
Posts: 524
Joined: Tue May 15, 2018 1:55 pm

Re: Total returns

Post by vineviz » Wed Jul 11, 2018 4:34 pm

alex_686 wrote:
Wed Jul 11, 2018 4:25 pm

I get where you are going with this but the 2 examples you gave are wrong. You can only charge expenses to a class where that expense can be specially associated with that class. Publishing, postage, and marketing are the classic examples. Trading costs - you are correct here that it is not counted - can't be specifically associated with a specific share class.
It can be, both legally and practically. Trading costs are a function, in part at least, of fund flows. If one share class is generating more net flows then that class is going to bear the weight of the associated trading costs.

Trades due to rebalancing etc will be allocated in proportion to assets, of course.

Same for security lending income, which offsets expenses. In fiscal 2017, the admiral shares had income a few bps greater than the investor class. The difference was small, but enough to show up in the annual report.

alex_686
Posts: 3604
Joined: Mon Feb 09, 2015 2:39 pm

Re: Total returns

Post by alex_686 » Wed Jul 11, 2018 9:16 pm

vineviz wrote:
Wed Jul 11, 2018 4:34 pm
It can be, both legally and practically. Trading costs are a function, in part at least, of fund flows. If one share class is generating more net flows then that class is going to bear the weight of the associated trading costs.

Trades due to rebalancing etc will be allocated in proportion to assets, of course.

Same for security lending income, which offsets expenses. In fiscal 2017, the admiral shares had income a few bps greater than the investor class. The difference was small, but enough to show up in the annual report.
I will have to look into this. This does not square with my understanding of the industry, and I did this stuff for 10 years.

coffeecup333
Posts: 32
Joined: Wed May 17, 2017 9:00 pm

Re: Total returns

Post by coffeecup333 » Thu Jul 12, 2018 4:58 pm

https://investor.vanguard.com/mutual-fu ... ve-returns
https://investor.vanguard.com/mutual-fu ... ve-returns

According to the links above, the benchmark returns were 21.41%.
Expense ratios:
.11% vtiax and .17% for vgtsx

Returns:
27.55% vtiax and 27.40% vgtsx

Subtracting the expense ratios off the top of these returns gives me 27.44% and 27.23% which are no where near the benchmarks. So there must be some other math in here as suggested.

Regarding the question about the AAs and returns, all the calculators out there have a rate of return as an input to figure out retirement, savings, etc. but my investments are divided based on an AA into the 3 fund categories. What I'm trying to figure out is what rate of return gets put into these calculators as the "one number" if each one of my stocks/bonds/international stock index funds' returns are different? What happens if I change my AA to different ones to see different outcomes? Does this then change the overall rate of return again since it would change my index holdings returns? This would help with setting some short, mid or long term goals.

Thank you all for the clarifications!

FactualFran
Posts: 668
Joined: Sat Feb 21, 2015 2:29 pm

Re: Total returns

Post by FactualFran » Thu Jul 12, 2018 5:10 pm

coffeecup333 wrote:
Thu Jul 12, 2018 4:58 pm
https://investor.vanguard.com/mutual-fu ... ve-returns
https://investor.vanguard.com/mutual-fu ... ve-returns

According to the links above, the benchmark returns were 21.41%.
Expense ratios:
.11% vtiax and .17% for vgtsx

Returns:
27.55% vtiax and 27.40% vgtsx

Subtracting the expense ratios off the top of these returns gives me 27.44% and 27.23% which are no where near the benchmarks. So there must be some other math in here as suggested.
The reported total returns are after expense ratio values. Subtracting the expense ratio from the total return gives a meaningless result. Adding the expense ratio the total return results in a return that excludes the expense ratio.

User avatar
jeffyscott
Posts: 6997
Joined: Tue Feb 27, 2007 9:12 am
Location: Wisconsin

Re: Total returns

Post by jeffyscott » Thu Jul 12, 2018 5:21 pm

coffeecup333 wrote:
Thu Jul 12, 2018 4:58 pm
According to the links above, the benchmark returns were 21.41%.
Expense ratios:
.11% vtiax and .17% for vgtsx

Returns:
27.55% vtiax and 27.40% vgtsx

Subtracting the expense ratios off the top of these returns gives me 27.44% and 27.23% which are no where near the benchmarks. So there must be some other math in here as suggested.
Benchmark is 27.41%, subtracting expense ratios from that gives 27.30% and 27.24%. So both funds had higher returns than would be expected from that calculation, but they are close.
press on, regardless - John C. Bogle

VaR
Posts: 552
Joined: Sat Dec 05, 2015 11:27 pm

Re: Total returns

Post by VaR » Thu Jul 12, 2018 6:23 pm

jeffyscott wrote:
Thu Jul 12, 2018 5:21 pm
coffeecup333 wrote:
Thu Jul 12, 2018 4:58 pm
According to the links above, the benchmark returns were 21.41%.
Expense ratios:
.11% vtiax and .17% for vgtsx

Returns:
27.55% vtiax and 27.40% vgtsx

Subtracting the expense ratios off the top of these returns gives me 27.44% and 27.23% which are no where near the benchmarks. So there must be some other math in here as suggested.
Benchmark is 27.41%, subtracting expense ratios from that gives 27.30% and 27.24%. So both funds had higher returns than would be expected from that calculation, but they are close.
That's not really how you're supposed to do it either. A better approximation is to add the expense ratio back into the return:
VTIAX 27.55 + .11 = 27.66
VGTSX 27.40 + .17 = 27.57

But I see that you used the year-end 2017 investment return numbers.

Off of those same pages, if you use the cumulative total returns as of 6/30/2018, you get:
VTIAX 7.10 + .11 = 7.21
VGTSX 7.03 + .17 = 7.20
Spl Total International Stock Index = 7.58

In this case, the difference in returns between the Investor share class and the Admiral share class is entirely explained by the difference in expense ratio.

User avatar
jeffyscott
Posts: 6997
Joined: Tue Feb 27, 2007 9:12 am
Location: Wisconsin

Re: Total returns

Post by jeffyscott » Thu Jul 12, 2018 7:17 pm

VaR wrote:
Thu Jul 12, 2018 6:23 pm
That's not really how you're supposed to do it either.
I'm not the arbiter of what is "supposed" to be done, but it's the exact same comparison, whether you subtract ERs from the index or add them to the funds.
press on, regardless - John C. Bogle

MIretired
Posts: 702
Joined: Fri Sep 06, 2013 12:35 pm

Re: Total returns

Post by MIretired » Thu Jul 12, 2018 9:04 pm

VaR wrote:
Thu Jul 12, 2018 6:23 pm
jeffyscott wrote:
Thu Jul 12, 2018 5:21 pm
coffeecup333 wrote:
Thu Jul 12, 2018 4:58 pm
According to the links above, the benchmark returns were 21.41%.
Expense ratios:
.11% vtiax and .17% for vgtsx

Returns:
27.55% vtiax and 27.40% vgtsx

Subtracting the expense ratios off the top of these returns gives me 27.44% and 27.23% which are no where near the benchmarks. So there must be some other math in here as suggested.
Benchmark is 27.41%, subtracting expense ratios from that gives 27.30% and 27.24%. So both funds had higher returns than would be expected from that calculation, but they are close.
That's not really how you're supposed to do it either. A better approximation is to add the expense ratio back into the return:
VTIAX 27.55 + .11 = 27.66
VGTSX 27.40 + .17 = 27.57

But I see that you used the year-end 2017 investment return numbers.

Off of those same pages, if you use the cumulative total returns as of 6/30/2018, you get:
VTIAX 7.10 + .11 = 7.21
VGTSX 7.03 + .17 = 7.20
Spl Total International Stock Index = 7.58

In this case, the difference in returns between the Investor share class and the Admiral share class is entirely explained by the difference in expense ratio.
A better way is to see that the ER is taken out of the yield.
VTIAX TTM yield is 2.91%
VGTSX is 2.85% >>=.06% difference; same as .17% ER - .11% ER
Then you can look at price comparison chart (chart of div not re-invested) and they are identically (+/- a rounding error)
Note: they do not have same NAV because investor class started earlier than adm. usually.
http://quotes.morningstar.com/chart/fun ... A%5B%5D%7D

VaR
Posts: 552
Joined: Sat Dec 05, 2015 11:27 pm

Re: Total returns

Post by VaR » Fri Jul 13, 2018 1:29 am

jeffyscott wrote:
Thu Jul 12, 2018 7:17 pm
VaR wrote:
Thu Jul 12, 2018 6:23 pm
That's not really how you're supposed to do it either.
I'm not the arbiter of what is "supposed" to be done, but it's the exact same comparison, whether you subtract ERs from the index or add them to the funds.
Apologies for that, I was going to do it differently but then changed my mind and decided to abandon that calculation. You're right that the calculation is essentially equivalent either way.

User avatar
House Blend
Posts: 4425
Joined: Fri May 04, 2007 1:02 pm

Re: Total returns

Post by House Blend » Fri Jul 13, 2018 9:36 am

coffeecup333 wrote:
Thu Jul 12, 2018 4:58 pm
Subtracting the expense ratios off the top of these returns gives me 27.44% and 27.23% which are no where near the benchmarks. So there must be some other math in here as suggested.
Others have already pointed out that you should have *added* the expense ratio to the fund returns.

Let me add that for an index fund that invests in international stocks, comparing fund returns with the returns of the benchmark is an exercise in futility, whether you adjust for ER or not.

The reason involves fair value pricing. In brief, because the US market is open when overseas markets are closed and vice-versa, the prices of international funds get adjusted daily at close in anticipation of the effects of the day's events on overseas markets. In the olden days, before these adjustments became standard, traders would profit from it.

But benchmarks don't make this adjustment. So for any time interval you pick, the start and end values are subject to this effect. Which makes it much more challenging to usefully measure tracking error for an international fund. The fuzziness can easily be big enough to swamp the effects of a 0.10% expense ratio.

See: https://www.bogleheads.org/wiki/Fair_value_pricing

FWIW, this doesn't explain why the spread in returns between Investor and Admiral shares seems to be wider than the ER for the period you were looking at. Echoing what FactualFran mentioned upthread, with one fund trading at ~$18/share and the other at $30/share, the fuzziness from penny rounding alone can add up to roughly 5 basis points in either direction.

User avatar
jeffyscott
Posts: 6997
Joined: Tue Feb 27, 2007 9:12 am
Location: Wisconsin

Re: Total returns

Post by jeffyscott » Fri Jul 13, 2018 9:44 am

VaR wrote:
Fri Jul 13, 2018 1:29 am
jeffyscott wrote:
Thu Jul 12, 2018 7:17 pm
VaR wrote:
Thu Jul 12, 2018 6:23 pm
That's not really how you're supposed to do it either.
I'm not the arbiter of what is "supposed" to be done, but it's the exact same comparison, whether you subtract ERs from the index or add them to the funds.
Apologies for that, I was going to do it differently but then changed my mind and decided to abandon that calculation. You're right that the calculation is essentially equivalent either way.
No problem, I would typically look at it the way you did, add the ER to the funds. But since the OP had mistakenly subtracted the ERs from the funds, I was indicating that if you are going to subtract from something, it would be from the index. :beer
press on, regardless - John C. Bogle

coffeecup333
Posts: 32
Joined: Wed May 17, 2017 9:00 pm

Re: Total returns

Post by coffeecup333 » Fri Jul 13, 2018 10:46 am

House Blend wrote:
Fri Jul 13, 2018 9:36 am
coffeecup333 wrote:
Thu Jul 12, 2018 4:58 pm
Subtracting the expense ratios off the top of these returns gives me 27.44% and 27.23% which are no where near the benchmarks. So there must be some other math in here as suggested.
Others have already pointed out that you should have *added* the expense ratio to the fund returns.

Let me add that for an index fund that invests in international stocks, comparing fund returns with the returns of the benchmark is an exercise in futility, whether you adjust for ER or not.

The reason involves fair value pricing. In brief, because the US market is open when overseas markets are closed and vice-versa, the prices of international funds get adjusted daily at close in anticipation of the effects of the day's events on overseas markets. In the olden days, before these adjustments became standard, traders would profit from it.

But benchmarks don't make this adjustment. So for any time interval you pick, the start and end values are subject to this effect. Which makes it much more challenging to usefully measure tracking error for an international fund. The fuzziness can easily be big enough to swamp the effects of a 0.10% expense ratio.

See: https://www.bogleheads.org/wiki/Fair_value_pricing

FWIW, this doesn't explain why the spread in returns between Investor and Admiral shares seems to be wider than the ER for the period you were looking at. Echoing what FactualFran mentioned upthread, with one fund trading at ~$18/share and the other at $30/share, the fuzziness from penny rounding alone can add up to roughly 5 basis points in either direction.

Thank you for such a clear explanation. I'm actually surprised at the discrepancy in the share prices and never knew about fair value pricing. So the take away is for international stocks, the returns will never be in line completely with the benchmark. But now looking at my vtsax, the difference between the benchmark of 21.19% and 21.17% of vtsax's total return in 2017, is a difference of .02%, not the ER of .04%. Do you know why could be? https://investor.vanguard.com/mutual-fu ... ve-returns

User avatar
House Blend
Posts: 4425
Joined: Fri May 04, 2007 1:02 pm

Re: Total returns

Post by House Blend » Fri Jul 13, 2018 10:58 am

coffeecup333 wrote:
Fri Jul 13, 2018 10:46 am
But now looking at my vtsax, the difference between the benchmark of 21.19% and 21.17% of vtsax's total return in 2017, is a difference of .02%, not the ER of .04%. Do you know why could be? https://investor.vanguard.com/mutual-fu ... ve-returns
I would wager that most of that is due to securities lending. Which overall is a pretty good deal for investors, assuming the profits get plowed back into the fund.

See: https://www.bogleheads.org/wiki/Securities_lending

FactualFran
Posts: 668
Joined: Sat Feb 21, 2015 2:29 pm

Re: Total returns

Post by FactualFran » Fri Jul 13, 2018 1:12 pm

Just to reinforce the fact that the reported total return is after expenses, if the only transactions for an account in Admiral shares of the fund were a purchase of $10,000 at the end of 2016 and reinvesting distributions, then the account balance at the end of 2017 would have been $12,755. The account balance have increased by the reported total return of 27.55%.

In more detail, a purchase of $10,000.00 at the end of 2016 would have bought 406.009 shares at an NAV of 24.63. Due to reinvesting each of the four distributions during 2017, the account would have had 417.935 shares at the end of 2017. Those shares at the end of 2017 NAV of 30.52 would have been worth $12,775.38.

FactualFran
Posts: 668
Joined: Sat Feb 21, 2015 2:29 pm

Re: Total returns

Post by FactualFran » Fri Jul 13, 2018 1:48 pm

House Blend wrote:
Fri Jul 13, 2018 10:58 am
coffeecup333 wrote:
Fri Jul 13, 2018 10:46 am
But now looking at my vtsax, the difference between the benchmark of 21.19% and 21.17% of vtsax's total return in 2017, is a difference of .02%, not the ER of .04%. Do you know why could be? https://investor.vanguard.com/mutual-fu ... ve-returns
I would wager that most of that is due to securities lending. Which overall is a pretty good deal for investors, assuming the profits get plowed back into the fund.

See: https://www.bogleheads.org/wiki/Securities_lending
According to the 2017 annual report for VTSAX, the fund received a net interest of $109,770,000 from Securities Lending and had net assets at the end of 2017 of 662,577,171,000. The net interest from Securities Lending was 0.02% of the net assets.

Post Reply