Yield on Vanguard Int. Term Tax Exempt Fund

Have a question about your personal investments? No matter how simple or complex, you can ask it here.
Wagnerjb
Posts: 7015
Joined: Mon Feb 19, 2007 8:44 pm
Location: Houston, Texas

Yield on Vanguard Int. Term Tax Exempt Fund

Post by Wagnerjb » Wed Sep 23, 2015 8:51 pm

I am trying to wrap my head around the performance of the Vanguard Intermediate Term Tax Exempt bond fund. I own this (Admiral shares) in my taxable account.

I understand that the SEC yield is 1.89% and I fully understand the yield to maturity concept. The Fund Annual report shows the average coupon yield for the portfolio is around 4.5% - meaning to me that the fund purchased the typical bond at a premium. The payout yield has been around 3.0% on average for the past 12 months, and that is why I am wondering what this means.

If the bond market did not change, this would mean that I should expect to receive a 3% cash distribution yield but a 1.11% loss of principal (on average) each year. Right? We haven't seen exactly that loss of principal recently, but in theory that is what one would expect.

If that is what one should expect, isn't that the perfect situation for a taxable investor? You get a higher yield (3.0% distribution yield vs. 1.89% yield to maturity) and you get to tax loss harvest your shares after the inevitable decline in value. (Yes, I am aware of the 6 month rule for TLH on Tax Exempt funds).

The typical investor in the 39% tax bracket who invests $100,000 in a TE fund with bonds at par and yielding 1.89% will receive $1,890 on after-tax basis. The same guy who invests the same amount in this Vanguard fund will also get a TLH benefit of $1,110 - worth $400, which increases his after-tax return by 21%.

Am I missing anything in this thought process?

Thanks.
Andy

User avatar
baw703916
Posts: 6631
Joined: Sun Apr 01, 2007 1:10 pm
Location: Seattle

Re: Yield on Vanguard Int. Term Tax Exempt Fund

Post by baw703916 » Wed Sep 23, 2015 9:54 pm

Well almost, except one wrinkle I that an intermediate fund probably isn't going to hold bonds all the way to maturity. What I think Vanguard tends to do is sell bonds from one duration fund to another at the maturity date approaches. Assuming a non-inverted yield curve, this usually produces a capital gain of the fund buying longer-dated bonds and selling shorter-dated ones. So if IT TE buys bonds from LT TE and then later sells them to ST TE, and interest rates remain unchanged with a non-inverted yield curve, it's likely to have a capital gain.
Most of my posts assume no behavioral errors.

User avatar
abuss368
Posts: 11641
Joined: Mon Aug 03, 2009 2:33 pm
Location: Where the water is warm, the drinks are cold, and I don't know the names of the players!

Re: Yield on Vanguard Int. Term Tax Exempt Fund

Post by abuss368 » Wed Sep 23, 2015 11:55 pm

Wagnerjb wrote:I am trying to wrap my head around the performance of the Vanguard Intermediate Term Tax Exempt bond fund. I own this (Admiral shares) in my taxable account.

I understand that the SEC yield is 1.89% and I fully understand the yield to maturity concept. The Fund Annual report shows the average coupon yield for the portfolio is around 4.5% - meaning to me that the fund purchased the typical bond at a premium. The payout yield has been around 3.0% on average for the past 12 months, and that is why I am wondering what this means.

If the bond market did not change, this would mean that I should expect to receive a 3% cash distribution yield but a 1.11% loss of principal (on average) each year. Right? We haven't seen exactly that loss of principal recently, but in theory that is what one would expect.

If that is what one should expect, isn't that the perfect situation for a taxable investor? You get a higher yield (3.0% distribution yield vs. 1.89% yield to maturity) and you get to tax loss harvest your shares after the inevitable decline in value. (Yes, I am aware of the 6 month rule for TLH on Tax Exempt funds).

The typical investor in the 39% tax bracket who invests $100,000 in a TE fund with bonds at par and yielding 1.89% will receive $1,890 on after-tax basis. The same guy who invests the same amount in this Vanguard fund will also get a TLH benefit of $1,110 - worth $400, which increases his after-tax return by 21%.

Am I missing anything in this thought process?

Thanks.


What is the 6 month rule?
John C. Bogle: "You simply do not need to put your money into 8 different mutual funds!" | | Disclosure: Three Fund Portfolio + U.S. & International REITs

User avatar
ogd
Posts: 4807
Joined: Thu Jun 14, 2012 11:43 pm

Re: Yield on Vanguard Int. Term Tax Exempt Fund

Post by ogd » Thu Sep 24, 2015 12:46 am

Wagnerjb wrote:If the bond market did not change, this would mean that I should expect to receive a 3% cash distribution yield but a 1.11% loss of principal (on average) each year. Right? We haven't seen exactly that loss of principal recently, but in theory that is what one would expect.

No. If the bond market truly did not change, you will get much more than SEC yield, perhaps 3.0% with no loss of capital whatsoever. The reason is, the SEC yield is a YtM, implying bonds are allowed to steadily decline in duration and mature. But the fund does not do that -- instead, it replenishes duration. If the duration difference between, say, 5 and 4 years happens to be well-rewarded, like it is presently, then you will reap those rewards. This is the "yield curve riding effect", explored at length in years past when bonds were the worry, not the thing celebrated: viewtopic.php?t=132601

The gotcha is the bond market is in fact expected to change. It would be weird and unexpected for the market to keep rewarding the first year or two of the bond that much more than the rest, forever. My personal expectation is a flattening in such a way as to make the returns from various maturities indifferent and fund returns roughly in line with SEC yield (these two descriptions are equivalent). If you don't buy that theory, you probably expect something inbetween the present state and a flat state.

Wagnerjb wrote:If that is what one should expect, isn't that the perfect situation for a taxable investor? You get a higher yield (3.0% distribution yield vs. 1.89% yield to maturity) and you get to tax loss harvest your shares after the inevitable decline in value. (Yes, I am aware of the 6 month rule for TLH on Tax Exempt funds).

The other gotcha is that it's well clear to all taxable investors that a high coupon tax exempt bond is more valuable than a low coupon bond of the same yield. So you expect the high coupon bonds to make less than low coupon bonds, just enough to make the average taxable investor indifferent to either (think about it: wouldn't you value a high coupon bond more richly because of taxation? you sort of just said so). The real reason why funds have high coupon bonds is that a few years ago munis were much higher yielding and that muni bond turnover is expensive. This will not last, at present rates.

User avatar
Toons
Posts: 12013
Joined: Fri Nov 21, 2008 10:20 am
Location: Hills of Tennessee

Re: Yield on Vanguard Int. Term Tax Exempt Fund

Post by Toons » Thu Sep 24, 2015 5:31 am

ogd wrote:
Wagnerjb wrote:If the bond market did not change, this would mean that I should expect to receive a 3% cash distribution yield but a 1.11% loss of principal (on average) each year. Right? We haven't seen exactly that loss of principal recently, but in theory that is what one would expect.

No. If the bond market truly did not change, you will get much more than SEC yield, perhaps 3.0% with no loss of capital whatsoever. The reason is, the SEC yield is a YtM, implying bonds are allowed to steadily decline in duration and mature. But the fund does not do that -- instead, it replenishes duration. If the duration difference between, say, 5 and 4 years happens to be well-rewarded, like it is presently, then you will reap those rewards. This is the "yield curve riding effect", explored at length in years past when bonds were the worry, not the thing celebrated: viewtopic.php?t=132601

The gotcha is the bond market is in fact expected to change. It would be weird and unexpected for the market to keep rewarding the first year or two of the bond that much more than the rest, forever. My personal expectation is a flattening in such a way as to make the returns from various maturities indifferent and fund returns roughly in line with SEC yield (these two descriptions are equivalent). If you don't buy that theory, you probably expect something inbetween the present state and a flat state.

Wagnerjb wrote:If that is what one should expect, isn't that the perfect situation for a taxable investor? You get a higher yield (3.0% distribution yield vs. 1.89% yield to maturity) and you get to tax loss harvest your shares after the inevitable decline in value. (Yes, I am aware of the 6 month rule for TLH on Tax Exempt funds).

The other gotcha is that it's well clear to all taxable investors that a high coupon tax exempt bond is more valuable than a low coupon bond of the same yield. So you expect the high coupon bonds to make less than low coupon bonds, just enough to make the average taxable investor indifferent to either (think about it: wouldn't you value a high coupon bond more richly because of taxation? you sort of just said so). The real reason why funds have high coupon bonds is that a few years ago munis were much higher yielding and that muni bond turnover is expensive. This will not last, at present rates.


Thanks for the insight :happy
"One does not accumulate but eliminate. It is not daily increase but daily decrease. The height of cultivation always runs to simplicity" –Bruce Lee

User avatar
neurosphere
Posts: 2599
Joined: Sun Jan 17, 2010 1:55 pm
Location: NYC

Re: Yield on Vanguard Int. Term Tax Exempt Fund

Post by neurosphere » Thu Sep 24, 2015 6:58 am

abuss368 wrote:What is the 6 month rule?


Here is how someone explains it somewhere else on the internet (sentence stolen without attribution :D ):

If an investor buys mutual fund shares in a tax-exempt fund, receives an exempt-interest dividend, and then sells the shares at a loss within six months after the purchase, the loss is disallowed to the extent of the exempt-interest dividend.
-- Real name: Sotirios Keros. If you have to ask "Is a Target Retirement fund right for me?", the answer is yes.

User avatar
abuss368
Posts: 11641
Joined: Mon Aug 03, 2009 2:33 pm
Location: Where the water is warm, the drinks are cold, and I don't know the names of the players!

Re: Yield on Vanguard Int. Term Tax Exempt Fund

Post by abuss368 » Thu Sep 24, 2015 10:33 am

neurosphere wrote:
abuss368 wrote:What is the 6 month rule?


Here is how someone explains it somewhere else on the internet (sentence stolen without attribution :D ):

If an investor buys mutual fund shares in a tax-exempt fund, receives an exempt-interest dividend, and then sells the shares at a loss within six months after the purchase, the loss is disallowed to the extent of the exempt-interest dividend.


Hi neurosphere,

Thank you for that response. Would you happen to be aware of the Internal Revenue Code Section or Regulation to support that?

Best.
John C. Bogle: "You simply do not need to put your money into 8 different mutual funds!" | | Disclosure: Three Fund Portfolio + U.S. & International REITs

User avatar
neurosphere
Posts: 2599
Joined: Sun Jan 17, 2010 1:55 pm
Location: NYC

Re: Yield on Vanguard Int. Term Tax Exempt Fund

Post by neurosphere » Thu Sep 24, 2015 3:17 pm

abuss368 wrote:
Hi neurosphere,

Thank you for that response. Would you happen to be aware of the Internal Revenue Code Section or Regulation to support that?

Best.


26 U.S. Code § 852

(4) Loss on sale or exchange of stock held 6 months or less
(A) Loss attributable to capital gain dividend
If—
(i)subparagraph (B) or (D) of paragraph (3) provides that any amount with respect to any share is to be treated as long-term capital gain, and
(ii)such share is held by the taxpayer for 6 months or less,
then any loss (to the extent not disallowed under subparagraph (B)) on the sale or exchange of such share shall, to the extent of the amount described in clause (i), be treated as a long-term capital loss.
(B) Loss attributable to exempt-interest dividend
If—
(i)a shareholder of a regulated investment company receives an exempt-interest dividend with respect to any share, and
(ii)such share is held by the taxpayer for 6 months or less,
then any loss on the sale or exchange of such share shall, to the extent of the amount of such exempt-interest dividend, be disallowed.


But then see further below:

(E) Exception to holding period requirement for certain regularly declared exempt-interest dividends
(i)Daily dividend companies Except as otherwise provided by regulations, subparagraph (B) shall not apply with respect to a regular dividend paid by a regulated investment company which declares exempt-interest dividends on a daily basis in an amount equal to at least 90 percent of its net tax-exempt interest and distributes such dividends on a monthly or more frequent basis.


I'm actually not sure of this exception applies to the tax-free bond funds such as those at Vanguard which pay monthly dividends. I just don't sell tax-exempt shares which have less than a 6 month holding period so I don't have to worry about it, and I've never had a tax-prep client to whom this situation was relevant so I've never had to answer the question. :D I'm sure the answer is on the internets somewhere, perhaps even right here on Bogleheads. :D :D

Neurosphere
-- Real name: Sotirios Keros. If you have to ask "Is a Target Retirement fund right for me?", the answer is yes.

User avatar
Kevin M
Posts: 8866
Joined: Mon Jun 29, 2009 3:24 pm
Contact:

Re: Yield on Vanguard Int. Term Tax Exempt Fund

Post by Kevin M » Thu Sep 24, 2015 4:02 pm

neurosphere wrote:
(E) Exception to holding period requirement for certain regularly declared exempt-interest dividends
(i)Daily dividend companies Except as otherwise provided by regulations, subparagraph (B) shall not apply with respect to a regular dividend paid by a regulated investment company which declares exempt-interest dividends on a daily basis in an amount equal to at least 90 percent of its net tax-exempt interest and distributes such dividends on a monthly or more frequent basis.


I'm actually not sure of this exception applies to the tax-free bond funds such as those at Vanguard which pay monthly dividends. I just don't sell tax-exempt shares which have less than a 6 month holding period so I don't have to worry about it, and I've never had a tax-prep client to whom this situation was relevant so I've never had to answer the question. :D I'm sure the answer is on the internets somewhere, perhaps even right here on Bogleheads. :D :D

Wow, I've never noticed this exception before, and I think it does apply to Vanguard TE bond funds, since they do declare the dividend on a daily basis. You can see the accrued dividend in your account by looking at Balances by date. The six month rule always seemed nonsensical for Vanguard TE funds for this reason, but the exemption makes it rational.

Kevin
||.......|| Suggested format for Asking Portfolio Questions (edit original post)

User avatar
Kevin M
Posts: 8866
Joined: Mon Jun 29, 2009 3:24 pm
Contact:

Re: Yield on Vanguard Int. Term Tax Exempt Fund

Post by Kevin M » Thu Sep 24, 2015 4:10 pm

Wagnerjb wrote:If the bond market did not change, this would mean that I should expect to receive a 3% cash distribution yield but a 1.11% loss of principal (on average) each year. Right? We haven't seen exactly that loss of principal recently, but in theory that is what one would expect.

I've posted about this apparent anomaly as long as four or five years ago. It is most evident in the shorter-term bond funds, for which you see the distribution yield remaining significantly above SEC yield with no decline in price for time periods exceeding the duration. Price gain due to bonds moving down the positive yield curve seems to be a reasonable explanation.

Kevin
||.......|| Suggested format for Asking Portfolio Questions (edit original post)

User avatar
neurosphere
Posts: 2599
Joined: Sun Jan 17, 2010 1:55 pm
Location: NYC

Re: Yield on Vanguard Int. Term Tax Exempt Fund

Post by neurosphere » Thu Sep 24, 2015 4:11 pm

Kevin M wrote:Wow, I've never noticed this exception before, and I think it does apply to Vanguard TE bond funds, since they do declare the dividend on a daily basis. You can see the accrued dividend in your account by looking at Balances by date. The six month rule always seemed nonsensical for Vanguard TE funds for this reason, but the exemption makes it rational.

Kevin


You know what's funny is that I HAVE noticed this exception, but assumed it must not apply for some reason to VG funds, or else someone would have brought it up, and wouldn't keep cautioning people about the 6-month rule. :D Or, someone has brought it up and I didn't know it or didn't realize what I was reading.

I wonder if anyone can confirm whether the VG funds which pay monthly tax-exempt interest are exempt from the 6-month tax loss rule regarding interest/dividends.

Anyone?

NS

User avatar
Kevin M
Posts: 8866
Joined: Mon Jun 29, 2009 3:24 pm
Contact:

Re: Yield on Vanguard Int. Term Tax Exempt Fund

Post by Kevin M » Thu Sep 24, 2015 4:33 pm

neurosphere wrote:I wonder if anyone can confirm whether the VG funds which pay monthly tax-exempt interest are exempt from the 6-month tax loss rule regarding interest/dividends.

While we're waiting, let's parse the regulation:

Except as otherwise provided by regulations, subparagraph (B) shall not apply with respect to a regular dividend paid by:

a regulated investment company
Check--that's what Vanguard is.
declares exempt-interest dividends on a daily basis
Check, assuming that what we see as accrued dividends in our accounts, which increase daily, counts as "declares".
in an amount equal to at least 90 percent of its net tax-exempt interest
Check. You will see that the monthly distribution equals the accrued interest on the day before the distribution plus an amount equal to one day of interest. Example for int-term TE: accrued interest on 8/30/15 = 119.45 , distribution on 8/31/15 = 123.46.

and distributes such dividends on a monthly or more frequent basis.
Check. Dividends are distributed monthly.

Another way we can tell that Vanguard is declaring dividends daily is that if you sell shares of a fund mid-month, you will receive the accrued dividend in addition to the capital amount that equals price times number of shares. In other words, the accrued dividend is not included in the share price, which I think is what the 60 day rule is about (taking a tax-exempt dividend and then selling at a loss because of drop in share price due to dividend distribution, which does not happen with Vanguard funds). This is why the exemption makes total sense, and all is now right with the world.

Kevin
||.......|| Suggested format for Asking Portfolio Questions (edit original post)

User avatar
abuss368
Posts: 11641
Joined: Mon Aug 03, 2009 2:33 pm
Location: Where the water is warm, the drinks are cold, and I don't know the names of the players!

Re: Yield on Vanguard Int. Term Tax Exempt Fund

Post by abuss368 » Thu Sep 24, 2015 4:40 pm

neurosphere wrote:
abuss368 wrote:
Hi neurosphere,

Thank you for that response. Would you happen to be aware of the Internal Revenue Code Section or Regulation to support that?

Best.


26 U.S. Code § 852

(4) Loss on sale or exchange of stock held 6 months or less
(A) Loss attributable to capital gain dividend
If—
(i)subparagraph (B) or (D) of paragraph (3) provides that any amount with respect to any share is to be treated as long-term capital gain, and
(ii)such share is held by the taxpayer for 6 months or less,
then any loss (to the extent not disallowed under subparagraph (B)) on the sale or exchange of such share shall, to the extent of the amount described in clause (i), be treated as a long-term capital loss.
(B) Loss attributable to exempt-interest dividend
If—
(i)a shareholder of a regulated investment company receives an exempt-interest dividend with respect to any share, and
(ii)such share is held by the taxpayer for 6 months or less,
then any loss on the sale or exchange of such share shall, to the extent of the amount of such exempt-interest dividend, be disallowed.


But then see further below:

(E) Exception to holding period requirement for certain regularly declared exempt-interest dividends
(i)Daily dividend companies Except as otherwise provided by regulations, subparagraph (B) shall not apply with respect to a regular dividend paid by a regulated investment company which declares exempt-interest dividends on a daily basis in an amount equal to at least 90 percent of its net tax-exempt interest and distributes such dividends on a monthly or more frequent basis.


I'm actually not sure of this exception applies to the tax-free bond funds such as those at Vanguard which pay monthly dividends. I just don't sell tax-exempt shares which have less than a 6 month holding period so I don't have to worry about it, and I've never had a tax-prep client to whom this situation was relevant so I've never had to answer the question. :D I'm sure the answer is on the internets somewhere, perhaps even right here on Bogleheads. :D :D

Neurosphere


Thanks Neurosphere!
John C. Bogle: "You simply do not need to put your money into 8 different mutual funds!" | | Disclosure: Three Fund Portfolio + U.S. & International REITs

User avatar
neurosphere
Posts: 2599
Joined: Sun Jan 17, 2010 1:55 pm
Location: NYC

Re: Yield on Vanguard Int. Term Tax Exempt Fund

Post by neurosphere » Thu Sep 24, 2015 4:41 pm

Kevin M wrote:
a regulated investment company
Check--that's what Vanguard is.


Kinda but not quite. In this case the regulated investment company refers to the mutual fund itself, in which one owns shares. The definition of regulated investment company comes as a result of the Investment Company Act of 1940. So the "company" is the mutual fund.

https://www.law.cornell.edu/uscode/text/26/851

I agree with the rest of your statements. It seems as if this would count as an exemption. But I'm not yet willing to publicly go on record. The internet never forgets, and I've been burned by assuming somewhat was reasonable only to find out some quirk which invalidated my conclusion. 8-)
-- Real name: Sotirios Keros. If you have to ask "Is a Target Retirement fund right for me?", the answer is yes.

User avatar
abuss368
Posts: 11641
Joined: Mon Aug 03, 2009 2:33 pm
Location: Where the water is warm, the drinks are cold, and I don't know the names of the players!

Re: Yield on Vanguard Int. Term Tax Exempt Fund

Post by abuss368 » Thu Sep 24, 2015 4:45 pm

neurosphere wrote:
Kevin M wrote:
a regulated investment company
Check--that's what Vanguard is.


Kinda but not quite. In this case the regulated investment company refers to the mutual fund itself, in which one owns shares. The definition of regulated investment company comes as a result of the Investment Company Act of 1940. So the "company" is the mutual fund.

https://www.law.cornell.edu/uscode/text/26/851

I agree with the rest of your statements. It seems as if this would count as an exemption. But I'm not yet willing to publicly go on record. The internet never forgets, and I've been burned by assuming somewhat was reasonable only to find out some quirk which invalidated my conclusion. 8-)


My understanding is the mutual fund itself is the Regulated Investment Company.

Is this correct?

Best.
John C. Bogle: "You simply do not need to put your money into 8 different mutual funds!" | | Disclosure: Three Fund Portfolio + U.S. & International REITs

User avatar
neurosphere
Posts: 2599
Joined: Sun Jan 17, 2010 1:55 pm
Location: NYC

Re: Yield on Vanguard Int. Term Tax Exempt Fund

Post by neurosphere » Thu Sep 24, 2015 4:50 pm

abuss368 wrote:
neurosphere wrote:
Kevin M wrote:
a regulated investment company
Check--that's what Vanguard is.


Kinda but not quite. In this case the regulated investment company refers to the mutual fund itself, in which one owns shares. The definition of regulated investment company comes as a result of the Investment Company Act of 1940. So the "company" is the mutual fund.

https://www.law.cornell.edu/uscode/text/26/851

I agree with the rest of your statements. It seems as if this would count as an exemption. But I'm not yet willing to publicly go on record. The internet never forgets, and I've been burned by assuming somewhat was reasonable only to find out some quirk which invalidated my conclusion. 8-)


My understanding is the mutual fund itself is the Regulated Investment Company.

Is this correct?

Best.


See the part I underlined and bolded and put in red in the quote above from my initial reply. :wink:

Yes, the mutual fund is the regulated investment company in this case, not "Vanguard".

Wagnerjb
Posts: 7015
Joined: Mon Feb 19, 2007 8:44 pm
Location: Houston, Texas

Re: Yield on Vanguard Int. Term Tax Exempt Fund

Post by Wagnerjb » Thu Sep 24, 2015 4:56 pm

neurosphere wrote:
Kevin M wrote:Wow, I've never noticed this exception before, and I think it does apply to Vanguard TE bond funds, since they do declare the dividend on a daily basis. You can see the accrued dividend in your account by looking at Balances by date. The six month rule always seemed nonsensical for Vanguard TE funds for this reason, but the exemption makes it rational.

Kevin


You know what's funny is that I HAVE noticed this exception, but assumed it must not apply for some reason to VG funds, or else someone would have brought it up, and wouldn't keep cautioning people about the 6-month rule. :D Or, someone has brought it up and I didn't know it or didn't realize what I was reading.

I wonder if anyone can confirm whether the VG funds which pay monthly tax-exempt interest are exempt from the 6-month tax loss rule regarding interest/dividends.

Anyone?

NS


I had part of a realized loss disallowed by Vanguard ( in the annual tax report) about 8-10 years ago and I posted about it
Andy

User avatar
abuss368
Posts: 11641
Joined: Mon Aug 03, 2009 2:33 pm
Location: Where the water is warm, the drinks are cold, and I don't know the names of the players!

Re: Yield on Vanguard Int. Term Tax Exempt Fund

Post by abuss368 » Thu Sep 24, 2015 4:56 pm

neurosphere wrote:
abuss368 wrote:
neurosphere wrote:
Kevin M wrote:
a regulated investment company
Check--that's what Vanguard is.


Kinda but not quite. In this case the regulated investment company refers to the mutual fund itself, in which one owns shares. The definition of regulated investment company comes as a result of the Investment Company Act of 1940. So the "company" is the mutual fund.

https://www.law.cornell.edu/uscode/text/26/851

I agree with the rest of your statements. It seems as if this would count as an exemption. But I'm not yet willing to publicly go on record. The internet never forgets, and I've been burned by assuming somewhat was reasonable only to find out some quirk which invalidated my conclusion. 8-)


My understanding is the mutual fund itself is the Regulated Investment Company.

Is this correct?

Best.


See the part I underlined and bolded and put in red in the quote above from my initial reply. :wink:

Yes, the mutual fund is the regulated investment company in this case, not "Vanguard".


Thanks!
John C. Bogle: "You simply do not need to put your money into 8 different mutual funds!" | | Disclosure: Three Fund Portfolio + U.S. & International REITs

User avatar
Kevin M
Posts: 8866
Joined: Mon Jun 29, 2009 3:24 pm
Contact:

Re: Yield on Vanguard Int. Term Tax Exempt Fund

Post by Kevin M » Thu Sep 24, 2015 6:08 pm

neurosphere wrote:In this case the regulated investment company refers to the mutual fund itself, in which one owns shares.

OK, but is the distinction relevant to the point being made? Just change my statement to "Check. That's what the Vanguard tax-exempt bond fund is".

Kevin
||.......|| Suggested format for Asking Portfolio Questions (edit original post)

User avatar
Kevin M
Posts: 8866
Joined: Mon Jun 29, 2009 3:24 pm
Contact:

Re: Yield on Vanguard Int. Term Tax Exempt Fund

Post by Kevin M » Thu Sep 24, 2015 6:28 pm

Wagnerjb wrote:I had part of a realized loss disallowed by Vanguard ( in the annual tax report) about 8-10 years ago and I posted about it

But Vanguard's tax report is not definitive, right? They also would have shown your gain/loss based on average cost, and you may have been using specific ID, keeping track of your own basis, in which case the basis and gain/loss shown on the tax report would be inaccurate.

I don't believe Vanguard gives tax advice, and they have been known to make mistakes. I wouldn't be surprised if Vanguard took a conservative approach on this.

In all the discussions I've seen here about the six-month rule, this is the first time I've seen anyone bring up this exception. Our own Wiki article on tax-exempt bonds does not mention the exception: Municipal bonds - Bogleheads. Maybe Vanguard doesn't know about it, or is too cautious to suggest that people take advantage of it?

Is this exception mentioned in the IRS publications or tax form instructions?

Kevin
||.......|| Suggested format for Asking Portfolio Questions (edit original post)

User avatar
neurosphere
Posts: 2599
Joined: Sun Jan 17, 2010 1:55 pm
Location: NYC

Re: Yield on Vanguard Int. Term Tax Exempt Fund

Post by neurosphere » Thu Sep 24, 2015 7:53 pm

Kevin M wrote:
neurosphere wrote:In this case the regulated investment company refers to the mutual fund itself, in which one owns shares.

OK, but is the distinction relevant to the point being made? Just change my statement to "Check. That's what the Vanguard tax-exempt bond fund is".

Kevin


Yes, agreed!

I was just focusing on a minor semantic detail (and sometimes a source of confusion) that in modern times, many investors would never consider an "investment company" as referring to a mutual fund. But the code very specifically is referring to the mutual fund, and not to Vanguard as most people consider it. At least for me, the term "company" refers to a "firm" or an "business". That's all.

User avatar
grabiner
Advisory Board
Posts: 20838
Joined: Tue Feb 20, 2007 11:58 pm
Location: Columbia, MD

Re: Yield on Vanguard Int. Term Tax Exempt Fund

Post by grabiner » Thu Sep 24, 2015 8:32 pm

Kevin M wrote:
Wagnerjb wrote:I had part of a realized loss disallowed by Vanguard ( in the annual tax report) about 8-10 years ago and I posted about it

But Vanguard's tax report is not definitive, right? They also would have shown your gain/loss based on average cost, and you may have been using specific ID, keeping track of your own basis, in which case the basis and gain/loss shown on the tax report would be inaccurate.


And I know Vanguard missed when I ran into the other version of the six-month rule, also with non-covered shares. I bought an ETF in November 2011, received a distribution of long-term gains in December 2011, and sold the ETF in March 2012 for a capital loss. According to the six-month rule, an amount equal to the 2011 capital gain distribution should be converted from a short-term loss to a long-term loss. (Here, the purpose of the rule is to prevent buying a fund just before a distribution, in order to get a long-term gain and a short-term loss, then use the short-term loss to offset a short-term gain elsewhere.)
David Grabiner

User avatar
Electron
Posts: 1656
Joined: Sat Mar 10, 2007 8:46 pm

Re: Yield on Vanguard Int. Term Tax Exempt Fund

Post by Electron » Fri Sep 25, 2015 2:15 pm

Wagnerjb wrote:If the bond market did not change, this would mean that I should expect to receive a 3% cash distribution yield but a 1.11% loss of principal (on average) each year. Right? We haven't seen exactly that loss of principal recently, but in theory that is what one would expect.

Another factor to consider is amortization of bond premium. I believe mutual funds follow the same rules as individual investors.

http://www.irs.gov/publications/p550/ch03.html

"If the bond yields tax-exempt interest, you must amortize the premium. This amortized amount is not deductible in determining taxable income. However, each year you must reduce your basis in the bond (and tax-exempt interest otherwise reportable on Form 1040, line 8b) by the amortization for the year."

It looks as though unrealized losses associated with buying bonds at a premium will decline over time as the bonds approach maturity.
Electron

User avatar
grabiner
Advisory Board
Posts: 20838
Joined: Tue Feb 20, 2007 11:58 pm
Location: Columbia, MD

Re: Yield on Vanguard Int. Term Tax Exempt Fund

Post by grabiner » Fri Sep 25, 2015 6:51 pm

Electron wrote:
Wagnerjb wrote:If the bond market did not change, this would mean that I should expect to receive a 3% cash distribution yield but a 1.11% loss of principal (on average) each year. Right? We haven't seen exactly that loss of principal recently, but in theory that is what one would expect.

Another factor to consider is amortization of bond premium. I believe mutual funds follow the same rules as individual investors.

http://www.irs.gov/publications/p550/ch03.html

"If the bond yields tax-exempt interest, you must amortize the premium. This amortized amount is not deductible in determining taxable income. However, each year you must reduce your basis in the bond (and tax-exempt interest otherwise reportable on Form 1040, line 8b) by the amortization for the year."

It looks as though unrealized losses associated with buying bonds at a premium will decline over time as the bonds approach maturity.


However, if a fund bought a bond at par and the bond is now trading at a premium, the amortization will not apply; the bond will pay out its higher coupon rate and the fund will have capital losses as the bond value declines back to par at maturity. (Most funds will sell such bonds before maturity, producing a capital gain unless there are capital losses to offset the gain.)
David Grabiner

User avatar
Kevin M
Posts: 8866
Joined: Mon Jun 29, 2009 3:24 pm
Contact:

Re: Yield on Vanguard Int. Term Tax Exempt Fund

Post by Kevin M » Fri Sep 25, 2015 7:33 pm

We've gotten off on a few tangents, and there have been some critiques of some of the details, but I haven't seen a wholehearted rejection of Andy's main point, which is that the potential TLH value offsets, at least to some extent, the expected capital loss as distribution yield approaches SEC yield. It's an interesting observation, and I hadn't thought about it in these terms before.

My only critique of the main idea is that the TLH benefit is not as quantifiable as Andy indicates. I believe his assumption is that he gets the benefit at the 39% marginal tax rate due to the $3,000/year deduction against ordinary income. For those of us who still have large losses carried forward from 2008/2009, we do not necessarily realize the benefit this way, since we may have more than enough losses carried forward to last us the rest of our lives at $3,000/year. In this case, the benefit is less tangible, and harder to quantify.

Kevin
||.......|| Suggested format for Asking Portfolio Questions (edit original post)

User avatar
Electron
Posts: 1656
Joined: Sat Mar 10, 2007 8:46 pm

Re: Yield on Vanguard Int. Term Tax Exempt Fund

Post by Electron » Mon Sep 28, 2015 4:44 pm

The typical investor in the 39% tax bracket who invests $100,000 in a TE fund with bonds at par and yielding 1.89% will receive $1,890 on after-tax basis. The same guy who invests the same amount in this Vanguard fund will also get a TLH benefit of $1,110 - worth $400, which increases his after-tax return by 21%.

I'm not sure that the portfolio of bonds purchased at par would necessarily have an SEC Yield of 1.89%. To have a fair comparison, the two portfolios should offer the same level of risk. With different bond characteristics and call features, the portfolio might be constructed in a somewhat different manner.

In addition, the bond market is reasonably efficient. If two similar portfolios have different expected total returns, the risk levels would very likely be different.

Also note that the SEC Yield has limitations and was really intended for comparing the yields of two or more bond funds. If you look at the SEC formula, interest income is determined based on a Yield-to-Maturity for each bond that is recalculated typically once per month. As a result, the SEC Yield can make step changes with each recalculation.

Here is something else of interest to think about. The Vanguard fund profile for VWIUX shows a Yield-to-Maturity of 2.0% on 8-31-15. If you now go to the Vanguard Price History Search tool, you will find that the SEC Yield was listed as 1.86% on the same date. It looks like the SEC Yield may deviate from the true Yield-to-Maturity of the portfolio.
Electron

User avatar
Kevin M
Posts: 8866
Joined: Mon Jun 29, 2009 3:24 pm
Contact:

Re: Yield on Vanguard Int. Term Tax Exempt Fund

Post by Kevin M » Mon Sep 28, 2015 8:43 pm

Electron wrote:It looks like the SEC Yield may deviate from the true Yield-to-Maturity of the portfolio.

Yes, these numbers usually are somewhat different (which we can only observe on the dates for which average YTM is published). My understanding is that the two main factors for the discrepancy are the 30-day averaging and the subtraction of expenses for SEC yield. Average YTM is a snapshot on the day of record, and fund expenses are not subtracted.

Anyway, I still thinks these are details that are not really getting at Andy's main point.

Kevin
||.......|| Suggested format for Asking Portfolio Questions (edit original post)

User avatar
Electron
Posts: 1656
Joined: Sat Mar 10, 2007 8:46 pm

Re: Yield on Vanguard Int. Term Tax Exempt Fund

Post by Electron » Wed Sep 30, 2015 2:20 pm

We've gotten off on a few tangents, and there have been some critiques of some of the details, but I haven't seen a wholehearted rejection of Andy's main point, which is that the potential TLH value offsets, at least to some extent, the expected capital loss as distribution yield approaches SEC yield.

The concept appears to be valid for any bond fund that invests in premium bonds. Here is a quote from the book Bogle on Mutual Funds.

"But the higher income returns tend to be offset, dollar for dollar, by lower capital returns."

The SEC Yield goes back to 1988. Many bond funds prior to that date would load up on premium bonds and advertise high yields. Unsuspecting investors might not be aware that some loss of capital would result over time. Some bond funds also sold call options.

https://books.google.com/books?id=JwD0B ... ds&f=false

However, I think we have to be careful in comparing VWIUX to a similar fund investing in par bonds. We don't know if the distribution yield will be maintained. The two funds would not have the exact same yield-to-maturity, duration, and average maturity. We don't know how many premium bonds would be called.

If I understand the concept, one would buy VWIUX and hold for tax exempt income. If a decline in NAV results in a useful capital loss the shares would be sold. The loss would in theory be greater than the loss from a similar fund holding fewer premium bonds. The investor's tax bracket would now come into play. The loss could offset capital gains or ordinary income depending on the specific case. If tax loss harvesting, the investor would purchase shares in another tax exempt fund at current NAV. A rebound in municipal bond prices would then result in unrealized capital gains. The investor would want to be aware of any six month rules related to capital losses or capital gains distributions.
Electron

User avatar
Kevin M
Posts: 8866
Joined: Mon Jun 29, 2009 3:24 pm
Contact:

Re: Yield on Vanguard Int. Term Tax Exempt Fund

Post by Kevin M » Wed Sep 30, 2015 3:42 pm

Electron wrote:However, I think we have to be careful in comparing VWIUX to a similar fund investing in par bonds. We don't know if the distribution yield will be maintained.

Correct--as a matter of fact, I think we're assuming that it won't be maintained, but would tend to converge to SEC yield, all other things held constant. Andy actually assumes that the higher distribution yield will be associated with a decline in NAV due to declining bond premiums, but that the decline in NAV will be more than offset by a TLH premium (which I think is debatable).

Kevin
||.......|| Suggested format for Asking Portfolio Questions (edit original post)

User avatar
Electron
Posts: 1656
Joined: Sat Mar 10, 2007 8:46 pm

Re: Yield on Vanguard Int. Term Tax Exempt Fund

Post by Electron » Wed Sep 30, 2015 4:32 pm

Andy actually assumes that the higher distribution yield will be associated with a decline in NAV due to declining bond premiums, but that the decline in NAV will be more than offset by a TLH premium (which I think is debatable).

The question might be answered by selecting an appropriate premium bond and a par bond. One could then model a one year period where rates remain unchanged and the bonds are sold. Normally the premium bond would be amortized but that could be skipped for this analysis.
Electron

User avatar
Electron
Posts: 1656
Joined: Sat Mar 10, 2007 8:46 pm

Re: Yield on Vanguard Int. Term Tax Exempt Fund

Post by Electron » Sat Oct 03, 2015 1:17 pm

Modeling two bonds in Excel might provide some answers. Here is some data using the Yield and Price functions in Excel.

Premium Bond: $110.00, 5.0% coupon, 10 year maturity, yield is 3.789%, pays $4.545 per year on $100 investment

Par Bond: $100.00, 3.789% coupon, 10 year maturity, yield is 3.789%, pays $3.789 per year on $100 investment

One issue is that the two bonds would probably not trade at the same yield in the actual market. Let's look at the case where rates remain unchanged and the bonds are sold after one year. The holding period can be adjusted for a short term or long term loss.

The Price of the Premium Bond is now $109.162.

Income Return = $4.545, Capital Loss = $0.838, Total Return = $3.707 + Tax Benefit from Capital Loss

Par Bond - Price remains $100.00 - Total Return = $3.789

After one year the Premium bond has a lower Total Return not counting any benefit of the tax loss. The benefit of the capital loss will depend on the investor's tax bracket and whether the loss offsets capital gains or ordinary income.

These results would only apply if bond premium is not amortized. If a tax exempt bond fund amortizes premium for all bonds purchased at a premium, the situation is quite different. The cost basis for each premium bond would decline and reach par at maturity.

The bonds selected in this example may not be the best choices, but hopefully the overall concept is valid.
Electron

User avatar
Kevin M
Posts: 8866
Joined: Mon Jun 29, 2009 3:24 pm
Contact:

Re: Yield on Vanguard Int. Term Tax Exempt Fund

Post by Kevin M » Sat Oct 03, 2015 5:37 pm

Did you use a yield curve assumption in your calculations (looks like flat yield curve)? You want to use YTM of a 9-year bond for pricing a year later, assuming a static yield curve (a risky assumption, but you've got to start somewhere). Interpolating between 7-year and 10-year rates for current yield curve, there's about 10 basis points per year of yield change, so I would use about 2% YTM for 10-year and 1.9% YTM for 9-year, assuming Treasuries.

A fund will amortize the premium (the premium when the bond was purchased, not the premium at current market price), which will subtract from the dividend distributions. As we know from previous discussions, it's pretty hard to model a fund with a single bond, since there are many more moving parts in the fund. However, looking at single bonds is a good starting point to understand the contributions of price change and interest payments to return, but the big unknown always is the future yield curve.

Of course the yield curve for munis looks different than that for Treasuries. You could grab quotes for munis from Fidelity or Vanguard to approximate the muni yield curve if you want to model something more relevant to Andy's scenario.

Kevin
||.......|| Suggested format for Asking Portfolio Questions (edit original post)

stlutz
Posts: 4011
Joined: Fri Jan 02, 2009 1:08 am

Re: Yield on Vanguard Int. Term Tax Exempt Fund

Post by stlutz » Sat Oct 03, 2015 8:52 pm

I'd like to point everybody back up to ogd's post. There is no way that the fund managers can engineer a capital loss absent a rise in interest rates. That is because the distribution yield is based on what the YTM of the bonds were at the time that the fund purchased them. It makes no difference whether they were purchased at par or at a premium.

Think about holding a rolling 10 year bond ladder. For the purposes of illustration, assume the yield curve is upward sloping with a 5 year bond yielding 2% and a 10 year bond yielding 3%, and that the yield curve remains unchanged for 10 years. In this scenario, you are buying a new 10 year bond yielding 3% every year. Guess what the return of your portfolio will be every year--3%! Yet, at any given point, your average YTM will be about what the yield of the 5 year bond is--i.e. 2%. As you go down the yield curve, the bonds will appreciate relative to their cost basis until the last year or two at which point their prices will revert to par.

For this portfolio, your distribution yield would be 3% and your SEC Yield would be 2%. If interest rates and the yield curve never changed, the distribution yield would never converge to the SEC Yield.

Distribution and SEC yields tend to be very close when the portfolio has a high turnover ratio. If instead of this 10 year ladder you bought a 5 year bond every year, held for one year, then sold, your total return would be very close to the bond ladder scenario, but your distribution and SEC Yields would be very much the same. The ladder would get all of its return from income while the high-turnover portfolio would get its return from a combination of income and capital gains.

Vanguard's IT muni bond fund has a very low turnover rate--around 10-15% per year. This indicates that they manage the fund more like my ladder scenario than my high turnover scenario. This is good for investors because you want all of your return to come from tax exempt interest as opposed to capital gains, but there will not be any capital losses unless rates go up.

In contract, VG's IT Treasury fund has a 100% turnover rates. For that fund, the SEC Yield and the distribution yield tend to track much more closely.

User avatar
Kevin M
Posts: 8866
Joined: Mon Jun 29, 2009 3:24 pm
Contact:

Re: Yield on Vanguard Int. Term Tax Exempt Fund

Post by Kevin M » Sun Oct 04, 2015 12:06 am

stlutz wrote:There is no way that the fund managers can engineer a capital loss absent a rise in interest rates.

That would be a very sweet deal if true! I would much rather continue earning the higher distribution yields than get a questionable return from tax-loss harvesting.

Why would we want the fund managers to "engineer" a capital loss anyway? I don't think the main point of Andy's post was that capital losses are a good thing, but that there's a silver lining if they do occur.

I think there are some errors in some of the details you posted, but no time right now to delve into it. But quickly, I don't think the distribution yield would be 3% in your scenario--I think it would be less. Assuming bonds are bought at par, only newly purchased bonds and bonds very close to maturity would be priced close to par, and have a current yield of about 3%. All other maturities would be priced at a premium, so current yield would be less than 3%. For bonds purchased at par, distribution yield (as Vanguard defines it) is very much like current yield for an individual bond, so distribution yield will be the weighted average of the bonds' current yields.

No?

However, as many of us have agreed, SEC yield or average yield to maturity of a bond ladder understates expected return assuming a static yield curve, even if bonds are held to maturity (unless I made some mistakes when I modeled this).

Kevin
||.......|| Suggested format for Asking Portfolio Questions (edit original post)

stlutz
Posts: 4011
Joined: Fri Jan 02, 2009 1:08 am

Re: Yield on Vanguard Int. Term Tax Exempt Fund

Post by stlutz » Sun Oct 04, 2015 1:10 am

I don't think the distribution yield would be 3% in your scenario--I think it would be less. Assuming bonds are bought at par, only newly purchased bonds and bonds very close to maturity would be priced close to par, and have a current yield of about 3%. All other maturities would be priced at a premium, so current yield would be less than 3%


Agreed.

IPer
Posts: 1639
Joined: Sun Jul 28, 2013 8:51 pm

Re: Yield on Vanguard Int. Term Tax Exempt Fund

Post by IPer » Sun Oct 04, 2015 5:29 am

I started buying it early this year in varying chunks and now it is worth (market value) more than I paid for it (dividends reinvested) and
show a potential gain of 0.45%, seems to be doing what it is supposed to do, better to study the underlying holdings and their characteristics
when you would consider such a fund.
Read the Wiki !

User avatar
ogd
Posts: 4807
Joined: Thu Jun 14, 2012 11:43 pm

Re: Yield on Vanguard Int. Term Tax Exempt Fund

Post by ogd » Sun Oct 04, 2015 12:52 pm

Kevin M wrote:I think there are some errors in some of the details you posted, but no time right now to delve into it. But quickly, I don't think the distribution yield would be 3% in your scenario--I think it would be less. Assuming bonds are bought at par, only newly purchased bonds and bonds very close to maturity would be priced close to par, and have a current yield of about 3%. All other maturities would be priced at a premium, so current yield would be less than 3%. For bonds purchased at par, distribution yield (as Vanguard defines it) is very much like current yield for an individual bond, so distribution yield will be the weighted average of the bonds' current yields.

It's a little less, but the difference due to the price denominator is considerably smaller than the coupon-yield difference. E.g. for a 10% premium due to yield quickly dropping (on the yield curve or otherwise) from 3% to 2%, the current yield is still 2.7%. The vast majority of the coupon-yield difference comes from the future capital loss from premium to par, something that in this example we assume occurs late in the life of the bond.

Kevin M wrote:Why would we want the fund managers to "engineer" a capital loss anyway?

It would be better for tax payers to get higher tax-exempt interest PLUS capital loss rather than lower tax-exempt interest. The second point in my earlier post is that this is probably priced in to some degree so it might be only better for high-bracket tax payers.

User avatar
Electron
Posts: 1656
Joined: Sat Mar 10, 2007 8:46 pm

Re: Yield on Vanguard Int. Term Tax Exempt Fund

Post by Electron » Sun Oct 04, 2015 1:34 pm

Kevin M wrote:Did you use a yield curve assumption in your calculations (looks like flat yield curve)? You want to use YTM of a 9-year bond for pricing a year later, assuming a static yield curve (a risky assumption, but you've got to start somewhere).

Thanks for your comments. The example I posted did maintain the same yield from 10 years to 9 years. A better way to model the portfolio might be a rolling ladder with perhaps ten bonds. It would be necessary to replicate the current yield curve and come up with representative prices and yields. It wouldn't be that difficult in Excel to then see the changes after one year. I did notice the effect mentioned where a bond can actually trade at a higher premium as it rides down the upper portions of the yield curve.

I just looked at the portfolio listing for VWIUX and noticed something very interesting. A large fraction of the bonds have a 5% or higher coupon. In just about all cases the Market Value is above the Face Amount meaning that most bonds in the portfolio are trading at a premium. There are also a few zero coupon bonds trading well below face value. It could be that many new issues come out with a standardized coupon and trade at a premium at the initial offering.

Here is another interesting thought. Taxable bond funds in the 1980s advertised high yields prior to the use of the SEC yield and apparently did see erosion in NAV as premium bonds matured. One could speculate that taxable bond funds did not amortize premium in those days. If they did amortize, the payouts would have been less and the NAV would not have declined to the same extent. The situation is different today and I believe tax exempt bond funds must amortize premium.
Electron

User avatar
grabiner
Advisory Board
Posts: 20838
Joined: Tue Feb 20, 2007 11:58 pm
Location: Columbia, MD

Re: Yield on Vanguard Int. Term Tax Exempt Fund

Post by grabiner » Sun Oct 04, 2015 2:11 pm

Electron wrote:I just looked at the portfolio listing for VWIUX and noticed something very interesting. A large fraction of the bonds have a 5% or higher coupon. In just about all cases the Market Value is above the Face Amount meaning that most bonds in the portfolio are trading at a premium. There are also a few zero coupon bonds trading well below face value. It could be that many new issues come out with a standardized coupon and trade at a premium at the initial offering.


The reason for the high coupon is that many bonds which are intermediate-term now were originally issued as long-term bonds; they had higher yields at the time both because long-term bonds originally have higher yields and because interest rates were much higher when they were issued.

The face value of a zero-coupon bond is its value at maturity, so zeros always trade at a discount. For tax purposes, the discount is amortized and taxed as if it were interest; if you buy a $10,000 zero for $5000 and hold it to maturity, you will report $5000 of interest over the years and no capital gain.
David Grabiner

User avatar
Electron
Posts: 1656
Joined: Sat Mar 10, 2007 8:46 pm

Re: Yield on Vanguard Int. Term Tax Exempt Fund

Post by Electron » Sun Oct 04, 2015 2:48 pm

grabiner wrote:The reason for the high coupon is that many bonds which are intermediate-term now were originally issued as long-term bonds; they had higher yields at the time both because long-term bonds originally have higher yields and because interest rates were much higher when they were issued.

If you look through the portfolio holdings, the striking thing is that most of the bonds have a coupon of 5.000%. That same coupon is also predominant in the Limited Term and Long Term Tax Exempt portfolios. It does appear that standardized coupons have been used with bonds issued at a premium.

Here is an article discussing why Municipal Bonds are issued at a premium.

http://www8.gsb.columbia.edu/cbs-direct ... remium.pdf
Electron

User avatar
Kevin M
Posts: 8866
Joined: Mon Jun 29, 2009 3:24 pm
Contact:

Re: Yield on Vanguard Int. Term Tax Exempt Fund

Post by Kevin M » Sun Oct 04, 2015 3:31 pm

I think we're continuing to go down paths that may not be getting at the key point of the OP, but Andy's not complaining, and this is what people seem to want to do, so I'll continue with the crowd ...

It had been awhile since I looked at my model that convinced me that average YTM for a bond ladder (which is roughly SEC yield for a bond fund, but with a 30-day averaging and ER subtracted) underestimates the bond ladder return with a static yield curve assumption (again, even holding bonds to maturity), so I worked with it a bit more today to refresh my understanding. Here's a summary of the model, and some insights. Feel free to poke holes in it.

It's a 10-year ladder using YTMs for the current yield curve (most recent available from FRED, which is one day behind what's available on Treasury.gov, so as of today, 10/1), and assuming coupon rate for all bonds in the ladder is the coupon rate of the 10-year bond. This is basically stlutz's scenario in which you buy a new 10-year bond every year for 10 years, with no change in yield curve. I can set the coupon rate so the 10-year bond is purchased at par (100, with coupon rate = YTM), or with at a premium (price > 100, coupon rate > YTM), or of course at a discount.

One-year return is calculated for each bond, basically using formula (P1+C)/P0, where C is annual coupon payment, P0 is current bond price, and P1 is price of the bond one year later. Price is calculated using the PRICE function, based on the YTM from the yield curve for the given maturity and the fixed coupon rate. I can provide an example calculation if anyone wants to see it.

Assuming the bonds are purchased at par, which currently would by YTM = coupon = 2.05%, average YTM is 1.35%, while 1-year return is 2.01%. So this basically verifies that the return is pretty much determined by the YTM of the 10-year bond in this simulated scenario, and that average YTM significantly underestimates 1-year return.

Assuming the maturing bond is rolled into a new 10-year bond each year, and yield curve is static, this is the steady state, and the average YTM will always underestimate the 1-year return.

For this scenario the average current yield (roughly distribution yield for a bond fund) also is 2.01% (same as average YTM), but the current yield of individual bonds ranges from 1.98% to 2.05%. Also in this scenario, price change is positive until the bond rolls from year 4 to year 3, at which point price starts dropping toward par.

Now change scenario so that coupon rate is 1% higher than YTM for 10-year bond, so 3.05% at current rates. Average YTM is the same at 1.35%; this is a tautology, since YTM is taken directly from the Treasury yield curve, but this is not exactly right, because the Treasury yield curve is a par yield curve, which basically assumes that coupon rate = YTM at each maturity, and that is not the case in this scenario. One-year return is slightly lower at 1.99%, but this slight difference may just be an artifact of the mismatch between par-yield curve and the actual yield curve for the bonds in this scenario, and/or other simplifications in the model.

However, in this scenario average current yield (roughly distribution yield for a bond fund) is much higher at 2.84%, with a range of 2.80% to 2.97% for the individual bonds. The current yield of the 10-year bond is 2.80%, so in this scenario neither the average current yield nor the the current yield of the 10-year bond is representative of the 1-year return, but is much higher. This is explained by the fact that in this scenario, bond value decreases over one year for every bond except the 10-year bond rolling to 9-years, which has a small increase of 0.04 (per 100).

For a rolling ladder with static yield curve, this also is a static scenario, so distribution yield will remain higher than average YTM, but will continue to overestimate 1-year return.

Note that this bond ladder scenario assumes that equal face value of each bond is held, which is not the same as equal amount of each different maturity bond, so the average YTM is not weighted assuming equal value for each year of maturity, but is a simple average of the YTMs assuming that par value of each bond is held.

As always, the important caveat is that assuming a static yield curve is naive, as it leads to the paradox that one can easily beat the market with no risk by simply trading around a steep part of the yield curve at longer maturities. My working paradigm is that the yield curve reflects an unknown combination of expectations for future rates and term premiums. The the extent to which buying longer term bonds is likely to pay off depends on to what extent each of these variables dominates the yield curve, and this is not observable. And of course expectations may not reflect what actually happens.

EDIT: A quick addendum. If we take the first scenario above, but instead roll the bonds as they reach 3-year maturity (so a 3-10 year ladder), the average YTM increases from 1.35% to 1.67% and the 1-year return increases from 2.01% to 2.48%. In this case the average YTM underestimate of return increases from 66 basis points to 81 basis points. What I think this indicates is that the biggest source of average YTM (or SEC yield) underestimating return is not from selling the bonds before maturity, but simply from owning a ladder of bonds that were originally purchased with 10-year maturities, or perhaps from using a par yield curve for non-par bonds. I think a next step is to do a simulation in which the coupon rate = YTM for each term to maturity.

Kevin
||.......|| Suggested format for Asking Portfolio Questions (edit original post)

User avatar
Kevin M
Posts: 8866
Joined: Mon Jun 29, 2009 3:24 pm
Contact:

Re: Yield on Vanguard Int. Term Tax Exempt Fund

Post by Kevin M » Sun Oct 04, 2015 4:25 pm

Kevin M wrote:<snip> ... but this is not exactly right, because the Treasury yield curve is a par yield curve, which basically assumes that coupon rate = YTM at each maturity, and that is not the case in this scenario.
<snip>

What I think this indicates is that the biggest source of average YTM (or SEC yield) underestimating return is not from selling the bonds before maturity, but simply from owning a ladder of bonds that were originally purchased with 10-year maturities, or perhaps from using a par yield curve for non-par bonds. I think a next step is to do a simulation in which the coupon rate = YTM for each term to maturity.

After a quick modification to the model, it looks like the key is indeed using a par-yield curve for non-par bonds. I changed the model to assume that all bonds are purchased at par (one bond at each term to maturity) and held for one year, so that the coupon rate is equal to the YTM for each bond at time of purchase. This model is more closely aligned with using the Treasury par yield curve. So the price of the 9-year bond is based on the 9-year YTM, but the coupon rate is equal to the YTM of the 10-year bond (which would be the case for a 10-year bond one year after purchase at par).

With this modification, average YTM pretty much perfectly predicts 1-year return for the 1-10 year ladder, both at 1.35%. For the 3-10 year ladder, average YTM underestimates 1-year return by only 8 basis points, with average YTM at 1.67% and 1-year return at 1.75%.

So other insights are welcome, but it looks to me like an error in the way some of us have modeled this is to use a par yield curve, but with bonds that have coupon rates above par. This is an internally inconsistent model, which may be the main reason for the results that do not seem intuitive--at least not to me.

If anyone doubts that the Treasury par yield curve is inappropriate for non-par bonds, try building a yield curve from quotes from the WSJ. I've done it, and it's a mess!

Kevin
||.......|| Suggested format for Asking Portfolio Questions (edit original post)

User avatar
Electron
Posts: 1656
Joined: Sat Mar 10, 2007 8:46 pm

Re: Yield on Vanguard Int. Term Tax Exempt Fund

Post by Electron » Tue Oct 06, 2015 5:22 pm

Kevin - Thanks for the detailed analysis which I am still reviewing. I found the yield curve on the Treasury site which is a good reference to bookmark.

http://www.treasury.gov/resource-center ... data=yield

Your post finally got me to review the Riding the Yield Curve threads which are quite intriguing.

In regards to Andy's original post, the document I posted should be of interest to everyone.

http://www8.gsb.columbia.edu/cbs-direct ... remium.pdf

It does appear that most Municipal Bonds are issued at a premium. That explains why we see so many bonds with a 5.000% coupon in the portfolio holdings. The article states that the premium is typically higher with longer maturity bonds. Investors apparently do not like buying Municipal bonds below par which can result in some of the return taxed as ordinary income. As a result, issuers have an incentive for the bond to trade at a premium as much as possible. If the market discount is below the De Minimis amount, the discount is treated as a capital gain.

The article also includes some very technical information on tax regimes and tax options. Loss Harvesting is mentioned. The conclusions on page 36 are an excellent summary of the key points in the article. One interesting comment is that many institutional bond buyers are very sophisticated. The result is that market prices can transmit information very efficiently even in the municipal bond market.
Electron

User avatar
ogd
Posts: 4807
Joined: Thu Jun 14, 2012 11:43 pm

Re: Yield on Vanguard Int. Term Tax Exempt Fund

Post by ogd » Tue Oct 06, 2015 8:24 pm

Kevin M wrote:With this modification, average YTM pretty much perfectly predicts 1-year return for the 1-10 year ladder, both at 1.35%. For the 3-10 year ladder, average YTM underestimates 1-year return by only 8 basis points, with average YTM at 1.67% and 1-year return at 1.75%.

So other insights are welcome, but it looks to me like an error in the way some of us have modeled this is to use a par yield curve, but with bonds that have coupon rates above par. This is an internally inconsistent model, which may be the main reason for the results that do not seem intuitive--at least not to me.

I don't understand this Kevin -- can you post some more details? Perhaps specific numbers?

Kevin M wrote:What I think this indicates is that the biggest source of average YTM (or SEC yield) underestimating return is not from selling the bonds before maturity, but simply from owning a ladder of bondsthat were originally purchased with 10-year maturities, or perhaps from using a par yield curve for non-par bonds.

I agree with the part in bold, with the caveat that whether "biggest source" or not probably depends on the exact yield curve. E.g. if the last 5 years are flat at zero, selling before maturity is probably a significant win.

I am not sure about the part in italic. Isn't the contention that SEC yield / YtM would underestimate the returns unaffected?

User avatar
neurosphere
Posts: 2599
Joined: Sun Jan 17, 2010 1:55 pm
Location: NYC

Re: Yield on Vanguard Int. Term Tax Exempt Fund

Post by neurosphere » Tue Oct 06, 2015 8:50 pm

Kevin M wrote:
neurosphere wrote:
(E) Exception to holding period requirement for certain regularly declared exempt-interest dividends
(i)Daily dividend companies Except as otherwise provided by regulations, subparagraph (B) shall not apply with respect to a regular dividend paid by a regulated investment company which declares exempt-interest dividends on a daily basis in an amount equal to at least 90 percent of its net tax-exempt interest and distributes such dividends on a monthly or more frequent basis.


I'm actually not sure of this exception applies to the tax-free bond funds such as those at Vanguard which pay monthly dividends. I just don't sell tax-exempt shares which have less than a 6 month holding period so I don't have to worry about it, and I've never had a tax-prep client to whom this situation was relevant so I've never had to answer the question. :D I'm sure the answer is on the internets somewhere, perhaps even right here on Bogleheads. :D :D

Wow, I've never noticed this exception before, and I think it does apply to Vanguard TE bond funds, since they do declare the dividend on a daily basis. You can see the accrued dividend in your account by looking at Balances by date. The six month rule always seemed nonsensical for Vanguard TE funds for this reason, but the exemption makes it rational.

Kevin


Was this issue/question ever resolved? Perhaps we should start a separate thread for it? It would be nice to be able to tell folks definitively that this "6-month rule" does not apply to Vanguard tax-exempt funds.
-- Real name: Sotirios Keros. If you have to ask "Is a Target Retirement fund right for me?", the answer is yes.

User avatar
Kevin M
Posts: 8866
Joined: Mon Jun 29, 2009 3:24 pm
Contact:

Re: Yield on Vanguard Int. Term Tax Exempt Fund

Post by Kevin M » Tue Oct 06, 2015 9:30 pm

neurosphere wrote:Was this issue/question ever resolved? Perhaps we should start a separate thread for it? It would be nice to be able to tell folks definitively that this "6-month rule" does not apply to Vanguard tax-exempt funds.

I haven't seen any additional evidence to support it or refute it, but a note about the exception was added to the Wiki article: Tax loss harvesting - Bogleheads. Apparently the forum member who did the update (Rkhusky) thought it was definitive:

Note that Vanguard Tax Exempt funds (and others like them) are not subject to the preceding 6 month rule because of the way they accrue and pay out interest (dividends). The 6 month rule for tax exempt interest comes from 26 U.S. Code § 852(b)(4)(B), but there is an exception in 852(b)(4)(E) for funds that declare dividends daily and pay them monthly or more frequently

Of course this is the way I interpret it as well, but the reference for the quote above was this thread, so not definitive unless you think the discussion in this thread has so determined.

I don't know that we can provide definitive tax advice here or in the Wiki, but my working assumption is that the exception applies to VG TE funds, and I'd do my tax return based on that assumption, unless perhaps Vanguard submitted different info to the IRS.

Kevin
||.......|| Suggested format for Asking Portfolio Questions (edit original post)

User avatar
neurosphere
Posts: 2599
Joined: Sun Jan 17, 2010 1:55 pm
Location: NYC

Re: Yield on Vanguard Int. Term Tax Exempt Fund

Post by neurosphere » Wed Oct 07, 2015 7:29 am

Kevin M wrote:
neurosphere wrote:Was this issue/question ever resolved? Perhaps we should start a separate thread for it? It would be nice to be able to tell folks definitively that this "6-month rule" does not apply to Vanguard tax-exempt funds.

I haven't seen any additional evidence to support it or refute it, but a note about the exception was added to the Wiki article: Tax loss harvesting - Bogleheads. Apparently the forum member who did the update (Rkhusky) thought it was definitive:

Note that Vanguard Tax Exempt funds (and others like them) are not subject to the preceding 6 month rule because of the way they accrue and pay out interest (dividends). The 6 month rule for tax exempt interest comes from 26 U.S. Code § 852(b)(4)(B), but there is an exception in 852(b)(4)(E) for funds that declare dividends daily and pay them monthly or more frequently

Of course this is the way I interpret it as well, but the reference for the quote above was this thread, so not definitive unless you think the discussion in this thread has so determined.

I don't know that we can provide definitive tax advice here or in the Wiki, but my working assumption is that the exception applies to VG TE funds, and I'd do my tax return based on that assumption, unless perhaps Vanguard submitted different info to the IRS.

Kevin


Here is some additional information. According to SLIDE 17 of this pdf from the Investment Company Institute (https://www.ici.org/pdf/webinar_11_modact_slides.pdf) it appears this exclusion is relatively new (2010). The slide reads:

Loss Disallowance on Muni Fund Shares
[Bill §309 – IRC §852(b)(4)(E)]

• For daily dividend muni funds, repeals disallowance of losses on
shares held for 6 months or less to extent of exempt-interest
dividends.

 Applies only to “regular” dividends.
 Effective Date – RIC shares for which taxpayer’s holding period begins
after 12/22/10.

• Observations
 § 852(b)(4) designed to prevent “buying” an exempt-interest dividend and
then realizing a non-economic capital loss.

But not possible to buy a dividend if fund declares daily.


The bolding and underlining is mine. Have we confirmed that the dividend in VG TE funds is indeed "declared" daily? They are certainly paid out monthly. And those are the two necessary criteria (along with the requirement to pay out more than 90% of the dividends).

NEVERMIND. I found confirmation (in addition to what was When in doubt, consult the prospectus. Here is a paragraph from the prospectus (or perhaps the supplement) for VG's national (i.e. non-state specific) municipal bond funds:

Each Fund distributes to shareholders virtually all of its net income [i.e. more than the required 90%] (interest less
expenses) as well as any net short-term or long-term capital gains realized from the
sale of its holdings. The Fund’s income dividends are declared daily and distributed
monthly;


Bolding and brackets are mine.

I have a question though. Can someone clarify why it's "not possible to buy a dividend if a fund declares daily" if the dividends are paid out monthly?

I've looked at dozens of websites trying to find information about this 6-month rule (i.e. Fairmark and other similar tax-info sites) and did not find mention of this exception. It appears that perhaps this new exception is obscure enough that it has gone largely unnoticed. And I have reviewed dozens of recent threads on BH where posters are warned about the loss-exclusion specifically on VG funds which seem to meet the criteria on an exception.

If I have time, maybe I'll start a separate thread, mostly just to call attention to this issue. I know that many have adjusted the timing of sales and/or reinvestment of TE dividends in order to steer clear of the 6-month rule. It would be nice to let people know this is not something to worry about for many traditional TE fund. It's just a little weird (given all of the super knowledgeable posters on BH who know in aggregate much more about taxes and investing than I do) that I was the one to stumble across this exception (thanks to the prompting of abuss368 who requested a citation from the tax code!). I'm not patting myself on the back or anything but rather the opposite, I'm a little insecure about making any new definitive statements. However, thanks to KevinM and others, I think I'm convinced. :)

NS
-- Real name: Sotirios Keros. If you have to ask "Is a Target Retirement fund right for me?", the answer is yes.

User avatar
Kevin M
Posts: 8866
Joined: Mon Jun 29, 2009 3:24 pm
Contact:

Re: Yield on Vanguard Int. Term Tax Exempt Fund

Post by Kevin M » Wed Oct 07, 2015 11:13 am

neurosphere wrote:Here is some additional information. According to SLIDE 17 of this pdf from the Investment Company Institute (https://www.ici.org/pdf/webinar_11_modact_slides.pdf) it appears this exclusion is relatively new (2010). The slide reads:<snip>

Another good find!

neurosphere wrote:I have a question though. Can someone clarify why it's "not possible to buy a dividend if a fund declares daily" if the dividends are paid out monthly?


As I explained above, the dividends are accrued to your account, which you can see by looking at the daily balance view. When someone buys the fund, they do not receive the accrued dividends, therefore they are not buying the dividend.

The empirical evidence, from personal experience, is that if you buy a fund mid-month, you will not receive the full dividend distribution at month end, and if you sell a fund mid-month, you will receive the accrued dividends as a separate transaction, and sometimes even as a check mailed to you.

Kevin
||.......|| Suggested format for Asking Portfolio Questions (edit original post)

User avatar
Kevin M
Posts: 8866
Joined: Mon Jun 29, 2009 3:24 pm
Contact:

Re: Yield on Vanguard Int. Term Tax Exempt Fund

Post by Kevin M » Wed Oct 07, 2015 12:21 pm

ogd wrote:I don't understand this Kevin -- can you post some more details? Perhaps specific numbers?

I don't think I understand it yet either. I may have made a mistake in the model--I'm still working on it. I just made what I think is a fix, and it generates results closer to the original model, but I'm still scratching my head over it.

One thing that is true is that if you look at WSJ quotes, you'll see different YTMs for same or close maturity due to different coupon rates. A blatant example is given by two bonds maturing on 11/15/2015, with calculated bid yields of 0.00% for the 4.5% coupon bond and 0.71% for the 9.875% coupon bond. There are many examples of such discrepancies, but none this large.

The Treasury.gov yield curve is based on on-the-run Treasuries, for which coupon rate tends to be close to YTM, and using this yield curve for premium bonds will introduce some error. However, the differences in YTM generally aren't huge, so it may not be nearly as bad as I thought.

Kevin
||.......|| Suggested format for Asking Portfolio Questions (edit original post)

User avatar
Electron
Posts: 1656
Joined: Sat Mar 10, 2007 8:46 pm

Re: Yield on Vanguard Int. Term Tax Exempt Fund

Post by Electron » Wed Oct 07, 2015 1:14 pm

neurosphere wrote:I have a question though. Can someone clarify why it's "not possible to buy a dividend if a fund declares daily" if the dividends are paid out monthly?

Compare a Vanguard bond fund with any bond fund that includes daily interest in NAV and then goes ex-dividend at the end of the month. The six month rule prevented an investor from buying a tax exempt bond fund one day before the ex-dividend date and then selling the next day at the lower ex-dividend price. If the tax loss was allowed, one could collect the tax exempt income and take an offsetting short term capital loss for a net overall gain.

There is also a six month rule related to capital gains distributions from mutual funds. That also prevents an investor from buying a mutual fund one day before the ex-dividend date, selling the fund the next day, and getting favorable tax treatment as a result.
Electron

Post Reply