vanguard VS.. fidelity GNMA

I'm trying to decide between starting to invest in vanguard gnma in a
taxable account OR fidelity gnma in IRA account. I know it sounds like
a bizarre question, but....
I know that ideally I would put bond funds (like gnma) in a
tax-efficient account (such as my ira account). And ideally I would
buy the lowest cost fund, like vanguard. But my IRA account is with
fidelity and i want to keep it there. And I don't want to pay the $75
or $100 that fidelity would charge me to open a brokerage account and
put the vang. fund in there.

So can one of you investment whizes tell me, off the top of your head
without doing alot of intricate calculations (which I can't pay you
for), whether the following statement seems plausible?:
PUTTING THE LOWER-COST VANG. GNMA ACCOUNT IN YOUR TAXABLE ACCOUNT
PROBABLY IS ABOUT THE SAME AS PUTTING THE HIGHER-COST FIDELITY FUND IN
YOUR IRA ACCOUNT, with respect to overall costs involved in the long
run (tha is, tax and fund expense costs). And I think that a collorary
(sp?) of this hypothesis would be that: If the NAV performance of these
funds were equal over a long period of time, the total returns I would
receive (after taxes, at least for the vanguard fund) would be roughly
the same in the above 2 scenarios. My tax bracket is LOW, by the way
(maybe 15% at the most), which is an important factor here and makes me
think the above hypothesis is likely to be close to true.

Re: vanguard VS.. fidelity GNMA

p.s. Forgot to mention that the ER for Vang gnma is .20, but it's .60
for fidelity gnma. I'm also thinking about the same question for
Vang. vs. Fidelity TIPS funds....probably about the same ER's., and so
the answer would be the same.

wrote:
> I'm trying to decide between starting to invest in vanguard gnma in a
> taxable account OR fidelity gnma in IRA account. I know it sounds
like
> a bizarre question, but....
> I know that ideally I would put bond funds (like gnma) in a
> tax-efficient account (such as my ira account). And ideally I would
> buy the lowest cost fund, like vanguard. But my IRA account is with
> fidelity and i want to keep it there. And I don't want to pay the $75
> or $100 that fidelity would charge me to open a brokerage account and
> put the vang. fund in there.
>
> So can one of you investment whizes tell me, off the top of your head
> without doing alot of intricate calculations (which I can't pay you
> for), whether the following statement seems plausible?:
> PUTTING THE LOWER-COST VANG. GNMA ACCOUNT IN YOUR TAXABLE ACCOUNT
> PROBABLY IS ABOUT THE SAME AS PUTTING THE HIGHER-COST FIDELITY FUND
IN
> YOUR IRA ACCOUNT, with respect to overall costs involved in the long
> run (tha is, tax and fund expense costs). And I think that a
collorary
> (sp?) of this hypothesis would be that: If the NAV performance of
these
> funds were equal over a long period of time, the total returns I
would
> receive (after taxes, at least for the vanguard fund) would be
roughly
> the same in the above 2 scenarios. My tax bracket is LOW, by the
way
> (maybe 15% at the most), which is an important factor here and makes
me
> think the above hypothesis is likely to be close to true.

Re: vanguard VS.. fidelity GNMA

<> wrote in message
news:
> I'm trying to decide between starting to invest in vanguard gnma in a
> taxable account OR fidelity gnma in IRA account. I know it sounds like
> a bizarre question, but....
> I know that ideally I would put bond funds (like gnma) in a
> tax-efficient account (such as my ira account). And ideally I would
> buy the lowest cost fund, like vanguard. But my IRA account is with
> fidelity and i want to keep it there. And I don't want to pay the $75
> or $100 that fidelity would charge me to open a brokerage account and
> put the vang. fund in there.
>
> So can one of you investment whizes tell me, off the top of your head
> without doing alot of intricate calculations (which I can't pay you
> for), whether the following statement seems plausible?:

I've little time to crunch numbers now, so I'll just give some qualitative
comments/suggestions:

> PUTTING THE LOWER-COST VANG. GNMA ACCOUNT IN YOUR TAXABLE ACCOUNT
> PROBABLY IS ABOUT THE SAME AS PUTTING THE HIGHER-COST FIDELITY FUND IN
> YOUR IRA ACCOUNT, with respect to overall costs involved in the
> long run (tha is, tax and fund expense costs).

Three other alternatives:
- Open a second IRA at Vanguard; this will cost you $10/year, until you
reach $5,000. So it can come out $55-$65 cheaper than buying the fund
through Fidelity. Note that you'd be paying the $75 fee every time you add
money, so it is really not a great idea to use non-NTF funds for periodic
investments at Fidelity.


- If you anticipate a one-time only contribution to the fund, you could open
the IRA at Vanguard, and then transfer it to Fidelity (assuming there is no
fee for transferring the mutual fund/IRA account; I haven't checked this.)

- Consider Fidelity Mortgage Securities instead of Fidelity GNMA fund. You
are already considering the Fidelity GNMA, which, unlike Vanguard's, invests
in Fredie Macs as well as GNMAs. So you are not adverse to mortgage
investments that are not pure GNMA. Fidelity Mortgage Securities is yet
more aggressive, investing more in CMOs, but its returns have been
historically greater as well. Note that Lipper classifies this as a U.S.
mortgage fund, as contrasted with the others, that it classifies as GNMA
funds.

> And I think that a collorary
> (sp?) of this hypothesis would be that: If the NAV performance of these
> funds were equal over a long period of time, the total returns I would
> receive (after taxes, at least for the vanguard fund) would be roughly
> the same in the above 2 scenarios.

I'm not going to rehash everything in another thread on how NAV does not
reflect performance - suffice to say that you should look at total return,
not NAV. For bond funds, over time, NAV should be pretty flat, because on
average, the total return is interest, not price appreciation.

> My tax bracket is LOW, by the way
> (maybe 15% at the most), which is an important factor here and makes
> me think the above hypothesis is likely to be close to true.

I hope you are using a Roth IRA - while they generally work out better in
most cases, this is especially true for someone in a low current tax
bracket, because it costs you relatively little to make after-tax
contributions.

--
Mark Freeland

Re: vanguard VS.. fidelity GNMA

<> wrote in message
news:
> p.s. Forgot to mention that the ER for Vang gnma is .20, but it's .60
> for fidelity gnma. I'm also thinking about the same question for
> Vang. vs. Fidelity TIPS funds....probably about the same ER's., and so
> the answer would be the same.

On a hundred grand, you're talking $400. You know that, right?

>
> wrote:
>> I'm trying to decide between starting to invest in vanguard gnma in a
>> taxable account OR fidelity gnma in IRA account. I know it sounds
> like
>> a bizarre question, but....
>> I know that ideally I would put bond funds (like gnma) in a
>> tax-efficient account (such as my ira account). And ideally I would
>> buy the lowest cost fund, like vanguard. But my IRA account is with
>> fidelity and i want to keep it there. And I don't want to pay the $75
>> or $100 that fidelity would charge me to open a brokerage account and
>> put the vang. fund in there.
>>
>> So can one of you investment whizes tell me, off the top of your head
>> without doing alot of intricate calculations (which I can't pay you
>> for), whether the following statement seems plausible?:
>> PUTTING THE LOWER-COST VANG. GNMA ACCOUNT IN YOUR TAXABLE ACCOUNT
>> PROBABLY IS ABOUT THE SAME AS PUTTING THE HIGHER-COST FIDELITY FUND
> IN
>> YOUR IRA ACCOUNT, with respect to overall costs involved in the long
>> run (tha is, tax and fund expense costs). And I think that a
> collorary
>> (sp?) of this hypothesis would be that: If the NAV performance of
> these
>> funds were equal over a long period of time, the total returns I
> would
>> receive (after taxes, at least for the vanguard fund) would be
> roughly
>> the same in the above 2 scenarios. My tax bracket is LOW, by the
> way
>> (maybe 15% at the most), which is an important factor here and makes
> me
>> think the above hypothesis is likely to be close to true.
>

Re: vanguard VS.. fidelity GNMA

Thanks Mark. But I cannot open any more IRA accounts, in ANY company.
I'm not eligible to, and the one's I already opened at Fidelity (one
being an ira beneficiary account) are relatively small anyway (less
than 10% of my portfolio). So my choices
are as stated in my original post (though I will consider the Fid.
mortgage fund you
mentioned...I vaguely remember reading about that a few months ago in
the M-Star 500
annual book; i think they also said they prefer it. But i'm still
curious about my
original question...I guess you're implying that the 2 scenarios
probably will be
pretty close in end results, if you're saying that time-consuming
calculations would
have to be done (otherwise, you would have given an off the top of the
head answer).

Mark Freeland wrote:
> <> wrote in message
> news:
> > I'm trying to decide between starting to invest in vanguard gnma in
a
> > taxable account OR fidelity gnma in IRA account. I know it sounds
like
> > a bizarre question, but....
> > I know that ideally I would put bond funds (like gnma) in a
> > tax-efficient account (such as my ira account). And ideally I
would
> > buy the lowest cost fund, like vanguard. But my IRA account is
with
> > fidelity and i want to keep it there. And I don't want to pay the
$75
> > or $100 that fidelity would charge me to open a brokerage account
and
> > put the vang. fund in there.
> >
> > So can one of you investment whizes tell me, off the top of your
head
> > without doing alot of intricate calculations (which I can't pay you
> > for), whether the following statement seems plausible?:
>
> I've little time to crunch numbers now, so I'll just give some
qualitative
> comments/suggestions:
>
> > PUTTING THE LOWER-COST VANG. GNMA ACCOUNT IN YOUR TAXABLE ACCOUNT
> > PROBABLY IS ABOUT THE SAME AS PUTTING THE HIGHER-COST FIDELITY FUND
IN
> > YOUR IRA ACCOUNT, with respect to overall costs involved in the
> > long run (tha is, tax and fund expense costs).
>
> Three other alternatives:
> - Open a second IRA at Vanguard; this will cost you $10/year, until
you
> reach $5,000. So it can come out $55-$65 cheaper than buying the
fund
> through Fidelity. Note that you'd be paying the $75 fee every time
you add
> money, so it is really not a great idea to use non-NTF funds for
periodic
> investments at Fidelity.
>

>
> - If you anticipate a one-time only contribution to the fund, you
could open
> the IRA at Vanguard, and then transfer it to Fidelity (assuming there
is no
> fee for transferring the mutual fund/IRA account; I haven't checked
this.)
>
> - Consider Fidelity Mortgage Securities instead of Fidelity GNMA
fund. You
> are already considering the Fidelity GNMA, which, unlike Vanguard's,
invests
> in Fredie Macs as well as GNMAs. So you are not adverse to mortgage
> investments that are not pure GNMA. Fidelity Mortgage Securities is
yet
> more aggressive, investing more in CMOs, but its returns have been
> historically greater as well. Note that Lipper classifies this as a
U.S.
> mortgage fund, as contrasted with the others, that it classifies as
GNMA
> funds.
>
> > And I think that a collorary
> > (sp?) of this hypothesis would be that: If the NAV performance of
these
> > funds were equal over a long period of time, the total returns I
would
> > receive (after taxes, at least for the vanguard fund) would be
roughly
> > the same in the above 2 scenarios.
>
> I'm not going to rehash everything in another thread on how NAV does
not
> reflect performance - suffice to say that you should look at total
return,
> not NAV. For bond funds, over time, NAV should be pretty flat,
because on
> average, the total return is interest, not price appreciation.
>
> > My tax bracket is LOW, by the way
> > (maybe 15% at the most), which is an important factor here and
makes
> > me think the above hypothesis is likely to be close to true.
>
> I hope you are using a Roth IRA - while they generally work out
better in
> most cases, this is especially true for someone in a low current tax
> bracket, because it costs you relatively little to make after-tax
> contributions.
>
> --
> Mark Freeland
>

Re: vanguard VS.. fidelity GNMA

Higher cost bond and bond-like funds often take on higher risk in order
to increase yield and compensate for their higher expenses. In the
early 1990s this caused many of them to experience larger losses than
similar Vanguard funds.

Re: vanguard VS.. fidelity GNMA

<> wrote in message
news:
> Higher cost bond and bond-like funds often take on higher risk in order
> to increase yield and compensate for their higher expenses. In the
> early 1990s this caused many of them to experience larger losses than
> similar Vanguard funds.

Many of Vanguard's bond index funds didn't begin until 1994. Their Treasury
bond funds did quite well in the early 1990's as did the funds of many other
companies.

I agree that low expenses are important for bond and money market funds but
don't recall large losses in bond funds in the early 1990's.

Re: vanguard VS.. fidelity GNMA

wrote:
>
> Thanks Mark. [...]
> But i'm still curious about my
> original question...I guess you're implying that the 2 scenarios
> probably will be
> pretty close in end results, if you're saying that time-consuming
> calculations would
> have to be done (otherwise, you would have given an off the top of the
> head answer).

Please don't read too much into my not doing calculations. It would
have to be set up algebraically - too many factors that you haven't
provided - and I wasn't going to work through everything needed.

For instance - how much money are we talking about (in the current post
you said not too much). Obviously, if we are talking $100, then
spending $75 dollars to get the Vanguard fund through Fidelity will come
out the loser, while if we were talking $10M, then spending $75, vs.
saving 0.40%/year in expenses would be a wise choice.

Another factor - how long you expect to hold the fund - that could (but
might not be) affected by the fact that this is a beneficiary IRA.
Again to take extremes - hold it for one day before liquidating, and
paying $75 doesn't make sense. Hold it for 30 years, and that might pay
off (depending on the size of the account).

Another factor - expected rate of return - this one I could make a fair
approximation on.

You get the idea.

--
Mark Freeland

Re: vanguard VS.. fidelity GNMA

Ed wrote:
>
> <> wrote in message
> news:
> > Higher cost bond and bond-like funds often take on higher risk in
> > order to increase yield and compensate for their higher expenses.
> > In the early 1990s this caused many of them to experience larger
> > losses than similar Vanguard funds.

Funds *may* take on higher risk to compensate for higher costs - true as
far as that goes. But since the question was about specific funds, it
would be a good idea to see if this applies for these funds in
particular.

There are various forms of risk. The risk generally implied by the
statement you gave is credit risk. In my previous post, I already noted
that Vanguard's fund is a pure GNMA, while the Fidelity fund takes on
FNMA and other quasi-government debt. I implied that this was some
additional risk, and if the poster was open to that, returns over time
could be improved over Vanguard's fund. Not much different from saying
that short term bond funds, over time, will outperform money market
funds, even though they didn't in 1994.

Then there is volatility risk. (Personally, I'm not thrilled with
standard deviation as a measure of volatility risk, but that's what many
people use.) Vanguard's GNMA fund has a three year standard deviation
(per Morningstar) of 3.13%, vs. Fidelity's Ginnie Mae std. dev. of
2.22%, and Fidelity's Mortgage Securities' 2.56%. Vanguard appears more
risky.

There is interest rate risk - how vulnerable the fund is to rises in
interest rates (resulting in loss of principal). For callable debt,
duration can be misleading, but FWIW, Vanguard's fund has the highest
duration, at 3.3 years, tied with Fidelity's Ginnie Mae fund, but higher
than Fidelity's Mortgage Securities at 3.0 years. The "riskiest" fund
(by credit risk) - Mortgage Securities - appears to be the the most
protected against interest rate risk.

So, while the rule of thumb you gave is true in the abstract, there is a
question about whether it applies to these particular funds.

> Many of Vanguard's bond index funds didn't begin until 1994.

That is true, but their GNMA (which is not an index fund) goes back much
further, and its 1994 return (fiscal year ending Jan 31, 1995) was
+0.36%.


Unfortunately, Fidelity's funds had a fiscal calendar year ending July
31, but their figures for FY94 (ending July 31, 1994) and FY95 were:
1994 1995
Ginnie Mae -0.63% 10.26%
Mtg. Sec. 3.13% 10.88%


Even in the early 90s, it looks like the "higher risk" Mortgage
Securities Fund was rewarded for the risk it took.

--
Mark Freeland

Re: vanguard VS.. fidelity GNMA

Thanks, but getting the Vang. fund though Fid. was not one of the 2
scenarios I was interested in (I said I didn't want to do that).

Mark Freeland wrote:
> wrote:
> >
> > Thanks Mark. [...]
> > But i'm still curious about my
> > original question...I guess you're implying that the 2 scenarios
> > probably will be
> > pretty close in end results, if you're saying that time-consuming
> > calculations would
> > have to be done (otherwise, you would have given an off the top of
the
> > head answer).
>
> Please don't read too much into my not doing calculations. It would
> have to be set up algebraically - too many factors that you haven't
> provided - and I wasn't going to work through everything needed.
>
> For instance - how much money are we talking about (in the current
post
> you said not too much). Obviously, if we are talking $100, then
> spending $75 dollars to get the Vanguard fund through Fidelity will
come
> out the loser, while if we were talking $10M, then spending $75, vs.
> saving 0.40%/year in expenses would be a wise choice.
>
> Another factor - how long you expect to hold the fund - that could
(but
> might not be) affected by the fact that this is a beneficiary IRA.
> Again to take extremes - hold it for one day before liquidating, and
> paying $75 doesn't make sense. Hold it for 30 years, and that might
pay
> off (depending on the size of the account).
>
> Another factor - expected rate of return - this one I could make a
fair
> approximation on.
>
> You get the idea.
>
> --
> Mark Freeland
>

Re: vanguard VS.. fidelity GNMA

<> wrote in message
news:
> Thanks, but getting the Vang. fund though Fid. was not one of the 2
> scenarios I was interested in (I said I didn't want to do that).

You are right - my error. So size of investment doesn't matter (all figures
are in percentages that are independent of size - this would not be true if
Fidelity charged a fixed annual fee for IRAs, or if you were investing in an
index fund with a fixed $10/year fee.)

The other variable factors still apply - expected rate of return and number
of years the investment would be held. (For example, if one were to
liquidate within a year, then there is no advantage in keeping money in an
IRA, because no taxes would be deferred, but if one were to hold the IRA for
decades, the tax deferral could compensate for the additional expenses of
the higher cost fund.)

--
Mark Freeland

Re: vanguard VS.. fidelity GNMA

"Mark Freeland" <> wrote in message
news:
> Ed wrote:
>>
>> <> wrote in message
>> news:
>> > Higher cost bond and bond-like funds often take on higher risk in
>> > order to increase yield and compensate for their higher expenses.
>> > In the early 1990s this caused many of them to experience larger
>> > losses than similar Vanguard funds.
>
> Funds *may* take on higher risk to compensate for higher costs - true as
> far as that goes. But since the question was about specific funds, it
> would be a good idea to see if this applies for these funds in
> particular.

If they didn't then why would anyone buy them? Lack of information?

>
> There are various forms of risk. The risk generally implied by the
> statement you gave is credit risk. In my previous post, I already noted
> that Vanguard's fund is a pure GNMA, while the Fidelity fund takes on
> FNMA and other quasi-government debt. I implied that this was some
> additional risk, and if the poster was open to that, returns over time
> could be improved over Vanguard's fund. Not much different from saying
> that short term bond funds, over time, will outperform money market
> funds, even though they didn't in 1994.

The less diversified a fund, the more risk the investor will have to absorb
and manage.

>
> Then there is volatility risk. (Personally, I'm not thrilled with
> standard deviation as a measure of volatility risk, but that's what many
> people use.) Vanguard's GNMA fund has a three year standard deviation
> (per Morningstar) of 3.13%, vs. Fidelity's Ginnie Mae std. dev. of
> 2.22%, and Fidelity's Mortgage Securities' 2.56%. Vanguard appears more
> risky.

Is this really a distinct risk or just a symptom of more or less
diversification?

>
> There is interest rate risk - how vulnerable the fund is to rises in
> interest rates (resulting in loss of principal). For callable debt,
> duration can be misleading, but FWIW, Vanguard's fund has the highest
> duration, at 3.3 years, tied with Fidelity's Ginnie Mae fund, but higher
> than Fidelity's Mortgage Securities at 3.0 years. The "riskiest" fund
> (by credit risk) - Mortgage Securities - appears to be the the most
> protected against interest rate risk.

Is not one of the reasons for a bond mutual fund to diversify away credit
risk to a certain degree?

>
> So, while the rule of thumb you gave is true in the abstract, there is a
> question about whether it applies to these particular funds.
>
>> Many of Vanguard's bond index funds didn't begin until 1994.
>
> That is true, but their GNMA (which is not an index fund) goes back much
> further, and its 1994 return (fiscal year ending Jan 31, 1995) was
> +0.36%.
>
>
> Unfortunately, Fidelity's funds had a fiscal calendar year ending July
> 31, but their figures for FY94 (ending July 31, 1994) and FY95 were:
> 1994 1995
> Ginnie Mae -0.63% 10.26%
> Mtg. Sec. 3.13% 10.88%
>
>
> Even in the early 90s, it looks like the "higher risk" Mortgage
> Securities Fund was rewarded for the risk it took.
>
> --
> Mark Freeland
>

Re: vanguard VS.. fidelity GNMA

"NoEd" <> wrote

> The less diversified a fund, the more risk the investor will have to
> absorb and manage.

Not always true. A fund that invests entirely in U.S. Treasury securities is
not diversified.

> Is not one of the reasons for a bond mutual fund to diversify away credit
> risk to a certain degree?

Depends on the type of bonds/fund. See above. All bond funds have interest
rate risk but not all of them have much in the way of credit risk.

Re: vanguard VS.. fidelity GNMA

"Ed" <> wrote in message
news:
>
> "NoEd" <> wrote
>
>> The less diversified a fund, the more risk the investor will have to
>> absorb and manage.
>
> Not always true. A fund that invests entirely in U.S. Treasury securities
> is not diversified.

True.

>
>> Is not one of the reasons for a bond mutual fund to diversify away credit
>> risk to a certain degree?
>
> Depends on the type of bonds/fund. See above. All bond funds have interest
> rate risk but not all of them have much in the way of credit risk.
>

Does Vanguard's total bond fund have more or less credit risk than their
GNMA fund, esp. given the crazy, new fangled ways to finance real estate
purchases?

Re: vanguard VS.. fidelity GNMA

"NoEd" <> wrote in message
news:
>
> "Mark Freeland" <> wrote in message
> news:
> > Funds *may* take on higher risk to compensate for higher costs - true
> >as far as that goes. But since the question was about specific funds,
it
> > would be a good idea to see if this applies for these funds in
> > particular.
>
> If they didn't then why would anyone buy them? Lack of information?

I'm not sure what you are asking here. If the question is why people would
buy a higher cost fund if it didn't take on more risk (in an attempt to
achieve comparable raw performance), then there are serveral possibilities.

Limited options, e.g. in a 401(k) plan - you choose from what is available,
not from what you would like.

Convenience - if I'm a Vanguard investor, I'll invest in VFINX, even though
Fidelity's S&P 500 index fund charges a few basis points less.

Different management style - I may prefer a bond fund that shortens
maturities in (what is perceived to be) a rising interest rate market, over
say, an index fund that doesn't have that flexibility. For example,
Fidelity's intermediate bond fund has moved to an average duration of 3.4
years, while Vanguard's (and I would assume other companies') intermediate
bond index fund is fixed around 6 years. Data from Morningstar.

I'm sure you can think of other reasons as well.

> > There are various forms of risk. The risk generally implied by the
> > statement [the OP] gave is credit risk. In my previous post, I already
> > noted that Vanguard's fund is a pure GNMA, while the Fidelity fund
> > takes on FNMA and other quasi-government debt. I implied that
> > this was some additional risk, and if the poster was open to that,
> > returns over time could be improved over Vanguard's fund. Not
> > much different from saying that short term bond funds, over time,
> > will outperform money market funds, even though they didn't in 1994.
>
> The less diversified a fund, the more risk the investor will have to
absorb
> and manage.

I'm not sure how that relates to my comment, unless you are suggesting that
a GNMA fund is inherently more risky, because it limits itself to GNMAs, as
compared with a fund that is open to all mortage debt. Ed has already
responded to the general principle, by pointing out that a pure Treasury
fund has less credit risk than a total bond fund that includes junk.

> > Then there is volatility risk. (Personally, I'm not thrilled with
> > standard deviation as a measure of volatility risk, but that's what many
> > people use.) Vanguard's GNMA fund has a three year standard
> > deviation (per Morningstar) of 3.13%, vs. Fidelity's Ginnie Mae std.
> > dev. of 2.22%, and Fidelity's Mortgage Securities' 2.56%.
> > Vanguard appears more risky.
>
> Is this really a distinct risk or just a symptom of more or less
> diversification?

We can study that by looking at the typical volatility of the two classes of
funds - Lipper cleanly differentiates these as U.S. Mortgage funds and GNMA
funds.

Ignoring multiple share classes, there are 36 distinct funds in the first
category, of which Morningstar reports standard deviations for thirty.
These have an average standard deviation of 2.954.

Likewise, in the second category (GNMA), there are 25 distinct funds, of
which Morningstar reports standard deviations for 24. The have an average
std deviation of 2.654.

So it looks like GNMA funds, on average, are less volatile, despite being
less diversified (they are restricted to GNMAs, while mortgage funds can
invest in GNMAs or other mortgage securities).

What this says is that Vanguard's GNMA is more volatile than the Fidelity
funds, despite being in a less volatile category (or, despite any diversity
effects of its category).

> Is not one of the reasons for a bond mutual fund to diversify away credit
> risk to a certain degree?

Absolutely. The "conventional wisdom" is that you need at least $50K to
adequately diversify (I assume for credit risk) in a corporate bond
portfolio, and at least $100K for a muni bond portfolio. Mutual funds give
you instant diversification to reduce various forms of risk, including
credit risk.

I have always considered Treasuries an exception to this rule. Individual
issues are essentially "riskless" as far as credit risk goes. I don't see
the point in paying any management fee for a fund to simply buy Treasuries
for me. How much research is needed?
--
Mark Freeland

Re: vanguard VS.. fidelity GNMA

"NoEd" <> wrote in message
news:
>
> Does Vanguard's total bond fund have more or less credit risk than
> their GNMA fund, esp. given the crazy, new fangled ways to finance
> real estate purchases?

According to Vanguard, 100% of their GNMA fund assets are AAA rated, while
the average bond rating in their Total Bond Index Fund is AA rated - almost
1/4 of its holdings are below AAA. So the GNMA fund has less credit risk.




--
Mark Freeland

Re: vanguard VS.. fidelity GNMA

Another option:
Skip the GNMA fund altogether and start an investment grade bond (and/or CD)
ladder in your Fidelity IRA. Go out five years, rungs six months apart, and
the yield will probably be competitive with FGMNX now and then gradually
rise, eventually plateauing. Dump the interest in Fidelity's municipal money
market fund FTEXX, currently yielding about 2.4% and also on the rise.

The NAVs of investment grade bond funds like FGMNX are on the high side now.
When they come down, start transferring the rungs of the bond ladder to it.


<> wrote
> I'm trying to decide between starting to invest in vanguard gnma in a
> taxable account OR fidelity gnma in IRA account. I know it sounds like
> a bizarre question, but....

Re: vanguard VS.. fidelity GNMA

Elle wrote:
>
> Another option:
> Skip the GNMA fund altogether and start an investment grade bond
> (and/or CD) ladder in your Fidelity IRA.
^^^^^^^^^^^^^^^^^^^^^
> [...]
> Dump the interest in Fidelity's municipal money
> market fund FTEXX, currently yielding about 2.4% and also on the rise.

I hope Fidelity is a little smarter than this, and prohibits investors
from dumping interest into a muni fund in their IRA.


> The NAVs of investment grade bond funds like FGMNX are on the high
> side now. When they come down, start transferring the rungs of the
> bond ladder to it.

The NAV of a mutual fund is meaningless (e.g. several ETFs are about to
split 2-1 or 3-1, reducing their NAVs - this doesn't make them
"cheaper", or "better buys":
)

Perhaps you mean that the prices of the underlying securities are high.
But then why would you recommend purchasing those securities (investment
grade bonds) directly?

What are you trying to say here?

--
Mark Freeland

Re: vanguard VS.. fidelity GNMA

"Mark Freeland" <> wrote

> I hope Fidelity is a little smarter than this, and prohibits investors
> from dumping interest into a muni fund in their IRA.

That came to my mind as well. I went back and read the original post though:
"I'm trying to decide between starting to invest in vanguard gnma in a
taxable account OR fidelity gnma in IRA account."

The poster sounds undecided on taxable or IRA.

Re: vanguard VS.. fidelity GNMA

Ed wrote:
>
> "Mark Freeland" <> wrote
>
> > I hope Fidelity is a little smarter than this, and prohibits
> > investors from dumping interest into a muni fund in their IRA.
>
> That came to my mind as well. I went back and read the original post
> though:
> "I'm trying to decide between starting to invest in vanguard gnma in a
> taxable account OR fidelity gnma in IRA account."
>
> The poster sounds undecided on taxable or IRA.

The issue being whether the tax shelter benefit of an IRA investment at
Fidelity outweighed the extra cost of a Fidelity fund. The alternative
was having a mortgage-backed security investment in a lower expense,
taxable Vanguard fund account. There was no question that if the
account was at Fidelity, it was in an IRA (so one shouldn't even be
thinking about a Fidelity muni fund).

As to which is better, if one makes lots of "ideal world" assumptions
(few of which are likely true in reality):

- 15% tax bracket (OP said this applied),

- No change in tax rates in the future (neither due to income changes
nor due to government raising tax rates)

- No state income taxes

- All profit comes from interest (this is fairly close to true, since
the 10 year annualized return for Vanguard GNMA including tax on sale of
shares after 10 years is nearly identical to the 10 year annualized
return including taxes on distributions but not sale of shares),

- Annual return of Vanguard fund (before taxes) is 7% (annual over past
5 years is 6.84%, over last 10 is 6.85%, over last 25 is 8.87%),

- Every year has the same return (i.e. return is absolutely consistent,
year after year)

- Taxes paid out of distributions; the remainder is reinvested,

- Taxes paid on Dec. 31 (no time value of interest between Dec. 31 and
April 15th of next year),

- Fidelity fund return is exactly 40 basis points less (due to 40 basis
point difference in expenses)

- No annuitization; all assets (less taxes) remain in account, until
account is liquidated after N years,

then:
Total return of Vanguard fund = (1.07 - 15% * 7%) ^ N - 1
= 1.0595 ^ N - 1

Total return of Fidelity IRA = ((1.066) ^ N - 1) * 85%

Obviously if N = 1, then Vanguard comes out better, because it has the
higher yield, and both funds get taxed 15% on the one year gain.

But if N is high enough, the Fidelity IRA comes out ahead. This happens
around the 18th year. After 30 years, the Vanguard investment will be
worth only 4 2/3 the original amount, while the Fidelity IRA will be
worth (after liquidating and paying taxes) 4 15/16 the original amount,
i.e. the Fidelity IRA will be worth more by about 1/4 of the original
amount.

If the average rate of return is 7%, but it varies year by year, then I
believe that it will take longer for the Fidelity IRA to come out
ahead. (For instance, in the extreme case, suppose the investment is
flat for N-1 years, and all the gain comes in the last year. Then the
Vanguard fund, pre-tax, will be ahead (lower expenses), and it will be
taxed simply 15% on the total gain - the same as the tax on the Fidelity
IRA gain.)

This is the type of analysis that, to do correctly, would take more time
than I want to dedicate. As the assumptions are relaxed (to reflect
more of a real-world situation) the question becomes even harder to
answer.

--
Mark Freeland

Re: vanguard VS.. fidelity GNMA

"Mark Freeland" <> wrote

> The issue being whether the tax shelter benefit of an IRA investment at
> Fidelity outweighed the extra cost of a Fidelity fund. The alternative
> was having a mortgage-backed security investment in a lower expense,
> taxable Vanguard fund account. There was no question that if the
> account was at Fidelity, it was in an IRA (so one shouldn't even be
> thinking about a Fidelity muni fund).

Ok, he did say the Fidelity fund would be in the IRA.

About your comment "so one shouldn't even be thinking about a Fidelity muni
fund", I have to admit that I have thought about it before. I'm not sure if
Fidelity would allow it, I don't think so. While I agree that at first
glance it would not be wise to put a tax free investment into an IRA there
are periods, sometimes lengthy, that muni's outperform taxable bonds of
approximately equal risk. I think they could make sense in a Roth IRA.

Re: vanguard VS.. fidelity GNMA

thanks for the analysis. Now i know it wasn't the kind of question
that would have an obvious (off the top of the head) answer.

Mark Freeland wrote:
> Ed wrote:
> >
> > "Mark Freeland" <> wrote
> >
> > > I hope Fidelity is a little smarter than this, and prohibits
> > > investors from dumping interest into a muni fund in their IRA.
> >
> > That came to my mind as well. I went back and read the original post
> > though:
> > "I'm trying to decide between starting to invest in vanguard gnma in a
> > taxable account OR fidelity gnma in IRA account."
> >
> > The poster sounds undecided on taxable or IRA.
>
> The issue being whether the tax shelter benefit of an IRA investment at
> Fidelity outweighed the extra cost of a Fidelity fund. The alternative
> was having a mortgage-backed security investment in a lower expense,
> taxable Vanguard fund account. There was no question that if the
> account was at Fidelity, it was in an IRA (so one shouldn't even be
> thinking about a Fidelity muni fund).
>
> As to which is better, if one makes lots of "ideal world" assumptions
> (few of which are likely true in reality):
>
> - 15% tax bracket (OP said this applied),
>
> - No change in tax rates in the future (neither due to income changes
> nor due to government raising tax rates)
>
> - No state income taxes
>
> - All profit comes from interest (this is fairly close to true, since
> the 10 year annualized return for Vanguard GNMA including tax on sale of
> shares after 10 years is nearly identical to the 10 year annualized
> return including taxes on distributions but not sale of shares),
>
> - Annual return of Vanguard fund (before taxes) is 7% (annual over past
> 5 years is 6.84%, over last 10 is 6.85%, over last 25 is 8.87%),
>
> - Every year has the same return (i.e. return is absolutely consistent,
> year after year)
>
> - Taxes paid out of distributions; the remainder is reinvested,
>
> - Taxes paid on Dec. 31 (no time value of interest between Dec. 31 and
> April 15th of next year),
>
> - Fidelity fund return is exactly 40 basis points less (due to 40 basis
> point difference in expenses)
>
> - No annuitization; all assets (less taxes) remain in account, until
> account is liquidated after N years,
>
> then:
> Total return of Vanguard fund = (1.07 - 15% * 7%) ^ N - 1
> = 1.0595 ^ N - 1
>
> Total return of Fidelity IRA = ((1.066) ^ N - 1) * 85%
>
> Obviously if N = 1, then Vanguard comes out better, because it has the
> higher yield, and both funds get taxed 15% on the one year gain.
>
> But if N is high enough, the Fidelity IRA comes out ahead. This happens
> around the 18th year. After 30 years, the Vanguard investment will be
> worth only 4 2/3 the original amount, while the Fidelity IRA will be
> worth (after liquidating and paying taxes) 4 15/16 the original amount,
> i.e. the Fidelity IRA will be worth more by about 1/4 of the original
> amount.
>
> If the average rate of return is 7%, but it varies year by year, then I
> believe that it will take longer for the Fidelity IRA to come out
> ahead. (For instance, in the extreme case, suppose the investment is
> flat for N-1 years, and all the gain comes in the last year. Then the
> Vanguard fund, pre-tax, will be ahead (lower expenses), and it will be
> taxed simply 15% on the total gain - the same as the tax on the Fidelity
> IRA gain.)
>
> This is the type of analysis that, to do correctly, would take more time
> than I want to dedicate. As the assumptions are relaxed (to reflect
> more of a real-world situation) the question becomes even harder to
> answer.
>
> --
> Mark Freeland
>

Re: vanguard VS.. fidelity GNMA

"Mark Freeland" <> wrote
> Elle wrote:
> >
> > Another option:
> > Skip the GNMA fund altogether and start an investment grade bond
> > (and/or CD) ladder in your Fidelity IRA.
> ^^^^^^^^^^^^^^^^^^^^^
> > [...]
> > Dump the interest in Fidelity's municipal money
> > market fund FTEXX, currently yielding about 2.4% and also on the rise.
>
> I hope Fidelity is a little smarter than this, and prohibits investors
> from dumping interest into a muni fund in their IRA.

Where do you think many IRA stock dividends go?

Keep this in context, if you are able. A bond ladder has advantages over
investment grade bond funds, in the eyes of many experts now.

> > The NAVs of investment grade bond funds like FGMNX are on the high
> > side now. When they come down, start transferring the rungs of the
> > bond ladder to it.
>
> The NAV of a mutual fund is meaningless (e.g. several ETFs are about to
> split 2-1 or 3-1, reducing their NAVs - this doesn't make them
> "cheaper", or "better buys":
> )

That's ridiculous. The NAV of a mutual fund has meaning in ways similar to
individual stock prices. Online charting tools take into account stock
splits, etc.

> Perhaps you mean that the prices of the underlying securities are high.
> But then why would you recommend purchasing those securities (investment
> grade bonds) directly?

As we have discussed, but you don't seem to get, a bond/CD ladder (where the
bonds/CDs are held to maturity) preserves principal. Not necessarily so with
an IG bond fund.

Google for info on why so many experts are recommending bond ladders over
bond funds these days.

Re: vanguard VS.. fidelity GNMA

"Elle" <> wrote

>> > Dump the interest in Fidelity's municipal money
>> > market fund FTEXX, currently yielding about 2.4% and also on the rise.
>>
>> I hope Fidelity is a little smarter than this, and prohibits investors
>> from dumping interest into a muni fund in their IRA.
>
> Where do you think many IRA stock dividends go?
>
> Keep this in context, if you are able. A bond ladder has advantages over
> investment grade bond funds, in the eyes of many experts now.

What the hell does a bond ladder have to do with FTEXX?
Talk about keeping things in context, if you are able. You are not able.
You are the stereotype for why a woman will never be president.

Perhaps you 'misread' yourself above. You are so stupid that I can't even
believe it. You don't have a brother named Greg Hennessey do you?

Re: vanguard VS.. fidelity GNMA

"Mark Freeland" <> wrote

>> Dump the interest in Fidelity's municipal money
>> market fund FTEXX, currently yielding about 2.4% and also on the rise.
>
> I hope Fidelity is a little smarter than this, and prohibits investors
> from dumping interest into a muni fund in their IRA.

Fidelity makes the info on this fund unclear.
Minimum Initial Investment $ 5,000

Minimum Retirement $ 2,500

Retirement Accounts No


I'll go with the "no".

Re: vanguard VS.. fidelity GNMA

"Elle" <> wrote in message
news:Sp_ne.71$
> "Mark Freeland" <> wrote
> > Elle wrote:
> > >
> > > Another option:
> > > Skip the GNMA fund altogether and start an investment grade bond
> > > (and/or CD) ladder in your Fidelity IRA.
> > ^^^^^^^^^^^^^^^^^^^^^
> > > [...]
> > > Dump the interest in Fidelity's municipal money
> > > market fund FTEXX, currently yielding about 2.4% and also on the rise.
> >
> > I hope Fidelity is a little smarter than this, and prohibits investors
> > from dumping interest into a muni fund in their IRA.
>
> Where do you think many IRA stock dividends go?

Is this a serious question, or mere rhetoric?

Giving you the benefit of the doubt, where I think many IRA stock dividends
go is:
a) Back into the mutual funds within the IRAs that hold the dividend-paying
stocks (note the name of this group)
b) Into a cash equivalent account within the IRA

Stock dividends, unlike bond coupons, are smaller, typically less frequent,
and more erratic. Thus they are a relatively poor direct source for a cash
flow. (You should have stuck with bonds here.) Because of these
attributes, they will either generate too much income - which would be a
waste to draw out of an IRA - or not enough income, in which case the
retiree would have the headache of managing a separate cash flow to be
similarly erratic, in order to complement the dividend stream and produce a
combined smooth flow.

As a Fidelity booster, I'm sure you understand that this is one of the
problems that Fidelity is addressing with their Income Management
Account(SM). One can keep funds within an IRA, and draw on them as needed,
not as they haphazardly generate cash.

> Keep this in context, if you are able.

The poster asked about which GNMA fund to use, given various tax
consequences. The group is misc.invest.mutual-funds. Pretty clear context.
Check the subject line.

> A bond ladder has advantages over
> investment grade bond funds, in the eyes of many experts now.

Appeal to authority. Not a particularly sound rhetorical technique.

Be that as it may, since the context is GNMAs, and we do want to stay within
context, you must be talking about laddering GNMAs. Lousy idea. Mortgage
backed securities, like various other categories of bonds, can be great when
heavily diversified in a fund, but extremely high risk when held
individually. Negative convexity and all (prepayment risk).

Besides, how would you ladder bonds with nebulous durations?

> > > The NAVs of investment grade bond funds like FGMNX are on the high
> > > side now. When they come down, start transferring the rungs of the
> > > bond ladder to it.
> >
> > The NAV of a mutual fund is meaningless (e.g. several ETFs are about to
> > split 2-1 or 3-1, reducing their NAVs - this doesn't make them
> > "cheaper", or "better buys":
> > )
>
> That's ridiculous. The NAV of a mutual fund has meaning in ways similar to
> individual stock prices. Online charting tools take into account stock
> splits, etc.

The NAV of an open end fund exactly equals the book value of the investment
company (i.e. fund). So by definition, it cannot be high (or low). In
contrast, Graham noted that one should look for stocks priced below net
current asset value - an impossibility for an open ended fund (and you
seemed to get freaked out by the idea of buying an exchange traded fund
where that could actually happen - if it were a bond fund, you would still
be calling it overpriced).

With respect to online charting tools, we had a whole thread on the problems
with applying them to funds (tracking NAVs). Just another way in which the
meaning of the NAV of a mutual fund is dissimilar from individual stock
prices.

> > Perhaps you mean that the prices of the underlying securities are high.
> > But then why would you recommend purchasing those securities (investment
> > grade bonds) directly?
>
> As we have discussed, but you don't seem to get, a bond/CD ladder (where
the
> bonds/CDs are held to maturity) preserves principal. Not necessarily so
with
> an IG bond fund.

As we have discussed, but you don't seem to get, there are a half dozen
different forms of risk, and focusing on a single one to the exclusion of
others is not sound investing. People who cannot tolerate any risk to
principal shouldn't be investing. They should stick to their MMAs, their
CDs, their savings bonds.

> Google for info on why so many experts are recommending bond ladders over
> bond funds these days.

Ah yes, the appeal to authority again. Had you suggested researching
explanations (independent of source), then you might have been inviting
reasoned study. The keyword "expert" is the clear giveaway.

--
Mark Freeland

Re: vanguard VS.. fidelity GNMA

"Mark Freeland" <> wrote
E wrote
> > "Mark Freeland" <> wrote
> > > Elle wrote:
> > > >
> > > > Another option:
> > > > Skip the GNMA fund altogether and start an investment grade bond
> > > > (and/or CD) ladder in your Fidelity IRA.
> > > ^^^^^^^^^^^^^^^^^^^^^
> > > > [...]
> > > > Dump the interest in Fidelity's municipal money
> > > > market fund FTEXX, currently yielding about 2.4% and also on the
rise.
> > >
> > > I hope Fidelity is a little smarter than this, and prohibits investors
> > > from dumping interest into a muni fund in their IRA.
> >
> > Where do you think many IRA stock dividends go?
>
> Is this a serious question, or mere rhetoric?
>
> Giving you the benefit of the doubt, where I think many IRA stock
dividends
> go is:
> a) Back into the mutual funds within the IRAs that hold the
dividend-paying
> stocks (note the name of this group)
> b) Into a cash equivalent account within the IRA

What is a "cash equivalent account" Mark? Something that earns _no_ interest
or dividends?

Oh that's real smart. If anything should be prohibited, it should be this.

> Stock dividends, unlike bond coupons, are smaller, typically less
frequent,
> and more erratic.

You're just being ridiculous. Stock dividends typically occur 4 times a
year. Bonds and CDs may pay interest 12x or 2x a year or possibly some other
rate. The difference in frequency is not significant. And erratic? Again,
you're splitting micro-hairs.

> Thus they are a relatively poor direct source for a cash
> flow.

Dividends are a poor direct source for cash flow, huh... now I've heard it
all.

Have you any idea how many people live off the income from dividends?

snip junk
> > A bond ladder has advantages over
> > investment grade bond funds, in the eyes of many experts now.
>
> Appeal to authority. Not a particularly sound rhetorical technique.

I appealed to logic on this issue in the past. It was beyond your abilities.

Enough. You're dug into another untenable position.

Anyone interested in why bond ladders, as opposed to bond funds, may be a
good idea, just ask here, google, or ask at misc.invest.financial-plan ,
where this comes up often enough, and where indeed the consensus of the
regulars there is that there certainly are advantages of a bond ladder over
a bond fund in a rising interest rate environment.

Re: vanguard VS.. fidelity GNMA

"Elle" <> wrote in message
news:2j8oe.408$
> "Mark Freeland" <> wrote
> E [sic] wrote
> > > Where do you think many IRA stock dividends go?
> >
> > Is this a serious question, or mere rhetoric?
> >
> > Giving you the benefit of the doubt, where I think many IRA stock
> > dividends go is:
> > a) Back into the mutual funds within the IRAs that hold the
> > dividend-paying stocks (note the name of this group)
> > b) Into a cash equivalent account within the IRA
>
> What is a "cash equivalent account" Mark? Something that earns _no_
interest
> or dividends?

Only if you stick it under a mattress; and how you get that mattress into
the IRA, I'll never know.

"Many planners recommend that people in or near retirement stash between one
and four years' worth of retirment income in a cash or cash-equivalent
account."


Are these the authorities that you are asking me to trust (below)?

> Oh that's real smart. If anything should be prohibited, it should be this.

Pulling IRA money out of a cash equivalent account just to put it (net of
taxes) into into a lower yielding account? Absolutely, that should be
prohibited.

> > Stock dividends, unlike bond coupons, are smaller, typically less
> > frequent and more erratic.
>
> You're just being ridiculous. Stock dividends typically occur 4 times a
> year. Bonds and CDs may pay interest 12x or 2x a year or possibly some
other
> rate. The difference in frequency is not significant. And erratic? Again,
> you're splitting micro-hairs.

Correct as far as that goes. I overstated some items. Being primarily a
fund investor, I tend to think in terms of annual dividends.

> > Thus they are a relatively poor direct source for a cash
> > flow.
>
> Dividends are a poor direct source for cash flow, huh... now I've heard it
> all.
>
> Have you any idea how many people live off the income from dividends?

Note the word "direct", which you dropped from your rhetorical question.
Dividends are buffered in a cash equivalent account, and people live off
drawings from the accounts - a smooth cash flow, unlike the erratic arrival
of dividend checks. People who live directly from check to check are
typically not looking at dividend checks.

> > > A bond ladder has advantages over
> > > investment grade bond funds, in the eyes of many experts now.
> >
> > Appeal to authority. Not a particularly sound rhetorical technique.
>
> I appealed to logic on this issue in the past. It was beyond your
abilities.

You appealed to me to go beyond what Herb wrote. Where's the logic in that?

> Enough. You're dug into another untenable position.

"Another" one? What was the first? Oh, I remember. It was my assertion
that there's more to buying bonds than checking that they are "investment
grade". Specifically, I pointed out when you recommended GM SmartNotes (as
an examplar), that you weren't considering the risks inherent in (a)
illiquid bonds, and (b) BBB-rated bonds.

Since you like referring to "experts", here's Jaffe's column on that little
gem:

"This is a real concern for a lot of people, *most of whom should have
avoided GM's paper in the first place*. ... The SmartNotes ... can't be
redeemed before maturity and have a limited secondary market that could dry
up if GM gets into trouble. (By comparison, GMAC DemandNotes - a similar
product - have no secondary market whatsoever, which means note holders
could be in an even worse position than folks owning SmartNotes.)" Emphasis
added.



The position I took looks tenable from here.

Me, I've been thinking about some of the shorter term GM bonds. But I go
into an investment cognizant of the various risks, including reinvestment
risk, credit risk, interest rate risk, etc., and don't just rely on what S&P
says.

> Anyone interested in why bond ladders, as opposed to bond funds, may be a
> good idea, just ask here, google, or ask at misc.invest.financial-plan ,
> where this comes up often enough, and where indeed the consensus of the
> regulars there is that there certainly are advantages of a bond ladder
over
> a bond fund in a rising interest rate environment.

Do you have anything to add that you haven't already said? Perhaps a
numeric illustration, including underlying assumptions? Nah, too much work,
no one would put in that kind of effort :-).

Meanwhile, here's someone who at least put a few numbers into his
calculations:
"The Myth Behind Laddering"

--
Mark Freeland

Re: vanguard VS.. fidelity GNMA

"Mark Freeland" <> wrote

> Meanwhile, here's someone who at least put a few numbers into his
> calculations:
> "The Myth Behind Laddering"

Buying and holding longer term bonds should outperform a ladder made up of
shorter term bonds. This is important but only if those dollars will always
be allocated to bonds. I think the advantage of a bond ladder is that it
periodically frees up money and makes it available for investment and
spending choices of all types.

Anyway, my favorite part of the article:
Although we are not in the forecasting business, we believe that as long as
economists and investors are convinced that rates have nowhere to go but up,
we wouldn't be surprised to see them head even lower.

Profunds inverse bond funds anyone?

Re: vanguard VS.. fidelity GNMA

Mark Freeland wrote:
>
>

> The position I took looks tenable from here.
>
> Me, I've been thinking about some of the shorter term GM bonds. But I go
> into an investment cognizant of the various risks, including reinvestment
> risk, credit risk, interest rate risk, etc., and don't just rely on what S&P
> says.

I understand (but don't know) that the GMAC bonds may be better on a
risk-return basis (lower yield but much safer and have been hit with
the general downgrade). Ditto the Ford bonds.

If I remember correctly, default rates of junk bonds do not fall with
time (individual issues), which means a shorter term junk bond is a
better bet.

>

Re: vanguard VS.. fidelity GNMA

Reading that, I believe his argument against laddering rests on 2
principles (or 3):

1. long term fall in interest rates since 1980 (and also upward sloping
yield curves)

2. sufficient post consumption income to reinvest bond coupons
(basically a point about duration and reinvestment)

My conclusion is the suggested strategy works well in a bull market for
bonds, and for investors who expect to reinvest their cash flows from
bonds.

In a bear market for bonds, laddering (lower duration) is the preferred
strategy.

Do you agree?

Re: vanguard VS.. fidelity GNMA

darkness39 wrote:
>
>

That's one of their older columns (1999), that focuses on falling
rates. The later one (strat_laddering3.html) discusses rising rates
also.

> Reading that, I believe his argument against laddering rests on 2
> principles (or 3):
>
> 1. long term fall in interest rates since 1980 (and also upward sloping
> yield curves)

The first part (long term falling rates) is clearly not relevant if
rates are rising. The second part (normal, i.e. rising, yield curve) is
the key to his argument in the case of rising rates.

I believe his point is that if you buy short term bonds (as you would in
a ladder), the amount of interst you lose until the bond matures and you
reinvest at a presumably higher rate is not made up with the replacement
bond.

For example - if you buy a 5 year bond, and renew at five years, as
opposed to investing in a 10 year bond, the 5 year rates will have had
to rise twice the size of the gap just to get even. And you'll have to
get that jump with every run on the ladder.

"Some experts in the US vociferously argue that a strategy of buying and
holding long-term bonds will outperform a bond ladder consistently over
longer periods."


(That's for the benefit of people who feel that appealing to "expert
opinion" proves a point.)

Cycling through shorter term bonds (as done with a ladder) may reduce
liquidity risk, but if that doesn't have value to you (e.g. if you are
investing long term for retirement), then laddering simply forces you
into lower yielding bonds (assuming the yield curve is not inverted).

> [...]
>
> My conclusion is the suggested strategy works well in a bull market for
> bonds, and for investors who expect to reinvest their cash flows from
> bonds.
>
> In a bear market for bonds, laddering (lower duration) is the preferred
> strategy.
>
> Do you agree?

Laddering is similar to dollar cost averaging with a lump sum - people
will invest a fraction at a time because they are afraid of entering the
equity market at the wrong time. The expected return is less than
investing everything at once, since there is an upward bias to the stock
market. But the risk of investing all at once is higher as well.

Unless rates are rising so fast that the long bond investment the wrong
choice (see above), then you can improve expected return by bullet
investing for your target date (analagous to investing the lump sum at
once). If, on the other hand, you know that rates are rising too
quickly, then you are better off holding cash until the rate increases
slow.

Laddering is more a way of hedging bets than improving total return.

One last page of interest - PIMCO's article from a year ago titled "The
Limitations of Bond Maturity Ladders"


--
Mark Freeland