Sec yield vs distributions

[hoops777]

I have noticed the distributions have been much higher than the sec yields over the the past year and longer.If you look you will also see the distributions have gone down every month on a consistent basis.If you look at the nav of say the total bond today and find the same nav 8 or 9months ago the distribution is still much smaller.Can someone explain how we should evaluate this.If in 3 years the nav of say the GNMA fund is down to 10 and the sec yield is up to 3.5 will the actual distribution be less than the sec yield in a rising interest rate environment?

[Occupier]

I have forgotten the exact formula that the SEC requires, but it looks backward a short period. Thus on things like TIPS it excludes things like future inflation based yield. So you should take it with a grain of salt just like things like P/E ratios, Book value, etc. They all can vary with the way they are calculated, but they still have meaning. Dave

[magellan]

Generally, distribution yield is just a tax/accounting fallacy and SEC yield is the much more relevant number. Distribution yield is mostly a historical artifact, having to do with the prices that a fund paid in the past for the bonds that it owns, coupled with IRS rules on how bond premium/discount amortization is handled w.r.t. taxes.

IMO, most people should ignore distribution yield and focus on SEC yield, except when considering tax implications and expected cash flows from a fund.

In a falling interest rate environment, many funds have a distribution yield that's higher than their SEC yield. For taxable investments, this is tax inefficient compared to a similar fund (in terms of duration and risk) that has a distribution yield that matches its SEC yield. In the fund with a distribution yield that's higher than the SEC yield, you're basically getting back part of your principal each month and getting taxed on it.

Jim

[YDNAL]

If in 3 years the nav of say the GNMA fund is down to 10 and the sec yield is up to 3.5 will the actual distribution be less than the sec yield in a rising interest rate environment?

Yes!

Distribution yield = holdings and coupons (regardless of age)

SEC yield = total expected annual yield for all holdings for the past 30 days assuming that each bond is held until maturity, and that all dividends are reinvested.
http://online.wsj.com/article/SB1000142 ... 46756.html

[hoops777]

I think it is time to take advantage of PenFed cd's.

[555]

What use is SEC yield if the fund does not hold the bonds to maturity? For example an intermediate fund will sell bonds when their maturity gets too short. Maybe SEC yield systematically underestimates the return of such a fund. Is that right?

[YDNAL]

What use is SEC yield if the fund does not hold the bonds to maturity?

You would never know, so speculating that a fund doesn't hold to maturity is just what is it is. SEC yield, based on last 30 days data, is as good as you can get to project future income (yield).

[555]

What use is SEC yield if the fund does not hold the bonds to maturity?

"You would never know, so speculating that a fund doesn't hold to maturity is just what is it is. SEC yield, based on last 30 days data, is as good as you can get to project future income (yield)."

You just took a partial quote and missed the point. Here's what I said.

What use is SEC yield if the fund does not hold the bonds to maturity? For example an intermediate fund will sell bonds when their maturity gets too short. Maybe SEC yield systematically underestimates the return of such a fund. Is that right?

I don't see where the speculation is. An intermediate (or long) term bond fund will sell (most) bonds when their maturity gets too short. That makes SEC yield a questionable estimate of projected future income.

[YDNAL]

Here's what I said.

What use is SEC yield if the fund does not hold the bonds to maturity? For example an intermediate fund will sell bonds when their maturity gets too short. Maybe SEC yield systematically underestimates the return of such a fund. Is that right?

I don't see where the speculation is. An intermediate (or long) term bond fund will sell (most) bonds when their maturity gets too short. That makes SEC yield a questionable estimate of projected future income.

Unless a Bond Fund experiences no growth in total assets, that is untrue, 555. For instance, 20-30 years to maturity become 10-20 years, become 5-10 years, become 3-5 years, etc. and replaced (not sold) in order to maintain the intended mix.

Distribution by maturity (% of fund) as of 12/31/2012
Total Bond Mkt Index Inv
Under 1 Year 1.8%
1 - 3 Years 26.4%
3 - 5 Years 29.5%
5 - 10 Years 28.0%
10 - 20 Years 4.1%
20 - 30 Years 9.8%
Over 30 Years 0.4%
Total 100.0%

Distribution by maturity (% of fund) as of 12/31/2012
Long-Term Bond Index
Under 1 Year 0.0%
1 - 3 Years 0.2%
3 - 5 Years 0.1%
5 - 10 Years 0.8%
10 - 20 Years 25.7%
20 - 30 Years 70.7%
Over 30 Years 2.5%
Total 100.0%

[555]

@ YDNAL
What exactly is your point?
Are you actually disagreeing with anything I said? If so, what part?

[YDNAL]

@ YDNAL
What exactly is your point?
Are you actually disagreeing with anything I said? If so, what part?

There is no "point"... In case you forgot your first post:

What use is SEC yield if the fund does not hold the bonds to maturity?

The fund, for the most part, has no specific need NOT to hold bonds to maturity (addressed below). "If the fund....." is conjecture.

For example an intermediate fund will sell bonds when their maturity gets too short.

A distribution to maturity for Vanguard VBMFX has been provided. There is no "too short" since a fund like VBMFX holds a full spectrum of maturities from <1 year to >30 years. If bond A with 20+ years to maturity slides to 10+ year to maturity, a new 20+ year bond is bought with new money (new assets or matured bonds).

Maybe SEC yield systematically underestimates the return of such a fund. Is that right?

No.

Time to move on.

[dkturner]

Here's what I said.

What use is SEC yield if the fund does not hold the bonds to maturity? For example an intermediate fund will sell bonds when their maturity gets too short. Maybe SEC yield systematically underestimates the return of such a fund. Is that right?

I don't see where the speculation is. An intermediate (or long) term bond fund will sell (most) bonds when their maturity gets too short. That makes SEC yield a questionable estimate of projected future income.

Unless a Bond Fund experiences no growth in total assets, that is untrue, 555. For instance, 20-30 years to maturity become 10-20 years, become 5-10 years, become 3-5 years, etc. and replaced (not sold) in order to maintain the intended mix.

Distribution by maturity (% of fund) as of 12/31/2012
Total Bond Mkt Index Inv
Under 1 Year 1.8%
1 - 3 Years 26.4%
3 - 5 Years 29.5%
5 - 10 Years 28.0%
10 - 20 Years 4.1%
20 - 30 Years 9.8%
Over 30 Years 0.4%
Total 100.0%

Distribution by maturity (% of fund) as of 12/31/2012
Long-Term Bond Index
Under 1 Year 0.0%
1 - 3 Years 0.2%
3 - 5 Years 0.1%
5 - 10 Years 0.8%
10 - 20 Years 25.7%
20 - 30 Years 70.7%
Over 30 Years 2.5%
Total 100.0%

Help me out here. Why would the manager of a long-term bond fund hold 20 or 30 year bonds to maturity? Wouldn't he unload them when they cease to be long-term and use the proceeds to buy more 20 to 30 year bonds? As 19 year bonds are replaced with 30 year bonds doesn't this operate to increase the effective yield above what it would be if the bonds were held to maturity?

[YDNAL]

Help me out here. Why would the manager of a long-term bond fund hold 20 or 30 year bonds to maturity? Wouldn't he unload them when they cease to be long-term and use the proceeds to buy more 20 to 30 year bonds? As 19 year bonds are replaced with 30 year bonds doesn't this operate to increase the effective yield above what it would be if the bonds were held to maturity?

A long-term Bond fund, as Vanguard VBLTX (Barclays US Long Gov/Cr Flt Adj Idx) holds the majority of assets in 20-30 year maturities. However, the Index also includes 10-20 year maturies and even 5-10 year maturities. Of course, and we are talking VBLTX (not VBMFX), there is certain need to get rid of anything getting much below 10-yr maturity in order to follow the Index. However, that doesn't mean there is widespread selling when 70% of assets held in 20-30 maturity move to 10-20 maturity (about 1/3 of assets). With regards to SEC yield, the topic in this thread, we must think in terms of materiality and not whether a 20 year, $100K specific bond is sold when it is 5 years to maturity.

My $0.02.

[ofcmetz]

What use is SEC yield if the fund does not hold the bonds to maturity? For example an intermediate fund will sell bonds when their maturity gets too short. Maybe SEC yield systematically underestimates the return of such a fund. Is that right?

I think it's usefulness is in telling you what you are getting in the here and now. I've read that the best predictor of future bond returns is what their yield is at time of purchase and the SEC yield gives us this.

When I look at the Intermediate-term Bond Index, I see that it holds mainly bonds which fall between the 5 and 10 year level.

Distribution by maturity (% of fund) as of 12/31/2012
Inter-Term Bond Index Adm
Under 1 Year 0.2%
1 - 3 Years. 0.2%
3 - 5 Years 2.7%
5 - 10 Years 96.9%
10 - 20 Years 0.0%
20 - 30 Years 0.0%
Over 30 Years 0.0%
Total 100.0%
Portfolio holdings

This is quite different from the Total Bond Market index fund and I think this is what 555 is referring too. I really don't know how it changes things. On a side not it seems that most of its holdings were purchased at a premium to par. (I think this is the right word.)

I don't see why the SEC yield would overestimate or underestimate the return of the fund. But it's helpful to understand that it is always changing as interest rates rise and fall. It's also helpful when comparing bond funds with one another.

[Akiva]

What use is SEC yield if the fund does not hold the bonds to maturity?

You would never know, so speculating that a fund doesn't hold to maturity is just what is it is. SEC yield, based on last 30 days data, is as good as you can get to project future income (yield).

The yield curve is a good predictor of what yields will be in the future. So if the yield curve is rising SEC yield should *on average* underestimate the returns to a fund that sells bonds before maturity.

[YDNAL]

What use is SEC yield if the fund does not hold the bonds to maturity?

You would never know, so speculating that a fund doesn't hold to maturity is just what is it is. SEC yield, based on last 30 days data, is as good as you can get to project future income (yield).

The yield curve is a good predictor of what yields will be in the future. So if the yield curve is rising SEC yield should *on average* underestimate the returns to a fund that sells bonds before maturity.

These conversations turn into a %!$$!#@ contest when "magnitude" is ignored. For instance, a fund with 26% of total assets in 10-20 year maturities and 70% in 20-30 year maturities should see no impactful effect (magnitude) from selling a bond here/there.

Lets see, these are top 20 holdings in Vanguard LT Index VBLTX.

United States Treasury Note/Bond 4.375% 11/15/2039 $120,225,000 $155,898,162
United States Treasury Note/Bond 4.250% 11/15/2040 $117,550,000 $149,637,624
United States Treasury Note/Bond 2.750% 08/15/2042 $129,235,000 $124,327,947
United States Treasury Note/Bond 4.375% 05/15/2040 $91,578,000 $118,807,803
United States Treasury Note/Bond 4.750% 02/15/2041 $86,205,000 $118,464,635
United States Treasury Note/Bond 4.375% 05/15/2041 $91,200,000 $118,431,408
United States Treasury Note/Bond 3.125% 11/15/2041 $112,730,000 $117,574,008
United States Treasury Note/Bond 4.625% 02/15/2040 $80,151,000 $107,928,131
United States Treasury Note/Bond 3.000% 05/15/2042 $105,130,000 $106,723,771
United States Treasury Note/Bond 4.250% 05/15/2039 $79,952,500 $101,677,193
United States Treasury Note/Bond 3.875% 08/15/2040 $66,970,000 $80,228,051
United States Treasury Note/Bond 5.250% 11/15/2028 $56,905,000 $78,475,409
United States Treasury Note/Bond 4.500% 08/15/2039 $57,831,200 $76,382,292
United States Treasury Note/Bond 6.875% 08/15/2025 $49,150,000 $75,575,989
United States Treasury Note/Bond 6.125% 08/15/2029 $43,840,000 $66,294,410
United States Treasury Note/Bond 2.750% 11/15/2042 $64,060,000 $61,517,459
United States Treasury Note/Bond 6.250% 08/15/2023 $41,550,000 $59,507,495
United States Treasury Note/Bond 6.000% 02/15/2026 $36,755,000 $53,099,581
United States Treasury Note/Bond 3.125% 02/15/2042 $45,073,000 $46,960,657
United States Treasury Note/Bond 4.500% 05/15/2038 $34,900,000 $45,953,528

I see 1 Bond maturing 2023 and two 2025/26 yet everything else is beyond that. By the way, this fund holds 26% of assets in 10-20 year maturities and some even 5-10 year maturities. If we wish to think in terms of "what if there are sale of bonds;" then, what is the "impact" of selling the 2023 maturity (10-yr to mature) in 5 years?

Answer: nada, nil, ничто!

[Akiva]

What use is SEC yield if the fund does not hold the bonds to maturity?

You would never know, so speculating that a fund doesn't hold to maturity is just what is it is. SEC yield, based on last 30 days data, is as good as you can get to project future income (yield).

The yield curve is a good predictor of what yields will be in the future. So if the yield curve is rising SEC yield should *on average* underestimate the returns to a fund that sells bonds before maturity.

These conversations turn into a %!$$!#@ contest when "magnitude" is ignored. For instance, a fund with 26% of total assets in 10-20 year maturities and 70% in 20-30 year maturities should see no impactful effect (magnitude) from selling a bond here/there.

You are conveniently picking the long end of the yield curve to keep the magnitude low.

Let's take an example from the shorter end of the yield curve.

Right now, 7 year bonds are yielding 1.43%. 3 year bonds are yielding .44%. If you buy the 7 year bond now at face value and hold it for 4 years (and the yield curve doesn't move), you'll get the interest from the 1.43% yield AND you'll get the enormous capital appreciation as the yield falls to .44% while the coupon is locked in at the 1.43% rate. (If the 7 year bond is worth 100, a 3 year bond with the same coupon is worth 256.76). So this effect *can be* substantial.

Edit: I should have cleared the spreadsheet and rerun the numbers because that 256.76 figure was clearly wrong. As I say below once the error is pointed out, the correct price is 102.95, which adds 74 basis points per year to your returns. This is a 50% increase over the returns implied by the yield alone, which is a significant magnitude.

[YDNAL]

You are conveniently picking the long end of the yield curve to keep the magnitude low.

Let's take an example from the shorter end of the yield curve.

Right now, 7 year bonds are yielding 1.43%. 3 year bonds are yielding .44%. If you buy the 7 year bond now at face value and hold it for 4 years (and the yield curve doesn't move), you'll get the interest from the 1.43% yield AND you'll get the enormous capital appreciation as the yield falls to .44% while the coupon is locked in at the 1.43% rate. (If the 7 year bond is worth 100, a 3 year bond with the same coupon is worth 256.76). So this effect *can be* substantial.

No, not picking anything.

1. Apparently you seem to have not read previous posts when the "selling bonds and SEC effect" issue was brought-up by a poster discussing Intermediate (and Long term) Bond Indices.

2. The Long-term Index was brought-up - a second time - by yet a different poster.

Nothing further.

[Akiva]

You are conveniently picking the long end of the yield curve to keep the magnitude low.

Let's take an example from the shorter end of the yield curve.

Right now, 7 year bonds are yielding 1.43%. 3 year bonds are yielding .44%. If you buy the 7 year bond now at face value and hold it for 4 years (and the yield curve doesn't move), you'll get the interest from the 1.43% yield AND you'll get the enormous capital appreciation as the yield falls to .44% while the coupon is locked in at the 1.43% rate. (If the 7 year bond is worth 100, a 3 year bond with the same coupon is worth 256.76). So this effect *can be* substantial.

No, not picking anything.

1. Apparently you seem to have not read previous posts when the "selling bonds and SEC effect" issue was brought-up by a poster discussing Intermediate (and Long term) Bond Indices.

2. The Long-term Index was brought-up - a second time - by yet a different poster.

Nothing further.

I read the post where this came up and he asked about "intermediate" bonds, which typically have an average maturity of 5 years and range between 7 and 3, so the fund would buy the bond at 7 years and sell it at 3 just like my above example. So what I said holds, in situations where the yield curve is steep, the capital appreciation on the bonds as they "roll down" can be significant and would cause SEC yield to be too low.

[magellan]

Right now, 7 year bonds are yielding 1.43%. 3 year bonds are yielding .44%. If you buy the 7 year bond now at face value and hold it for 4 years (and the yield curve doesn't move), you'll get the interest from the 1.43% yield AND you'll get the enormous capital appreciation as the yield falls to .44% while the coupon is locked in at the 1.43% rate. (If the 7 year bond is worth 100, a 3 year bond with the same coupon is worth 256.76). So this effect *can be* substantial.

I must be missing something. I don't see how your 3 year bond could possibly be worth $256.76. I get $102.95 for a price on a $100 par bond three years before maturity with a coupon of 1.43% and a yield of .44%.

In excel with the analysis tool-pak addon I used:
=price(date(2013,2,14), date (2016,2,14), 1.43%, .44%, 100, 2)

Jim

[Akiva]

Right now, 7 year bonds are yielding 1.43%. 3 year bonds are yielding .44%. If you buy the 7 year bond now at face value and hold it for 4 years (and the yield curve doesn't move), you'll get the interest from the 1.43% yield AND you'll get the enormous capital appreciation as the yield falls to .44% while the coupon is locked in at the 1.43% rate. (If the 7 year bond is worth 100, a 3 year bond with the same coupon is worth 256.76). So this effect *can be* substantial.

I must be missing something. I don't see how your 3 year bond could possibly be worth $256.76. I get $104.31 for a price on a $100 par bond three years before maturity with a coupon of 1.53% and a yield of .44%.

In excel with the analysis tool-pak addon I used:
=price(date(2013,2,14), date (2017,2,14), 1.53%, .44%, 100, 2)

Jim

Yeah, I thought that was too high at the time, but couldn't figure out why it was giving me funny numbers. I reran it in a new sheet and my numbers match yours. Doing it with 1.43%, shows that if you bought it at 7 and sold it after 4 years, you'd be expected to get 102.95 for it. So it adds 74 basis points per year to your returns; i.e., your returns would be 50% higher than the yield alone would imply.

NB: The regression of future yields on current yields is very significant (statistically and economically), but the R^2 is low, so this effect only holds true on average over a large number of years and isn't a sure thing in any given year.

[magellan]

Doing it with 1.43%, shows that if you bought it at 7 and sold it after 4 years, you'd be expected to get 102.95 for it. So it adds 74 basis points per year to your returns.

That makes sense and yes, $102.95 is correct. You'll notice I fixed my previous post between the time you hit "quote" and the time you hit submit. I saw my error as soon as I posted, but not soon enough apparently.

Jim

[Akiva]

Doing it with 1.43%, shows that if you bought it at 7 and sold it after 4 years, you'd be expected to get 102.95 for it. So it adds 74 basis points per year to your returns.

That makes sense and yes, $102.95 is correct. You'll notice I fixed my previous post between the time you hit "quote" and the time you hit submit. I saw my error as soon as I posted, but not soon enough apparently.

Jim

Anyway, thanks for pointing out the error. I appreciate it.