Duration of Vanguards TBM?

Discuss all general (i.e. non-personal) investing questions and issues, investing news, and theory.

Duration of Vanguards TBM?

Postby Jay69 » Thu Jan 31, 2013 10:47 am

The following is from here http://www.bogleheads.org/forum/viewtopic.php?f=10&t=109780 I thought it would be best to start another thread on the subject.

I have been thinking about this for a few days now, whats the boards take on this, what is the real duration of the fund?


Doc wrote:
Jay69 wrote:One of those little pesky things I don't care for about Vanguards Total Bond Fund VBTLX/BND with its 5.2 years average duration is those pesky longer bonds.

10 - 20 Years 4.1%
20 - 30 Years 9.8%
Over 30 Years 0.4%

Overall will it make a difference, no clue?



Something is really really screwy here. Morningstar has as of 9/30/12 for VBTIX and 1/25/13 for AGG. Jay's number are from Vanguad as of 12/31/12.

Code: Select all
                  VBTIX    AGG
10-20     years   8.87%     9.55
20-30     years  29.41     34.19
over 30   years   5.07      2.83
Duration          4.97      4.59

AGG is iShares Core Total US Bond Market ETF AGG (Bar Cap Agg Index)

If the older M* data and the more current AGG data are correct the problem of the long bonds is even worse than you think.

In any case you could use IShares GVI - Bar Cap Intermediate (1-10) Gov/Credit - with a duration of only 3.85 years. You would have to add an MBS fund like GNMA if you wanted mrotgage backed. (I don't.)
"Out of clutter, find simplicity” Albert Einstein
User avatar
Jay69
 
Posts: 1801
Joined: Thu Feb 17, 2011 10:42 pm

Re: Duration of Vanguards TBM?

Postby rkhusky » Thu Jan 31, 2013 10:57 am

When I had a similar question, here was the response for the difference between Vanguard and Morningstar:

Langkawi wrote:Many of those bonds are VRDOs (variable rate demand obligations). These bonds have an interest rate that resets periodiocally. Each time it resets, the holder has the opportunity to redeem the bond to the issuer. Therefore, Vanguard considers the maturity of these bonds to be the time remaining until the next interest rate reset.


I have also seen this explanation:

dkturner wrote:Probably depends on how the person doing the calculating approaches mortgage pass through securities. One that doesn't have a firm grip on the concept of how a mortgage is amortized might classify a 30 year mortgage just like a 30 year Treasury bond, because that how time will pass before it goes away. When the guy in the next cubicle calculates the duration for this 30 year mortgage he tracks the actual cash flow and determines it has a duration of only, say, 7 years. We still haven't got to the guy in the third cubicle who adjusts the duration calculation for the impact of prepayment risk.
rkhusky
 
Posts: 1513
Joined: Thu Aug 18, 2011 9:09 pm

Re: Duration of Vanguards TBM?

Postby Jay69 » Thu Jan 31, 2013 11:37 am

The question still remains, what is the real duration today, I would be happy to know within a year or two. I can understand that its going to move around a some but from I see here its seems to move a fair bit.
"Out of clutter, find simplicity” Albert Einstein
User avatar
Jay69
 
Posts: 1801
Joined: Thu Feb 17, 2011 10:42 pm

Re: Duration of Vanguards TBM?

Postby Cut-Throat » Thu Jan 31, 2013 11:46 am

I'd believe what Vanguard has on their website.

Fixed income characteristics as of 12/31/2012
Total Bond Mkt Index Adm
Number of bonds 5583
Yield to maturity 1.7%
Average coupon 3.7%
Average maturity 7.1 years
Average duration 5.2 years
Short-term reserves 0.7% —
Fund total net assets $116.7 billion —
Share class total net assets $35.5 billion
User avatar
Cut-Throat
 
Posts: 1880
Joined: Sun Oct 17, 2010 10:46 am

Re: Duration of Vanguards TBM?

Postby Doc » Thu Jan 31, 2013 11:49 am

rkhusky wrote:When I had a similar question, here was the response for the difference between Vanguard and Morningstar:

Langkawi wrote:Many of those bonds are VRDOs (variable rate demand obligations). These bonds have an interest rate that resets periodiocally. Each time it resets, the holder has the opportunity to redeem the bond to the issuer. Therefore, Vanguard considers the maturity of these bonds to be the time remaining until the next interest rate reset.


I have also seen this explanation:

dkturner wrote:Probably depends on how the person doing the calculating approaches mortgage pass through securities. One that doesn't have a firm grip on the concept of how a mortgage is amortized might classify a 30 year mortgage just like a 30 year Treasury bond, because that how time will pass before it goes away. When the guy in the next cubicle calculates the duration for this 30 year mortgage he tracks the actual cash flow and determines it has a duration of only, say, 7 years. We still haven't got to the guy in the third cubicle who adjusts the duration calculation for the impact of prepayment risk.


I don't see a big difference: M* 4.97 as of 9/30/12 and Vg 5.2 as of 12/31/12.

IShares is showing 4.73 for its AGG ETF and 4.97 for the index as of 1/29/13 while M* has the duration as 4.59 a day earlier - 1/28/13. http://us.ishares.com/product_info/fund/fixedincome.htm

FWIW:

The Bar Cap Agg Index excludes variable rate unless rate(s) are fixed:

Coupon 􀂄 Must be fixed-rate.
􀂄 Step-up coupons and those that change according to a predetermined schedule are also included.


This does not negate Langkawi's explaination but casts some doubt on it's magnitude.

Re dk's comment: The duration calculation for MBS seems to be "left to the decision of the holder". :(
A scientist looks for THE answer to a problem, an engineer looks for AN answer. Investing is not a science.
User avatar
Doc
 
Posts: 4763
Joined: Sat Feb 24, 2007 2:10 pm

Re: Duration of Vanguards TBM?

Postby Jay69 » Thu Jan 31, 2013 12:09 pm

I did not word it very well I guess, I'm not liking if 35% of the bonds a are 20 years+ in duration.
"Out of clutter, find simplicity” Albert Einstein
User avatar
Jay69
 
Posts: 1801
Joined: Thu Feb 17, 2011 10:42 pm

Re: Duration of Vanguards TBM?

Postby Cut-Throat » Thu Jan 31, 2013 12:12 pm

Jay69 wrote:I did not word it very well I guess, I'm not liking if 35% of the bonds a are 20 years+ in duration.


Here again is what Vanguard has on their website.

Distribution by maturity (% of fund) as of 12/31/2012
Total Bond Mkt Index Adm
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%
User avatar
Cut-Throat
 
Posts: 1880
Joined: Sun Oct 17, 2010 10:46 am

Re: Duration of Vanguards TBM?

Postby Doc » Thu Jan 31, 2013 12:23 pm

Jay69 wrote:I did not word it very well I guess, I'm not liking if 35% of the bonds a are 20 years+ in duration.


I didn't even like it when it was only 20% plus ten years. :wink:

The Bar Cap Intermediate (1-10) Government/Credit Index has a duration of 3.82 years and "no" long bonds while the Aggregate Index has a duration of 4.97 and like you say some 35% long bonds.

http://us.ishares.com/product_info/fund/fixedincome.htm
A scientist looks for THE answer to a problem, an engineer looks for AN answer. Investing is not a science.
User avatar
Doc
 
Posts: 4763
Joined: Sat Feb 24, 2007 2:10 pm

Re: Duration of Vanguards TBM?

Postby YDNAL » Thu Jan 31, 2013 12:53 pm

Doc wrote:
Jay69 wrote:I did not word it very well I guess, I'm not liking if 35% of the bonds a are 20 years+ in duration.

I didn't even like it when it was only 20% plus ten years. :wink:

iShares Total US Bond ETF (AGG)
http://us.ishares.com/product_info/fund ... ew/AGG.htm
AGG wrote:Maturity as of 1/29/2013

0-1 Years 6.67%
1-5 Years 32.79%
5-10 Years 17.10%
10-15 Years 4.73%
15-20 Years 3.01%
20-25 Years 6.50%
25 Years and Over 27.21%

Average 6.50 years

Where/how is Vanguard coming up with maturities?
https://personal.vanguard.com/us/funds/ ... =INT#tab=2
VBMFX wrote:Distribution by maturity (% of fund) as of 12/31/2012

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%
Landy
Be yourself, everyone else is already taken -- Oscar Wilde
YDNAL
 
Posts: 13349
Joined: Tue Apr 10, 2007 5:04 pm
Location: Biscayne Bay

Re: Duration of Vanguards TBM?

Postby Doc » Thu Jan 31, 2013 2:02 pm

YDNAL wrote: Where/how is Vanguard coming up with maturities?
https://personal.vanguard.com/us/funds/ ... =INT#tab=2
VBMFX wrote:Distribution by maturity (% of fund) as of 12/31/2012

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%


There is a big discrepancy between M*'s VBMFX data as of 9/30/12 and Vg's website as of 12/31/12. Much more than can be explained by a three month lag in my opinion.

See the other thread in the OP.
A scientist looks for THE answer to a problem, an engineer looks for AN answer. Investing is not a science.
User avatar
Doc
 
Posts: 4763
Joined: Sat Feb 24, 2007 2:10 pm

Re: Duration of Vanguards TBM?

Postby Jay69 » Thu Jan 31, 2013 2:23 pm

Doc wrote:
There is a big discrepancy between M*'s VBMFX data as of 9/30/12 and Vg's website as of 12/31/12. Much more than can be explained by a three month lag in my opinion.

See the other thread in the OP.



This the piece I'm having a hard time with
"Out of clutter, find simplicity” Albert Einstein
User avatar
Jay69
 
Posts: 1801
Joined: Thu Feb 17, 2011 10:42 pm

Re: Duration of Vanguards TBM?

Postby YDNAL » Thu Jan 31, 2013 2:32 pm

Doc wrote:There is a big discrepancy between M*'s VBMFX data as of 9/30/12 and Vg's website as of 12/31/12. Much more than can be explained by a three month lag in my opinion.

See the other thread in the OP.

There's a big discrepancy every Quarter. This is Vanguard's Semi-Annual report on VBMFX as of 6/30/12 (not huge changes at 12/31/2012).
https://personal.vanguard.com/funds/rep ... 2210067894
VBMFX Prospectus wrote:Distribution by Maturity (% of portfolio)
Under 1 Year 3.7%
1 - 3 Years 24.4
3 - 5 Years 28.4
5 - 10 Years 29.3
10 - 20 Years 4.1
20 - 30 Years 9.8
Over 30 Years 0.3

That's why agreed with DK originally.
dkturner wrote:Probably depends on how the person doing the calculating approaches mortgage pass through securities. One that doesn't have a firm grip on the concept of how a mortgage is amortized might classify a 30 year mortgage just like a 30 year Treasury bond, because that how time will pass before it goes away. When the guy in the next cubicle calculates the duration for this 30 year mortgage he tracks the actual cash flow and determines it has a duration of only, say, 7 years. We still haven't got to the guy in the third cubicle who adjusts the duration calculation for the impact of prepayment risk.
Landy
Be yourself, everyone else is already taken -- Oscar Wilde
YDNAL
 
Posts: 13349
Joined: Tue Apr 10, 2007 5:04 pm
Location: Biscayne Bay

Re: Duration of Vanguards TBM?

Postby Doc » Thu Jan 31, 2013 7:13 pm

YDNAL wrote:
Doc wrote:There is a big discrepancy between M*'s VBMFX data as of 9/30/12 and Vg's website as of 12/31/12. Much more than can be explained by a three month lag in my opinion.

See the other thread in the OP.

There's a big discrepancy every Quarter. This is Vanguard's Semi-Annual report on VBMFX as of 6/30/12 (not huge changes at 12/31/2012).
https://personal.vanguard.com/funds/rep ... 2210067894
VBMFX Prospectus wrote:Distribution by Maturity (% of portfolio)
Under 1 Year 3.7%
1 - 3 Years 24.4
3 - 5 Years 28.4
5 - 10 Years 29.3
10 - 20 Years 4.1
20 - 30 Years 9.8
Over 30 Years 0.3

That's why agreed with DK originally.
dkturner wrote:Probably depends on how the person doing the calculating approaches mortgage pass through securities. One that doesn't have a firm grip on the concept of how a mortgage is amortized might classify a 30 year mortgage just like a 30 year Treasury bond, because that how time will pass before it goes away. When the guy in the next cubicle calculates the duration for this 30 year mortgage he tracks the actual cash flow and determines it has a duration of only, say, 7 years. We still haven't got to the guy in the third cubicle who adjusts the duration calculation for the impact of prepayment risk.


And in between these two Vg reports is Morningstar data that originates with Vanguard that shows something very much different. It is obviously too large a difference to be a change in portfolio. There is an apparent difference in the calculation or maybe just an error. But the differences in various fund portfolios brought up in the other thread is based on all M* data. I am in the believe that the various funds provide M* with the raw data i.e. the holdings as stated in the semi and annual reports. The duration question is unrelated. There are various definitions of duration and some firms use one and some another but they are relatively close except when you have MBS involved and then they can be very different. In the OP Jay use the term "duration" but he really meant maturity profiles. The differences in the funds and the time frames when looking at the Vg data are about the amount of bonds with 20+ year maturities not duration or even "effective" average maturity. I don't think there are even enough MBS in the portfolio to make up the differences we see. M* will have updated there data in the next few weeks and then at least we can eliminate the time frame problem.
A scientist looks for THE answer to a problem, an engineer looks for AN answer. Investing is not a science.
User avatar
Doc
 
Posts: 4763
Joined: Sat Feb 24, 2007 2:10 pm


Return to Investing - Theory, News & General

Who is online

Users browsing this forum: heyyou, james99, Kevin M, Yahoo [Bot] and 57 guests