VBMFX Duration

Have a question about your personal investments? No matter how simple or complex, you can ask it here.
Post Reply
Topic Author
antiqueman
Posts: 697
Joined: Thu Mar 12, 2009 5:22 pm

VBMFX Duration

Post by antiqueman »

I am considering moving all of my non TIPS investment into the Admiral Shares of VBMFX. The duration is 5.1 years.


I know that if interest rates rise 1 percent then the investment will decline about 5%. So if 100k is purchase and rates rise 2 % this would equate to about 12k dollar loss. Of course over some period of time interest payments on the fund would increase to offset the loss.


My question is , assuming a 5. 1 duration of the fund, if rates rise 1% will it take 5 years to break even or 5 years from the LAST interest rate increase, or neither. I am getting close to retirement and I am trying to think through how much hit I will take if rates rise.


Thanks.
kenner
Posts: 3128
Joined: Sat Mar 01, 2008 7:45 am

Re: VBMFX Duration

Post by kenner »

Duration is but one factor that comes into play when trying to predict future bond fund NAV declines when interest rates rise. Other factors include bond maturity dates in the fund's portfolio (how rapidly the fund replaces current securities), where on the yield curve rates are actually rising (Vanguard opined a few months back that rates will probably rise most on the short end), etc.

VBMFX captures what many investors feel is the "sweet spot" for balancing relative safety with overall return. If your investment time horizon is longer than 5 - 10 years, you should be fine.
Topic Author
antiqueman
Posts: 697
Joined: Thu Mar 12, 2009 5:22 pm

Re: VBMFX Duration

Post by antiqueman »

". If your investment time horizon is longer than 5 - 10 years, you should be fine."

Kenner, by your statement "if your investment time is longer than 5-10 years," do you mean before one starts to withdraw from the fund or if you just plan to always have VBMFX in my portfolio. I may withdraw in about 5 years, when I plan to retire.



Thanks.
kenner
Posts: 3128
Joined: Sat Mar 01, 2008 7:45 am

Re: VBMFX Duration

Post by kenner »

antiqueman wrote:".

Kenner, by your statement "if your investment time is longer than 5-10 years," do you mean before one starts to withdraw from the fund or if you just plan to always have VBMFX in my portfolio. I may withdraw in about 5 years, when I plan to retire.



Thanks.
I'm referring to how long you intend to hold the specific investment. Sounds like you may have other plans for the money in 5 years. If so, you might want to consider a bond fund that has a bit shorter duration, perhaps VG's Short Term Bond Index Fund. The two funds obviously have different risk/reward characteristics. Or maybe split the two 50/50. Only you know your risk tolerance, your eventual needs and whether preservation of capital trumps potential higher total return. You may not have a need to take even moderate risk of loss of capital.

If you plan to start withdrawals in 5 years, but keep the bond fund itself for many years, then VBMFX should be a very good fund for you. My AA calls for keeping VBMFX for the long haul, even after starting withdrawals.
User avatar
grabiner
Advisory Board
Posts: 35265
Joined: Tue Feb 20, 2007 10:58 pm
Location: Columbia, MD

Re: VBMFX Duration

Post by grabiner »

antiqueman wrote:I My question is , assuming a 5. 1 duration of the fund, if rates rise 1% will it take 5 years to break even or 5 years from the LAST interest rate increase, or neither. I am getting close to retirement and I am trying to think through how much hit I will take if rates rise.
If rates rise by 1%, it will take 5 years to get back to the same spot as if they had never changed. That is, if rates rise by 1% tomorrow and never change again, your returns will be the same as if rates were fixed. If rates stay the same for 4 years and then rise by 1%, then it will take 9 years to get the return you were originally expecting. Therefore, if your time horizon is less than 9 years, you should switch to shorter-term bonds after 4 years.

But as a long-term investor, your time horizon is the average of the time for your withdrawals. If you are 5 years from retirement, your time horizon is about 15 years, because you will be using some of the money every year from 5 to 25 if you live 20 years in retirement.
Wiki David Grabiner
Chipmaster
Posts: 24
Joined: Sun Jan 30, 2011 4:10 pm

Re: VBMFX Duration

Post by Chipmaster »

A really important point should be made on duration, that I don't see get made often enough at bogleheads. While it's units happen to be in years, interpreting it as anything more than the change in bond value relative to a change in interest rates is very misleading. Want a good example? There are certain bonds that have negative duration (inverse floaters come to mind). So when rates go up I made my money back a few months ago? What?

Furthermore, it should be stressed how inaccurate a measure duration is. Anything short of full blown scenario analysis will be only a weak approximation of what would happen when given a change in rates. There are several reasons why. Your duration approximation assumes a parallel shift in the yield curve. Furthermore, duration is an EXTREMELY poor measure of bond value change with any sizable change in interest rates. This is because the bond value vs interest rate graph is not linear - it's CONVEX (a term you've probably heard mentioned before). This means that our linear approximation of the slope of this curve is technically never 100% accurate, and differs largely from the curve when for any decent shock in interest rates - but I would only recommend doing this if you understand the pitfalls of duration.

So this begs the question, what good is duration? There are several other issues with duration, but that said if used appropriately, it is a useful tool in that it's easy to interpret, simple to calculate, and above all else, easy to communicate. So how should you interpret duration? All other things equal, a higher duration means a higher interest rate risk. Also, you can compute a rough approximation of how significant the value of your bonds can fluctuate to interest rate changes.

I see so often on this forum the teaching that duration is some exact time measure that you can understand some exact time period under when you'll get your money back if you suffer an unfortunate interest rate rise. Don't fall for this. This forum is great for teaching investing discipline and common sense, simple techniques, but if you want to learn about more complex investing calculations (which, frankly, this forum discourages as it's generally a sign that you're trying to outsmart the market), pick up a finance text book.

It would be nice if financial institutions chose to report other, less misleading metrics instead like DV01. Then we could avoid this confusion.
User avatar
Bustoff
Posts: 2033
Joined: Sat Mar 03, 2012 5:45 pm

Re: VBMFX Duration

Post by Bustoff »

grabiner wrote:
antiqueman wrote:I My question is , assuming a 5. 1 duration of the fund, if rates rise 1% will it take 5 years to break even or 5 years from the LAST interest rate increase, or neither. I am getting close to retirement and I am trying to think through how much hit I will take if rates rise.
If rates rise by 1%, it will take 5 years to get back to the same spot as if they had never changed. That is, if rates rise by 1% tomorrow and never change again, your returns will be the same as if rates were fixed. If rates stay the same for 4 years and then rise by 1%, then it will take 9 years to get the return you were originally expecting. Therefore, if your time horizon is less than 9 years, you should switch to shorter-term bonds after 4 years.

But as a long-term investor, your time horizon is the average of the time for your withdrawals. If you are 5 years from retirement, your time horizon is about 15 years, because you will be using some of the money every year from 5 to 25 if you live 20 years in retirement.
grabiner,
If I retired today with a million and invested the entire million in Total Bond Fund with the intention of living off the dividends, would I be able to rely on my annual dividends of $20,700 per year (@ 2.07% SEC yield) not decreasing in a rising rate environment ?
thanks,
bustoff
dbr
Posts: 46137
Joined: Sun Mar 04, 2007 8:50 am

Re: VBMFX Duration

Post by dbr »

grabiner wrote:
But as a long-term investor, your time horizon is the average of the time for your withdrawals. If you are 5 years from retirement, your time horizon is about 15 years, because you will be using some of the money every year from 5 to 25 if you live 20 years in retirement.
I think I would prefer to say that the concept of a time horizon does not apply to the long term investor. The reason I say that is that there is too much of a tendency of long term investors to get needlessly exercised over the issue of duration and "horizon." Also, of course, if one combines your figure with the mantra of choose duration to match your horizon, it would put all long term investors in long bonds, and I am not sure that is good advice.
steve_14
Posts: 1507
Joined: Wed Jun 20, 2012 12:05 am

Re: VBMFX Duration

Post by steve_14 »

Bustoff wrote:If I retired today with a million and invested the entire million in Total Bond Fund with the intention of living off the dividends, would I be able to rely on my annual dividends of $20,700 per year (@ 2.07% SEC yield) not decreasing in a rising rate environment ?
Old thread alert. Your nominal dividends will increase if rates rise, barring some unlikely event (defaults, higher fund costs). You'd be taking some serious inflation risk though.
User avatar
Taylor Larimore
Posts: 32839
Joined: Tue Feb 27, 2007 7:09 pm
Location: Miami FL

Bond myth.

Post by Taylor Larimore »

I know that if interest rates rise 1 percent then the investment will decline about 5%.
Antiqueman:

This often repeated statement is simply untrue. There are many periods in which inflation (and interest rates) went up and bond returns also went up.

Listed below are the historical U.S. inflation rates (CPI-U) and total returns for the Aggregate Bond Index since the Index's inception in 1976:

YEAR--INFLATION--BOND INDEX
1976-------4.9%--------15.6%
1977-------6.7-----------3.0
1978-------9.0-----------1.4
1979------13.3-----------1.9
1980------12.5-----------2.7
1980------12.5-----------2.7
1981-------8.9-----------6.3
1982-------3.8----------32.6
1983-------3.8-----------8.4
1984-------3.9----------15.2
1985-------3.8----------22.1
1986-------1.1----------15.2
1987-------4.4-----------2.8
1988-------4.4-----------7.9
1989-------4.6----------14.5
1990-------6.1-----------8.9
1991-------3.1----------16.0
1992-------2.9-----------7.4
1993-------2.7-----------9.7
1994-------2.7---------(-2.9)
1995-------2.5----------18.5
1996-------3.3-----------3.6
1997-------1.7-----------9.7
1998-------1.6-----------8.7
1999-------2.7---------(-0.8)
2000-------3.4----------11.6
2001-------1.6-----------8.4
2002-------2.4----------10.3
2003-------1.9-----------4.1
2004-------3.3-----------4.3
2005-------3.4-----------2.4
2006-------2.5-----------4.3
2007-------4.1-----------7.0
2008-------0.1-----------5.2
2009-------2.7-----------5.9
2010-------1.5-----------6.5
2011-------3.0-----------7.7
2012-------1.7-----------4.3
Source: U.S. Department of Labor and Barclays

Best wishes
Taylor
"Simplicity is the master key to financial success." -- Jack Bogle
User avatar
Bustoff
Posts: 2033
Joined: Sat Mar 03, 2012 5:45 pm

Re: VBMFX Duration

Post by Bustoff »

steve_14 wrote:
Bustoff wrote:If I retired today with a million and invested the entire million in Total Bond Fund with the intention of living off the dividends, would I be able to rely on my annual dividends of $20,700 per year (@ 2.07% SEC yield) not decreasing in a rising rate environment ?
Old thread alert. Your nominal dividends will increase if rates rise, barring some unlikely event (defaults, higher fund costs). You'd be taking some serious inflation risk though.
Thanks Steve.
For some reason I can't get my head around the back and forth regarding the danger of bond fund losses and the rising rate environment. I get it that it's possible to lock in losses selling a bond fund at a lower NAV than the original cost. On the other hand, the way the word losses is thrown around in the context of rising rates without regard to whether one holds or sells a bond fund gives the impression that losses will occur regardless of whether one holds or sells the fund. Isn't that like saying buying the Total Stock Index will result in losses during a period of falling stock prices.
steve_14
Posts: 1507
Joined: Wed Jun 20, 2012 12:05 am

Re: VBMFX Duration

Post by steve_14 »

Bustoff wrote:Thanks Steve.
For some reason I can't get my head around the back and forth regarding the danger of bond fund losses and the rising rate environment. I get it that it's possible to lock in losses selling a bond fund at a lower NAV than the original cost. On the other hand, the way the word losses is thrown around in the context of rising rates without regard to whether one holds or sells a bond fund gives the impression that losses will occur regardless of whether one holds or sells the fund. Isn't that like saying buying the Total Stock Index will result in losses during a period of falling stock prices.
Yes, stocks are where you should worry about NAV losses, not in short/int term low risk bond funds, assuming you're holding for the long term. Existing bond prices will generally rise as they approach maturity in the fund, and your div payment will increase as well.
dbr
Posts: 46137
Joined: Sun Mar 04, 2007 8:50 am

Re: VBMFX Duration

Post by dbr »

Bustoff wrote: For some reason I can't get my head around the back and forth regarding the danger of bond fund losses and the rising rate environment. I get it that it's possible to lock in losses selling a bond fund at a lower NAV than the original cost. On the other hand, the way the word losses is thrown around in the context of rising rates without regard to whether one holds or sells a bond fund gives the impression that losses will occur regardless of whether one holds or sells the fund. Isn't that like saying buying the Total Stock Index will result in losses during a period of falling stock prices.
The issue here is discussing on the one hand what is the present case and on the other hand what are the prospects for the future.

When one talks about "locking in losses" or, alternatively, "it is only paper losses" that is a conversation about prospects for the future. Having sold, one's prospects for the future no longer involve that investment; they involve some other investment where the money is now placed. Not having sold, one's prospects for the future do involve that investment. The conversation is really about the difference between two choices of the asset allocation going forward.

On the other hand, the conversation about the present case is very simple. The investment can be valued in the market today, and that is its value, pure and simple. That value is more,less or the same as the value at some time in the past. We can characterize that difference as a loss or a gain, again a pure and simple concept.

The question is which of these two conversations are you having. If you try to have both of them at once, then one can get very confused.
Don46
Posts: 198
Joined: Sun Feb 24, 2013 5:53 pm

Re: Bond myth.

Post by Don46 »

Taylor Larimore wrote:
I know that if interest rates rise 1 percent then the investment will decline about 5%.
Antiqueman:

This often repeated statement is simply untrue. There are many periods in which inflation (and interest rates) went up and bond returns also went up.

Listed below are the historical U.S. inflation rates (CPI-U) and total returns for the Aggregate Bond Index since the Index's inception in 1976:

YEAR--INFLATION--BOND INDEX
...
Taylor
I am confused, not well-versed in these matters, but I want to understand it better. I thought it was a hike in the Fed interest rate that posed the risk to bond prices, not the CPI per se. Are the Fed interest rates and the Consumer Price Index perfectly correlated? Can Taylor or someone help us on this?
I just put a bunch of my retirement funds into VBMFX and I want to understand this better.
Many thanks.
--Don
User avatar
grabiner
Advisory Board
Posts: 35265
Joined: Tue Feb 20, 2007 10:58 pm
Location: Columbia, MD

Re: VBMFX Duration

Post by grabiner »

Bustoff wrote:[If I retired today with a million and invested the entire million in Total Bond Fund with the intention of living off the dividends, would I be able to rely on my annual dividends of $20,700 per year (@ 2.07% SEC yield) not decreasing in a rising rate environment ?
Almost yes. The dividends would eventually increase, as rising rates allowed maturing bonds to be replaced with higher-yielding bonds. However, the distribution yield may not equal the SEC yield.

It's easiest to understand this with an individual bond. If you buy a 10-year bond for $10,000 with a 3% yield, you will receive $300 every year and get $10,000 back at maturity. You will always get that $300 dividend. If interest rates rise, the bond price may fall to $9000. Your distribution yield is then 3.33%, but the SEC yield will be higher because it reflects the fact that the bond, worth $9000 now, will pay $10,000 when it matures, so you will get that $1000 back.
When the bond matures in 10 years, you have $10,000 in cash, which you can use to buy a new bond; if that new bond has a 5% yield, you will receive $500 every year for the next 10 years.

The real risk of "living off the dividends" is the risk of inflation; the $20,700 which pays twelve months' rent on your apartment today might be only three months' rent in ten years.
Wiki David Grabiner
User avatar
Salty1
Posts: 89
Joined: Mon Apr 28, 2008 6:26 pm

Re: Bond myth.

Post by Salty1 »

Antiqueman:

Listed below are the historical U.S. inflation rates (CPI-U) and total returns for the Aggregate Bond Index since the Index's inception in 1976
with the addition of Effective Federal Funds Rate to Taylor's info. Effective Federal Funds data from http://research.stlouisfed.org/fred2/data/FEDFUNDS.txt
:

YEAR--INFLATION--BOND INDEX--Date----------Effective Federal Funds Rate
1976-------4.9%--------15.6%-------1977-01-01---4.61
1977-------6.7-----------3.0 --------1978-01-01---6.70
1978-------9.0-----------1.4 --------1979-01-01--10.07
1979------13.3-----------1.9 --------1980-01-01--13.82
1980------12.5-----------2.7 --------1981-01-01--19.08
1981-------8.9-----------6.3 --------1982-01-01--13.22
1982-------3.8----------32.6 --------1983-01-01---8.68
1983-------3.8-----------8.4 --------1984-01-01---9.56
1984-------3.9----------15.2 --------1985-01-01---8.35
1985-------3.8----------22.1 --------1986-01-01---8.14
1986-------1.1----------15.2 --------1987-01-01---6.43
1987-------4.4-----------2.8 --------1988-01-01---6.83
1988-------4.4-----------7.9 --------1989-01-01---9.12
1989-------4.6----------14.5 --------1990-01-01---8.23
1990-------6.1-----------8.9 --------1991-01-01---6.91
1991-------3.1----------16.0 --------1992-01-01---4.03
1992-------2.9-----------7.4 --------1993-01-01---3.02
1993-------2.7-----------9.7 --------1994-01-01---3.05
1994-------2.7---------(-2.9) -------1995-01-01---5.53
1995-------2.5----------18.5 -------1996-01-01---5.56
1996-------3.3-----------3.6 --------1997-01-01---5.25
1997-------1.7-----------9.7 --------1998-01-01---5.56
1998-------1.6-----------8.7 --------1999-01-01---4.63
1999-------2.7---------(-0.8) -------2000-01-01---5.45
2000-------3.4----------11.6 --------2001-01-01---5.98
2001-------1.6-----------8.4 --------2002-01-01---1.73
2002-------2.4----------10.3 --------2003-01-01---1.24
2003-------1.9-----------4.1 --------2004-01-01---1.00
2004-------3.3-----------4.3 --------2005-01-01---2.28
2005-------3.4-----------2.4 --------2006-01-01---4.29
2006-------2.5-----------4.3 --------2007-01-01---5.25
2007-------4.1-----------7.0 --------2008-01-01---3.94
2008-------0.1-----------5.2 --------2009-01-01---0.15
2009-------2.7-----------5.9 --------2010-01-01---0.11
2010-------1.5-----------6.5 --------2011-01-01---0.17
2011-------3.0-----------7.7 --------2012-01-01---0.08
2012-------1.7-----------4.3 --------2013-01-01---0.14

Source: U.S. Department of Labor and Barclays
Don46
Posts: 198
Joined: Sun Feb 24, 2013 5:53 pm

Re: VBMFX Duration

Post by Don46 »

I was able to put Salty1's data into an excel graph, after some manipulation, but I don't know how to post a graph on this site. Any suggestions?
User avatar
Salty1
Posts: 89
Joined: Mon Apr 28, 2008 6:26 pm

Re: VBMFX Duration

Post by Salty1 »

Don46 wrote:I was able to put Salty1's data into an excel graph, after some manipulation, but I don't know how to post a graph on this site. Any suggestions?
basically I believe you need to create the chart as an image and post it at a site like tinypic.com
see link below for some guidance
http://www.bogleheads.org/forum/posting ... 24#preview
linuxizer
Posts: 1783
Joined: Wed Jan 02, 2008 6:55 am

Re: VBMFX Duration

Post by linuxizer »

Taylor wrote: This often repeated statement is simply untrue. There are many periods in which inflation (and interest rates) went up and bond returns also went up.
"Untrue" is a bit strong here. It's true for the interest rate component. The Agg index includes credit risk as well, which was improving in those periods.
Chipmaster wrote: A really important point should be made on duration, that I don't see get made often enough at bogleheads. While it's units happen to be in years, interpreting it as anything more than the change in bond value relative to a change in interest rates is very misleading.
...
Furthermore, duration is an EXTREMELY poor measure of bond value change with any sizable change in interest rates.
...
This forum is great for teaching investing discipline and common sense, simple techniques, but if you want to learn about more complex investing calculations (which, frankly, this forum discourages as it's generally a sign that you're trying to outsmart the market), pick up a finance text book.
What is it about bonds that tends to bring out such passions in people? :-) And yes, guilty as charged.

I think the three points I've pulled out are all over-stated or inaccurate. Duration is a term higher in the Taylor Series expansion than convexity, so it's appropriate that it's talked about more as a summary statistic. You could equally argue that asset returns are not linear and that skew (third moment) or kurtosis (fourth moment) matter in addition to the commonly discussed mean (first moment) and standard deviation (second moment). But the first two moments suffice in general--that's the whole point of using a Taylor Series.

I've argued in the past (and even asked Vanguard for the data, which they were unable to provide) that bond funds should report convexity as a standard item alongside their duration and SEC yield. This would actually be to Vanguard's benefit, since from looking at past returns from unexpected events they clearly do not "juice" returns by buying lower convexity bonds to the same degree as others. So it has been discussed, ad nauseum if you go back a few years (geez I sound old now!).

Also, if you look at the bond pages in the wiki, you'll note plenty of in-depth discussion.
Don46
Posts: 198
Joined: Sun Feb 24, 2013 5:53 pm

Re: VBMFX Duration

Post by Don46 »

Here is a chart of the rates of Inflation, Bond Index, and Fed Rates, 1976 to 2012, which is based on the data posted earlier by antiqueman.

I didn't run any statistical correlation tests, but the graph lines seem to indicate bonds have a mind of their own from year to year.

Image
simple man
Posts: 185
Joined: Sun Nov 22, 2009 9:44 am

Re: VBMFX Duration

Post by simple man »

Thanks so much Taylor, Salty1 and Don46! I have really been struggling to stay the course - given the constant drumbeat that I will be losing 5% for each point rise in interest rates. This is good evidence that nothing is guaranteed in this respect, and that I should definitely stay the course with my AA. (which we always say is easier said than done). :beer
User avatar
Electron
Posts: 2656
Joined: Sat Mar 10, 2007 7:46 pm

Re: VBMFX Duration

Post by Electron »

antiqueman wrote:I know that if interest rates rise 1 percent then the investment will decline about 5%. So if 100k is purchase and rates rise 2 % this would equate to about 12k dollar loss. Of course over some period of time interest payments on the fund would increase to offset the loss.
Here is a thread and chart illustrating the Point of Indifference concept in bond funds.

http://www.bogleheads.org/forum/viewtop ... 0&t=120947
Enjoying the Outdoors
Post Reply