Another GNMA post

Discuss all general (i.e. non-personal) investing questions and issues, investing news, and theory.
Post Reply
re@51.5
Posts: 196
Joined: Thu Sep 02, 2010 7:51 am
Location: West Michigan

Another GNMA post

Post by re@51.5 » Fri Apr 27, 2012 11:10 am

In this thread, The Two Fund Portfolio
nisiprius wrote:GNMAs have a special risk factor not present in most bonds.
abuss368 wrote:GNMA are a narrow part of the bond market with additional risks that Vanguard discloses (i.e. prepayment risks).
pkcrafter wrote:GNMAs are riskier than other high quality bonds as has been noted by others.
TBM also has "prepayment risk".

Vanguard's "Mid-Term Treasury Grade" bond funds

(taxable funds as of 03/31/2012)

VFIIX (GNMA Fund Investor Shares)
"The fund is broadly diversified across the universe of GNMA mortgage-backed securities. The advisor seeks to mitigate prepayment risk by moderately adjusting the coupon or maturity structure in anticipation of interest rate changes. Historically the fund’s volatility has been lower relative to that of the Barclays Capital U.S. Aggregate Bond Index due to the fund’s short- to intermediate-term average duration. The fund’s investments in U.S. government agency securities do not prevent fluctuations in share price."
10yr Standard deviation 2.91% <<-- LOWEST
10yr Sharpe ratio 1.35 <<-- HIGHEST

VIPSX (Inflation-Protect Sec Inv)
"The fund’s long-term volatility is expected to be lower than that of conventional bond funds :?: , because prices of inflation-protected securities respond to changes in real interest rates, not to changes in expected inflation. Although the fund’s average maturity is intermediate-term, the fund’s lower sensitivity to changes in nominal interest rates is consistent with a shorter duration. Credit risk is low because the fund invests primarily in U.S. government and agency securities. The income component of returns is likely to fluctuate more than in conventional bond funds because interest payments are adjusted for inflation."
10yr Standard deviation 6.85% <<-- HIGHEST
10yr Sharpe ratio 0.79 <<-- LOWEST

VBIIX (Inter-Term Bond Index Inv)
"The fund’s risk profile is similar to that of the intermediate-term, investment-grade U.S. fixed income market. Interest rate risk and income risk are moderate, reflecting the fund’s intermediate duration. Credit risk is low because the fund purchases only bonds issued by the U.S. Treasury or by corporations whose securities are rated as investment-grade."
10yr Standard deviation 5.75%
10yr Sharpe ratio 0.86

VFITX (Inter-Term Treasury Inv)
"The fund is subject to moderate interest-rate risk, given its intermediate duration. The advisor will maintain intermediate duration, while making adjustments to duration based on factors such as interest rates and the yield curve. Credit default risk is very low because the fund owns primarily U.S. government and agency securities. The fund’s investments in U.S. Treasury or agency securities do not prevent fluctuations in share price."
10yr Standard deviation 5.22%
10yr Sharpe ratio 0.86

VBMFX (Total Bond Mkt Index Inv)
"The fund’s risk profile is similar to that of the broad, investment-grade U.S. fixed income market. Interest rate risk and income risk are moderate, reflecting the fund’s intermediate duration. Credit risk is low because the fund primarily purchases bonds issued by the U.S. Treasury or by corporations whose securities are rated as investment-grade. Additionally, to the extent that it invests in mortgage-backed securities, the fund is subject to moderate prepayment/call risk."
10yr Standard deviation 3.67%
10yr Sharpe ratio 1.01

Of these five bond funds, GNMA has LOWEST SD & HIGHEST Sharpe ratio.

Confused,
Mike
As Merton Miller, a Nobel laureate at the University of Chicago, puts it, "I'll never understand why they call bonds 'fixed' income."

User avatar
Taylor Larimore
Advisory Board
Posts: 27097
Joined: Tue Feb 27, 2007 8:09 pm
Location: Miami FL

Confused.

Post by Taylor Larimore » Fri Apr 27, 2012 11:21 am

Of these five bond funds, GNMA has LOWEST SD & HIGHEST Sharpe ratio.

Confused,
Mike
Hi Mike:

"Confused" You shouldn't be. Your figures show that ALL of these funds provided income and safety in a portfolio.

But don't forget: "Past performance does not forecast future performance."

Best wishes.
Taylor
"Simplicity is the master key to financial success." -- Jack Bogle

boglestan
Posts: 193
Joined: Wed Mar 30, 2011 1:52 pm

Re: Another GNMA post

Post by boglestan » Fri Apr 27, 2012 11:42 am

Of these five bond funds, GNMA has LOWEST SD & HIGHEST Sharpe ratio.
... for that one specific 10 year period.

winterescape
Posts: 261
Joined: Sun Apr 15, 2007 2:53 pm
Location: Upstate NY

Re: Another GNMA post

Post by winterescape » Fri Apr 27, 2012 2:48 pm

Yes, yet another GNMA “discussion”

1) Prepayment risk: When interest rates fall people refinance to capture a portion of this benefit. However, the cost to refinance limits this activity. Also people move all the time, I believe that 30 year mortgages are held on average for 12 years. All this is interesting but look at the performance of the GNMA fund during the extended falling interest rate environment we have experienced.

2) GNMA bonds are still backed by “the full faith and credit of the US government”. So the question of diversification?

User avatar
nisiprius
Advisory Board
Posts: 36012
Joined: Thu Jul 26, 2007 9:33 am
Location: The terrestrial, globular, planetary hunk of matter, flattened at the poles, is my abode.--O. Henry

Re: Another GNMA post

Post by nisiprius » Fri Apr 27, 2012 9:17 pm

re@51.5 wrote:TBM also has "prepayment risk".
Of course it does, because it holds mortgage-backed bonds, because they are a big part of the total bond market. But they only constitute about 30% of Total Bond Market, so the prepayment risk is proportionately less.

I agree that the GNMA fund has a reasonably long history and during that period of time it has outperformed Total Bond, slightly, and the extra risk that is theoretically there is not visible to the eyeball. (Much the same thing can be said about the PIMCO Total Return bond fund, an actively managed bond fund). Do you think you've found a case of extra return without extra risk? In both cases, I can understand someone saying "I see where the extra risk is and for reasons A, B, and C I am willing to take it." But it's dangerous if you can't see the risk, because it's more likely that you're overlooking it than that it really isn't there.

I don't think it's a big risk, but I do think Total Bond is the default choice and the person who wants to tilt has the burden of explaining why, and it should be a better explanation than "it has superior past performance."
Annual income twenty pounds, annual expenditure nineteen nineteen and six, result happiness; Annual income twenty pounds, annual expenditure twenty pounds ought and six, result misery.

winterescape
Posts: 261
Joined: Sun Apr 15, 2007 2:53 pm
Location: Upstate NY

Re: Another GNMA post

Post by winterescape » Sat Apr 28, 2012 12:01 pm

nisiprius wrote: Do you think you've found a case of extra return without extra risk?
Nisiprius, Please quantify the “increased risk” of GNMA

Credit risk:
GNMA =“backed by the full faith and credit of the US government”
TBM= Diversification, with 70% of the holdings backed by the US government

Interest rate risk:
Average duration GNMA = 4.0 years
Average duration TBM = 5.1 years

convexity of the GNMA bonds, considering the current interest rate?
The proportionality higher prepayment risk, again considering the current interest rate?

I would hypothesize that when we purchase the TBM we are accepting bond transaction and diversification expenses that are not of value in comparison to the GNMA fund.

Or maybe it is all explained by the negative convexity, in which case GNMA looks to me like a great bargain.

User avatar
nisiprius
Advisory Board
Posts: 36012
Joined: Thu Jul 26, 2007 9:33 am
Location: The terrestrial, globular, planetary hunk of matter, flattened at the poles, is my abode.--O. Henry

Re: Another GNMA post

Post by nisiprius » Sat Apr 28, 2012 2:01 pm

winterescape wrote:
nisiprius wrote: Do you think you've found a case of extra return without extra risk?
Nisiprius, Please quantify the “increased risk” of GNMA... Credit risk... Interest rate risk... convexity... [don't hold up]. I would hypothesize that when we purchase the TBM we are accepting bond transaction and diversification expenses that are not of value.
Look, I've said repeatedly that I don't think it amounts to much. At one extreme, at least I think it's an extreme, Larry Swedroe is so adamant about the extra risk that he doesn't think people should even buy Total Bond. I look at the growth chart and I say if there's any extra risk it should hasn't shown up visibly in the last thirty years. But you could say exactly the same thing about PIMCO Total Return.

It's hard for me to believe that GNMA bonds should be just plain better, why you there should be one simple well-defined category of bonds that just plumb outperforms the rest of the market, yet in 32 years nobody has cottoned on to that and bid up the price. I think it's much more likely that there's risk there even if I don't know what it is. And frankly that's the most dangerous situation--when you don't know what the risk is. But in this (just as with PIMCO Total Bond!) the outperformance isn't that much so likely the risk doesn't amount to much, either.

I think anyone who's been in VFIIX for a long time might well consider sticking with it, why not? And anyone who does think it's better, well, why not? There have been arguments for using a 100% Treasuries fund in place of Total Bond, and 100% Inflation Protected Securities instead of Total Bond. They'd all work fine. Intermediate-term investment-grade bonds are intermediate-term investment grade bonds.

What I guess I don't understand is why all of a sudden we've had four or five threads on GNMA bonds in just a couple of weeks. Are they the flavor-of-the-month now for some reason?
Annual income twenty pounds, annual expenditure nineteen nineteen and six, result happiness; Annual income twenty pounds, annual expenditure twenty pounds ought and six, result misery.

Saphomd
Posts: 165
Joined: Tue Apr 03, 2012 6:11 pm
Location: SF, CA

Re: Another GNMA post

Post by Saphomd » Sat Apr 28, 2012 2:32 pm

In another post, there was a nice chart posted by Nisiprius comparing GNMA and Total bond market index over 20-25 years, and both funds had similar returns during that period of time. In my opinion, just pick your favorite bond fund, and stay the course and you will be fine. :happy

winterescape
Posts: 261
Joined: Sun Apr 15, 2007 2:53 pm
Location: Upstate NY

Re: Another GNMA post

Post by winterescape » Sat Apr 28, 2012 4:05 pm

Actually I was agnostic between TBM and GNMA until 2008. Take a look at this five year comparison chart, check out in particular 2008. You may call it past performance, I might call it a “real world test during extreme market conditions”

http://quote.morningstar.com/fund/chart ... %2C0%22%7D

Not sure what has prompted the rash of recent threads, also I do not believe it has to be all or nothing. I would consider all of the bond funds that Taylor has posted the performance of as great candidate bond funds.

re@51.5
Posts: 196
Joined: Thu Sep 02, 2010 7:51 am
Location: West Michigan

Re: Confused.

Post by re@51.5 » Sun Apr 29, 2012 8:44 am

Taylor Larimore wrote:But don't forget: "Past performance does not forecast future performance."
John Bogle's Book, "Common Sense on Mutual Funds, 10th Edition"
page 127
Rule 4: Use Past Performance to Determine Consistency and Risk

... there is an important role that past performance can play in helping you to make your fund selections. While you should disregard a single aggregate number showing a fund's past long-term return, you can learn a great deal by studying the nature of its past returns. Above all, look for consistency."
page 128
Morningstar Mutual Funds makes these comparisons easy. It shows, in a simple chart, whether a fund was in the first, second, third, or forth quartile of its group during each of the preceding 12 years. The chart gives a fair reflection of both the consistency of a fund's policies and the relative success of its managers. For a fund to earn a top performance evaluation, it should have, in my opinion, at least six to nine years in the top two quartiles and no more than one or two years in the bottom quartile.
(need Premium Membership to see "PDF Report")

VFIIX (GNMA Fund Investor Shares) 12 years in top two quartiles, 0 3rd quartile, 0 in 4th <<-- Best
VIPSX (Inflation-Protect Sec Inv) 9 years in top two quartiles, 3 in 3rd quartile, 0 in 4th
VBIIX (Inter-Term Bond Index Inv) 8 years in top two quartiles, 2 in 3rd quartile, 2 in 4th
VFITX (Inter-Term Treasury Inv) 9 years in top two quartiles, 1 in 3rd quartile, 2 in 4th
VBMFX (Total Bond Mkt Index Inv) 7 years in top two quartiles, 3 in 3rd quartile, 2 in 4th

Capture Ratio as a Tool to Measure Investment Performance
We examine the concepts of up-capture and down-capture, two widely used statistics for measuring investment performance. Our research indicates that each one on their own is an ineffective tool to accurately describe investment performance. However, the use of the Capture Ratio, the relationship between up-capture and down-capture, provides a much more useful tool.
.
.
.
To summarize, these findings make a very compelling case for the use of the Capture Ratio as opposed to either up-capture or down-capture alone when analyzing investments. This ratio is also very useful in comparing investments with different absolute values for their up-capture and down-capture ratios, since it normalizes those values putting all of the investments on a common scale.

Capture Ratio is Upside Capture divided by Downside Capture.
Capture Ratio 5yr/15yr
VFIIX (GNMA Fund Investor Shares) 2.52 1.58 <<-- Best
VBIIX (Inter-Term Bond Index Inv) .73 .76
VFITX (Inter-Term Treasury Inv) .87 .80
VBMFX (Total Bond Mkt Index Inv) .96 .96

Downside Capture 1yr 3yr 5yr 10yr 15yr (as compared to the BarCap US Agg Bond TR USD)
VFIIX (GNMA Fund Investor Shares) -49.08 35.29 34.20 49.38 54.68 <<-- Best
VBIIX (Inter-Term Bond Index Inv) 224.95 193.98 195.13 172.74 167.82
VFITX (Inter-Term Treasury Inv) 138.93 202.09 138.54 139.76 144.68
VBMFX (Total Bond Mkt Index Inv) 139.00 111.71 103.48 102.65 102.25

Mike
As Merton Miller, a Nobel laureate at the University of Chicago, puts it, "I'll never understand why they call bonds 'fixed' income."

Valuethinker
Posts: 35015
Joined: Fri May 11, 2007 11:07 am

Re: Another GNMA post

Post by Valuethinker » Tue May 01, 2012 6:43 am

winterescape wrote:Actually I was agnostic between TBM and GNMA until 2008. Take a look at this five year comparison chart, check out in particular 2008. You may call it past performance, I might call it a “real world test during extreme market conditions”

http://quote.morningstar.com/fund/chart ... %2C0%22%7D

Not sure what has prompted the rash of recent threads, also I do not believe it has to be all or nothing. I would consider all of the bond funds that Taylor has posted the performance of as great candidate bond funds.
It's really important to understand what you are seeing:

- non US government bonds have had credit risk concerns especially in 08-09. GNMAs did not have that. You see that equity risk blow out during the financial crisis

- the prepayment/ extension risk on Mortgage Backed Securities has just not shown up. People cannot refi (prepayment) because of declining housing values and extension risk comes when interest rates *rise*

So the period just does not show you a period when MBS experienced their worst risks (the Freddie and Fannie thing did not affect GNMA bonds which are legally distinct and carry a 'full faith and credit' guarantee by the US Treasury)

Valuethinker
Posts: 35015
Joined: Fri May 11, 2007 11:07 am

Re: Another GNMA post

Post by Valuethinker » Tue May 01, 2012 6:45 am

nisiprius wrote:
winterescape wrote:
nisiprius wrote: Do you think you've found a case of extra return without extra risk?
Nisiprius, Please quantify the “increased risk” of GNMA... Credit risk... Interest rate risk... convexity... [don't hold up]. I would hypothesize that when we purchase the TBM we are accepting bond transaction and diversification expenses that are not of value.
Look, I've said repeatedly that I don't think it amounts to much. At one extreme, at least I think it's an extreme, Larry Swedroe is so adamant about the extra risk that he doesn't think people should even buy Total Bond. I look at the growth chart and I say if there's any extra risk it should hasn't shown up visibly in the last thirty years.
What is of actual concern is that the known higher risks have not materialized as higher returns. I would have expected GNMA funds to outperform US Treasuries by 0.5-1.0% a year over those 30 years.

Either:

- the risk is just less than I think
- we just haven't seen the risk yet given what has basically been a 30 year bull market in bonds
I think anyone who's been in VFIIX for a long time might well consider sticking with it, why not? And anyone who does think it's better, well, why not? There have been arguments for using a 100% Treasuries fund in place of Total Bond, and 100% Inflation Protected Securities instead of Total Bond. They'd all work fine. Intermediate-term investment-grade bonds are intermediate-term investment grade bonds.

What I guess I don't understand is why all of a sudden we've had four or five threads on GNMA bonds in just a couple of weeks. Are they the flavor-of-the-month now for some reason?
TIPS have empirically shown a lot greater volatility than theory or expectations at introduction would suggest. These should be the most boring of bonds (given guaranteed real returns) and they are not. I still recommend high weightings in them, but that volatility has to be taken into account.

Post Reply