Why should bond duration = investment horizon?

Discuss all general (i.e. non-personal) investing questions and issues, investing news, and theory.
Post Reply
Chief_Engineer
Posts: 34
Joined: Wed Jan 15, 2014 9:32 pm

Why should bond duration = investment horizon?

Post by Chief_Engineer » Wed Sep 19, 2018 9:17 am

In a number of recent bond related threads I've seen people suggest that your bond portfolio should have a duration roughly equal to your investment horizon. However, my searching has not provided much rational as to why. So I thought I would ask here.

As I understand, bonds are for 'stability'. Short term funds will have volatile interest rates but the NAV will be relatively constant. Long term funds on the other hand will have interest rates that change very slowly, as a result the NAV will be more volatile in times of changing interest rates. If bonds are for stability, shouldn't I go for short term bonds where the NAV is stable and yields will track changing interest rates? I welcome your wisdom.

BogleMelon
Posts: 1514
Joined: Mon Feb 01, 2016 11:49 am

Re: Why should bond duration = investment horizon?

Post by BogleMelon » Wed Sep 19, 2018 9:20 am

This is my understanding. Other bond experts here could correct me if I am wrong or not accurate..
Bonds NAV are changing through the life of bond because of so many factors, but at the maturity date you should get what you have promised to get. So in other words, selling premature bonds involving NAV risk, while selling mature bonds doesn't involve that risk unless something went wrong with the business.
"One of the funny things about stock market, every time one is buying another is selling, and both think they are astute" - William Feather

bgf
Posts: 658
Joined: Fri Nov 10, 2017 9:35 am

Re: Why should bond duration = investment horizon?

Post by bgf » Wed Sep 19, 2018 9:33 am

the assumption is that you will be required to satisfy some liability at some particular future time. so, say you anticipate having a $50,000 liability 5 years from now. you purchase a bond that pays "x" and matures in 5 years. when that liability comes due in 5 years, you get back the principal and use it to satisfy the liability. along the way, you have collected coupon payments.

edit* if you are NOT purchasing bonds for this particular reason (liability matching), but instead you are purchasing bonds to decrease the overall volatility of your portfolio, the cost you are paying for the decrease in volatility is equal to the difference between your coupon payment and the return of your equity allocation over that time period. this may not be a cost at all (bonds outperform equities) or it may be incredibly costly (equities return far more than bonds).
Last edited by bgf on Wed Sep 19, 2018 9:43 am, edited 1 time in total.
“TE OCCIDERE POSSUNT SED TE EDERE NON POSSUNT NEFAS EST"

asif408
Posts: 1500
Joined: Sun Mar 02, 2014 8:34 am
Location: Florida

Re: Why should bond duration = investment horizon?

Post by asif408 » Wed Sep 19, 2018 9:43 am

I think it depends on what you mean by "stability". If you mean something that goes up when stocks are down, bonds and bond funds are more likely to do that, particularly high quality Treasuries, but no guarantees. If you mean by "stability" having money to spend at a certain point in the near future, individual T-bills and notes will do that, but bond funds won't.

The more compelling argument for short-term bonds and bond funds is that the difference in yields between T-bills and 30 year bonds is around 1%. 4 years ago that difference was 4%. So if you believe the yield represents a good approximation of what you will receive from the bond or bond fund, you don't gain much by extending maturities, and you could potentially lose a lot more going long. Short term bonds should be fine in either a deflationary or inflationary environment, whereas long bonds, while a great hedge in a deflationary environment, will almost certainly do worse in an inflationary environment.

User avatar
vineviz
Posts: 2310
Joined: Tue May 15, 2018 1:55 pm

Re: Why should bond duration = investment horizon?

Post by vineviz » Wed Sep 19, 2018 10:04 am

Chief_Engineer wrote:
Wed Sep 19, 2018 9:17 am
In a number of recent bond related threads I've seen people suggest that your bond portfolio should have a duration roughly equal to your investment horizon. However, my searching has not provided much rational as to why. So I thought I would ask here.

As I understand, bonds are for 'stability'. Short term funds will have volatile interest rates but the NAV will be relatively constant. Long term funds on the other hand will have interest rates that change very slowly, as a result the NAV will be more volatile in times of changing interest rates. If bonds are for stability, shouldn't I go for short term bonds where the NAV is stable and yields will track changing interest rates? I welcome your wisdom.
I'll work on a separate reply to the question in the title (why bond duration should roughly approximate the investment horizon) in a separate reply, but first let me address the premise that "bonds are for 'stability'".

This premise is, indeed, widely held and often repeated on this forum. There is an element of truth to it, but there's enough wrong with it to potentially lead an investor astray.

First, I think you should clearly draw a distinction between an investment portfolio (a combination of risky assets, constructed to provide the investor with a return commensurate with the risk it takes) and savings vehicles (like bank accounts, CDs, money market funds, etc.).

I would say that your savings would be your "emergency fund" plus any assets you expect to spend in the next three years or so. Most people would agree that your savings should be invested in very stable and safe accounts, ones where the risk of principal loss is low (and possibly where the account is insured by the FDIC).

What's left is your investment portfolio, the total risk of which should reflect your ability and willingness to endure market volatility (among other things). Setting aside emotional considerations, what matters is what happens to the portfolio as a whole not what happens to the individual pieces. This is important because of the way that portfolio diversification works: combining two very risky assets that are poorly (or even negatively) correlated can result in a portfolio that in total is less risky than a combination of two moderately risky assets.

An example: take the following funds (two stock funds, two bond funds) ranked in descending order of annualized standard deviation of monthly returns from 2007 to present:
  • IJS iShares S&P Small-Cap 600 Value ETF (StDev=19.11%)
  • VYM Vanguard High Dividend Yield ETF (StDev=14.01%)
  • TLH iShares 10-20 Year Treasury Bond ETF (StDev=8.74%)
  • SHV iShares Short Treasury Bond ETF (StDev=0.48%)
Combine the least volatile stock fund (VYM) with the least volatile bond fund (SHV) to form Portfolio 1. Combine the most volatile stock fund (IJS) with the most volatile bond fund (TLH) to form Portfolio 2. The result?

Image

Portfolio 2 (despite containing the riskiest assets) has the lowest volatility and the smallest drawdown (it lost only 17.69% during the subprime crisis virus 33.64% for the "safe" portfolio).

So while the role of bonds might be described as adding stability to a stock portfolio, the bonds themselves don't need to be (and probably shouldn't be) be the most stable bonds you can find.
"Far more money has been lost by investors preparing for corrections than has been lost in corrections themselves." ~~ Peter Lynch

rkhusky
Posts: 5903
Joined: Thu Aug 18, 2011 8:09 pm

Re: Why should bond duration = investment horizon?

Post by rkhusky » Wed Sep 19, 2018 10:07 am

When interest rates go up, the NAV of a bond fund goes down. The NAV will go down approximately an amount equal to the duration x interest rate change. You will recoup the loss in NAV in increased interest payments, if you hold the fund for a time approximately equal to the duration. Therefore, if you have to sell the fund in a time shorter than the duration, you will not get all the extra interest. That is why the duration of your bond fund should be less than the time you expect to hold the fund.

There is probably also some relation related to duration versus expected return for why long term bond funds are not usually recommended on Bogleheads. Perhaps the extra interest in that case does not follow the duration vs risk vs interest relation. I'm not sure on that.

User avatar
vineviz
Posts: 2310
Joined: Tue May 15, 2018 1:55 pm

Re: Why should bond duration = investment horizon?

Post by vineviz » Wed Sep 19, 2018 10:23 am

Chief_Engineer wrote:
Wed Sep 19, 2018 9:17 am
In a number of recent bond related threads I've seen people suggest that your bond portfolio should have a duration roughly equal to your investment horizon. However, my searching has not provided much rational as to why. So I thought I would ask here.
There are two main sources (aka factors) in bond returns: credit risk and term risk.

Credit risk refers to chance that the company or government to which you loaned your money will go bankrupt or otherwise fail to pay the promised interest and/or repay the principal of the bond.

Term risk refers to the chance that interest rates will change during the time you hold the bond. Interest rate changes will usually affect a bond’s value.

As with any investment, the more risk the investor is willing to bear the higher their expected return. Because long-term bonds are more exposed to term risk, by definition, than short-term bonds investors can over the long-run expect a higher return from long-term bonds than from short-term bonds. Like stocks, though, the value of the bonds will typically go up and down while the investor owns them.

What bond investors often overlook is that term risk actually has two important components, price risk (the gains/losses from changes in the bonds value due to interst rate changes) and coupon reinvestment risk (gains/losses from being able to reinvest interest and principal from the initial bonds at higher or lower rates in the future).

These risks move opposite to each other, and bond duration is an estimate of the time at which price gains/losses will be offset by coupon reinvestment gains/losses.

From http://analystnotes.com/ :
The duration gap is the difference between the Macaulay duration and the investment horizon.
  • When the investment horizon is greater than the Macaulay duration of the bond, coupon reinvestment risk dominates price risk. The investor's risk is to lower interest rates. The duration gap is negative.
  • When the investment horizon is equal to the Macaulay duration of the bond, coupon reinvestment risk offsets price risk. The investor is hedged against interest rate risk. The duration gap is zero.
  • When the investment horizon is less than the Macaulay duration of the bond, price risk dominates coupon reinvestment risk. The investor's risk is to higher interest rates. The duration gap is positive.
So the notion of having your average effective duration roughly match (ideally be slightly shorter than) your expected investment horizon is to keep these two countervailing risks in balance.

This is just another form of diversification, or ensuring your portfolio's risk are not unnecessarily concentrated.

It's my opinion that most investors, including many Bogleheads, overestimate the power of price risk and underestimate the compounding power of reinvestment risk. This leads to an overemphasis on short-term bonds in cases where they aren't necessarily appropriate. An investor who has just retired might very well need to have a good portion of their bond allocation in short-term bonds, but an investor who is still saving for a retirement that is still 25+ years away should certainly not be holding short-term bonds in their investment portfolio.
Last edited by vineviz on Wed Sep 19, 2018 10:30 am, edited 1 time in total.
"Far more money has been lost by investors preparing for corrections than has been lost in corrections themselves." ~~ Peter Lynch

Thesaints
Posts: 1664
Joined: Tue Jun 20, 2017 12:25 am

Re: Why should bond duration = investment horizon?

Post by Thesaints » Wed Sep 19, 2018 10:25 am

If bond funds did not existed and one had to use individual bonds, wouldnn’t make a lot of sense to pick maturity dates close to one’s time horizon ?

People who posted better results in downturns by using longer dated bonds are simply showing the effect of negative correlation between stocks and bonds, but today valuations of both classes are strongly dependent on monetary policies and it would be naive not to expect increased correlation instead.

User avatar
nisiprius
Advisory Board
Posts: 37060
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: Why should bond duration = investment horizon?

Post by nisiprius » Wed Sep 19, 2018 10:47 am

It's a rough rule-of-thumb relationship.

Following your comment, "If bonds are for stability, shouldn't I go for short term bonds where the NAV is stable and yields will track changing interest rates?" The NAV is only relatively stable, of course. I would ask, "according to that reasoning, why shouldn't you go for a money market mutual fund or a bank account, where the NAV is perfectly stable?"

The answer for me is that I am balancing risk and return. And that I feel that in the past, intermediate-term bond funds have offered noticeably higher return than cash-like or very-short-term bonds. I feel that in the context of a portfolio that contains any normal allocation to stocks, portfolio risk is dominated by the stocks and there's very little gain in stability by going shorter than intermediate.

The idea of matching bond duration to investment horizon is based on the fact that unlike stocks bonds have a maturity date. Stocks really do seem to have a sort of vague, loosey-goosey tendency to mean reversion, that can be expressed by saying that over holding periods of 20 years or so, the standard deviation--uncertainty of final value--of a stock holding is only about 3/4 of what it would be without mean reversion. But with bonds, you are talking about a contractual commitment. Given investment-grade bonds, the chance of default is very small, and the value is guaranteed to rise or fall to face value at maturity. With a bond, by legal contract, barring default, what goes down must come up--and on schedule.

In practice in a bond fund, if you just visually inspect the growth chart of something like Vanguard Total Bond, what you will see visually is that ups and downs of as much as 5% or so are superimposed on a steady upward loft generated by coupon interest payments from the bond fund. The ups and downs are market-based, and have the same mysteries as any other market, but they are limited in size.

What the bond duration = investment horizon is saying that, both in theory and pragmatically, the bond duration is about right to insure that the steady upward loft from interest payments will exceed that from market ups and downs.

Thus, if you look at Vanguard Total Bond, whose duration has varied but has typically been around six years, you won't find many places where the fund would have lost money if held for six years. (I don't think there are any, in fact).

Over the course of one year, yes. Check in the vicinity of 1994 or 2013. Not a huge loss, but a loss. Over six years, no.

Source
Image

So, hold for the duration and you are relatively safe in terms of dollar loss over that holding period. Hold for a much shorter period of time and you will encounter a moderate chance of having less money at the end than you start with, and the amount of that loss for an investment-grade intermediate-term bond fund could be on the order of 5-10%.

Further comparisons, which you can do for yourself, of the actual returns of intermediate-term bond funds, short-term bond funds, and money market funds should convince you that if you'd chosen to invest in a short-term bond fund instead of intermediate term in 2009 (the first posting I've noted in which someone said "interest rates can only go up, shouldn't we be shortening up duration), you would have foregone quite a lot of return.
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.

User avatar
vineviz
Posts: 2310
Joined: Tue May 15, 2018 1:55 pm

Re: Why should bond duration = investment horizon?

Post by vineviz » Wed Sep 19, 2018 11:38 am

nisiprius wrote:
Wed Sep 19, 2018 10:47 am
Thus, if you look at Vanguard Total Bond, whose duration has varied but has typically been around six years, you won't find many places where the fund would have lost money if held for six years. (I don't think there are any, in fact).
Going back to 1930, intermediate treasuries have had a positive nominal return in 100% of rolling five-year periods. They've had positive real returns in 70% of of rolling five-year periods.

Side note: on average, a constant maturity portfolio of intermediate bonds has historically had a total return over five years that is about 10-20% higher than the portfolio's initial YTM.
"Far more money has been lost by investors preparing for corrections than has been lost in corrections themselves." ~~ Peter Lynch

User avatar
Phineas J. Whoopee
Posts: 7699
Joined: Sun Dec 18, 2011 6:18 pm

Re: Why should bond duration = investment horizon?

Post by Phineas J. Whoopee » Wed Sep 19, 2018 1:54 pm

Thesaints wrote:
Wed Sep 19, 2018 10:25 am
If bond funds did not existed and one had to use individual bonds, wouldnn’t make a lot of sense to pick maturity dates close to one’s time horizon ?
...
Only if one has a known obligation at some fixed point in the future.

Many, not all, members have as one of their objectives funding retirement. If one is 60 today, and plans to retire at 65, it is not the case they will need everything in cash all at once.

The value of less-risky assets in a portfolio is to damp the wild fluctuations of more risky, higher expected but not by any means guaranteed return ones.

To continue with the example, if one expects to retire in five years, and has an unknown remaining lifetime as is the case for at least most of us, one's time horizon may be anywhere from five years to fifty years (assuming medical advances).

Performance of the portfolio as a whole, structured to take account of the financial risks its owner faces, is what's important.

The proposition in the thread title is simply not true for assets to be drawn from over the long term.

PJW

aristotelian
Posts: 4902
Joined: Wed Jan 11, 2017 8:05 pm

Re: Why should bond duration = investment horizon?

Post by aristotelian » Wed Sep 19, 2018 2:01 pm

The idea is that if interest rates increase, the price will drop but the yield will increase. If you sell prematurely, you could take a loss. If your horizon is long enough to hold through the fund duration, the higher yield will eventually compensate for the lower price.

ThrustVectoring
Posts: 583
Joined: Wed Jul 12, 2017 2:51 pm

Re: Why should bond duration = investment horizon?

Post by ThrustVectoring » Wed Sep 19, 2018 2:36 pm

vineviz wrote:
Wed Sep 19, 2018 10:23 am
an investor who is still saving for a retirement that is still 25+ years away should certainly not be holding short-term bonds in their investment portfolio.
That investor should probably not be holding any bonds. The higher CAGR of equity investments over long enough timeframes dwarfs the short-term volatility of equities; over 30 years, even at 5th percentile market returns, equities still come out ahead of long-term bonds.

IMO, the correct way to do bonds is to do liability matching with a 10-year TIPS a decade before your planned retirement date. That's my desired retirement portfolio: 18 years of expenses in stocks, 10 in a TIPS ladder, 2 in cash. Refill the cash account from either selling stocks or not rolling the TIPS ladder periodically, depending on whether the stock account is above or below 18x expenses.
Current portfolio: 60% VTI / 40% VXUS

User avatar
vineviz
Posts: 2310
Joined: Tue May 15, 2018 1:55 pm

Re: Why should bond duration = investment horizon?

Post by vineviz » Wed Sep 19, 2018 2:56 pm

ThrustVectoring wrote:
Wed Sep 19, 2018 2:36 pm
vineviz wrote:
Wed Sep 19, 2018 10:23 am
an investor who is still saving for a retirement that is still 25+ years away should certainly not be holding short-term bonds in their investment portfolio.
That investor should probably not be holding any bonds. The higher CAGR of equity investments over long enough timeframes dwarfs the short-term volatility of equities; over 30 years, even at 5th percentile market returns, equities still come out ahead of long-term bonds.
I happen to agree with you, but we are in the distinct minority here.
"Far more money has been lost by investors preparing for corrections than has been lost in corrections themselves." ~~ Peter Lynch

User avatar
Phineas J. Whoopee
Posts: 7699
Joined: Sun Dec 18, 2011 6:18 pm

Re: Why should bond duration = investment horizon?

Post by Phineas J. Whoopee » Wed Sep 19, 2018 3:03 pm

vineviz wrote:
Wed Sep 19, 2018 2:56 pm
...
I happen to agree with you, but we are in the distinct minority here.
I agree it's the minority, and I fully understand the reasoning. I disagree with the conclusion because it makes assumptions about the individual's financial circumstances each and every year for the next many years. Nobody likes to contemplate a loss of employment-related income, but it can and does happen.

I don't mean one should live in fear. I merely opine that it's a risk a plan should account for.

Naturally there will exist sets of conditions which are not individually recoverable from, financially.

I've no doubt you fully understand my reasoning, too. We look at the same set of facts and draw different conclusions. That's the way it goes.

PJW

ThrustVectoring
Posts: 583
Joined: Wed Jul 12, 2017 2:51 pm

Re: Why should bond duration = investment horizon?

Post by ThrustVectoring » Wed Sep 19, 2018 3:15 pm

Phineas J. Whoopee wrote:
Wed Sep 19, 2018 3:03 pm
vineviz wrote:
Wed Sep 19, 2018 2:56 pm
...
I happen to agree with you, but we are in the distinct minority here.
I agree it's the minority, and I fully understand the reasoning. I disagree with the conclusion because it makes assumptions about the individual's financial circumstances each and every year for the next many years. Nobody likes to contemplate a loss of employment-related income, but it can and does happen.

I don't mean one should live in fear. I merely opine that it's a risk a plan should account for.

Naturally there will exist sets of conditions which are not individually recoverable from, financially.

I've no doubt you fully understand my reasoning, too. We look at the same set of facts and draw different conclusions. That's the way it goes.

PJW
Oh I agree with you there, I didn't say that they shouldn't be holding any fixed income products whatsoever, just that they shouldn't be investing in bonds as a percentage of their overall portfolio. Disability insurance to cover issues that would cause long-term unemployment, cash emergency fund to cover unexpected expenses/short-term unemployment, and then 100% equities afterwards until you're considering retiring in a decade.

Emergency cash should be in, uhh, cash. Or other short-term products. If you become unemployed you'll need to replace your next few months of paychecks. Not paychecks two to five years from now.
Current portfolio: 60% VTI / 40% VXUS

User avatar
patrick013
Posts: 2427
Joined: Mon Jul 13, 2015 7:49 pm

Re: Why should bond duration = investment horizon?

Post by patrick013 » Wed Sep 19, 2018 3:45 pm

I wouldn't use duration or investment horizon for any
decisions regarding bond investments. I still refer
to maturity, YTM, and target dating expected rate or
yield curve changes. Doing just fine doing just that.
I'd learn some bond strategies or buy TBM.

Duration isn't accurate enough for me especially when
the curve changes. The more nonlinear the relationship
between interest rates and bond price, the more inaccurate
duration is for measuring interest rate sensitivity. That
is happening right now.

I think duration is just journalistic verbiage it can be such a
joke.
age in bonds, buy-and-hold, 10 year business cycle

Thesaints
Posts: 1664
Joined: Tue Jun 20, 2017 12:25 am

Re: Why should bond duration = investment horizon?

Post by Thesaints » Wed Sep 19, 2018 3:48 pm

vineviz wrote:
Wed Sep 19, 2018 11:38 am
Side note: on average, a constant maturity portfolio of intermediate bonds has historically had a total return over five years that is about 10-20% higher than the portfolio's initial YTM.
Can you quote a source for that ?

Thesaints
Posts: 1664
Joined: Tue Jun 20, 2017 12:25 am

Re: Why should bond duration = investment horizon?

Post by Thesaints » Wed Sep 19, 2018 3:50 pm

Phineas J. Whoopee wrote:
Wed Sep 19, 2018 1:54 pm
Thesaints wrote:
Wed Sep 19, 2018 10:25 am
If bond funds did not existed and one had to use individual bonds, wouldnn’t make a lot of sense to pick maturity dates close to one’s time horizon ?
...
Only if one has a known obligation at some fixed point in the future.
Isn't that the definition of "time horizon" ?
Last edited by Thesaints on Wed Sep 19, 2018 3:57 pm, edited 1 time in total.

Thesaints
Posts: 1664
Joined: Tue Jun 20, 2017 12:25 am

Re: Why should bond duration = investment horizon?

Post by Thesaints » Wed Sep 19, 2018 3:57 pm

ThrustVectoring wrote:
Wed Sep 19, 2018 2:36 pm
vineviz wrote:
Wed Sep 19, 2018 10:23 am
an investor who is still saving for a retirement that is still 25+ years away should certainly not be holding short-term bonds in their investment portfolio.
That investor should probably not be holding any bonds. The higher CAGR of equity investments over long enough timeframes dwarfs the short-term volatility of equities; over 30 years, even at 5th percentile market returns, equities still come out ahead of long-term bonds.

IMO, the correct way to do bonds is to do liability matching with a 10-year TIPS a decade before your planned retirement date. That's my desired retirement portfolio: 18 years of expenses in stocks, 10 in a TIPS ladder, 2 in cash. Refill the cash account from either selling stocks or not rolling the TIPS ladder periodically, depending on whether the stock account is above or below 18x expenses.
So, until T-10 years one has to be 100% in stocks ? That's a whole lot of risk that not everybody is able to bear. Certainly not everybody on a psychological level.
As the short-term bonds are concerned, I agree that at first sight it could look odd to have 10% of ST bonds in a 70/30 portfolio with a 10 years time horizon. But what if the other 20% is made of high-yield bonds ?
Barbell strategies are not something new. After all it is a lot easier to assume intermediate risk by mixing riskless and very risky assets, than trying to correctly identify mid-range risks.

User avatar
bobcat2
Posts: 5247
Joined: Tue Feb 20, 2007 3:27 pm
Location: just barely Outside the Beltway

Re: Why should bond duration = investment horizon?

Post by bobcat2 » Wed Sep 19, 2018 4:04 pm

vineviz wrote:
Wed Sep 19, 2018 10:23 am
It's my opinion that most investors, including many [most] Bogleheads, overestimate the power of price risk and underestimate the compounding power of reinvestment risk. This leads to an overemphasis on short-term bonds in cases where they aren't necessarily appropriate. An investor who has just retired might very well need to have a good portion of their bond allocation in short-term bonds, but an investor who is still saving for a retirement that is still 25+ years away should certainly not be holding short-term bonds in their investment portfolio.
:thumbsup :thumbsup

BobK
In finance risk is defined as uncertainty that is consequential (nontrivial). | The two main methods of dealing with financial risk are the matching of assets to goals & diversifying.

User avatar
vineviz
Posts: 2310
Joined: Tue May 15, 2018 1:55 pm

Re: Why should bond duration = investment horizon?

Post by vineviz » Wed Sep 19, 2018 5:18 pm

patrick013 wrote:
Wed Sep 19, 2018 3:45 pm
I think duration is just journalistic verbiage it can be such a joke.
I think that bond investors, most of the time, have good reason to treat duration as the single most important descriptive measure of a bond or bond portfolio.

Is it the ONLY descriptive measure an investor should pay attention to? I'm certainly not suggesting that, and I don't think anyone else here is suggesting it either.

If you are trying to match the payments from specific bonds to specific liabilities, maturity and coupon rate are definitely things you should also be managing. And managing convexity in addition to duration can offer benefits, albeit with some additional portfolio complexity.

Duration has drawbacks, and these are pretty well understood and discussed in the academic and practitioner literature. That doesn't change the reality that duration represents an accurate measure of interest rate elasticity of the price of a bond.
"Far more money has been lost by investors preparing for corrections than has been lost in corrections themselves." ~~ Peter Lynch

User avatar
vineviz
Posts: 2310
Joined: Tue May 15, 2018 1:55 pm

Re: Why should bond duration = investment horizon?

Post by vineviz » Wed Sep 19, 2018 5:43 pm

Thesaints wrote:
Wed Sep 19, 2018 3:48 pm
vineviz wrote:
Wed Sep 19, 2018 11:38 am
Side note: on average, a constant maturity portfolio of intermediate bonds has historically had a total return over five years that is about 10-20% higher than the portfolio's initial YTM.
Can you quote a source for that ?
I had to run my own analysis, but it wasn’t complicated.

I’ll write it up when I have a chance.
"Far more money has been lost by investors preparing for corrections than has been lost in corrections themselves." ~~ Peter Lynch

Thesaints
Posts: 1664
Joined: Tue Jun 20, 2017 12:25 am

Re: Why should bond duration = investment horizon?

Post by Thesaints » Wed Sep 19, 2018 5:48 pm

vineviz wrote:
Wed Sep 19, 2018 5:43 pm
Thesaints wrote:
Wed Sep 19, 2018 3:48 pm
vineviz wrote:
Wed Sep 19, 2018 11:38 am
Side note: on average, a constant maturity portfolio of intermediate bonds has historically had a total return over five years that is about 10-20% higher than the portfolio's initial YTM.
Can you quote a source for that ?
I had to run my own analysis, but it wasn’t complicated.

I’ll write it up when I have a chance.
I asked because it sounds rather surprising and
https://www.wsj.com/articles/how-to-pre ... 78?tesla=y

User avatar
patrick013
Posts: 2427
Joined: Mon Jul 13, 2015 7:49 pm

Re: Why should bond duration = investment horizon?

Post by patrick013 » Wed Sep 19, 2018 6:04 pm

vineviz wrote:
Wed Sep 19, 2018 5:18 pm
patrick013 wrote:
Wed Sep 19, 2018 3:45 pm
I think duration is just journalistic verbiage it can be such a joke.
I think that bond investors, most of the time, have good reason to treat duration as the single most important descriptive measure of a bond or bond portfolio.
Why even bother with it. Textbook writers must need something
to fill the pages if you ask me. Can do the same with maturity
and YTM with less potential inaccuracy. Just put the numbers in
a yield to maturity calculator. Voila, the answer is there. Can
do the same with spread analysis even better predicting interest
rate fluctuations. All the literature about it just isn't even
needed. It doesn't benefit any strategy I use and I haven't had
a loss in a long long time. Every time somebody figures something
out somebody backtests it with duration is just plain distracting.

Sorry, but I can't see any practical use for it.
age in bonds, buy-and-hold, 10 year business cycle

Chief_Engineer
Posts: 34
Joined: Wed Jan 15, 2014 9:32 pm

Re: Why should bond duration = investment horizon?

Post by Chief_Engineer » Wed Sep 19, 2018 6:27 pm

OP, here. I want to thank everyone for their replies. I think I primarily have difficulty wrapping my head around bond funds. This is also complicated by the number of diverse opinions on Bogleheads regarding fixed income strategy.

Does the argument boil down to that there is a duration risk premium for longer term bonds, and for that reason you should pick bonds that mature exactly when you will need the money?

I understand stocks. I understand small and value tilt and why people do that. I don't think I have a good grasp of all of the different types of risk in bonds, and how they are associated with term. I guess this is why Total Bond Market exists.

MotoTrojan
Posts: 2613
Joined: Wed Feb 01, 2017 8:39 pm

Re: Why should bond duration = investment horizon?

Post by MotoTrojan » Wed Sep 19, 2018 6:37 pm

Chief_Engineer wrote:
Wed Sep 19, 2018 6:27 pm
OP, here. I want to thank everyone for their replies. I think I primarily have difficulty wrapping my head around bond funds. This is also complicated by the number of diverse opinions on Bogleheads regarding fixed income strategy.

Does the argument boil down to that there is a duration risk premium for longer term bonds, and for that reason you should pick bonds that mature exactly when you will need the money?

I understand stocks. I understand small and value tilt and why people do that. I don't think I have a good grasp of all of the different types of risk in bonds, and how they are associated with term. I guess this is why Total Bond Market exists.
Yes I believe you are correct, there is a risk premium for longer durations, just as there is for corporate, emerging market, or junk bonds.

I have never liked the saying in your OP title though as it is tough to implement. If I am 10 years from retirement, what duration do I want? I'll withdraw some assets in 10 years, but some others in 20, 30, etc... Do I need a ladder of bonds (excessive, I would never actually) that is shifting down a year in duration as I approach needing those assets? Of course I won't know how stocks will do and thus how many bonds I'll actually withdraw...

That is why I will eventually just go with Total Bond and not worry much beyond that (I currently hold zero). Only deviation that makes sense to me is tax-exempt bonds, if you are forced to invest them in taxable and are in a high bracket.

stlutz
Posts: 4839
Joined: Fri Jan 02, 2009 1:08 am

Re: Why should bond duration = investment horizon?

Post by stlutz » Wed Sep 19, 2018 7:14 pm

Thesaints wrote:
Wed Sep 19, 2018 3:48 pm
vineviz wrote:
Wed Sep 19, 2018 11:38 am
Side note: on average, a constant maturity portfolio of intermediate bonds has historically had a total return over five years that is about 10-20% higher than the portfolio's initial YTM.
Can you quote a source for that ?
Skip the source and just go to the logic of it. Suppose I have a rolling 10 year Treasury bond ladder. Right now the *average* yield of my portfolio would essentially equal what the 5 year note is yielding, which is 2.96%.

However, the return of a bond from issuance to redemption is determined by the yield when I purchase it, not the yield it might have along the way. A new 10 year bond yields 3.04%.

Suppose interest rates never change again. The annual return of my bond ladder would be 3.04%, not 2.96%.

Now that is not much of a different right now. But if I roll back the clock a few years, the 10 year was yielding about a half a percent more per year than the 5 year. It's those times that drive the result that vineviz refers to. No magic--just remember that the return of a bond you hold to maturity is the YTM at the time you buy it.

Thesaints
Posts: 1664
Joined: Tue Jun 20, 2017 12:25 am

Re: Why should bond duration = investment horizon?

Post by Thesaints » Wed Sep 19, 2018 7:17 pm

stlutz wrote:
Wed Sep 19, 2018 7:14 pm
Skip the source and just go to the logic of it. Suppose I have a rolling 10 year Treasury bond ladder. Right now the *average* yield of my portfolio would essentially equal what the 5 year note is yielding, which is 2.96%.

However, the return of a bond from issuance to redemption is determined by the yield when I purchase it, not the yield it might have along the way. A new 10 year bond yields 3.04%.

Suppose interest rates never change again. The annual return of my bond ladder would be 3.04%, not 2.96%.

Now that is not much of a different right now. But if I roll back the clock a few years, the 10 year was yielding about a half a percent more per year than the 5 year. It's those times that drive the result that vineviz refers to. No magic--just remember that the return of a bond you hold to maturity is the YTM at the time you buy it.
If you have a rolling 10 year ladder, your average yield would be the 10-year bond yield (assuming it remained constant during the time you built your ladder), not the 5-year bond yield. As you observe later, you only own 10-year bonds.
Problem is your initial YTM is also equal to the 10-year bond yield, so I don't see how rolling may increase return for equal average maturity. Rates oscillations may increase it, but also decrease it.

stlutz
Posts: 4839
Joined: Fri Jan 02, 2009 1:08 am

Re: Why should bond duration = investment horizon?

Post by stlutz » Wed Sep 19, 2018 7:30 pm

The preference for short-term bonds among many has been driven by those who lived through the high inflation periods of the 60s and 70s. If I compare 60/40 portfolios for the 1960 to 1990 period, short-term bonds proved to be the notably superior holding--giving you a portfolio with higher returns and lower volatility.

Conversely, if I compare the 1990-present period, the reverse was the case--longer term bonds gave you higher returns and less volatility.

Hence the preference among many (including me) to split the difference and take something in between (or just use a total bond market fund)--instead of betting on one scenario, pick something that will do reasonably well in all scenarios.

However, I think the interesting point that vineviz has made in various threads is that, on average, long-term bonds will outperform short-term bonds because of the term premium. And when bonds have done poorly, the return of a diversified portfolio with LT bonds wasn't *that* much worse than one with ST bonds. As such, why would you skip the chance to earn that premium? Is bond volatility really that bad?

User avatar
JoMoney
Posts: 6284
Joined: Tue Jul 23, 2013 5:31 am

Re: Why should bond duration = investment horizon?

Post by JoMoney » Wed Sep 19, 2018 7:37 pm

Personally, I want my bond fund duration to match the duration of a bond ladder I'd be willing to hold.
If I was willing to hold a bond ladder going out 1,2,3,4, and 5 years, the average duration of such a portfolio would be about 3 years.
Theoretically, it shouldn't make any difference between holding the portfolio of individual bonds myself, or a bond fund with similar duration, but the bond fund is easier to manage/reinvest/diversify.
"To achieve satisfactory investment results is easier than most people realize; to achieve superior results is harder than it looks." - Benjamin Graham

Thesaints
Posts: 1664
Joined: Tue Jun 20, 2017 12:25 am

Re: Why should bond duration = investment horizon?

Post by Thesaints » Wed Sep 19, 2018 7:37 pm

Since bond market and stock market are not perfectly uncorrelated, one opens the door to the possibility of stocks volatility and higher volatility of longer termed bonds to line up unfavorably.
It is perhaps a matter of avoiding risk tails, rather than just risk.

User avatar
patrick013
Posts: 2427
Joined: Mon Jul 13, 2015 7:49 pm

Re: Why should bond duration = investment horizon?

Post by patrick013 » Wed Sep 19, 2018 8:00 pm

stlutz wrote:
Wed Sep 19, 2018 7:30 pm
And when bonds have done poorly, the return of a diversified portfolio with LT bonds wasn't *that* much worse than one with ST bonds. As such, why would you skip the chance to earn that premium? Is bond volatility really that bad?
I think the numbers that portray that are based on prior coupons that were
much higher than today. And more recently a downward trend in interest
rates. Having some concern for rate levels, going long when rates are high
for example, should increase returns. Going long any time would not increase
returns then.
age in bonds, buy-and-hold, 10 year business cycle

User avatar
vineviz
Posts: 2310
Joined: Tue May 15, 2018 1:55 pm

Re: Why should bond duration = investment horizon?

Post by vineviz » Wed Sep 19, 2018 9:17 pm

Thesaints wrote:
Wed Sep 19, 2018 5:48 pm
I asked because it sounds rather surprising and
https://www.wsj.com/articles/how-to-pre ... 78?tesla=y
Okay, so I expanded the analysis I did earlier to include a longer time frame than I initially used (1953 to 2018). I looked at 5-year treasury bonds rather than 10-year bonds (as in the WSJ article), primarily because the Ibbotson total return data I have is for 5-year bonds. I could extrapolate a 10-year total return by blending 5-year and 20-year numbers, but I don't think that'd be entirely accurate due to shifting yield curves.

Image

Similar to the WSJ analysis, I found that entry 5-year treasury yield explained about 88% of the annualized total return for the subsequent 5-year period. On average, the total return was higher than the entry yield by about 50 bps.

Obviously this represents bigger percentage boost when entry yields are lowest: total returns were about 15% higher than entry yield when the entry yield is less than 6.0%, but only about 5% higher when entry yields were 6% or higher.
"Far more money has been lost by investors preparing for corrections than has been lost in corrections themselves." ~~ Peter Lynch

stlutz
Posts: 4839
Joined: Fri Jan 02, 2009 1:08 am

Re: Why should bond duration = investment horizon?

Post by stlutz » Wed Sep 19, 2018 9:18 pm

patrick013 wrote:
Wed Sep 19, 2018 8:00 pm


I think the numbers that portray that are based on prior coupons that were
much higher than today. And more recently a downward trend in interest
rates. Having some concern for rate levels, going long when rates are high
for example, should increase returns. Going long any time would not increase
returns then.
In the mid 50s, the 10 year treasury was ~3%, just like today. Looking at what happened over the next 30 years is actually a good comparison for what *could* happen to folks today.

If I had a 60/40 portfolio starting in 1956 and invested for 30 years (up to 1985), the one with short-term bonds returned 8.75% per year and the one with long-term bonds returned 8.00% per year. Standard deviation was 10.86% vs. 11.65%, respectively.

I don't know that this is the worst case scenario, but that is a real-life pretty darn bad-case scenario.

Note: Data calculated using the Siamond/Simba spreadsheet.

User avatar
vineviz
Posts: 2310
Joined: Tue May 15, 2018 1:55 pm

Re: Why should bond duration = investment horizon?

Post by vineviz » Wed Sep 19, 2018 9:19 pm

vineviz wrote:
Wed Sep 19, 2018 9:17 pm
Thesaints wrote:
Wed Sep 19, 2018 5:48 pm
I asked because it sounds rather surprising and
https://www.wsj.com/articles/how-to-pre ... 78?tesla=y
Okay, so I expanded the analysis I did earlier to include a longer time frame than I initially used (1953 to 2018). I looked at 5-year treasury bonds rather than 10-year bonds (as in the WSJ article), primarily because the Ibbotson total return data I have is for 5-year bonds. I could extrapolate a 10-year total return by blending 5-year and 20-year numbers, but I don't think that'd be entirely accurate due to shifting yield curves.

Image

Similar to the WSJ analysis, I found that entry 5-year treasury yield explained about 88% of the annualized total return for the subsequent 5-year period. On average, the total return was higher than the entry yield by about 50 bps. Furthermore, total return exceeded entry yield in the majority (over 65%) of five-year periods since 1953 and in over 85% of five-year periods since 1977.

Obviously this represents bigger percentage boost when entry yields are lowest: total returns were about 15% higher than entry yield when the entry yield is less than 6.0%, but only about 5% higher when entry yields were 6% or higher.
"Far more money has been lost by investors preparing for corrections than has been lost in corrections themselves." ~~ Peter Lynch

User avatar
JoMoney
Posts: 6284
Joined: Tue Jul 23, 2013 5:31 am

Re: Why should bond duration = investment horizon?

Post by JoMoney » Wed Sep 19, 2018 9:26 pm

stlutz wrote:
Wed Sep 19, 2018 9:18 pm
patrick013 wrote:
Wed Sep 19, 2018 8:00 pm


I think the numbers that portray that are based on prior coupons that were
much higher than today. And more recently a downward trend in interest
rates. Having some concern for rate levels, going long when rates are high
for example, should increase returns. Going long any time would not increase
returns then.
In the mid 50s, the 10 year treasury was ~3%, just like today. Looking at what happened over the next 30 years is actually a good comparison for what *could* happen to folks today.

If I had a 60/40 portfolio starting in 1956 and invested for 30 years (up to 1985), the one with short-term bonds returned 8.75% per year and the one with long-term bonds returned 8.00% per year. Standard deviation was 10.86% vs. 11.65%, respectively.

I don't know that this is the worst case scenario, but that is a real-life pretty darn bad-case scenario.

Note: Data calculated using the Siamond/Simba spreadsheet.
Here's a adjustable chart using the Ibbotson data for 30day TBills vs Long Term Government Bonds
Morningstar Growth Chart
That 1956-1985 was a particularly nasty period to be holding long term bonds. :shock:
"To achieve satisfactory investment results is easier than most people realize; to achieve superior results is harder than it looks." - Benjamin Graham

User avatar
vineviz
Posts: 2310
Joined: Tue May 15, 2018 1:55 pm

Re: Why should bond duration = investment horizon?

Post by vineviz » Wed Sep 19, 2018 9:38 pm

stlutz wrote:
Wed Sep 19, 2018 9:18 pm
In the mid 50s, the 10 year treasury was ~3%, just like today. Looking at what happened over the next 30 years is actually a good comparison for what *could* happen to folks today.

If I had a 60/40 portfolio starting in 1956 and invested for 30 years (up to 1985), the one with short-term bonds returned 8.75% per year and the one with long-term bonds returned 8.00% per year. Standard deviation was 10.86% vs. 11.65%, respectively.

I don't know that this is the worst case scenario, but that is a real-life pretty darn bad-case scenario.
A scenario like this is definitely a possibility in the future, but it's also important I think to remember that the Federal Reserve's monetary policy was fundamentally different before 1981 than it has been since then.

From the early 1950s until the early 1980s, the Fed was also actively trying to moderate long-term interest rates. Now, rather than setting interest rate targets it defines its objective primarily as managing inflation and unemployment.

In the periods since the 1980s during which interest rates were rising, long-term bonds have outperformed short-term bonds which is something of a reversal from previous times.
"Far more money has been lost by investors preparing for corrections than has been lost in corrections themselves." ~~ Peter Lynch

pascalwager
Posts: 1272
Joined: Mon Oct 31, 2011 8:36 pm

Re: Why should bond duration = investment horizon?

Post by pascalwager » Wed Sep 19, 2018 9:44 pm

Phineas J. Whoopee wrote:
Wed Sep 19, 2018 1:54 pm
Thesaints wrote:
Wed Sep 19, 2018 10:25 am
If bond funds did not existed and one had to use individual bonds, wouldnn’t make a lot of sense to pick maturity dates close to one’s time horizon ?
...
Only if one has a known obligation at some fixed point in the future.

Many, not all, members have as one of their objectives funding retirement. If one is 60 today, and plans to retire at 65, it is not the case they will need everything in cash all at once.

The value of less-risky assets in a portfolio is to damp the wild fluctuations of more risky, higher expected but not by any means guaranteed return ones.

To continue with the example, if one expects to retire in five years, and has an unknown remaining lifetime as is the case for at least most of us, one's time horizon may be anywhere from five years to fifty years (assuming medical advances).

Performance of the portfolio as a whole, structured to take account of the financial risks its owner faces, is what's important.

The proposition in the thread title is simply not true for assets to be drawn from over the long term.

PJW
It's true if one wants to assure return of principal when making redemptions. It's not true if you don't care about return of principal or critically need to pay living expenses in retirement.

garlandwhizzer
Posts: 2063
Joined: Fri Aug 06, 2010 3:42 pm

Re: Why should bond duration = investment horizon?

Post by garlandwhizzer » Thu Sep 20, 2018 11:42 am

asif408 wrote:
The more compelling argument for short-term bonds and bond funds is that the difference in yields between T-bills and 30 year bonds is around 1%. 4 years ago that difference was 4%. So if you believe the yield represents a good approximation of what you will receive from the bond or bond fund, you don't gain much by extending maturities, and you could potentially lose a lot more going long. Short term bonds should be fine in either a deflationary or inflationary environment, whereas long bonds, while a great hedge in a deflationary environment, will almost certainly do worse in an inflationary environment.
1+

Totally agree, average bond duration IMO should be adjusted to some degree from time to time based on bond market/macroeconomic background changes. Those who held long bonds in 1940 - 1982 got killed in real inflation adjusted returns regardless of how long their investment horizon was. Cumulative inflation over the increased duration of long bonds can destroy returns producing considerable real losses over decades. Bond investors get seduced by higher yields and stretch for increased duration and increased credit risk to get them which is usually a mistake. That is the illusion of safety not real safety. If you want real safety get TIPS and accept the ridiculously low real returns up front. Real safety ain't cheap.

Intermediate duration is the default position for average duration of a bond portfolio IMO because it hedges risk/return in both directions, never optimal but never the worst. There are times, like now, when I adjust average bond duration based on interest rate spreads, FED policy, and macroeconomic forces. Long bonds look good now on backtesting of 3 - 4 decades because until recently that was a persistent period of decreasing inflation and interest rates. Long bond principal values increased consistently for decades, juicing returns. That period may be behind us now. If you did the 3 - 4 decade backtesting in real dollar returns (the only kind of returns that actually matter) in 1982, long bonds would have been nothing less than a total and complete disaster. If you're going to use long bonds at all, which I don't do, I suggest either a barbell approach (long bonds plus short bonds) or a total bond market fund to reduce inflation/interest rate risk whatever your time horizon is. Just my 2 cents worth.

Garland Whizzer

boglehawk02
Posts: 1
Joined: Thu Sep 20, 2018 12:26 pm

Re: Why should bond duration = investment horizon?

Post by boglehawk02 » Thu Sep 20, 2018 12:36 pm

patrick013 wrote:
Wed Sep 19, 2018 6:04 pm
vineviz wrote:
Wed Sep 19, 2018 5:18 pm
patrick013 wrote:
Wed Sep 19, 2018 3:45 pm
I think duration is just journalistic verbiage it can be such a joke.
I think that bond investors, most of the time, have good reason to treat duration as the single most important descriptive measure of a bond or bond portfolio.
Why even bother with it. Textbook writers must need something
to fill the pages if you ask me. Can do the same with maturity
and YTM with less potential inaccuracy. Just put the numbers in
a yield to maturity calculator. Voila, the answer is there. Can
do the same with spread analysis even better predicting interest
rate fluctuations. All the literature about it just isn't even
needed. It doesn't benefit any strategy I use and I haven't had
a loss in a long long time. Every time somebody figures something
out somebody backtests it with duration is just plain distracting.

Sorry, but I can't see any practical use for it.
Please keep in mind not all textbooks are written with your investment strategy in mind. I can assure you as someone who helps manage institutional fixed income portfolios that have been structured to back very specific types of liabilities, duration (and the many versions thereof) is still one of the most important metrics in bond investing.

User avatar
patrick013
Posts: 2427
Joined: Mon Jul 13, 2015 7:49 pm

Re: Why should bond duration = investment horizon?

Post by patrick013 » Thu Sep 20, 2018 7:59 pm

boglehawk02 wrote:
Thu Sep 20, 2018 12:36 pm
patrick013 wrote:
Wed Sep 19, 2018 6:04 pm
Sorry, but I can't see any practical use for it.
Please keep in mind not all textbooks are written with your investment strategy in mind. I can assure you as someone who helps manage institutional fixed income portfolios that have been structured to back very specific types of liabilities, duration (and the many versions thereof) is still one of the most important metrics in bond investing.
Well the YTM calc does the same thing so duration isn't absolutely
needed. The Advisor manual would use maturity date matching
so that looks like the better advice to me.
age in bonds, buy-and-hold, 10 year business cycle

User avatar
vineviz
Posts: 2310
Joined: Tue May 15, 2018 1:55 pm

Re: Why should bond duration = investment horizon?

Post by vineviz » Thu Sep 20, 2018 8:50 pm

patrick013 wrote:
Thu Sep 20, 2018 7:59 pm
Well the YTM calc does the same thing so duration isn't absolutely needed.
This isn’t even a little bit true.
"Far more money has been lost by investors preparing for corrections than has been lost in corrections themselves." ~~ Peter Lynch

User avatar
patrick013
Posts: 2427
Joined: Mon Jul 13, 2015 7:49 pm

Re: Why should bond duration = investment horizon?

Post by patrick013 » Thu Sep 20, 2018 9:09 pm

vineviz wrote:
Thu Sep 20, 2018 8:50 pm
patrick013 wrote:
Thu Sep 20, 2018 7:59 pm
Well the YTM calc does the same thing so duration isn't absolutely needed.
This isn’t even a little bit true.
Relatively easy actually, to derive the bond price change percentage
related to a 1% change in yield using a YTM calculator.
age in bonds, buy-and-hold, 10 year business cycle

NoHeat
Posts: 224
Joined: Sun Sep 18, 2016 10:13 am

Re: Why should bond duration = investment horizon?

Post by NoHeat » Thu Sep 20, 2018 9:37 pm

The OP asked a good question, and vineviz answered it.

Beyond that, there are many unfortunate posts in this thread.

User avatar
patrick013
Posts: 2427
Joined: Mon Jul 13, 2015 7:49 pm

Re: Why should bond duration = investment horizon?

Post by patrick013 » Fri Sep 21, 2018 4:28 pm

vineviz wrote:
Wed Sep 19, 2018 9:17 pm
I looked at 5-year treasury bonds rather than 10-year bonds (as in the WSJ article), primarily because the Ibbotson total return data I have is for 5-year bonds. I could extrapolate a 10-year total return by blending 5-year and 20-year numbers, but I don't think that'd be entirely accurate due to shifting yield curves.

Image
This all the data that I have for TRSY 10. IRR's are average 70 bps
over the annual rate. I used annual portfolio return data from
portfolio visualizer for a 10 year treasury portfolio. I think the
structure is like a 10 year ladder.

Image
age in bonds, buy-and-hold, 10 year business cycle

Post Reply