Short- and longer-term effects of rising interest rates on a bond fund

[nisiprius]

Some investors are being overly spooked by a misunderstanding of a) the appropriate holding time for a bond fund, and b) the short-term versus long-term relationship of interest rates to bond fund behavior.

"Interest up, bonds down" is a short-term relationship. Vanguard says that the Vanguard Total Bond Market Index Fund, for example, "may be appropriate for investors with medium-term investment horizons (4 to 10 years)," so that is the kind of time frame we should be looking at. "Interest up, bonds down" is not true in the longer term.

There are other considerations for bonds but they should be considered separately. (e.g. "But inflation." "But TIPS.") The purpose of this posting is to address one single point: even interest rates were certain to rise, that not mean you are certain to lose money if you keep holding the bond fund. In fact, it is the opposite. You will likely make money if you keep holding the bond fund for the appropriate holding period, roughly equal to the duration; you will lose money if you don't keep holding, and sell the fund during the period after the interest rate rises.

Here are three computer simulations of a rolling bond portfolio with a duration roughly similar to that of the Vanguard Total Bond Market Index Fund. In the simulation it's about six years. The current rate on a 10-year Treasury is about 1.45%. Because bond funds are not bond ladders, and because there is variation for different terms in the yield curve, this is no sense an accurate simulation, but it is qualitatively right.

Simulation #1 assumes that the interest rate stays at 1.45%.
Simulation #2 assumes that it rises to 3.45% over the next two years and then levels off.
Simulation #3 assumes that it rises to 5.45% over the next two years and then levels off.

Would a long-term investor rather have interest rates stay the same or rise?

There is a rule of thumb--for a bond, it is exact, but for a bond fund it is only a rule of thumb. It is: the "duration" of the bond fund is the period of time over which short- and long-term effects of an interest rate increase balance out. An interest rate increase is bad for an investor who holds for less than the duration, good an investor who holds for longer than that. Does the rule of thumb roughly hold in these simulations?

The recovery of a bond fund after an interest rate rise is strongly influenced by bond math, not on loose "mean reversion" effect. Assuming a high-quality bond that does not default, an individual bond's market value has a known value at one point in its future: it returns to face value at maturity. So what goes down not only must come up, but it does so on schedule.

#1, interest rate remains constant at 1.45%.


#2, interest rate rises to 3.45% over the next two years.


#3, interest rate rises to 5.45% over the next four years.


Comparison: black, no rise; green, 2% rise in rates; red, 4% rise in rates.

[whereskyle]

Some investors are being overly spooked by a misunderstanding of a) the appropriate holding time for a bond fund, and b) the short-term versus long-term relationship of interest rates to bond fund behavior.

"Interest up, bonds down" is a short-term relationship. Vanguard says that the Vanguard Total Bond Market Index Fund, for example, "may be appropriate for investors with medium-term investment horizons (4 to 10 years)," so that is the kind of time frame we should be looking at. "Interest up, bonds down" is not true in the longer term.

There are other considerations for bonds but they should be considered separately. (e.g. "But inflation." "But TIPS.") The purpose of this posting is to address one single point: even interest rates were certain to rise, that not mean you are certain to lose money if you keep holding the bond fund. In fact, it is the opposite. You will likely make money if you keep holding the bond fund for the appropriate holding period, roughly equal to the duration; you will lose money if you don't keep holding, and sell the fund during the period after the interest rate rises.

Here are three computer simulations of a rolling bond portfolio with a duration roughly similar to that of the Vanguard Total Bond Market Index Fund. In the simulation it's about six years. The current rate on a 10-year Treasury is about 1.45%. Because bond funds are not bond ladders, and because there is variation for different terms in the yield curve, this is no sense an accurate simulation, but it is qualitatively right.

Simulation #1 assumes that the interest rate stays at 1.45%.
Simulation #2 assumes that it rises to 3.45% over the next two years and then levels off.
Simulation #3 assumes that it rises to 5.45% over the next two years and then levels off.

Would a long-term investor rather have interest rates stay the same or rise?

There is a rule of thumb--for a bond, it is exact, but for a bond fund it is only a rule of thumb. It is: the "duration" of the bond fund is the period of time over which short- and long-term effects of an interest rate increase balance out. An interest rate increase is bad for an investor who holds for less than the duration, good an investor who holds for longer than that. Does the rule of thumb roughly hold in these simulations?

The recovery of a bond fund after an interest rate rise is strongly influenced by bond math, not on loose "mean reversion" effect. Assuming a high-quality bond that does not default, an individual bond's market value has a known value at one point in its future: it returns to face value at maturity. So what goes down not only must come up, but it does so on schedule.

#1, interest rate remains constant at 1.45%.


#2, interest rate rises to 3.45% over the next two years.


#3, interest rate rises to 5.45% over the next four years.


Comparison: black, no rise; green, 2% rise in rates; red, 4% rise in rates.

Excellent post, Nisi! I appreciate you.

[Robot Monster]

#3, interest rate rises to 5.45% over the next four years.

*two

[Bernmaster]

Some investors are being overly spooked by a misunderstanding of a) the appropriate holding time for a bond fund, and b) the short-term versus long-term relationship of interest rates to bond fund behavior.

"Interest up, bonds down" is a short-term relationship. Vanguard says that the Vanguard Total Bond Market Index Fund, for example, "may be appropriate for investors with medium-term investment horizons (4 to 10 years)," so that is the kind of time frame we should be looking at. "Interest up, bonds down" is not true in the longer term.

There are other considerations for bonds but they should be considered separately. (e.g. "But inflation." "But TIPS.") The purpose of this posting is to address one single point: even interest rates were certain to rise, that not mean you are certain to lose money if you keep holding the bond fund. In fact, it is the opposite. You will likely make money if you keep holding the bond fund for the appropriate holding period, roughly equal to the duration; you will lose money if you don't keep holding, and sell the fund during the period after the interest rate rises.

Here are three computer simulations of a rolling bond portfolio with a duration roughly similar to that of the Vanguard Total Bond Market Index Fund. In the simulation it's about six years. The current rate on a 10-year Treasury is about 1.45%. Because bond funds are not bond ladders, and because there is variation for different terms in the yield curve, this is no sense an accurate simulation, but it is qualitatively right.

Simulation #1 assumes that the interest rate stays at 1.45%.
Simulation #2 assumes that it rises to 3.45% over the next two years and then levels off.
Simulation #3 assumes that it rises to 5.45% over the next two years and then levels off.

Would a long-term investor rather have interest rates stay the same or rise?

There is a rule of thumb--for a bond, it is exact, but for a bond fund it is only a rule of thumb. It is: the "duration" of the bond fund is the period of time over which short- and long-term effects of an interest rate increase balance out. An interest rate increase is bad for an investor who holds for less than the duration, good an investor who holds for longer than that. Does the rule of thumb roughly hold in these simulations?

The recovery of a bond fund after an interest rate rise is strongly influenced by bond math, not on loose "mean reversion" effect. Assuming a high-quality bond that does not default, an individual bond's market value has a known value at one point in its future: it returns to face value at maturity. So what goes down not only must come up, but it does so on schedule.

#1, interest rate remains constant at 1.45%.


#2, interest rate rises to 3.45% over the next two years.


#3, interest rate rises to 5.45% over the next four years.


Comparison: black, no rise; green, 2% rise in rates; red, 4% rise in rates.

Thank you for this post!

[dorster]

You will likely make money if you keep holding the bond fund for the appropriate holding period, roughly equal to the duration; you will lose money if you don't keep holding, and sell the fund during the period after the interest rate rises.

For me, the logical extension of this is that (isolating for credit risk) we should pick the highest-yielding bond fund with a duration close to our investment horizon. Meaning anyone born after 1955 or thereabouts should likely be investing in Vanguard Long-Term Bond Index Fund (duration 16 years) rather than Vanguard Total Bond Market Index Fund (duration 6.8 years).

I suspect your response will be that you are a satisficer, with a respect for intertia when it comes to personal investing decisions, and that it will barely make a difference anyway (all of which make sense). But I am generally curious about your thoughts as a Vanguard Total Bond Market Index Fund investor who thinks about duration.

[nisiprius]

#3, interest rate rises to 5.45% over the next four years.

*two

Thanks, fixed.

[Candor]

As a relatively recent bond holder for the first time with a significant portion of my portfolio I appreciate the reminder to not focus on the short term. Interestingly, I am fine when it comes to equities but for some reason with bonds the same sentiment seems to elude me

[azanon]

I liked the way Bill McNabb of Vanguard used to put it - if your holding period is going to be longer than your bond duration, you actually WANT rates to go up as that bond holder. Simply worded, and quickly debunks the way too common myth that rising rates is always a bad thing if you own bonds.

[nisiprius]

I liked the way Bill McNabb of Vanguard used to put it - if your holding period is going to be longer than your bond duration, you actually WANT rates to go up as that bond holder. Simply worded, and quickly debunks the way too common myth that rising rates is always a bad thing if you own bonds.

Yeah, the naïve expectation is that rising interest rates should be good for bond holders, and the naïve expectation actually is broadly correct. But it's important to understand the short-term effect as well, and it a) catches peoples' attention because it's so seemingly paradoxical, b) gets undue attention because I suspect much of Wall Street is focussing on short-term effects, and c) is a stick that bond-haters use to beat bonds with.

You will likely make money if you keep holding the bond fund for the appropriate holding period, roughly equal to the duration; you will lose money if you don't keep holding, and sell the fund during the period after the interest rate rises.

For me, the logical extension of this is that (isolating for credit risk) we should pick the highest-yielding bond fund with a duration close to our investment horizon. Meaning anyone born after 1955 or thereabouts should likely be investing in Vanguard Long-Term Bond Index Fund (duration 16 years) rather than Vanguard Total Bond Market Index Fund (duration 6.8 years).

I suspect your response will be that you are a satisficer, with a respect for intertia when it comes to personal investing decisions, and that it will barely make a difference anyway (all of which make sense). But I am generally curious about your thoughts as a Vanguard Total Bond Market Index Fund investor who thinks about duration.

Well, I do think about inflation as well. I feel that the inflation risks of long-term nominal bonds ought to rule them out. But actually it's a good point, and I hadn't thought about it. If you say "long-term TIPS" it probably does make sense. Instead of using the (nominal) Total Bond Market fund in the Target Retirement series, introducing short-term TIPS only when close to retirement, it would probably make more sense for to glide from long-term TIPS to intermediate-term TIPS to short-term TIPS. But I haven't really thought about it.

[CWRadio]

For me, the logical extension of this is that (isolating for credit risk) we should pick the highest-yielding bond fund with a duration close to our investment horizon. Meaning anyone born after 1955 or thereabouts should likely be investing in Vanguard Long-Term Bond Index Fund (duration 16 years) rather than Vanguard Total Bond Market Index Fund (duration 6.8 years).

What duration should you use if you are retired and 76 years old? Thanks Paul

[RickyGold]

Wow, another great post from Nisiprius. Makes me feel better as I own a boatload of Total Bond Fund.

Also, check out this article by Mark Hulbert:

https://www.marketwatch.com/story/you-c ... =home-page

If my understanding is correct, and interest rates rise, I won't lose any money as long as I hold my Total Bond fund for about 10-years.

I am good with that, as my holding period for Total Bond is forever!

[ResearchMed]

Well, I do think about inflation as well. I feel that the inflation risks of long-term nominal bonds ought to rule them out. But actually it's a good point, and I hadn't thought about it. If you say "long-term TIPS" it probably does make sense. Instead of using the (nominal) Total Bond Market fund in the Target Retirement series, introducing short-term TIPS only when close to retirement, it would probably make more sense for to glide from long-term TIPS to intermediate-term TIPS to short-term TIPS. But I haven't really thought about it.

Is it possible to create anything remotely similar to these graphs for TIPS (or short term vs medium term, etc.) funds?

And for both nominal and TIPS funds, does the duration of the fund over time remain approximately the same... that a short term fund won't drift considerably and become a medium term fund? (I know there is no guarantee, but have there been such changes in the past, for example?)

Many thanks.
Your post and the graphs are terrific.

RM

[azanon]

I liked the way Bill McNabb of Vanguard used to put it - if your holding period is going to be longer than your bond duration, you actually WANT rates to go up as that bond holder. Simply worded, and quickly debunks the way too common myth that rising rates is always a bad thing if you own bonds.

Yeah, the naïve expectation is that rising interest rates should be good for bond holders, and the naïve expectation actually is broadly correct. But it's important to understand the short-term effect as well, and it a) catches peoples' attention because it's so seemingly paradoxical, b) gets undue attention because I suspect much of Wall Street is focussing on short-term effects, and c) is a stick that bond-haters use to beat bonds with.

You will likely make money if you keep holding the bond fund for the appropriate holding period, roughly equal to the duration; you will lose money if you don't keep holding, and sell the fund during the period after the interest rate rises.

For me, the logical extension of this is that (isolating for credit risk) we should pick the highest-yielding bond fund with a duration close to our investment horizon. Meaning anyone born after 1955 or thereabouts should likely be investing in Vanguard Long-Term Bond Index Fund (duration 16 years) rather than Vanguard Total Bond Market Index Fund (duration 6.8 years).

I suspect your response will be that you are a satisficer, with a respect for intertia when it comes to personal investing decisions, and that it will barely make a difference anyway (all of which make sense). But I am generally curious about your thoughts as a Vanguard Total Bond Market Index Fund investor who thinks about duration.

Well, I do think about inflation as well. I feel that the inflation risks of long-term nominal bonds ought to rule them out. But actually it's a good point, and I hadn't thought about it. If you say "long-term TIPS" it probably does make sense. Instead of using the (nominal) Total Bond Market fund in the Target Retirement series, introducing short-term TIPS only when close to retirement, it would probably make more sense for to glide from long-term TIPS to intermediate-term TIPS to short-term TIPS. But I haven't really thought about it.

I assume you meant bad, but just checking….

[dorster]

For me, the logical extension of this is that (isolating for credit risk) we should pick the highest-yielding bond fund with a duration close to our investment horizon. Meaning anyone born after 1955 or thereabouts should likely be investing in Vanguard Long-Term Bond Index Fund (duration 16 years) rather than Vanguard Total Bond Market Index Fund (duration 6.8 years).

What duration should you use if you are retired and 76 years old? Thanks Paul

Any fund with a 5-10 year duration should be reasonable. Shorter duration if you expect to die sooner than average (or have an older spouse) and longer duration if you expect to live longer (or plan to leave a legacy to the next generation, or have a much younger spouse). Lots of actuarial resources online, I like longevity illustrator. These can all be approximate calculations just getting you in the right ballpark.

So using vanguard funds you could use any of the following (which have a 5-10 year duration):
VBTLX (Vanguard Total Bond Market Index Fund) if you wanted a mix of bonds
VFITX (Vanguard Intermediate-Term Treasury Fund) if you wanted only treasuries and no credit risk
VIPSX (Vanguard Inflation-Protected Securities Fund) if you wanted no credit risk and no inflation risk
VFICX (Vanguard Intermediate-Term Investment-Grade Fund) if you wanted to take on more credit risk

n.b.: I'm just over half your age. My thoughts on this have been heavily influenced by frequent (former) poster vineviz. Here are some threads if you'd like to read up bond duration glidepath or investment horizon.

Well, I do think about inflation as well. I feel that the inflation risks of long-term nominal bonds ought to rule them out. But actually it's a good point, and I hadn't thought about it. If you say "long-term TIPS" it probably does make sense. Instead of using the (nominal) Total Bond Market fund in the Target Retirement series, introducing short-term TIPS only when close to retirement, it would probably make more sense for to glide from long-term TIPS to intermediate-term TIPS to short-term TIPS. But I haven't really thought about it.

Thanks for your thoughtful response. Switching to short term TIPS in the Target Retirement Series never made sense to me either.

[secondopinion]

Some investors are being overly spooked by a misunderstanding of a) the appropriate holding time for a bond fund, and b) the short-term versus long-term relationship of interest rates to bond fund behavior.

"Interest up, bonds down" is a short-term relationship. Vanguard says that the Vanguard Total Bond Market Index Fund, for example, "may be appropriate for investors with medium-term investment horizons (4 to 10 years)," so that is the kind of time frame we should be looking at. "Interest up, bonds down" is not true in the longer term.

I cannot wait for yields to be higher again; I certainly hope there is a downturn despite holding fixed income. It might mean inflation on top of that, but it does give some buffer in the long run because of the increased yield.

[whereskyle]

You will likely make money if you keep holding the bond fund for the appropriate holding period, roughly equal to the duration; you will lose money if you don't keep holding, and sell the fund during the period after the interest rate rises.

For me, the logical extension of this is that (isolating for credit risk) we should pick the highest-yielding bond fund with a duration close to our investment horizon. Meaning anyone born after 1955 or thereabouts should likely be investing in Vanguard Long-Term Bond Index Fund (duration 16 years) rather than Vanguard Total Bond Market Index Fund (duration 6.8 years).

I suspect your response will be that you are a satisficer, with a respect for intertia when it comes to personal investing decisions, and that it will barely make a difference anyway (all of which make sense). But I am generally curious about your thoughts as a Vanguard Total Bond Market Index Fund investor who thinks about duration.

I know you're not asking me but I'd say while the logic of the proposal is sound, the reality is that trusting investors to do nothing while holding a portfolio exclusively of highly volatile assets is foolish.

Intermediate-duration occupies a sweet spot not because it is perfect for every investor but because it dampens volatility while providing some (but not as much) upside.

[sureshoe]

Not arguing for market timing, as I've increased my bond holding,

BUT

If I knew interest rates are going up, aren't I better being OUT and buying on the dip?

[skierincolorado]

Your simulations clearly show if you expect rates to rise that much you should wait two years and then buy.

That’s incorrect. You should buy now. What you should simulate is interest rates rising about .6% in two years. That is what the fed and market expect. You will still make money.

[skierincolorado]

Not arguing for market timing, as I've increased my bond holding,

BUT

If I knew interest rates are going up, aren't I better being OUT and buying on the dip?

Maybe. Depends how fast they rise. See my post one up. By my calculations interest rate would have to rise well over .5% per year for you to lose money. That’s unlikely.

[skierincolorado]

Some investors are being overly spooked by a misunderstanding of a) the appropriate holding time for a bond fund, and b) the short-term versus long-term relationship of interest rates to bond fund behavior.

"Interest up, bonds down" is a short-term relationship. Vanguard says that the Vanguard Total Bond Market Index Fund, for example, "may be appropriate for investors with medium-term investment horizons (4 to 10 years)," so that is the kind of time frame we should be looking at. "Interest up, bonds down" is not true in the longer term.

There are other considerations for bonds but they should be considered separately. (e.g. "But inflation." "But TIPS.") The purpose of this posting is to address one single point: even interest rates were certain to rise, that not mean you are certain to lose money if you keep holding the bond fund. In fact, it is the opposite. You will likely make money if you keep holding the bond fund for the appropriate holding period, roughly equal to the duration; you will lose money if you don't keep holding, and sell the fund during the period after the interest rate rises.

Here are three computer simulations of a rolling bond portfolio with a duration roughly similar to that of the Vanguard Total Bond Market Index Fund. In the simulation it's about six years. The current rate on a 10-year Treasury is about 1.45%. Because bond funds are not bond ladders, and because there is variation for different terms in the yield curve, this is no sense an accurate simulation, but it is qualitatively right.

Simulation #1 assumes that the interest rate stays at 1.45%.
Simulation #2 assumes that it rises to 3.45% over the next two years and then levels off.
Simulation #3 assumes that it rises to 5.45% over the next two years and then levels off.

Would a long-term investor rather have interest rates stay the same or rise?

There is a rule of thumb--for a bond, it is exact, but for a bond fund it is only a rule of thumb. It is: the "duration" of the bond fund is the period of time over which short- and long-term effects of an interest rate increase balance out. An interest rate increase is bad for an investor who holds for less than the duration, good an investor who holds for longer than that. Does the rule of thumb roughly hold in these simulations?

The recovery of a bond fund after an interest rate rise is strongly influenced by bond math, not on loose "mean reversion" effect. Assuming a high-quality bond that does not default, an individual bond's market value has a known value at one point in its future: it returns to face value at maturity. So what goes down not only must come up, but it does so on schedule.

#1, interest rate remains constant at 1.45%.


#2, interest rate rises to 3.45% over the next two years.


#3, interest rate rises to 5.45% over the next two years.


Comparison: black, no rise; green, 2% rise in rates; red, 4% rise in rates.

These graphs have a couple problems for making your point. You have mostly or entirely neglected roll yield which is substantial and nearly equal to the coupon. The actual return would be higher, unless the yield curve was flat.

In addition, what you’ve shown is that if rates rise this much it would be better to sell and buy back at the dip. What you need to simulate is a more realistic rate rise of around .3% per year for two years. Thats how much the market expects 5 year interest rates to rise. And include roll yield. There won’t be a dip and you will have proved your point much better. Now is a good time to buy.. not two years from now.

[secondopinion]

Not arguing for market timing, as I've increased my bond holding,

BUT

If I knew interest rates are going up, aren't I better being OUT and buying on the dip?

Can one actually buy the dip? Furthermore, after real returns of cash, do we want to be waiting on the sidelines? One can find good short-term fixed income in taxable accounts; but I am not waiting for a bond dip with my short-term holdings. Cash is for short term uses, not long-term uses.

[skierincolorado]

Not arguing for market timing, as I've increased my bond holding,

BUT

If I knew interest rates are going up, aren't I better being OUT and buying on the dip?

Can one actually buy the dip? Furthermore, after real returns of cash, do we want to be waiting on the sidelines? One can find good short-term fixed income in taxable accounts; but I am not waiting for a bond dip with my short-term holdings. Cash is for short term uses, not long-term uses.

If Nisis graphs were correct, yes we would want to avoid the 10%++ dip and then buy. But the graphs are not very realistic. They seem to neglect roll yield. And interest rates on BND and VGIT are expected to rise 0.3% per year not 1-2%.

[skierincolorado]

In retrospect, the simulation looks like it might contain some roll yield but not much. But more importantly the interest rate increases are unrealistic. If the 5 or 10 year rate increases 2% per I will eat my hat.

[vitaflo]

In retrospect, the simulation looks like it might contain some roll yield but not much. But more importantly the interest rate increases are unrealistic. If the 5 or 10 year rate increases 2% per I will eat my hat.

I think it was purposefully an extreme example to prove a broader point. Even in an extreme case, you will be fine if you hold the duration.

[skierincolorado]

In retrospect, the simulation looks like it might contain some roll yield but not much. But more importantly the interest rate increases are unrealistic. If the 5 or 10 year rate increases 2% per I will eat my hat.

I think it was purposefully an extreme example to prove a broader point. Even in an extreme case, you will be fine if you hold the duration.

Sure, but it's very misleading because most will see that you lose money when rates go up and that they should wait until after the rate increaes to buy. That's usually not the case. That's only the case when rates rise rapidly. When the price decrease is greater than the coupon and roll yield.

I've been holding well over my net worth in 5 year bonds since April and I've done fine despite the rate increases being faster than I expected. VGIT which holds 5 year bonds was up 1% (2% annualized) from April to the end of September, **despite interest rates rising .1% and a coupon under 1%**. If the coupon was under 1% annualized, and the price of the bonds fell as interest rates rose, how did I return over 2% annualized during this period? Answer: Roll yield.

https://www.portfoliovisualizer.com/bac ... ion1_1=100

[dorster]

I know you're not asking me but I'd say while the logic of the proposal is sound, the reality is that trusting investors to do nothing while holding a portfolio exclusively of highly volatile assets is foolish.

Intermediate-duration occupies a sweet spot not because it is perfect for every investor but because it dampens volatility while providing some (but not as much) upside.

I don't disagree.

If someone understands that they're accepting lower yield and more interest rate risk but prefers to choose an intermediate bond fund because they're worried about volatility that seems fine. If someone decides they're not interested in bond minutiae (or don't really understand bond duration) and their takeaway is that intermediate bond funds won't be an unreasonable choice that also seems fine.

I've enjoyed learning more about bonds and was posting in that spirit.

[sycamore]

There are other considerations for bonds but they should be considered separately. (e.g. "But inflation." "But TIPS.") The purpose of this posting is to address one single point: even interest rates were certain to rise, that not mean you are certain to lose money if you keep holding the bond fund. In fact, it is the opposite. You will likely make money if you keep holding the bond fund for the appropriate holding period, roughly equal to the duration; you will lose money if you don't keep holding, and sell the fund during the period after the interest rate rises.
...

Thanks for the posting.

One comment on the underlined part of the above quote: "don't keep holding" applies to a retiree who's drawing down from the portfolio for spending needs. I think this where some investors start to get spooked -- they know they'll have to sell bonds at a loss (relative to the last rate hike) and they hate that feeling of losing out. This drives some investors to use only CDs or individual bonds that they hold to maturity. For many investors, that's too much work and so they stick with a bond fund. What keeps me in bond funds long term is knowing that even if I do sell bonds (for expenses), it's only a small part of my fixed income so I'll eventually get the benefit of rising rates, just slightly later than I would've otherwise.

[skierincolorado]

In retrospect, the simulation looks like it might contain some roll yield but not much. But more importantly the interest rate increases are unrealistic. If the 5 or 10 year rate increases 2% per I will eat my hat.

I think it was purposefully an extreme example to prove a broader point. Even in an extreme case, you will be fine if you hold the duration.

Or how about a real time test. From April 2021 to present the 5 year rate rose from .89 to 1.21%. And yet the returns of VGIT are still positive (barely)!

https://www.morningstar.com/etfs/xnas/vgit/performance

[secondopinion]

There are other considerations for bonds but they should be considered separately. (e.g. "But inflation." "But TIPS.") The purpose of this posting is to address one single point: even interest rates were certain to rise, that not mean you are certain to lose money if you keep holding the bond fund. In fact, it is the opposite. You will likely make money if you keep holding the bond fund for the appropriate holding period, roughly equal to the duration; you will lose money if you don't keep holding, and sell the fund during the period after the interest rate rises.
...

Thanks for the posting.

One comment on the underlined part of the above quote: "don't keep holding" applies to a retiree who's drawing down from the portfolio for spending needs. I think this where some investors start to get spooked -- they know they'll have to sell bonds at a loss (relative to the last rate hike) and they hate that feeling of losing out. This drives some investors to use only CDs or individual bonds that they hold to maturity. For many investors, that's too much work and so they stick with a bond fund. What keeps me in bond funds long term is knowing that even if I do sell bonds (for expenses), it's only a small part of my fixed income so I'll eventually get the benefit of rising rates, just slightly later than I would've otherwise.

Holding until maturity is a false sense of security for most investors. Unless one literally needs the money close to then to spend, it gains nothing to think in these terms. Also, that is often is in nominal dollars; so, holding to maturity is not entirely safe in real dollars, even if it will be spent at maturity.

CDs (those that are not marketable) have usually a put clause where you pay a penalty to break it. This can be helpful at times during sharp interest rate increases, since the penalty is fixed. CDs have been my safe money maker for fixed income, and some terms have made me a lot of money over bonds (about 3% on new cash added into the CD for 2 1/2 years if I will hold until maturity; it sure helps when yields are half that for BND).

[UpperNwGuy]

To me, any CD that matures more than 24 months into the future is a gamble. I'll stick to bonds for my investment portfolio and only use CDs for short term savings.

[nisiprius]

Your simulations clearly show if you expect rates to rise that much you should wait two years and then buy. That’s incorrect. You should buy now. What you should simulate is interest rates rising about .6% in two years. That is what the fed and market expect. You will still make money.

Let me be very clear.

My pictures are is a hopefully-correct simulation of an unrealistic, oversimplified model. It simulates a bond ladder, not a mutual fund. It assumes that there is a single interest rate for all terms. And I leaned pessimistic in all assumptions, because my goal was to counter the idea that an interest rate rise means you are certain to lose money--even "to lose principal."

1) They are not an illustration of what the optimum strategy would be if you had an accurate prediction of the future interest rate (or rates, or the yield curve). It's not a prediction of what I think interest rates will do.

2) They are not a planning tool.

3) They are not a market timing guide.

I believe that many investors think that bond fund values are a simple mirror image of the interest rate, and that if the interest rate goes up 2%, a fund with a 6-year duration will go down 12% and stay down forever. Hence frequent concern about "loss of principal," when (of course) you do not lose principal if the bond or bonds do not default.

[skierincolorado]

Your simulations clearly show if you expect rates to rise that much you should wait two years and then buy. That’s incorrect. You should buy now. What you should simulate is interest rates rising about .6% in two years. That is what the fed and market expect. You will still make money.

Let me be very clear.

My pictures are is a hopefully-correct simulation of an unrealistic, oversimplified model. It simulates a bond ladder, not a mutual fund. It assumes that there is a single interest rate for all terms. And I leaned pessimistic in all assumptions, because my goal was to counter the idea that an interest rate rise means you are certain to lose money--even "to lose principal."

1) They are not an illustration of what the optimum strategy would be if you had an accurate prediction of the future interest rate (or rates, or the yield curve). It's not a prediction of what I think interest rates will do.

2) They are not a planning tool.

3) They are not a market timing guide.

I believe that many investors think that bond fund values are a simple mirror image of the interest rate, and that if the interest rate goes up 2%, a fund with a 6-year duration will go down 12% and stay down forever. Hence frequent concern about "loss of principal," when (of course) you do not lose principal if the bond or bonds do not default.

I understand that but I think most people’s immediate reaction to those graphs will be why don’t I wait till rates stop going up since everyone expects rates to go up including the fed and the market itself.

For example see the post directly below this one! “ I don’t like the negative symbol in the short run” ... what negative symbol!?!?! I’ve been making money in bonds since April even despite the rapid run up in rates the last 3 weeks.

How did you incorporate roll return? It looks like either it was not included or was only included slightly. Actual returns should be higher.

Can you do them for a more realistic rate increase scenario to show that you will likely still make money even if rates go up?

Including both of these would make the point much stronger.

Or just replace your graph with a graph of VGIT since April. Rates going up. Returns still positive.

[kelway]

Some investors are being overly spooked by a misunderstanding of a) the appropriate holding time for a bond fund, and b) the short-term versus long-term relationship of interest rates to bond fund behavior.

I really appreciate you making this post. I've been spooked and in fact kind of knew that what you posted is true, but I was still considering stable value funds etc. as I don't like the negative symbol in the short term.

[secondopinion]

To me, any CD that matures more than 24 months into the future is a gamble. I'll stick to bonds for my investment portfolio and only use CDs for short term savings.

What kind of CDs are you looking at? If you buy those with dirt cheap rates, then yes it is bad. Remember, brokered CDs are usually not the deals.

But I assure you that some research will reveal some real gems.

[UpperNwGuy]

To me, any CD that matures more than 24 months into the future is a gamble. I'll stick to bonds for my investment portfolio and only use CDs for short term savings.

What kind of CDs are you looking at? If you buy those with dirt cheap rates, then yes it is bad. Remember, brokered CDs are usually not the deals.

But I assure you that some research will reveal some real gems.

If interest rates rise, I don't want to be locked into today's low interest rates. I don't buy brokered CDs. I can get decent ones, such as they are, from either of my two credit unions.

[skierincolorado]

Some investors are being overly spooked by a misunderstanding of a) the appropriate holding time for a bond fund, and b) the short-term versus long-term relationship of interest rates to bond fund behavior.

I really appreciate you making this post. I've been spooked and in fact kind of knew that what you posted is true, but I was still considering stable value funds etc. as I don't like the negative symbol in the short term.

See my post directly above yours.

[secondopinion]

To me, any CD that matures more than 24 months into the future is a gamble. I'll stick to bonds for my investment portfolio and only use CDs for short term savings.

What kind of CDs are you looking at? If you buy those with dirt cheap rates, then yes it is bad. Remember, brokered CDs are usually not the deals.

But I assure you that some research will reveal some real gems.

If interest rates rise, I don't want to be locked into today's low interest rates. I don't buy brokered CDs. I can get decent ones, such as they are, from either of my two credit unions.

You do not need to be locked into those low rates; take the break penalty and reinvest at better rates. It is a put option that saves you for that purpose.

[neurosphere]

On my to-do list is to make similar graphs, but instead of comparing funds with different interest rates scenarios, to illustrate the effect on funds of different durations in response to a given set of interest rates changes.

Nisi, I have bookmarked your post. Thanks.

[UpperNwGuy]

To me, any CD that matures more than 24 months into the future is a gamble. I'll stick to bonds for my investment portfolio and only use CDs for short term savings.

What kind of CDs are you looking at? If you buy those with dirt cheap rates, then yes it is bad. Remember, brokered CDs are usually not the deals.

But I assure you that some research will reveal some real gems.

I just looked at BankRate.com and don't see any real gems:
https://www.bankrate.com/banking/cds/cd-rates/

[rockstar]

In retrospect, the simulation looks like it might contain some roll yield but not much. But more importantly the interest rate increases are unrealistic. If the 5 or 10 year rate increases 2% per I will eat my hat.

I think it was purposefully an extreme example to prove a broader point. Even in an extreme case, you will be fine if you hold the duration.

Or how about a real time test. From April 2021 to present the 5 year rate rose from .89 to 1.21%. And yet the returns of VGIT are still positive (barely)!

https://www.morningstar.com/etfs/xnas/vgit/performance

It's down 2% after dividends. That's at least down 7% with inflation. And then, you're down whatever you end up paying in taxes.

The biggest drag on bonds now is taxes and inflation given the rates are so low. This is the biggest worry now. If inflation stays around 3ish percent for the next 5 years, which is about what the market is pricing in so far, then after taxes, bonds are going to be a big drag on a portfolio.

Now, let's add to this rising rates, and let's hope the OP is right and we get back to at least zero nominally because real and after taxes is going to hurt.

[cflannagan]

It's down 2% after dividends. That's at least down 7% with inflation. And then, you're down whatever you end up paying in taxes.

The biggest drag on bonds now is taxes and inflation given the rates are so low. This is the biggest worry now. If inflation stays around 3ish percent for the next 5 years, which is about what the market is pricing in so far, then after taxes, bonds are going to be a big drag on a portfolio.

Now, let's add to this rising rates, and let's hope the OP is right and we get back to at least zero nominally because real and after taxes is going to hurt.

I don't think I saw anyone advocating for bonds to go into a taxable account here? At least not from what you quoted anyway.

[make_a_better_world]

Some investors are being overly spooked by a misunderstanding of a) the appropriate holding time for a bond fund, and b) the short-term versus long-term relationship of interest rates to bond fund behavior.

The issue I have as a novice bond investor, is how am I supposed to use the bond allocation as a ballast for equities in an environment of increasing interest rates? In your bond ladder the rate increases over 2 years and then remains fixed.

If we take the same duration of bonds (4-10 years, avg 7) and increase the interest rate gradually over 4 years, one must hold the fund for 4.5 years to recover back to time 0.


If we take shorter duration bonds (3-5 years, avg 4), we can sell at any point without incurring a loss should interest rates rise the same amount over the same period:


If:
Today’s yield on a bond fund is an excellent proxy for its total return in the subsequent decade,
The yield between 5 year and 10 year treasuries is < 0.5% difference,
The real returns are < 0 for all of these bonds (these are not deals we're missing out on),
One plans to sell bonds at an unpredictable time to balance including next year if needed,
And interest rates rise as a trend over years (I do not pretend to be able to predict interest rates - so a big if)

Then: short term duration bonds are superior to long term bonds in an environment of rising interest rates to the investor planning to use this basket as a safety net instead of for returns (i.e. sell this portion to take advantage of equity declines regardless of investing time horizon). The long term bond holder will need to worry less about moving money from bonds to equities if the market contracts due to rising interest rates because both the equity and bond baskets of that investor should decline and the percentage of each in asset allocation will not change as much.

Here's the bond ladder tool I used: https://funds.eatonvance.com/laddered-tools.php

I'd love to hear your thoughts if you think my logic is wrong.

[nisiprius]

...How did you incorporate roll return? It looks like either it was not included or was only included slightly. Actual returns should be higher....

You'll need to explain roll return to me. I've only seen the term used in the context of commodity futures, where I haven't needed to understand it.

What is "roll return" and how does it come into play in a real-world bond fund? Does it have something to do with intentionally buying bonds after issue and selling them before maturity? As noted, no, my simulation is a simple bond ladder, with bonds being bought at issue and held to maturity.

The simulation involves a portfolio of bonds. There is no yield curve (or a level or constant yield curve), there is assumed to be a single prevailing interest rate at any point in time. Bonds are bought at issue at face value, with no premium or discount. A bond, once bought, is held to maturity. The bonds make coupon payments every six months. In this set of simulations, the bonds are have a 14-year term. Thus the portfolio always contains 28 bonds, maturing in 0.5, 1, 1.5, ... 14 years. This number was chosen by trial and error because it gives a duration for the whole ladder--calculated empirically by making a tiny instantaneous change in the interest rate and observing the resulting change in value for the whole portfolio--to give a duration of about 6.2 years.

As each bond makes a coupon payment the payment is reinvested into the identical bond issue i.e. used to buy more of the same bond issue at the current market value. As each bond matures, it is replaced with a freshly-issued bond.

If you have the capability of running a comparable but more accurate simulation, it would certainly be useful to see more precise results.

P.S. As I understand it, skierincolorado and make_a_better_world are not challenging the general idea--an interest rate rise causes a temporary loss which is made up if you hold for a holding period roughly equal to the duration. The question is has to do with the timing and depth of the loss. I understand skierincolorado is saying for a sufficiently gentle interest rate rise there may be no loss at all, and that for real-world bond funds the managers are able to reduce or eliminate those losses by better management of the timing of purchases and sales, and by exploiting unequal interest rates across the yield curve.... correct?

[tomsense76]

Thanks for writing this up Nisi! Have referenced other comments you have written on this, but they can be hard to find sometimes. It's great to have this in its own topic. Would it make sense to make this a sticky and/or link from the wiki?

[rockstar]

...How did you incorporate roll return? It looks like either it was not included or was only included slightly. Actual returns should be higher....

You'll need to explain roll return to me. I've only seen the term used in the context of commodity futures, where I haven't needed to understand it.

What is "roll return" and how does it come into play in a real-world bond fund? Does it have something to do with intentionally buying bonds after issue and selling them before maturity? As noted, no, my simulation is a simple bond ladder, with bonds being bought at issue and held to maturity.

The simulation involves a portfolio of bonds. There is no yield curve (or a level or constant yield curve), there is assumed to be a single prevailing interest rate at any point in time. Bonds are bought at issue at face value, with no premium or discount. A bond, once bought, is held to maturity. The bonds make coupon payments every six months. In this set of simulations, the bonds are have a 14-year term. Thus the portfolio always contains 28 bonds, maturing in 0.5, 1, 1.5, ... 14 years. This number was chosen by trial and error because it gives a duration for the whole ladder--calculated empirically by making a tiny instantaneous change in the interest rate and observing the resulting change in value for the whole portfolio--to give a duration of about 6.2 years.

As each bond makes a coupon payment the payment is reinvested into the identical bond issue i.e. used to buy more of the same bond issue at the current market value. As each bond matures, it is replaced with a freshly-issued bond.

If you have the capability of running a comparable but more accurate simulation, it would certainly be useful to see more precise results.

Help me understand how this relates to a bond fund, where bonds are sold and bought and not always held to maturity. And corporate bonds can be called before maturity too. It's not like a bond fund's duration decreases after you buy it as a single bond would as it ages out.

[nisiprius]

...Help me understand how this relates to a bond fund, where bonds are sold and bought and not always held to maturity. And corporate bonds can be called before maturity too. It's not like a bond fund's duration decreases after you buy it as a single bond would as it ages out...

My simulation is of a rolling bond ladder. The duration holds pretty steady at about 6.2 years.

And it's the best this autodidact can do, or is willing to spend the time doing. It's closer to reality than "interest rates can only go up, and when they do, you know what direction your bond fund is going to go? Down! That's what direction it's going to go in!"

I've also posted actual behavior of real-world bond funds during, for example, 2010-2011, when Jeremy Schwartz and Jeremy Siegel warned about the collapse of the "great American bond bubble," and interest rates actually rose into about the range they said would be disastrous, and the effect on Total Bond was... not dramatic. And I've explored the behavior of Total Bond over overlapping 48-month periods since inception, and so far it has yet to lose money in any 48-month holding period, including periods ending within the "bond massacre" of 1994, and the period from 2004-2007 when the Fed raised its rate by 4% in less than 4 years, and the interest rate rise of 2011, as well as the last 48 months.

[kelway]

For example see the post directly below this one! “ I don’t like the negative symbol in the short run” ... what negative symbol!?!?! I’ve been making money in bonds since April even despite the rapid run up in rates the last 3 weeks.

Color me confused because my comment came after this one, but I guess you're referring to mine since that was basically my quote about the negative symbol. I refer to, for example, SPTI's one year return. My fixed income exposure is limited to short and intermediate treasuries and short tips... Only the tips have been positive. After inflation, clearly they're all negative.

EDIT - ah I see now - you said edited it and said below this one. I thought I read above this one.

[quattro73]

I posted this in another thread a few weeks ago. It seems sort of related:

If you look at the period from 06/01/2012 to 12/31/2013, the 10 year TCM went from 1.47% to 3.04%. BND posted -0.76% return for those 19 months. But if you look at 05/01/2013 to 12/31/2013, 10 year went from 1.66% to the aforementioned 3.04%. BND posted -3.01% for those 8 months and almost all of it was straight up for the 10 year in that period. Reinvested dividends and redeployment of maturing cash flows into higher yielding issuances will blunt the rate increases, with a little time.

The next bottoming of rates was in 2016 (07/05/2016 was the low), with rates then peaking in 2018 (10/05/2018 was the high). If you run Portfolio Visualizer from 07/01/2016 where 10 year rates start at 1.46% through 10/31/2018 where rates end at 3.15%, BND returned -0.82%. But there was a sharp increase in the 10 year in there from July 2016 to December 2016 before going sideways in 2017, and resuming a more steady climb in 2018. That 07/01/2016 to 12/30/2016 had the 10 year going from 1.46% to 2.45%. BND returned -2.78%.

It is the sharp increases that get you, but the redeployment of maturing issuances and reinvestment of dividends ( if one is so inclined), act to rather quickly “self correct “ a bond fund, at least for these minor increases. A point and half increase seems pretty plausible in the next couple years.

[skierincolorado]

In retrospect, the simulation looks like it might contain some roll yield but not much. But more importantly the interest rate increases are unrealistic. If the 5 or 10 year rate increases 2% per I will eat my hat.

I think it was purposefully an extreme example to prove a broader point. Even in an extreme case, you will be fine if you hold the duration.

Or how about a real time test. From April 2021 to present the 5 year rate rose from .89 to 1.21%. And yet the returns of VGIT are still positive (barely)!

https://www.morningstar.com/etfs/xnas/vgit/performance

It's down 2% after dividends. That's at least down 7% with inflation. And then, you're down whatever you end up paying in taxes.

The biggest drag on bonds now is taxes and inflation given the rates are so low. This is the biggest worry now. If inflation stays around 3ish percent for the next 5 years, which is about what the market is pricing in so far, then after taxes, bonds are going to be a big drag on a portfolio.

Now, let's add to this rising rates, and let's hope the OP is right and we get back to at least zero nominally because real and after taxes is going to hurt.

From 3/31 to 10/20 it is not down 2% after dividends. It is up slightly. You have the dates wrong. During the dates I picked, the interest rate rose from .89 to 1.18% and yet VGIT had slightly positive returns.

So my stocks are down 5% after inflation too... guess I should sell them too huh

how is the return on cash these days?

oh wait

[skierincolorado]

...How did you incorporate roll return? It looks like either it was not included or was only included slightly. Actual returns should be higher....

You'll need to explain roll return to me. I've only seen the term used in the context of commodity futures, where I haven't needed to understand it.

What is "roll return" and how does it come into play in a real-world bond fund? Does it have something to do with intentionally buying bonds after issue and selling them before maturity? As noted, no, my simulation is a simple bond ladder, with bonds being bought at issue and held to maturity.

The simulation involves a portfolio of bonds. There is no yield curve (or a level or constant yield curve), there is assumed to be a single prevailing interest rate at any point in time. Bonds are bought at issue at face value, with no premium or discount. A bond, once bought, is held to maturity. The bonds make coupon payments every six months. In this set of simulations, the bonds are have a 14-year term. Thus the portfolio always contains 28 bonds, maturing in 0.5, 1, 1.5, ... 14 years. This number was chosen by trial and error because it gives a duration for the whole ladder--calculated empirically by making a tiny instantaneous change in the interest rate and observing the resulting change in value for the whole portfolio--to give a duration of about 6.2 years.

As each bond makes a coupon payment the payment is reinvested into the identical bond issue i.e. used to buy more of the same bond issue at the current market value. As each bond matures, it is replaced with a freshly-issued bond.

If you have the capability of running a comparable but more accurate simulation, it would certainly be useful to see more precise results.

P.S. As I understand it, skierincolorado and make_a_better_world are not challenging the general idea--an interest rate rise causes a temporary loss which is made up if you hold for a holding period roughly equal to the duration. The question is has to do with the timing and depth of the loss. I understand skierincolorado is saying for a sufficiently gentle interest rate rise there may be no loss at all, and that for real-world bond funds the managers are able to reduce or eliminate those losses by better management of the timing of purchases and sales, and by exploiting unequal interest rates across the yield curve.... correct?

So most bond funds would not hold to maturity, which actually substantially boosts returns via roll yield. If a bond fund holds a ladder from 8 years to 4 years and then sells them, it's not holding the bonds to maturity. The yield will therefore be more than the YTM.

As an example, if I buy a 5 year bond today the yield is 1.16% (for simplicity let's say the price was $1000 and the coupon was exactly 1.16%). If I hold it for a year, and interest rates don't change at all, I would now hold a 4 year bond with a .96% yield - that's the market rate for 4 year bonds. Since yield and price move inversely, the price of my bond is now higher. Same coupon as before, but higher price. The combination of the 1.16% coupon, ~$1001 price, and $1000 face redepmtion value, mathematically equals .96% yield.

If interest rates rose 0.1%, the 4 year yield would now be 1.06% - still lower than when I bought it, so the price would still be over $1000. Plus I have the coupon.


Since you are doing a ladder holding to maturity - this doesn't apply. But for most real bond funds roll yield is a significant part of returns. Rolling down the yield curve to shorter durations = price increases (all else equal). This is especially true where the curve is steep - less than 10 years of duration.




So yes I *strongly* agree with your overall point. Bond returns can be quite good even when rates rise, especially once you factor in the roll return on most bond funds, and if the rate increase is modest. The 5 year rate is 1.16% today. Even if it is 1.66% two years from now, holding and rolling 5 year bonds will still have significantly positive reutrns.