How to use duration matching in a stock/bond portfolio

Discuss all general (i.e. non-personal) investing questions and issues, investing news, and theory.
Topic Author
TN_Boy
Posts: 2945
Joined: Sat Jan 17, 2009 12:51 pm

How to use duration matching in a stock/bond portfolio

Post by TN_Boy »

I've been reading and sometimes participating in several threads that talk about duration matching.

See

viewtopic.php?p=6799004#p6799004
and
viewtopic.php?p=5331115#p5331115

And I confess, I don't know how to use duration in my personal situation, which I think is quite common.

Assume this scenario (which is not mine, but is similar enough to make the answers relevant)

1) Retiree with 50-50 stock bond portfolio
2) Retiree expects to pull *about* 4% out, adjusting for inflation. Maybe a bit less. The amount is likely to vary.
3) Retiree will follow the method used in most SWR studies, i.e. rebalancing portfolio yearly via withdrawals and explicit buys/sells
4) Retiree is 60, so figure the standard 30 year or thereabouts retirement.
5) Like many folks, Retiree has SS as a baseline income stream, so much of his/her needs are handled by SS.

Because of the harvesting method in 3) withdrawals in a given year might pull from stocks, bonds (including sales of bonds), or both.

My questions:

1) If I use duration matching in my bond portfolio, what duration should I have? (This is the least interesting question)
2) What difference to portfolio survival does duration matching provide given this situation?

How am I better off selecting a duration than simply using an intermediate term bond fund like BND? I'm NOT matching any particular bond maturing or bond interest to a specific liability in the future. I can pick a bond duration for the fixed income. And shorten it as the years go by (the mechanics of this are explained in some of the other threads).

But I can't see how the duration matching is going to make a clear positive difference in my overall wealth, nor have I have seen studies showing how this is better than ... simply using an intermediate bond fund.

Help me out here, I've read several of these threads, but I'm not getting it for the scenario I describe, which I believe is common.
Florida Orange
Posts: 40
Joined: Thu Jun 16, 2022 2:22 pm

Re: How to use duration matching in a stock/bond portfolio

Post by Florida Orange »

I think the point of matching duration to investment horizon is to balance reinvestment risk with NAV volatility risk so you're not taking too much of either one. It affects your overall wealth by ensuring that you get what you expected to get when you expected to get it.
Topic Author
TN_Boy
Posts: 2945
Joined: Sat Jan 17, 2009 12:51 pm

Re: How to use duration matching in a stock/bond portfolio

Post by TN_Boy »

Florida Orange wrote: Wed Aug 03, 2022 3:19 pm I think the point of matching duration to investment horizon is to balance reinvestment risk with NAV volatility risk so you're not taking too much of either one. It affects your overall wealth by ensuring that you get what you expected to get when you expected to get it.
It is not clear to me that duration matching the bond portion of the portfolio will reduce the *portfolio* NAV volatility. And I don't think that is what some of the duration matching folks argue is the main idea. Then again, I've said I'm not following the point very well .....
User avatar
vineviz
Posts: 12309
Joined: Tue May 15, 2018 1:55 pm
Location: Baltimore, MD

Re: How to use duration matching in a stock/bond portfolio

Post by vineviz »

TN_Boy wrote: Wed Aug 03, 2022 1:18 pm 1) If I use duration matching in my bond portfolio, what duration should I have? (This is the least interesting question)
A retiree with a planned 30-year retirement has, as a first order approximation, a 15 year investment horizon so that would be the duration you'd target (if you choose to do that).

It might be a little shorter if the withdraws are greater before age 70 than after (due, say, to Social Security), but that's the ballpark.

TN_Boy wrote: Wed Aug 03, 2022 1:18 pm 2) What difference to portfolio survival does duration matching provide given this situation?
I'd argue that the primary benefit of duration matching is one of risk reduction. Duration matching provides more upfront certainty about the level of portfolio duration, but the truth is that an investor is set on rebalancing and has a significant equity exposure (e.g. 50% in your example) is giving up some of the certainty anyway. For such a retiree, the difference might not be very great.
"Far more money has been lost by investors preparing for corrections than has been lost in corrections themselves." ~~ Peter Lynch
RubyTuesday
Posts: 1925
Joined: Fri Oct 19, 2012 11:24 am

Re: How to use duration matching in a stock/bond portfolio

Post by RubyTuesday »

I confess to not having searched for this, so I will attempt to find a prior thread on this topic, but posting here for thoughts while I look..

If an investor has a typical portfolio comprised of equity and fixed income, and hopes to duration match their portfolio to their investing horizon, to what extent should the duration of the equity portion come into play? Is there even a consensus on whether equity has a duration or what it means?
“Doing nothing is better than being busy doing nothing.” – Lao Tzu
User avatar
vineviz
Posts: 12309
Joined: Tue May 15, 2018 1:55 pm
Location: Baltimore, MD

Re: How to use duration matching in a stock/bond portfolio

Post by vineviz »

RubyTuesday wrote: Wed Aug 03, 2022 5:22 pm Is there even a consensus on whether equity has a duration or what it means?
There really isn't any consensus on this, though I'd say support for the concept of "equity duration" has arguably waned over the last decade or two.

I would ignore most investors to ignore it: it's mostly something economics grad students argue about in pubs.
"Far more money has been lost by investors preparing for corrections than has been lost in corrections themselves." ~~ Peter Lynch
User avatar
willthrill81
Posts: 32250
Joined: Thu Jan 26, 2017 3:17 pm
Location: USA

Re: How to use duration matching in a stock/bond portfolio

Post by willthrill81 »

TN_Boy wrote: Wed Aug 03, 2022 3:25 pm
Florida Orange wrote: Wed Aug 03, 2022 3:19 pm I think the point of matching duration to investment horizon is to balance reinvestment risk with NAV volatility risk so you're not taking too much of either one. It affects your overall wealth by ensuring that you get what you expected to get when you expected to get it.
It is not clear to me that duration matching the bond portion of the portfolio will reduce the *portfolio* NAV volatility. And I don't think that is what some of the duration matching folks argue is the main idea. Then again, I've said I'm not following the point very well .....
The market price volatility of the bonds between the time of purchase and redemption is irrelevant, at least from a completely objective standpoint.

As noted by vineviz in this post in a very relevant thread, there are two forms of interest rate risk: market price risk and reinvestment risk. Many investors focus exclusively on market price risk, likely because this is highly visible in day to day fluctuations in a bond's market value. But reinvestment risk is generally more important for long-term investors.
I have left the forum but occasionally check PMs.
RubyTuesday
Posts: 1925
Joined: Fri Oct 19, 2012 11:24 am

Re: How to use duration matching in a stock/bond portfolio

Post by RubyTuesday »

vineviz wrote: Wed Aug 03, 2022 6:56 pm
RubyTuesday wrote: Wed Aug 03, 2022 5:22 pm Is there even a consensus on whether equity has a duration or what it means?
There really isn't any consensus on this, though I'd say support for the concept of "equity duration" has arguably waned over the last decade or two.

I would ignore most investors to ignore it: it's mostly something economics grad students argue about in pubs.
Okay, now I’m motivated to learn (just) enough about this to “rub two grad students’ noses together” and tip a pint while they argue.

Thanks!
:sharebeer
“Doing nothing is better than being busy doing nothing.” – Lao Tzu
Topic Author
TN_Boy
Posts: 2945
Joined: Sat Jan 17, 2009 12:51 pm

Re: How to use duration matching in a stock/bond portfolio

Post by TN_Boy »

vineviz wrote: Wed Aug 03, 2022 5:09 pm
TN_Boy wrote: Wed Aug 03, 2022 1:18 pm 1) If I use duration matching in my bond portfolio, what duration should I have? (This is the least interesting question)
A retiree with a planned 30-year retirement has, as a first order approximation, a 15 year investment horizon so that would be the duration you'd target (if you choose to do that).

It might be a little shorter if the withdraws are greater before age 70 than after (due, say, to Social Security), but that's the ballpark.
TN_Boy wrote: Wed Aug 03, 2022 1:18 pm 2) What difference to portfolio survival does duration matching provide given this situation?
I'd argue that the primary benefit of duration matching is one of risk reduction. Duration matching provides more upfront certainty about the level of portfolio duration, but the truth is that an investor is set on rebalancing and has a significant equity exposure (e.g. 50% in your example) is giving up some of the certainty anyway. For such a retiree, the difference might not be very great.
And you've described in other threads how a weighted average of long and shorter term funds could implement the matching if one didn't wish to buy individual bonds.

Interesting point about greater withdrawals before SS, which is exactly the situation I'd be in. The withdrawal amount after SS would be more than cut in half.

The scenario I describe here - 50-50 stock/bond allocation, monies may be drawn from stock or bond side, via sales as well as interest and dividends, is not in the least unusual.

I guess if I had a bond first harvesting approach -- spend only from bonds initially, do not rebalance yearly -- one could pick a duration matched to the likely years needed for that approach (I'll bet the duration for that would wind up being sort of .. intermediate duration bonds). McClung finds this approach better than annual rebalancing. Mostly because it gives stocks a longer time to run.

The big problem of course is that you need a lot of bonds to get something like 4% from a portfolio. Which is why many of us consider 50-50 as sort of a lower bound for stock percentage. And once you are depending upon stock returns for a lot of your withdrawal ....

Thanks for answering the question -- this one here and others in the various threads, I am learning a lot -- but I'm still thinking that in my rather conventional situation, duration matching of the bond side of the portfolio isn't very useful. Or maybe a better way to say it, is that using intermediate duration bonds for my fixed income is very unlikely to be much inferior to other approaches.
User avatar
willthrill81
Posts: 32250
Joined: Thu Jan 26, 2017 3:17 pm
Location: USA

Re: How to use duration matching in a stock/bond portfolio

Post by willthrill81 »

TN_Boy wrote: Wed Aug 03, 2022 8:03 pm Thanks for answering the question -- this one here and others in the various threads, I am learning a lot -- but I'm still thinking that in my rather conventional situation, duration matching of the bond side of the portfolio isn't very useful. Or maybe a better way to say it, is that using intermediate duration bonds for my fixed income is very unlikely to be much inferior to other approaches.
The differences resulting from using intermediate-term bonds rather than duration matching depend greatly on the magnitude of interest rates changes along the way. The greater the changes in interest rates, the more interest rate risk the former will have compared to the latter. But, of course, no one knows how interest rates will change going forward.
I have left the forum but occasionally check PMs.
Florida Orange
Posts: 40
Joined: Thu Jun 16, 2022 2:22 pm

Re: How to use duration matching in a stock/bond portfolio

Post by Florida Orange »

TN-Boy, it reduces NAV volatility at the time when you care. It does nothing to reduce volatility in the interim.
CloseEnough
Posts: 319
Joined: Sun Feb 14, 2021 8:34 am

Re: How to use duration matching in a stock/bond portfolio

Post by CloseEnough »

TN_Boy wrote: Wed Aug 03, 2022 1:18 pm I've been reading and sometimes participating in several threads that talk about duration matching.

See

viewtopic.php?p=6799004#p6799004
and
viewtopic.php?p=5331115#p5331115

And I confess, I don't know how to use duration in my personal situation, which I think is quite common.



Help me out here, I've read several of these threads, but I'm not getting it for the scenario I describe, which I believe is common.
OP, I have been contemplating similar questions, following other threads that bring this issue to the surface. It's a good question.

I think part of the answer for the question you raise, on the facts pattern you present, is it probably won't make much of a difference over time. And, part of the reason you can't really use "duration matching" is that on your fact pattern you don't know the duration.

Good luck.
User avatar
willthrill81
Posts: 32250
Joined: Thu Jan 26, 2017 3:17 pm
Location: USA

Re: How to use duration matching in a stock/bond portfolio

Post by willthrill81 »

CloseEnough wrote: Thu Aug 04, 2022 10:02 am I think part of the answer for the question you raise, on the facts pattern you present, is it probably won't make much of a difference over time.
While true, that's not a compelling reason for not duration matching. It would be akin to saying 'I'll probably be fine by not carrying insurance on my house'.
I have left the forum but occasionally check PMs.
CloseEnough
Posts: 319
Joined: Sun Feb 14, 2021 8:34 am

Re: How to use duration matching in a stock/bond portfolio

Post by CloseEnough »

willthrill81 wrote: Thu Aug 04, 2022 10:07 am
CloseEnough wrote: Thu Aug 04, 2022 10:02 am I think part of the answer for the question you raise, on the facts pattern you present, is it probably won't make much of a difference over time.
While true, that's not a compelling reason for not duration matching. It would be akin to saying 'I'll probably be fine by not carrying insurance on my house'.
I'd say that analogy is pretty far off. I really don't see duration matching as making a huge difference overall in withdrawals of a portfolio on the general facts presented, if the FI is invested in an intermediate bond fund, or some reasonable mix. I have thought about not carrying home insurance, have decided to negotiate a lower rate instead (waiting for a call back now!).

How would you duration match when the duration is not known? Perhaps that can help answer the issue raised by the OP.
dbr
Posts: 41905
Joined: Sun Mar 04, 2007 9:50 am

Re: How to use duration matching in a stock/bond portfolio

Post by dbr »

It looks to me that most of the replies so far only repeat how duration matching is done for a portfolio of bonds. I don't think I see how that is supposed to work for a portfolio of stocks and bonds. The most relevant reply was the one that suggests duration is not a meaningful concept for stocks, so that kind of leaves the question without an answer.
Florida Orange
Posts: 40
Joined: Thu Jun 16, 2022 2:22 pm

Re: How to use duration matching in a stock/bond portfolio

Post by Florida Orange »

Good point, dbr. Duration matching only works for bonds. But by minimizing the risk in the bond portion of the portfolio, you decrease the risk for the whole portfolio.
dbr
Posts: 41905
Joined: Sun Mar 04, 2007 9:50 am

Re: How to use duration matching in a stock/bond portfolio

Post by dbr »

Florida Orange wrote: Thu Aug 04, 2022 11:52 am Good point, dbr. Duration matching only works for bonds. But by minimizing the risk in the bond portion of the portfolio, you decrease the risk for the whole portfolio.
Also a good point. Most of the cases, however, bond risk in a portfolio of stocks and bonds is far outweighed by stock risk, so there would be an issue how much practical effect is to be found there. I think someone mentioned somewhere else that they couldn't quite see how to assign a liability to bonds bought and sold as one rebalances a stock/bond portfolio even as withdrawals from the portfolio are on a known schedule. That is aside from the issue that portfolio withdrawals can be highly variable.

From a portfolio theory point of view one tends to attend to return and risk of the portfolio as a function of return, risk, and correlation of returns of the components, but that is using instantaneous (as in annual behavior of returns) volatility of the components and not a long term model.

I wonder what in practice people who construct bond portfolios used to support spending do with stock investments. The obvious examples I can think of separate the bonds asset completely away from the stock assets and enter them into the plan as income streams, such as a TIPS ladder or letting someone else design the bond portfolio by buying an SPIA.
Florida Orange
Posts: 40
Joined: Thu Jun 16, 2022 2:22 pm

Re: How to use duration matching in a stock/bond portfolio

Post by Florida Orange »

dbr, I understand your point. The value of duration matching is that you can know how much money you will have in bonds at a certain point in the future. That should immunize you, to some extent, from fluctuations on the stock side. Stocks and bonds go up and down, but by utilizing duration matching, at least there can be a known end point for the bond side. That doesn't completely solve the problem you ask about, but at least not everything is at risk. There is one aspect of the portfolio you can count on.
User avatar
vineviz
Posts: 12309
Joined: Tue May 15, 2018 1:55 pm
Location: Baltimore, MD

Re: How to use duration matching in a stock/bond portfolio

Post by vineviz »

dbr wrote: Thu Aug 04, 2022 10:34 am It looks to me that most of the replies so far only repeat how duration matching is done for a portfolio of bonds. I don't think I see how that is supposed to work for a portfolio of stocks and bonds. The most relevant reply was the one that suggests duration is not a meaningful concept for stocks, so that kind of leaves the question without an answer.
Right.

You duration-match the bonds. You do not duration-match the stocks.
"Far more money has been lost by investors preparing for corrections than has been lost in corrections themselves." ~~ Peter Lynch
Corvidae
Posts: 180
Joined: Tue May 10, 2022 1:11 am

Re: How to use duration matching in a stock/bond portfolio

Post by Corvidae »

Unless you are a bank, pension fund, insurance company, or similar, it's easy: you don't.
Topic Author
TN_Boy
Posts: 2945
Joined: Sat Jan 17, 2009 12:51 pm

Re: How to use duration matching in a stock/bond portfolio

Post by TN_Boy »

Florida Orange wrote: Thu Aug 04, 2022 9:55 am TN-Boy, it reduces NAV volatility at the time when you care. It does nothing to reduce volatility in the interim.
If you have longer duration bonds, the NAV volatility of the bond portion of your portfolio will be higher. The overall portfolio volatility (as you change bond duration) depends upon the bond volatility and correlation with the stock component of the portfolio.
Topic Author
TN_Boy
Posts: 2945
Joined: Sat Jan 17, 2009 12:51 pm

Re: How to use duration matching in a stock/bond portfolio

Post by TN_Boy »

Florida Orange wrote: Thu Aug 04, 2022 11:52 am Good point, dbr. Duration matching only works for bonds. But by minimizing the risk in the bond portion of the portfolio, you decrease the risk for the whole portfolio.
My question was, based upon the scenario described in the OP, how would one do duration matching in the bond component of the portfolio and how does it affect portfolio survivability. I don't care about the bonds really, I care about the portfolio.

I do not think I have a definitive answer yet, other than vineviz noting that given the scenario I described -- which was not something unusual -- duration matching probably won't make much of a difference.
Topic Author
TN_Boy
Posts: 2945
Joined: Sat Jan 17, 2009 12:51 pm

Re: How to use duration matching in a stock/bond portfolio

Post by TN_Boy »

dbr wrote: Thu Aug 04, 2022 12:39 pm
Florida Orange wrote: Thu Aug 04, 2022 11:52 am Good point, dbr. Duration matching only works for bonds. But by minimizing the risk in the bond portion of the portfolio, you decrease the risk for the whole portfolio.
Also a good point. Most of the cases, however, bond risk in a portfolio of stocks and bonds is far outweighed by stock risk, so there would be an issue how much practical effect is to be found there. I think someone mentioned somewhere else that they couldn't quite see how to assign a liability to bonds bought and sold as one rebalances a stock/bond portfolio even as withdrawals from the portfolio are on a known schedule. That is aside from the issue that portfolio withdrawals can be highly variable.

From a portfolio theory point of view one tends to attend to return and risk of the portfolio as a function of return, risk, and correlation of returns of the components, but that is using instantaneous (as in annual behavior of returns) volatility of the components and not a long term model.

I wonder what in practice people who construct bond portfolios used to support spending do with stock investments. The obvious examples I can think of separate the bonds asset completely away from the stock assets and enter them into the plan as income streams, such as a TIPS ladder or letting someone else design the bond portfolio by buying an SPIA.
Exactly. My scenario was a rebalanced constant (in real terms) withdrawal from a stock/bond portfolio, and I still don't get how duration matching would be useful for the bonds in the portfolio, as the withdrawal amount from the bond side would vary. And that's with a constant real withdrawal.

If you used a variable withdrawal method, the real withdrawal amount changes, along with whatever the harvesting method does.

I did try and leave a couple of doors open :-). For example, what if one's strategy was "bonds first". I.e. spend the bonds while the stocks run. This is not a commonly used strategy (McClung covers it) but it seems like a situation where duration matching might be interesting. Or McClung's Prime Harvesting which pulls from bonds, but also replenishes the bond side when stocks are up enough.

Thus I am kinda left still thinking that if you are living off a stock bond portfolio -- like a whole lot of folks -- that simply picking an intermediate bond fund is as good as anything.

If you are setting aside money to be placed in bonds to fund a specific real or nominal liability in the future (e.g. $25k in 10 years ..) I understand the point of duration matching. If you are not doing that, I remain unclear on the merits of worrying about this with a stock/bond portfolio used to fund retirement.
Topic Author
TN_Boy
Posts: 2945
Joined: Sat Jan 17, 2009 12:51 pm

Re: How to use duration matching in a stock/bond portfolio

Post by TN_Boy »

willthrill81 wrote: Thu Aug 04, 2022 10:07 am
CloseEnough wrote: Thu Aug 04, 2022 10:02 am I think part of the answer for the question you raise, on the facts pattern you present, is it probably won't make much of a difference over time.
While true, that's not a compelling reason for not duration matching. It would be akin to saying 'I'll probably be fine by not carrying insurance on my house'.
I don't think a very good case has been made in the scenario I give, which is a pretty standard scenario. The insurance analogy seems flawed; I don't follow.
Florida Orange
Posts: 40
Joined: Thu Jun 16, 2022 2:22 pm

Re: How to use duration matching in a stock/bond portfolio

Post by Florida Orange »

It helps with portfolio survivability because you know that the money in the bond fund will be there in a known amount at the time when you need it. Of course, to do duration matching, you first have to determine the duration, or time frame, for when you will need the money. If you are not matching to a specific future liability, you could figure out approximately when you are going to want the money. For example, if you know you are going to want a certain amount of money in about ten years, you could duration match to ten years. It helps because it greatly reduces that possibility that you have to sell shares when prices are down. By not having to sell at an inopportune time, you protect the bond portion of the portfolio thereby enhancing survivability of the overall portfolio. The catch is you do have to have some idea of when you're going to want the money. If you have absolutely no idea, then you can't duration match and I see nothing wrong with just putting it all in an intermediate bond fund.
User avatar
willthrill81
Posts: 32250
Joined: Thu Jan 26, 2017 3:17 pm
Location: USA

Re: How to use duration matching in a stock/bond portfolio

Post by willthrill81 »

CloseEnough wrote: Thu Aug 04, 2022 10:27 am
willthrill81 wrote: Thu Aug 04, 2022 10:07 am
CloseEnough wrote: Thu Aug 04, 2022 10:02 am I think part of the answer for the question you raise, on the facts pattern you present, is it probably won't make much of a difference over time.
While true, that's not a compelling reason for not duration matching. It would be akin to saying 'I'll probably be fine by not carrying insurance on my house'.
I'd say that analogy is pretty far off. I really don't see duration matching as making a huge difference overall in withdrawals of a portfolio on the general facts presented, if the FI is invested in an intermediate bond fund, or some reasonable mix. I have thought about not carrying home insurance, have decided to negotiate a lower rate instead (waiting for a call back now!).
We can make some decisions based on probabilities, but we need to consider downside risk too. If the downside risk would be financially detrimental to us (i.e., our house burns down), it makes good sense to at least consider insuring that risk.

The probability of duration matching making a 'huge difference' one's withdrawals is probably very small. However, it's certainly possible, so if you could take reasonable steps to greatly reduce that risk for little or no cost, why wouldn't you do it?
CloseEnough wrote: Thu Aug 04, 2022 10:27 am How would you duration match when the duration is not known? Perhaps that can help answer the issue raised by the OP.
We never know our remaining investment horizon with precision, but that doesn't mean that throwing up our hands is the best we can do. Rather, we can do something such as estimating our remaining life expectancy and then choose a duration that's half that.
I have left the forum but occasionally check PMs.
User avatar
willthrill81
Posts: 32250
Joined: Thu Jan 26, 2017 3:17 pm
Location: USA

Re: How to use duration matching in a stock/bond portfolio

Post by willthrill81 »

TN_Boy wrote: Thu Aug 04, 2022 2:28 pm
willthrill81 wrote: Thu Aug 04, 2022 10:07 am
CloseEnough wrote: Thu Aug 04, 2022 10:02 am I think part of the answer for the question you raise, on the facts pattern you present, is it probably won't make much of a difference over time.
While true, that's not a compelling reason for not duration matching. It would be akin to saying 'I'll probably be fine by not carrying insurance on my house'.
I don't think a very good case has been made in the scenario I give, which is a pretty standard scenario. The insurance analogy seems flawed; I don't follow.
I elaborated upon the insurance analogy in this post. To put it simply, if you could buy insurance that would almost completely protect you from interest rate risk, and that insurance cost you nothing, wouldn't you get it? That's what duration matching does.

What if a combination of something like rising interest rates and inflation results in intermediate-term bonds returning -1.6% real for the next 40 years, like what happened from 1941-1981? Losing almost half the buying power of one's existing bonds could be a very negative result for an investor with a significant bond allocation. Duration matching greatly reduces interest rate risk, and if this is done with inflation-linked bonds, the unexpected inflation risk is also greatly reduced.
I have left the forum but occasionally check PMs.
Kudu
Posts: 15
Joined: Sat Feb 05, 2022 1:16 pm

Re: How to use duration matching in a stock/bond portfolio

Post by Kudu »

I can't imagine how I would duration-match a stock/bond portfolio, but I can imagine separating my savings into a duration-matched portion and a separate traditional "risk portfolio," which by the way needn't be 100% stocks. It's similar to what people do when they annuitize a portion of their savings.

I'm attracted to the idea, but so far what's winning is my aversion to the extra complexity.
User avatar
willthrill81
Posts: 32250
Joined: Thu Jan 26, 2017 3:17 pm
Location: USA

Re: How to use duration matching in a stock/bond portfolio

Post by willthrill81 »

Kudu wrote: Thu Aug 04, 2022 5:25 pm I can't imagine how I would duration-match a stock/bond portfolio...
It's not very difficult.

1. Determine your investment horizon. Your life expectancy plus 10-15 years is pretty reasonable for this purpose.

2. Your bond duration should be half your investment horizon.

3. Find two bond funds, one with a duration longer than your investment horizon and one with a duration shorter than your investment horizon.

4. Determine the needed split between the two funds to achieve the desired duration. This can be done with the formula below or by trial-and-error in under a minute with a weighted average calculator like this one.

D = Desired duration
L = Duration of fund with the longer duration
S = Duration of fund with the shorter duration

Percentage allocation of the fund with the longer duration = (D-S)/(L-S)
The remainder of the allocation goes to the fund with the shorter duration.
I have left the forum but occasionally check PMs.
dbr
Posts: 41905
Joined: Sun Mar 04, 2007 9:50 am

Re: How to use duration matching in a stock/bond portfolio

Post by dbr »

Kudu wrote: Thu Aug 04, 2022 5:25 pm I can't imagine how I would duration-match a stock/bond portfolio, but I can imagine separating my savings into a duration-matched portion and a separate traditional "risk portfolio," which by the way needn't be 100% stocks. It's similar to what people do when they annuitize a portion of their savings.

I'm attracted to the idea, but so far what's winning is my aversion to the extra complexity.
That's what it comes down to for me. It is not clear that one can establish a clear preference for that kind of separation up to and including the decision to take an annuity. A related problem is deciding between a pension and a lump sum if offered. The decision was easy for me. My esteemed employer did not offer a lump sum. Another variation on the problem is financing delay in taking Social Security benefits, though SS is not bond like in any of its terms. There are differences between liability matching and annuitized assets, including the longevity aspect in particular.
Topic Author
TN_Boy
Posts: 2945
Joined: Sat Jan 17, 2009 12:51 pm

Re: How to use duration matching in a stock/bond portfolio

Post by TN_Boy »

willthrill81 wrote: Thu Aug 04, 2022 7:14 pm
Kudu wrote: Thu Aug 04, 2022 5:25 pm I can't imagine how I would duration-match a stock/bond portfolio...
It's not very difficult.

1. Determine your investment horizon. Your life expectancy plus 10-15 years is pretty reasonable for this purpose.

2. Your bond duration should be half your investment horizon.

3. Find two bond funds, one with a duration longer than your investment horizon and one with a duration shorter than your investment horizon.

4. Determine the needed split between the two funds to achieve the desired duration. This can be done with the formula below or by trial-and-error in under a minute with a weighted average calculator like this one.

D = Desired duration
L = Duration of fund with the longer duration
S = Duration of fund with the shorter duration

Percentage allocation of the fund with the longer duration = (D-S)/(L-S)
The remainder of the allocation goes to the fund with the shorter duration.
Sure. So let us take my scenario in the OP. Investment horizon about 30 years, so bond duration is 15 years. Now, you are saying there is a clear advantage in using the two bond fund approach above over simply picking an intermediate bond fund.

For example:

Year 1: stocks are up a lot, so I sell no bonds (probably take bond interest) but fund withdrawal via stock sales.
Year 2: stocks take a hit, so I sell bonds. Obviously, I'd sell the bonds in the shorter duration fund.
Year 3: maybe I sell some stocks and some bonds, etc

Each year I'm rebalancing the bond fund(s) to shorten the duration to match my investment horizon, regardless of whether or not I sell any bonds.

Right so far? Because the weighted duration of my bond funds is higher than that of an intermediate fund, I am (hopefully) getting more interest, but the portfolio itself is somewhat more volatile ... as my bond duration is more like 15 years, so an occurrence of rising interest rates really punishes the long bonds.

However, I'm always selling the shorter term bonds when selling bonds, so presumably I'm never selling sharply depressed bonds.

Now, this is not terribly complex (no more so than rebalancing stocks) though I'm grappling with really, how much difference is this going to make.
rossington
Posts: 1140
Joined: Fri Jun 07, 2019 2:00 am
Location: Florida

Re: How to use duration matching in a stock/bond portfolio

Post by rossington »

TN_Boy wrote: Thu Aug 04, 2022 7:37 pm
willthrill81 wrote: Thu Aug 04, 2022 7:14 pm
Kudu wrote: Thu Aug 04, 2022 5:25 pm I can't imagine how I would duration-match a stock/bond portfolio...
It's not very difficult.

1. Determine your investment horizon. Your life expectancy plus 10-15 years is pretty reasonable for this purpose.

2. Your bond duration should be half your investment horizon.

3. Find two bond funds, one with a duration longer than your investment horizon and one with a duration shorter than your investment horizon.

4. Determine the needed split between the two funds to achieve the desired duration. This can be done with the formula below or by trial-and-error in under a minute with a weighted average calculator like this one.

D = Desired duration
L = Duration of fund with the longer duration
S = Duration of fund with the shorter duration

Percentage allocation of the fund with the longer duration = (D-S)/(L-S)
The remainder of the allocation goes to the fund with the shorter duration.
Sure. So let us take my scenario in the OP. Investment horizon about 30 years, so bond duration is 15 years. Now, you are saying there is a clear advantage in using the two bond fund approach above over simply picking an intermediate bond fund.

For example:

Year 1: stocks are up a lot, so I sell no bonds (probably take bond interest) but fund withdrawal via stock sales.
Year 2: stocks take a hit, so I sell bonds. Obviously, I'd sell the bonds in the shorter duration fund.
Year 3: maybe I sell some stocks and some bonds, etc

Each year I'm rebalancing the bond fund(s) to shorten the duration to match my investment horizon, regardless of whether or not I sell any bonds.

Right so far? Because the weighted duration of my bond funds is higher than that of an intermediate fund, I am (hopefully) getting more interest, but the portfolio itself is somewhat more volatile ... as my bond duration is more like 15 years, so an occurrence of rising interest rates really punishes the long bonds.

However, I'm always selling the shorter term bonds when selling bonds, so presumably I'm never selling sharply depressed bonds.

Now, this is not terribly complex (no more so than rebalancing stocks) though I'm grappling with really, how much difference is this going to make.
This (underlined area) is what I am trying to figure out, it would seem to me that you'd want to sell the long bonds to get a higher percentage of short bonds as the investment horizon shortens. So if bonds and stocks are down what are the options other than selling the depressed bonds? Take all your dividends and interest?
(Time to go back to the TIPS threads... :oops: ).
"Success is going from failure to failure without loss of enthusiasm." Winston Churchill.
User avatar
willthrill81
Posts: 32250
Joined: Thu Jan 26, 2017 3:17 pm
Location: USA

Re: How to use duration matching in a stock/bond portfolio

Post by willthrill81 »

TN_Boy wrote: Thu Aug 04, 2022 7:37 pmBecause the weighted duration of my bond funds is higher than that of an intermediate fund, I am (hopefully) getting more interest, but the portfolio itself is somewhat more volatile ... as my bond duration is more like 15 years, so an occurrence of rising interest rates really punishes the long bonds.
Since 1978, a portfolio with 60% bonds and 40% in either short-term, intermediate-term, or long-term Treasuries all had very similar standard deviations, but the portfolios with longer term bonds had slightly smaller maximum drawdowns.
TN_Boy wrote: Thu Aug 04, 2022 7:37 pmNow, this is not terribly complex (no more so than rebalancing stocks) though I'm grappling with really, how much difference is this going to make.
The expected benefit of duration matching is zero because we don't know which way interest rate risk will go. It might benefit bondholders, or it might hurt them. Duration matching is about reducing downside risk resulting from interest rate risk.
I have left the forum but occasionally check PMs.
Kudu
Posts: 15
Joined: Sat Feb 05, 2022 1:16 pm

Re: How to use duration matching in a stock/bond portfolio

Post by Kudu »

willthrill81 wrote: Thu Aug 04, 2022 7:14 pm
Kudu wrote: Thu Aug 04, 2022 5:25 pm I can't imagine how I would duration-match a stock/bond portfolio...
It's not very difficult.
Right - I didn't mean that it's difficult, I meant that it's difficult to think about while stocks are in the mix. Hiving off a separate duration-matching portfolio adds complexity - I'm doing just this with money I manage for an older relative, but I'm hesitant to make a similar arrangement for myself because i can imagine reaching a point where I'd prefer something far simpler.
CloseEnough
Posts: 319
Joined: Sun Feb 14, 2021 8:34 am

Re: How to use duration matching in a stock/bond portfolio

Post by CloseEnough »

willthrill81 wrote: Thu Aug 04, 2022 10:07 am
CloseEnough wrote: Thu Aug 04, 2022 10:27 am How would you duration match when the duration is not known? Perhaps that can help answer the issue raised by the OP.
We never know our remaining investment horizon with precision, but that doesn't mean that throwing up our hands is the best we can do. Rather, we can do something such as estimating our remaining life expectancy and then choose a duration that's half that.
Willthrill,

I don't think estimating the remaining life expectancy and choosing a duration that's half that answers the question the OP raised. That would be the case if we knew that withdrawals would be made from the bond side of the portfolio over the life expectancy you estimated, but you really don't know that. And that is the real problem here, with a mixed stock/bond portfolio and withdrawals made to keep the allocation steady, you don't know which side you will be drawing from, so you really can't duration match. You don't know what the duration may be on the bond side, and you really can't estimate it. So I'm not "throwing up our hands", I'm just saying that on the OPs fact pattern (and it is fairly common to many on this site, I think) you really can't duration match in a meaningful way. As suggested above, you could separate out a portion of the fixed income, and duration match that, to for example cover expenses over some known period of time without adjusting the allocation in the rest of the portfolio. In the OPs fact pattern, you really don't know WHEN you will need to pull money from the fixed income side, and I see that as the crux of the problem. I agree we need to consider downside risk. If I didn't, I would not have any fixed income, and wouldn't be considering adding TIPS to my fixed income allocation (prompted in part by your posts, you'll be happy to know!).
Last edited by CloseEnough on Thu Aug 04, 2022 10:58 pm, edited 1 time in total.
friar1610
Posts: 2076
Joined: Sat Nov 29, 2008 9:52 pm
Location: MA South Shore

Re: How to use duration matching in a stock/bond portfolio

Post by friar1610 »

If someone uses a Life-Cycle fund in lieu of separate funds he has basically opted not to try to duration match the bonds, correct? So when he takes withdrawals, he is taking it proportionally from all the assets within the Life-Cycle wrapper, be they stocks or bonds. The bonds are likely to consist of a single intermediate term fund or perhaps that and another bond fund or two (a short-term or a TIPS fund, for example). Is there any evidence that this investor’s withdrawals are subject to more or less risk and volatility than the investor with the identical overall AA who rolls his own and tries to duration-match his withdrawals? It seems to me that the answer to this question could help answer the OP’s original question about whether duration-matching really makes any difference.
Friar1610 | | I have recently concluded that when it comes to investing I am a satisficer, not an optimizer.
User avatar
willthrill81
Posts: 32250
Joined: Thu Jan 26, 2017 3:17 pm
Location: USA

Re: How to use duration matching in a stock/bond portfolio

Post by willthrill81 »

CloseEnough wrote: Thu Aug 04, 2022 10:23 pm
willthrill81 wrote: Thu Aug 04, 2022 10:07 am
CloseEnough wrote: Thu Aug 04, 2022 10:27 am How would you duration match when the duration is not known? Perhaps that can help answer the issue raised by the OP.
We never know our remaining investment horizon with precision, but that doesn't mean that throwing up our hands is the best we can do. Rather, we can do something such as estimating our remaining life expectancy and then choose a duration that's half that.
Willthrill,

I don't think estimating the remaining life expectancy and choosing a duration that's half that answers the question the OP raised. That would be the case if we knew that withdrawals would be made from the bond side of the portfolio over the life expectancy you estimated, but you really don't know that. And that is the real problem here, with a mixed stock/bond portfolio and withdrawals made to keep the allocation steady, you don't know which side you will be drawing from, so you really can't duration match. You don't know what the duration may be on the bond side, and you really can't estimate it. So I'm not "throwing up our hands", I'm just saying that on the OPs fact pattern (and it is fairly common to many on this site, I think) you really can't duration match in a meaningful way. As suggested above, you could separate out a portion of the fixed income, and duration match that, to for example cover expenses over some known period of time without adjusting the allocation in the rest of the portfolio. In the OPs fact pattern, you really don't know WHEN you will need to pull money from the fixed income side, and I see that as the crux of the problem. I agree we need to consider downside risk. If I didn't, I would not have any fixed income, and wouldn't be considering adding TIPS to my fixed income allocation (prompted in part by your posts, you'll be happy to know!).
You're right that we don't know with precision how long anyone will live, precisely what their duration should be, etc. But frankly, those aren't the questions we should really be asking. The bigger point is that duration matching using the best information we have provides at least pretty good insulation from interest rate risk, whereas we know that the lion's share of investors who just put all their fixed income in intermediate-term bonds are seeking to minimize the highly visible market price risk of bonds while ramping up the largely invisible reinvestment risk of the future.

We must be careful to not let perfection be the enemy of good.

I'm glad to hear that my posts helped you! :beer
I have left the forum but occasionally check PMs.
User avatar
willthrill81
Posts: 32250
Joined: Thu Jan 26, 2017 3:17 pm
Location: USA

Re: How to use duration matching in a stock/bond portfolio

Post by willthrill81 »

friar1610 wrote: Thu Aug 04, 2022 10:38 pm If someone uses a Life-Cycle fund in lieu of separate funds he has basically opted not to try to duration match the bonds, correct?
Yes.
friar1610 wrote: Thu Aug 04, 2022 10:38 pmIs there any evidence that this investor’s withdrawals are subject to more or less risk and volatility than the investor with the identical overall AA who rolls his own and tries to duration-match his withdrawals?
Yes, and the evidence is math. Remember that there are two components to interest rate risk: market price risk and reinvestment risk. Most retail investors only consider market price risk, likely because it's highly visible in the short-term in the form of a fluctuating NAV, but for longer term investing, reinvestment risk is usually more impactful.
I have left the forum but occasionally check PMs.
Robot Monster
Posts: 4011
Joined: Sun May 05, 2019 11:23 am

Re: How to use duration matching in a stock/bond portfolio

Post by Robot Monster »

TN_Boy wrote: Thu Aug 04, 2022 2:22 pm For example, what if one's strategy was "bonds first". I.e. spend the bonds while the stocks run. This is not a commonly used strategy (McClung covers it) but it seems like a situation where duration matching might be interesting.
That's my strategy, I suppose. The longest TIPS is 30 years, so the money I'll use for 2050ish onwards is in stocks. I see my TIPS money as covering the next 30ish years, which would imply a 15 year desired duration. Well, my TIPS have a 13 year duration, so not far off.

Maybe this isn't the right way of looking at things. Dunno.
User avatar
vineviz
Posts: 12309
Joined: Tue May 15, 2018 1:55 pm
Location: Baltimore, MD

Re: How to use duration matching in a stock/bond portfolio

Post by vineviz »

TN_Boy wrote: Thu Aug 04, 2022 2:28 pm
willthrill81 wrote: Thu Aug 04, 2022 10:07 am
CloseEnough wrote: Thu Aug 04, 2022 10:02 am I think part of the answer for the question you raise, on the facts pattern you present, is it probably won't make much of a difference over time.
While true, that's not a compelling reason for not duration matching. It would be akin to saying 'I'll probably be fine by not carrying insurance on my house'.
I don't think a very good case has been made in the scenario I give, which is a pretty standard scenario. The insurance analogy seems flawed; I don't follow.
Projecting a difference between duration matching and some sort of constant duration strategy will largely depend on what the constant duration is. The difference will be smaller now, when total bond market has an average duration of 6.7 years, than it was when the average duration was less than 4 years.

Unsurprisingly, the biggest difference between the approaches was for the 1982 retiree cohort. Duration matching with a 60/40 portfolio increased retirement income by about 8%. Contrast that with a 1978 retiree cohort, when duration matching reduced retirement income by about 5%.

Duration matching is most appropriate for someone who wants to eliminate that interest rate risk, and such an investor is probably advised to undertake a more LMP-oriented approach rather than a total return approach
"Far more money has been lost by investors preparing for corrections than has been lost in corrections themselves." ~~ Peter Lynch
dbr
Posts: 41905
Joined: Sun Mar 04, 2007 9:50 am

Re: How to use duration matching in a stock/bond portfolio

Post by dbr »

Robot Monster wrote: Fri Aug 05, 2022 11:08 am
TN_Boy wrote: Thu Aug 04, 2022 2:22 pm For example, what if one's strategy was "bonds first". I.e. spend the bonds while the stocks run. This is not a commonly used strategy (McClung covers it) but it seems like a situation where duration matching might be interesting.
That's my strategy, I suppose. The longest TIPS is 30 years, so the money I'll use for 2050ish onwards is in stocks. I see my TIPS money as covering the next 30ish years, which would imply a 15 year desired duration. Well, my TIPS have a 13 year duration, so not far off.

Maybe this isn't the right way of looking at things. Dunno.
This is a possible choice. An alternative is to either hold a rebalanced portfolio of stocks and bonds and rely on being informed by the nature of safe withdrawal rates, which are largely insensitive to asset allocation, or one can pry out a liability matching tool of another type as exemplified by pensions, annuities, and Social Security.

A point is that almost everyone has a combination of all or many of these options. That could also include the liability matching nature of owning one's home.

It might be interesting to hear if there is anyone who is providing for retirement income almost exclusively from a bond portfolio that is designed for duration match. Things like SPIAs involve someone else duration matching obligations with also pooing risk, so that is not the same at the end of the agency holding the bonds.
dbr
Posts: 41905
Joined: Sun Mar 04, 2007 9:50 am

Re: How to use duration matching in a stock/bond portfolio

Post by dbr »

vineviz wrote: Fri Aug 05, 2022 11:16 am
Projecting a difference between duration matching and some sort of constant duration strategy will largely depend on what the constant duration is. The difference will be smaller now, when total bond market has an average duration of 6.7 years, than it was when the average duration was less than 4 years.

Unsurprisingly, the biggest difference between the approaches was for the 1982 retiree cohort. Duration matching with a 60/40 portfolio increased retirement income by about 8%. Contrast that with a 1978 retiree cohort, when duration matching reduced retirement income by about 5%.

Duration matching is most appropriate for someone who wants to eliminate that interest rate risk, and such an investor is probably advised to undertake a more LMP-oriented approach rather than a total return approach
That is a helpful perspective on the theory of the problem.
CloseEnough
Posts: 319
Joined: Sun Feb 14, 2021 8:34 am

Re: How to use duration matching in a stock/bond portfolio

Post by CloseEnough »

vineviz wrote: Fri Aug 05, 2022 11:16 am
TN_Boy wrote: Thu Aug 04, 2022 2:28 pm
willthrill81 wrote: Thu Aug 04, 2022 10:07 am
CloseEnough wrote: Thu Aug 04, 2022 10:02 am I think part of the answer for the question you raise, on the facts pattern you present, is it probably won't make much of a difference over time.
While true, that's not a compelling reason for not duration matching. It would be akin to saying 'I'll probably be fine by not carrying insurance on my house'.
I don't think a very good case has been made in the scenario I give, which is a pretty standard scenario. The insurance analogy seems flawed; I don't follow.


Duration matching is most appropriate for someone who wants to eliminate that interest rate risk, and such an investor is probably advised to undertake a more LMP-oriented approach rather than a total return approach
Isn't the OPs fact pattern that was used to present these questions based on a total return approach? Although not stated as such, it seems to be it is implied, and is not a LMP-oriented approach. It seems to me that much of the discussion and different views, as well as the inability to clearly answer OPs question, is because of the difference in "approach". Does duration matching to eliminate interest rate risk only work where a LMP-oriented approach is utilized by an investor? And if so is that because the liabilities and duration being matched are more well defined? I am, admittedly, confused by some of this discussion, as well as the discussion on using bond duration glide path in your other thread.
dbr
Posts: 41905
Joined: Sun Mar 04, 2007 9:50 am

Re: How to use duration matching in a stock/bond portfolio

Post by dbr »

CloseEnough wrote: Fri Aug 05, 2022 11:44 am
Isn't the OPs fact pattern that was used to present these questions based on a total return approach? Although not stated as such, it seems to be it is implied, and is not a LMP-oriented approach. It seems to me that much of the discussion and different views, as well as the inability to clearly answer OPs question, is because of the difference in "approach". Does duration matching to eliminate interest rate risk only work where a LMP-oriented approach is utilized by an investor? And if so is that because the liabilities and duration being matched are more well defined? I am, admittedly, confused by some of this discussion, as well as the discussion on using bond duration glide path in your other thread.
Here is the first sentence from the thread started by VineViz on duration matching: "It's a basic tenet of financial planning that one way an investor can minimize their interest rate risk is to match the average duration of their bond holdings with their investment horizon."

viewtopic.php?t=318412

If a person's portfolio holds both stocks and bonds then the questions become things like:

1. In a portfolio of stocks and bonds does it make sense to apply a duration matching method to reduce the risk of the bond component. Is this even possible if there are transactions in bonds to rebalance the portfolio. How does one direct the overall withdrawals from the portfolio to the bonds and to the stocks.

2. Is this or any other concept of duration matching useful to reduce the risk of the entire portfolio. How does this compare to setting the risk according to asset allocation and continuing to control the risk by rebalancing either through contributions, through withdrawals, or through explicit buying and selling of assets.

3. A possible third question is to ask whether a person holding a portfolio of stocks and bonds might be less interested in duration matching the bond component than in prying out an LMP, investing in delayed SS, buying an SPIA with or without COL increases, or opting for a pension if available. Even owning a home rather than selling the home, investing the assets, and paying rent is a piece of liability matching.
Florida Orange
Posts: 40
Joined: Thu Jun 16, 2022 2:22 pm

Re: How to use duration matching in a stock/bond portfolio

Post by Florida Orange »

Implicit in the technique of duration matching is the idea that the investor has some idea of how much money he will need or want and when he will need or want it. It's hard to imagine that someone has no idea whatsoever, except perhaps for a very young investor who probably doesn't need to worry about this anyway.
Parkinglotracer
Posts: 1382
Joined: Fri Dec 20, 2019 3:49 am
Location: Upstate NY

Re: How to use duration matching in a stock/bond portfolio

Post by Parkinglotracer »

Duration implies taking time into account … obviously one has has to make some assumptions as to when one needs their money in order to accomplish this. As to what affect it will have on their portfolio performance no one knows because one can’t predict what interest rates will do or won’t do - but if you don’t duration match you are leaving yourself open to having to get your bond money right after interest rates boom and your bond or bond funds take a hit.

Many would agree usually it is not a big hit; this year with bnd down 8% it might brother you to sell your bond or bond fund.

Tsp G fund, i bonds, CDs, MYGAs, ladder of treasury securities are alternatives to use that will require a little more work than bnd with a 7-8 yr duration.
Topic Author
TN_Boy
Posts: 2945
Joined: Sat Jan 17, 2009 12:51 pm

Re: How to use duration matching in a stock/bond portfolio

Post by TN_Boy »

rossington wrote: Thu Aug 04, 2022 8:00 pm
stuff deleted ...

However, I'm always selling the shorter term bonds when selling bonds, so presumably I'm never selling sharply depressed bonds.

Now, this is not terribly complex (no more so than rebalancing stocks) though I'm grappling with really, how much difference is this going to make.
This (underlined area) is what I am trying to figure out, it would seem to me that you'd want to sell the long bonds to get a higher percentage of short bonds as the investment horizon shortens. So if bonds and stocks are down what are the options other than selling the depressed bonds? Take all your dividends and interest?
(Time to go back to the TIPS threads... :oops: ).
[/quote]

Well now that's a good point. As one is shortening the duration of the bonds during retirement, clearly you have to sell shares in the long bond fund and buy shares in the short bond fund to maintain the desired duration. And of course if you are selling bonds then why not sell the long ones to get the rebalancing?

But this seems counterintuitive ... I'm always selling the longer, more volatile bonds.

Perhaps one of the duration matching experts could confirm what is supposed to happen.
Topic Author
TN_Boy
Posts: 2945
Joined: Sat Jan 17, 2009 12:51 pm

Re: How to use duration matching in a stock/bond portfolio

Post by TN_Boy »

vineviz wrote: Fri Aug 05, 2022 11:16 am
TN_Boy wrote: Thu Aug 04, 2022 2:28 pm
willthrill81 wrote: Thu Aug 04, 2022 10:07 am
CloseEnough wrote: Thu Aug 04, 2022 10:02 am I think part of the answer for the question you raise, on the facts pattern you present, is it probably won't make much of a difference over time.
While true, that's not a compelling reason for not duration matching. It would be akin to saying 'I'll probably be fine by not carrying insurance on my house'.
I don't think a very good case has been made in the scenario I give, which is a pretty standard scenario. The insurance analogy seems flawed; I don't follow.
Projecting a difference between duration matching and some sort of constant duration strategy will largely depend on what the constant duration is. The difference will be smaller now, when total bond market has an average duration of 6.7 years, than it was when the average duration was less than 4 years.

Unsurprisingly, the biggest difference between the approaches was for the 1982 retiree cohort. Duration matching with a 60/40 portfolio increased retirement income by about 8%. Contrast that with a 1978 retiree cohort, when duration matching reduced retirement income by about 5%.

Duration matching is most appropriate for someone who wants to eliminate that interest rate risk, and such an investor is probably advised to undertake a more LMP-oriented approach rather than a total return approach
And of course, the scenario in the OP was a total return approach.

As your response notes, duration matching (which at the start of a retirement means you will have longer duration bonds) may or may not help, mostly depending upon what interest rates do. It is difficult to predict interest rates......
Topic Author
TN_Boy
Posts: 2945
Joined: Sat Jan 17, 2009 12:51 pm

Re: How to use duration matching in a stock/bond portfolio

Post by TN_Boy »

CloseEnough wrote: Fri Aug 05, 2022 11:44 am
vineviz wrote: Fri Aug 05, 2022 11:16 am
TN_Boy wrote: Thu Aug 04, 2022 2:28 pm
willthrill81 wrote: Thu Aug 04, 2022 10:07 am
CloseEnough wrote: Thu Aug 04, 2022 10:02 am I think part of the answer for the question you raise, on the facts pattern you present, is it probably won't make much of a difference over time.
While true, that's not a compelling reason for not duration matching. It would be akin to saying 'I'll probably be fine by not carrying insurance on my house'.
I don't think a very good case has been made in the scenario I give, which is a pretty standard scenario. The insurance analogy seems flawed; I don't follow.


Duration matching is most appropriate for someone who wants to eliminate that interest rate risk, and such an investor is probably advised to undertake a more LMP-oriented approach rather than a total return approach
Isn't the OPs fact pattern that was used to present these questions based on a total return approach? Although not stated as such, it seems to be it is implied, and is not a LMP-oriented approach. It seems to me that much of the discussion and different views, as well as the inability to clearly answer OPs question, is because of the difference in "approach". Does duration matching to eliminate interest rate risk only work where a LMP-oriented approach is utilized by an investor? And if so is that because the liabilities and duration being matched are more well defined? I am, admittedly, confused by some of this discussion, as well as the discussion on using bond duration glide path in your other thread.
CloseEnough,

Sorry my OP wasn't more clear. I was absolutely defining a total return approach, which I believe is what most people actually do with their retirement portfolios. By the OP definition, I was not using any sort of LMP.*

Were I dedicating some monies to LMP (e.g. I want 20k *guaranteed* every year for the next 20 years ...) then sure, I'd be looking at all this duration matching and finding it useful for the monies used for the LMP.

But in a total return portfolio I am liquidating stocks and bonds in a manner that is usually not 100% predictable. And the amount withdrawn may be variable -- I used the standard 4% rule as a start, but noted you might want to use a variable withdrawal method, which makes the timings and amounts of bond withdrawals even more variable.

* Note that most people have SS income, which serves as guaranteed money for whatever expenses the SS will pay for.
Topic Author
TN_Boy
Posts: 2945
Joined: Sat Jan 17, 2009 12:51 pm

Re: How to use duration matching in a stock/bond portfolio

Post by TN_Boy »

dbr wrote: Fri Aug 05, 2022 11:57 am
CloseEnough wrote: Fri Aug 05, 2022 11:44 am
Isn't the OPs fact pattern that was used to present these questions based on a total return approach? Although not stated as such, it seems to be it is implied, and is not a LMP-oriented approach. It seems to me that much of the discussion and different views, as well as the inability to clearly answer OPs question, is because of the difference in "approach". Does duration matching to eliminate interest rate risk only work where a LMP-oriented approach is utilized by an investor? And if so is that because the liabilities and duration being matched are more well defined? I am, admittedly, confused by some of this discussion, as well as the discussion on using bond duration glide path in your other thread.
Here is the first sentence from the thread started by VineViz on duration matching: "It's a basic tenet of financial planning that one way an investor can minimize their interest rate risk is to match the average duration of their bond holdings with their investment horizon."

viewtopic.php?t=318412

If a person's portfolio holds both stocks and bonds then the questions become things like:

1. In a portfolio of stocks and bonds does it make sense to apply a duration matching method to reduce the risk of the bond component. Is this even possible if there are transactions in bonds to rebalance the portfolio. How does one direct the overall withdrawals from the portfolio to the bonds and to the stocks.

2. Is this or any other concept of duration matching useful to reduce the risk of the entire portfolio. How does this compare to setting the risk according to asset allocation and continuing to control the risk by rebalancing either through contributions, through withdrawals, or through explicit buying and selling of assets.

3. A possible third question is to ask whether a person holding a portfolio of stocks and bonds might be less interested in duration matching the bond component than in prying out an LMP, investing in delayed SS, buying an SPIA with or without COL increases, or opting for a pension if available. Even owning a home rather than selling the home, investing the assets, and paying rent is a piece of liability matching.
Yes. I started this thread because I wanted to explore in more detail the merits, if they exist, of duration matching in a typical stocks and bonds retirement portfolio. If there are some merits, great! If there is little to no merit in that scenario, let's not oversell the concept and confuse people, or at least stop confusing me.

As you note, doing things like deferring SS, buying SPIAs, etc. can be done to generate "guaranteed" income while continuing to rely upon a stock/bond portfolio's total return to fund more expenses.

I *think* everybody understands that while historically the total return of a stock/bond portfolio has been reasonably generous, there is nothing guaranteed about the amount of money you can pull from such a portfolio. Which is why supplementing portfolio returns with annuities, SS, pension, an LMP portfolio using bonds, etc is often considered desirable.
Post Reply