Target fund glide path observations

Discuss all general (i.e. non-personal) investing questions and issues, investing news, and theory.
Post Reply
Topic Author
marginal
Posts: 37
Joined: Thu Dec 10, 2020 6:26 pm

Target fund glide path observations

Post by marginal »

I was considering target fund 2030. Its glide path looks like this.
I tried to calculate AA using a ruler from graph, at age 40, 5 years before retirement date at 60, at retirement and +7 years later (at age 72).
I see two issues:
1) On glide, the pre-retirement US stock is decreasing from 41 ->37.7 in 5 years, almost -1% per year.
Basically if you are investing at the same time this can be noise. Is it worth having ramp down?

2) It makes sense to be more conservative before retirement. After retirement I also want to have growth so I can have a reserve for draw down later. Basically make up for slow down and build up for spending at age 72-79.
Do you think ramping up after retirement age 72+ for another 7 years (like V-shape) makes sense?

3) I'm not sure how to interpret slope down sharper after 7 years? AA of Target Income Fund has already reached.

Image

Code: Select all

+----------------------------------+-------------------+-------------------+--------------+--------------------+
|                                  | At -25, AA=90/10  | At (-5) AA=68/32  | At 0, 64/36  | At (+7), AA=30/70  | 
+----------------------------------+-------------------+-------------------+--------------+--------------------+
| Total Market Stock Index         | 54                | 41                | 37.7         | 18                 |
| Total In’l stock Index           | 36                | 27                | 26.7         | 12                 |
| Total Bond Market Index          | 7                 | 22                | 25           | 37                 |
| Total Int’l Bond Index           | 3                 | 10                | 10.6         | 16                 |
| short term Inflation Protected   | -                 | -                 | -            | 17                 |
+----------------------------------+-------------------+-------------------+--------------+--------------------+
-- Marginal
dbr
Posts: 43804
Joined: Sun Mar 04, 2007 8:50 am

Re: Target fund glide path observations

Post by dbr »

There are threads and studies on reverse glidepaths after retirement including the infamous "bond tent" around retirement date. You might try some searching on the forum and Google in general or other posters may have references handy.

Discussions of exactly why Vanguard follows the allocations they propose tend to be futile, but it might be someone will have some explanations that help. There are lots of opinions regarding what asset allocations should be, but it also tends to be true that nuances of asset allocation are not important in retirement except at extremes. Asset allocation matters because it can be varied over a wide range rather than because outcomes are highly sensitive to it.
NiceUnparticularMan
Posts: 4827
Joined: Sat Mar 11, 2017 6:51 am

Re: Target fund glide path observations

Post by NiceUnparticularMan »

I think one key to understanding glide paths like this is there is a rough sort of liability/duration matching going on.

Roughly speaking, stocks are typically modeled as having durations at or above very long term bonds. The "total" bond funds they use are on the short side of intermediate, and then there is the short-term TIPS fund.

That downward acceleration of the stock curve is basically necessitated because first they are increasing the rate at which intermediate bonds are substituting for stocks, and then eventually they need to make room for mixing in more of the ST TIPS too, all of which is collectively serving to decrease the overall portfolio duration. But then they reach a terminal duration, where basically they stop modeling any need to further decrease duration.

From this perspective, ramping up stocks at some point would seem to be implying an increasing desire to use wealth not only to fund the investor's remaining personal spending, but rather to also leave some of it as an inheritance to people or organizations with a much longer implied duration. Which you might well want to do (although I also think giving while alive is nice), but that is a goal which by hypothesis is largely outside of the "Target" framework.

So looking holistically at your personal portfolio, if things go reasonably well with your investments in early retirement, and/or your spending goals end up more moderate than you planned, you might become increasingly confident you have wealth you don't need for the purpose of providing income for you to spend on yourself. And then you might well start putting more of your wealth into something else besides a Target fund, including stock funds.

But Target funds typically are not designed to make that decision for you.
petulant
Posts: 2606
Joined: Thu Sep 22, 2016 1:09 pm

Re: Target fund glide path observations

Post by petulant »

Here's a quick rule of thumb: if you're savvy enough to question the math on the asset allocation of a target date retirement fund, you're savvy enough to identify your goals and create a better asset allocation, glide path, and distribution strategy without the target date retirement fund.
dbr
Posts: 43804
Joined: Sun Mar 04, 2007 8:50 am

Re: Target fund glide path observations

Post by dbr »

NiceUnparticularMan wrote: Tue Jan 24, 2023 8:16 am I think one key to understanding glide paths like this is there is a rough sort of liability/duration matching going on.

That makes sense, but are you aware of anywhere that Vanguard teaches or explains that investing concept and why one would use it.

Their white paper is here: https://institutional.vanguard.com/cont ... Online.pdf
User avatar
jeffyscott
Posts: 11973
Joined: Tue Feb 27, 2007 8:12 am

Re: Target fund glide path observations

Post by jeffyscott »

marginal wrote: Mon Jan 23, 2023 8:52 pm
3) I'm not sure how to interpret slope down sharper after 7 years? AA of Target Income Fund has already reached.
It must be mislabeled, since the prospectus says: Within seven years after 2030, the Fund’s asset allocation should become similar to that of the Target Retirement Income Fund.

And prospectus also says: Effective as of the close of business on July 8, 2022, the reorganization of Vanguard Target Retirement 2015 Fund (the 2015 Fund), a series of Vanguard Chester Funds (the Trust), with and into Vanguard Target Retirement Income Fund, a series of the Trust, is complete. Any references to the 2015 Fund in the Prospectus are hereby deleted.

If you want to know the glide path, maybe just look at the current allocation of target 2025 and target 2020 :?: . In 5 years the 2030 fund will have the 2025 allocation and in 10 years, the 2020 allocation. And in 2037 it will reach the target retirement income allocation.

So it looks like it will go from 64% stock to 55% in the next 5 years and to 43% in the next 10. Decreasing the stock allocation by around 2 percentage points per year on average. Then from 42% to 30% stock in another 4-5 years, so an average decrease closer to 3% per year.

Morningstar gives some additional data points:
The glide path starts with 90% equity exposure until 25 years to retirement and continues to decline until arriving at a 30% equity stake seven years after retirement. The series' equity glide path deviates the most from the peer average near retirement when account balances are likely near their peaks. Five years from retirement, the glide path has 59% equity exposure, 5 percentage points higher than the norm, and at retirement it has 50% equity exposure; 7 percentage points higher than the norm, leaving investors more vulnerable to sudden market drops near the retirement date.

So that also indicates about a 2 percentage point per year decline in the years just prior to target date and then, in going from 50% to 30% over the following 7 years, that is about 3 percentage points per year immediately after the target date.
And so it goes, And so it goes, And so it goes, And so it goes, But where it's goin' no one knows
NiceUnparticularMan
Posts: 4827
Joined: Sat Mar 11, 2017 6:51 am

Re: Target fund glide path observations

Post by NiceUnparticularMan »

petulant wrote: Tue Jan 24, 2023 8:24 am Here's a quick rule of thumb: if you're savvy enough to question the math on the asset allocation of a target date retirement fund, you're savvy enough to identify your goals and create a better asset allocation, glide path, and distribution strategy without the target date retirement fund.
Of course some savvy people might think coming up with something that will predictably do better than a reasonable Target approach is a lot harder of a task than some realize. And if you are very savvy you might conclude the behavioral risks inherent to a DIY approach outweigh any such predictable benefits.
student
Posts: 8457
Joined: Fri Apr 03, 2015 6:58 am

Re: Target fund glide path observations

Post by student »

Please note that different companies use different asset allocation. American Century is more conservative than Fidelity.
dbr
Posts: 43804
Joined: Sun Mar 04, 2007 8:50 am

Re: Target fund glide path observations

Post by dbr »

NiceUnparticularMan wrote: Tue Jan 24, 2023 8:41 am
petulant wrote: Tue Jan 24, 2023 8:24 am Here's a quick rule of thumb: if you're savvy enough to question the math on the asset allocation of a target date retirement fund, you're savvy enough to identify your goals and create a better asset allocation, glide path, and distribution strategy without the target date retirement fund.
Of course some savvy people might think coming up with something that will predictably do better than a reasonable Target approach is a lot harder of a task than some realize. And if you are very savvy you might conclude the behavioral risks inherent to a DIY approach outweigh any such predictable benefits.
In practice unless both spouses of a couple are entirely invested in accounts that can be allocated 100% to any particular TD fund, the "carefully optimized" glide path is blown up anyway. Especially the nuanced allocations to things like international bonds and short TIPS are going to be irrelevant. And that is before the whole life-cycle analysis is undercut by not allowing for Social Security, pensions, annuities, real estate investing and whatever small business enterprise people are engaged in.

That doesn't mean systematically shifting basic components with age does not make sense. It just means the nuanced precision is arbitrary and probably hubris more than analysis.
NiceUnparticularMan
Posts: 4827
Joined: Sat Mar 11, 2017 6:51 am

Re: Target fund glide path observations

Post by NiceUnparticularMan »

dbr wrote: Tue Jan 24, 2023 8:27 am That makes sense, but are you aware of anywhere that Vanguard teaches or explains that investing concept and why one would use it.

Their white paper is here: https://institutional.vanguard.com/cont ... Online.pdf
Sorta, but not in such explicit terms.

If you look Appendices 4 and 5 of that white paper, they provide an overall sketch of the VLCM (Vanguard Life-Cycle Investing Model). Critical inputs include the modeled retirement age, mortality rate, and replacement ratio. They also assume some amount of myopic loss aversion and what they call "income shortfall aversion sensitivity" (which they define in the text as "an investor’s fear of income falling below a certain level") , and finally "preferences toward timing of consumption" (which they don't really discuss in the text, but they list it as a "rational preference" along with risk aversion, and there is a lot of academic literature about why things like the diminishing marginal utility of consumption leads rationally to things like a preference for smoothed consumption).

They then combine all these assumptions with their forward-looking asset model to select a path which offers "the best balance between the amount and volatility of lifetime spending". If you look back at the main text, they explain their balancing approach is somewhat biased toward what they call "maintaining lifestyle" approaches after retirement, although prior to retirement they are more tracking what they call "enhancing lifestyle". See Figure 8 and surrounding discussion.

This all amounts to a type of implied liability/duration matching, not so much as an explicit rule but as a practical implication of using a model with such assumptions and optimization goals.

This paper goes into a little more detail on the VLCM, but I don't think really changes the big picture:

https://corporate.vanguard.com/content/ ... online.pdf

Ultimately, the way I would put it is that liability/duration matching is not a primary goal in itself. Rather, liability/duration matching is a strategy implied by certain assumptions about goals, risks, the relevant assets, and so on.

And really, what Vanguard is doing here is quite a bit more sophisticated than what the DIY community here tends to be doing. That could typically be said of any reputable company's Target fund design, and the federal Thrift Savings Plan's L Funds.

But of course the Vanguard literature also explains that the farther you vary from their modeling assumptions, the more some degree of customization might make sense for you. And something they didn't say, but I believe to be true, is some of of their asset decisions are constrained for other practical reasons, like say the actual size of the TIPS market.
NiceUnparticularMan
Posts: 4827
Joined: Sat Mar 11, 2017 6:51 am

Re: Target fund glide path observations

Post by NiceUnparticularMan »

dbr wrote: Tue Jan 24, 2023 9:44 am That doesn't mean systematically shifting basic components with age does not make sense. It just means the nuanced precision is arbitrary and probably hubris more than analysis.
Indeed.

The positive spin on this is that there is a decent range of approaches you can take with the understanding that proving some other approach within that range would be a better choice is more or less impossible. So you can choose A reasonable approach, and get on with living your life, comfortable in the knowledge that the never-ending quest for THE optimal approach is likely to be a futile one anyway.
User avatar
calmaniac
Posts: 1026
Joined: Fri Jan 30, 2015 2:32 pm

Re: Target fund glide path observations

Post by calmaniac »

marginal wrote: Mon Jan 23, 2023 8:52 pm 2) It makes sense to be more conservative before retirement. After retirement I also want to have growth so I can have a reserve for draw down later. Basically make up for slow down and build up for spending at age 72-79.
Do you think ramping up after retirement age 72+ for another 7 years (like V-shape) makes sense?
Search "Bond tent" on Bogleheads or check out Michael Kitces/Wade Pfau's initial work here: https://www.kitces.com/blog/managing-po ... -red-zone/
≈64yo. AA 75/25: 30% TSM, 19% value (VFVA/AVUV), 18% Int'l LC, 8% emerging, 25% GFund/VBTLX. Fed pensions now ≈60% of expenses. Taking SS @age 70--> pension+SS ≈100% of expenses. What me worry?
JBTX
Posts: 9959
Joined: Wed Jul 26, 2017 12:46 pm

Re: Target fund glide path observations

Post by JBTX »

NiceUnparticularMan wrote: Tue Jan 24, 2023 10:11 am
dbr wrote: Tue Jan 24, 2023 8:27 am That makes sense, but are you aware of anywhere that Vanguard teaches or explains that investing concept and why one would use it.

Their white paper is here: https://institutional.vanguard.com/cont ... Online.pdf
Sorta, but not in such explicit terms.

If you look Appendices 4 and 5 of that white paper, they provide an overall sketch of the VLCM (Vanguard Life-Cycle Investing Model). Critical inputs include the modeled retirement age, mortality rate, and replacement ratio. They also assume some amount of myopic loss aversion and what they call "income shortfall aversion sensitivity" (which they define in the text as "an investor’s fear of income falling below a certain level") , and finally "preferences toward timing of consumption" (which they don't really discuss in the text, but they list it as a "rational preference" along with risk aversion, and there is a lot of academic literature about why things like the diminishing marginal utility of consumption leads rationally to things like a preference for smoothed consumption).

They then combine all these assumptions with their forward-looking asset model to select a path which offers "the best balance between the amount and volatility of lifetime spending". If you look back at the main text, they explain their balancing approach is somewhat biased toward what they call "maintaining lifestyle" approaches after retirement, although prior to retirement they are more tracking what they call "enhancing lifestyle". See Figure 8 and surrounding discussion.

This all amounts to a type of implied liability/duration matching, not so much as an explicit rule but as a practical implication of using a model with such assumptions and optimization goals.

This paper goes into a little more detail on the VLCM, but I don't think really changes the big picture:

https://corporate.vanguard.com/content/ ... online.pdf

Ultimately, the way I would put it is that liability/duration matching is not a primary goal in itself. Rather, liability/duration matching is a strategy implied by certain assumptions about goals, risks, the relevant assets, and so on.

And really, what Vanguard is doing here is quite a bit more sophisticated than what the DIY community here tends to be doing. That could typically be said of any reputable company's Target fund design, and the federal Thrift Savings Plan's L Funds.

But of course the Vanguard literature also explains that the farther you vary from their modeling assumptions, the more some degree of customization might make sense for you. And something they didn't say, but I believe to be true, is some of of their asset decisions are constrained for other practical reasons, like say the actual size of the TIPS market.
You’ve done a very good job of explaining this. I’ve always thought that liability/duration component was kind of obvious and have always been surprised by the resistance TDF downward sloping allocations get on some of these forums. Assuming you are risk averse, and assuming typical longevity tables, for every year you age your life expectancy decreases, it would only make sense that it would be generally downward sloping.

The exceptions would be a lengthy pre social security period resulting in parking a greater portion of cash and bond assets for that period, or legacy/inheritance goals resulting in higher late state stock allocations.
dcabler
Posts: 2923
Joined: Wed Feb 19, 2014 10:30 am
Location: Austin, TX

Re: Target fund glide path observations

Post by dcabler »

A couple of additional resources for you

Bogleheads wiki: https://www.bogleheads.org/wiki/Glide_p ... %20horizon
The wiki hasn't been kept up to date and some of the links are broken, but still a good read.

A bit old, but explains Morningstar's views on this:
https://s21.q4cdn.com/198919461/files/d ... lebook.pdf

Cheers.
User avatar
Ben Mathew
Posts: 2226
Joined: Tue Mar 13, 2018 11:41 am
Location: Seattle

Re: Target fund glide path observations

Post by Ben Mathew »

Vanguard bases their glide path in part on the lifecycle model. From Vanguard's approach to target date funds:
The second strategic principle underlying Vanguard TDFs’ construction is inspired by the human capital theory, which holds that younger investors are better able to withstand portfolio risk (see Figure 2). It recognizes that total net worth consists of both current financial holdings and future work earnings. Most of the younger individuals’ wealth is in the form of what they will earn in the future. This human capital may be looked at as a more stable or bond-like asset, and therefore it may be appropriate for a younger person’s portfolio to have a larger commitment to stocks to balance and diversify risk exposure to work-related earnings (Viceira, 2001; Cocco, Gomes, and Maenhout, 2005).
The basic shape of the glidepath that emerges from a lifecycle model is 100/0 in early career (assuming no leverage), downward sloping in mid to late career, reaching target asset allocation at retirement age (because there is no more future savings coming in—assuming no pensions). Vanguard's target date funds seems to be reaching target asset allocation about 7 years after retirement. In an older version of this document, I remember them saying that they expect people to be able to take on more risk in early retirement because they have the option of going back to work. But I could not find that reference in a quick look at this new version. But it should be about 7 years after retirement, not around 17 as your graph seems to show. I think you may have mislabeled your X axis?

Pensions in retirement will call for a riskier asset allocation in retirement. A gap between retirement and pension start dates will call for asset allocation to increase during the gap—you're basically consuming the extra bonds that cover the gap.

You can calculate a glide path based on the lifecycle model specific to your circumstances using Total portfolio allocation and withdrawal (TPAW).
NiceUnparticularMan
Posts: 4827
Joined: Sat Mar 11, 2017 6:51 am

Re: Target fund glide path observations

Post by NiceUnparticularMan »

Ben Mathew wrote: Wed Jan 25, 2023 1:27 pm Vanguard bases their glide path in part on the lifecycle model. From Vanguard's approach to target date funds:
The second strategic principle underlying Vanguard TDFs’ construction is inspired by the human capital theory, which holds that younger investors are better able to withstand portfolio risk (see Figure 2). It recognizes that total net worth consists of both current financial holdings and future work earnings. Most of the younger individuals’ wealth is in the form of what they will earn in the future. This human capital may be looked at as a more stable or bond-like asset, and therefore it may be appropriate for a younger person’s portfolio to have a larger commitment to stocks to balance and diversify risk exposure to work-related earnings (Viceira, 2001; Cocco, Gomes, and Maenhout, 2005).
The basic shape of the glidepath that emerges from a lifecycle model is 100/0 in early career (assuming no leverage), downward sloping in mid to late career, reaching target asset allocation at retirement age (because there is no more future savings coming in—assuming no pensions). Vanguard's target date funds seems to be reaching target asset allocation about 7 years after retirement. In an older version of this document, I remember them saying that they expect people to be able to take on more risk in early retirement because they have the option of going back to work. But I could not find that reference in a quick look at this new version. But it should be about 7 years after retirement, not around 17 as your graph seems to show. I think you may have mislabeled your X axis?

Pensions in retirement will call for a riskier asset allocation in retirement. A gap between retirement and pension start dates will call for asset allocation to increase during the gap—you're basically consuming the extra bonds that cover the gap.

You can calculate a glide path based on the lifecycle model specific to your circumstances using Total portfolio allocation and withdrawal (TPAW).
Yeah, there is a very popular chart that goes way back that typically looks something like this:

Image

That does suggest the general shape of a downward glide path, but otherwise is not a very close fit for most modern Target glidepaths. Which is because their models are now quite a bit more complex.

We looked at the Vanguard materials, but here is an interesting recent piece from Blackrock:

https://www.blackrock.com/us/financial- ... -inflation

Some key charts from that article . . .

First, this one points out the basic problem with most real retirement scenarios:

Image

This one is about how wages looked for different age-cohorts versus inflation in four different decades. The green line is cohorts starting 55, so this is the last 10 years before an expected retirement at 65. The green lines are pretty darn chaotic sometimes, and generally not as favorable as earlier lines (indeed usually the youngest line is best):

Image

That's all leading to a rather more complex human capital model.

I note as an aside the rest of that Blackrock article is explaining how their model ends up showing people should probably not being using TIPS early in accumulation, as human capital appears very "TIP-like" itself at younger ages (which makes sense). That last period before retirement, though, not so much, so a hard shift into IP bonds during that period would seem to make sense. I note DFA does the same thing.

Anyway, point being all this gets complicated in more sophisticated models. The basic downward-sloping shape, at a high level of generality, remains the same, but the details, including things like the timing, the fixed-income mix, and so on can vary significantly.
User avatar
jeffyscott
Posts: 11973
Joined: Tue Feb 27, 2007 8:12 am

Re: Target fund glide path observations

Post by jeffyscott »

Ben Mathew wrote: Wed Jan 25, 2023 1:27 pm But it should be about 7 years after retirement, not around 17 as your graph seems to show. I think you may have mislabeled your X axis?
The graph and any mislabeling is Vanguard's, not the OP's:
https://advisors.vanguard.com/investmen ... d#overview
And so it goes, And so it goes, And so it goes, And so it goes, But where it's goin' no one knows
User avatar
Ben Mathew
Posts: 2226
Joined: Tue Mar 13, 2018 11:41 am
Location: Seattle

Re: Target fund glide path observations

Post by Ben Mathew »

jeffyscott wrote: Wed Jan 25, 2023 3:02 pm
Ben Mathew wrote: Wed Jan 25, 2023 1:27 pm But it should be about 7 years after retirement, not around 17 as your graph seems to show. I think you may have mislabeled your X axis?
The graph and any mislabeling is Vanguard's, not the OP's:
https://advisors.vanguard.com/investmen ... d#overview
Interesting, thanks.

The illustration on page 8 of Vanguard's approach to target date funds, which seems to show a glide down in retirement from age 65 (retirement) to about age 72—which is about seven years.

The old document that I quoted here also referred explicitly to a seven year glide post-retirement.

If anyone knows why there's a difference, please let me know.
User avatar
jeffyscott
Posts: 11973
Joined: Tue Feb 27, 2007 8:12 am

Re: Target fund glide path observations

Post by jeffyscott »

Ben Mathew wrote: Wed Jan 25, 2023 3:19 pm
jeffyscott wrote: Wed Jan 25, 2023 3:02 pm
Ben Mathew wrote: Wed Jan 25, 2023 1:27 pm But it should be about 7 years after retirement, not around 17 as your graph seems to show. I think you may have mislabeled your X axis?
The graph and any mislabeling is Vanguard's, not the OP's:
https://advisors.vanguard.com/investmen ... d#overview
Interesting, thanks.

The illustration on page 8 of Vanguard's approach to target date funds, which seems to show a glide down in retirement from age 65 (retirement) to about age 72—which is about seven years.

The old document that I quoted here also referred explicitly to a seven year glide post-retirement.

If anyone knows why there's a difference, please let me know.
It's definitely 7 years after retirement to the final allocations. Vanguard has put an erroneous chart on the fund pages there.
And so it goes, And so it goes, And so it goes, And so it goes, But where it's goin' no one knows
User avatar
Ben Mathew
Posts: 2226
Joined: Tue Mar 13, 2018 11:41 am
Location: Seattle

Re: Target fund glide path observations

Post by Ben Mathew »

jeffyscott wrote: Wed Jan 25, 2023 3:23 pm
Ben Mathew wrote: Wed Jan 25, 2023 3:19 pm
jeffyscott wrote: Wed Jan 25, 2023 3:02 pm
Ben Mathew wrote: Wed Jan 25, 2023 1:27 pm But it should be about 7 years after retirement, not around 17 as your graph seems to show. I think you may have mislabeled your X axis?
The graph and any mislabeling is Vanguard's, not the OP's:
https://advisors.vanguard.com/investmen ... d#overview
Interesting, thanks.

The illustration on page 8 of Vanguard's approach to target date funds, which seems to show a glide down in retirement from age 65 (retirement) to about age 72—which is about seven years.

The old document that I quoted here also referred explicitly to a seven year glide post-retirement.

If anyone knows why there's a difference, please let me know.
It's definitely 7 years after retirement to the final allocations. Vanguard has put an erroneous chart on the fund pages there.
Great, thanks.
Topic Author
marginal
Posts: 37
Joined: Thu Dec 10, 2020 6:26 pm

Re: Target fund glide path observations

Post by marginal »

Original documentation is incorrect (see link). As you noticed position of +7 is not in right place.
https://advisors.vanguard.com/investmen ... d#overview

I attempted to correct it from my observations and excellent comments by you. I really liked reading them.

Image

Following institutional web site glide path is different. Look for "Glidepath and asset allocation" in the page.
https://institutional.vanguard.com/inve ... /fund/0308

And Investor site doesn't have it, or I didn't find it.
Post Reply