A bond rule of thumb better than "age in bonds"

Discuss all general (i.e. non-personal) investing questions and issues, investing news, and theory.
User avatar
David Jay
Posts: 10714
Joined: Mon Mar 30, 2015 5:54 am
Location: Michigan

Re: A bond rule of thumb better than "age in bonds"

Post by David Jay »

marcopolo wrote: Mon May 11, 2020 10:57 pmI am still trying to understand the distinction you are making. I am in the same situation as you (did not start at 100% stock), but reduced equity exposure leading up to retirement, and plan to let the equity portion start to glide (drift, rebound, etc.) back up to probably in the 65-70% range eventually. So, I am interested in your approach.

I am not sure what you mean when you seem to imply in your response above that what you are doing is not the same as rising equity percentages. Is that because you are differentiating a passive drift/glide approach from a more active (i.e., powered) approach of forcing the equity percentage to rise?
If anyone wants further discussion of the bond tent, I have created a new thread here: viewtopic.php?f=10&t=314679

This discussion has badly hijacked "vineviz" rule-of-thumb discussion.
Prediction is very difficult, especially about the future - Niels Bohr | To get the "risk premium", you really do have to take the risk - nisiprius
User avatar
NerveDoc
Posts: 29
Joined: Sat Sep 21, 2019 8:58 am

Re: A bond rule of thumb better than "age in bonds"

Post by NerveDoc »

KEotSK66 wrote: Sat May 09, 2020 8:12 am If I recall, Bogle advised not having less than 30% bonds or more than 70% stocks. So, a 70/30 portfolio would be considered the most aggressive and 30/70 the most conservative. Anywhere in-between will come down to someone's own risk tolerance. Pretty simple really.

Interestingly, Mr Bogle gifted his grandchildren VBIAX (Balanced Index) which is 60/40.


I think this recommendation comes from the long-term efficient frontier (SP+ABI), which plots return vs volatility/SD

in the vicinity of 35% stocks you get the minimum variance portfolio, best efficiency in terms of unit return per unit risk

in the vicinity of 65% stocks the unit return per unit risk begins to flatten out so the efficiency begins to drop off rapidly as you assume additional SD, of course the thinking out here is the extra return is worth the extra risk

as your investment period shortens you become more susceptible to "random events" so your efficiency experience may vary considerably

i would just go about 35/65 or 65/35 depending on your needs, 30/70-40/60 or 60/40-70/30 are close enough. vg offers about half a dozen funds with these AAs

i think this glidepath stuff is overcomplicating things unnecessarily
There's a 2018 interview with Bogle where he says "I’ve come to conclude that a 60/40 portfolio is probably the best option, rather than going from 80/20 to 20/80 in a target retirement plan." That's good enough for me.
LeslieSmiley
Posts: 227
Joined: Fri May 08, 2015 7:43 pm

Re: A bond rule of thumb better than "age in bonds"

Post by LeslieSmiley »

120 minus age for stock allocation is what i have been using. the rest is not just "bond" but "fixed income" which is comprised with bond index fund, treasury i-bond, online CD ladder and cold hard cash.
klaus14
Posts: 706
Joined: Sun Nov 25, 2018 7:43 pm

Re: A bond rule of thumb better than "age in bonds"

Post by klaus14 »

vineviz wrote: Thu May 07, 2020 9:58 am For those who prefer to think in terms of wealth, rather than age, there's a roughly equivalent wealth-based rule of thumb.

Compute your expected retirement withdrawals (in current dollars) as a multiple of your current retirement savings: this is (in percentage terms) your bond allocation.

So, for example, imagine that your current retirement savaging are $200,000 and you expect your annual portfolio withdrawals in retirement will be $20,000. 200,000 divided by 20,000 is 10. Your bond allocation is 10%.

If your current retirement savaging are $800,000 and you expect your annual portfolio withdrawals in retirement will be $20,000 then your bond allocation is 40% ( 800,000 divided by 20,000 is 40).

This will produce a slightly different glide path than the age-based rule, but the shape of the glide path will be roughly similar for most investors.
Doesn't this contradict the rule in OP? In OP, you have 50% bonds at the age of retirement. With wealth rule you'll have 30% bonds (assuming 30x savings). And somewhere in thread you said 25% bonds is better at the age of 60 with 25x savings. How about age 50 or 40? Even less bonds?
Last edited by klaus14 on Wed Dec 30, 2020 9:24 pm, edited 1 time in total.
User avatar
abuss368
Posts: 24651
Joined: Mon Aug 03, 2009 2:33 pm
Location: Where the water is warm, the drinks are cold, and I don't know the names of the players!
Contact:

Re: A bond rule of thumb better than "age in bonds"

Post by abuss368 »

Generally speaking the allocation to bonds should probably increase with age. Our glide path and timeframe short ends. Our capitalized human capital decreases. While this may not be true for everyone as facts and circumstances can be unique it is probably true for most.

The best approach may simple be the allocation that allows one to sleep well at night.

Best.
Tony
John C. Bogle: “Simplicity is the master key to financial success."
User avatar
Topic Author
vineviz
Posts: 9837
Joined: Tue May 15, 2018 1:55 pm

Re: A bond rule of thumb better than "age in bonds"

Post by vineviz »

klaus14 wrote: Wed Dec 30, 2020 9:11 pm
vineviz wrote: Thu May 07, 2020 9:58 am For those who prefer to think in terms of wealth, rather than age, there's a roughly equivalent wealth-based rule of thumb.

Compute your expected retirement withdrawals (in current dollars) as a multiple of your current retirement savings: this is (in percentage terms) your bond allocation.

So, for example, imagine that your current retirement savaging are $200,000 and you expect your annual portfolio withdrawals in retirement will be $20,000. 200,000 divided by 20,000 is 10. Your bond allocation is 10%.

If your current retirement savaging are $800,000 and you expect your annual portfolio withdrawals in retirement will be $20,000 then your bond allocation is 40% ( 800,000 divided by 20,000 is 40).

This will produce a slightly different glide path than the age-based rule, but the shape of the glide path will be roughly similar for most investors.
Doesn't this contradict the rule in OP? In OP, you have 50% bonds at the age of retirement. With wealth rule you'll have 30% bonds (assuming 30x savings). And somewhere in thread you said 25% bonds is better at the age of 60 with 25x savings. How about age 50 or 40? Even less bonds?
I'm not sure it's a contradiction, per se, but the two rules would produce slightly different glide paths. Any "rule of thumb" is likely to be problematic, and the wealth-based "rule of thumb" is a bit less conservative than the modified "age-base rule of thumb" from the OP (which itself is like too conservative after age 60 or so).

The main thrust, I think, is that most investors should probably have equity allocations over 80% during most of their 40s. The "age in bonds" rule of thumb is FAR too conservative for people in this age group.
"Far more money has been lost by investors preparing for corrections than has been lost in corrections themselves." ~~ Peter Lynch
klaus14
Posts: 706
Joined: Sun Nov 25, 2018 7:43 pm

Re: A bond rule of thumb better than "age in bonds"

Post by klaus14 »

vineviz wrote: Wed Dec 30, 2020 9:47 pm
klaus14 wrote: Wed Dec 30, 2020 9:11 pm
vineviz wrote: Thu May 07, 2020 9:58 am For those who prefer to think in terms of wealth, rather than age, there's a roughly equivalent wealth-based rule of thumb.

Compute your expected retirement withdrawals (in current dollars) as a multiple of your current retirement savings: this is (in percentage terms) your bond allocation.

So, for example, imagine that your current retirement savaging are $200,000 and you expect your annual portfolio withdrawals in retirement will be $20,000. 200,000 divided by 20,000 is 10. Your bond allocation is 10%.

If your current retirement savaging are $800,000 and you expect your annual portfolio withdrawals in retirement will be $20,000 then your bond allocation is 40% ( 800,000 divided by 20,000 is 40).

This will produce a slightly different glide path than the age-based rule, but the shape of the glide path will be roughly similar for most investors.
Doesn't this contradict the rule in OP? In OP, you have 50% bonds at the age of retirement. With wealth rule you'll have 30% bonds (assuming 30x savings). And somewhere in thread you said 25% bonds is better at the age of 60 with 25x savings. How about age 50 or 40? Even less bonds?
I'm not sure it's a contradiction, per se, but the two rules would produce slightly different glide paths. Any "rule of thumb" is likely to be problematic, and the wealth-based "rule of thumb" is a bit less conservative than the modified "age-base rule of thumb" from the OP (which itself is like too conservative after age 60 or so).

The main thrust, I think, is that most investors should probably have equity allocations over 80% during most of their 40s. The "age in bonds" rule of thumb is FAR too conservative for people in this age group.
How much bond do you think is reasonable for someone entering a hopefully long (50+ years) of retirement with adequate savings (~30x annual expenses) ?
abc132
Posts: 546
Joined: Thu Oct 18, 2018 1:11 am

Re: A bond rule of thumb better than "age in bonds"

Post by abc132 »

These results are very close to what I have gotten in my own simulations. Well done!

I really like the graph with the optimal path.

I'm at 70/25/5 stock/bond/cash at 35x needed/current expenses, but 23x desired retirement expenses, which puts me right in the 23% to 35% range you would suggest. I'm young enough I can work as many years as is needed, so the portfolio value method is more appropriate for me than picking a retirement age. I was also over 40 before getting any bonds, based on my own simulations.

It might be worthwhile to repeat the simulation with 3% less than historical for both stock and bond returns and show that optimal path, just to address the complaint about low expected stock/bond returns.
abc132
Posts: 546
Joined: Thu Oct 18, 2018 1:11 am

Re: A bond rule of thumb better than "age in bonds"

Post by abc132 »

JBTX wrote: Mon May 11, 2020 11:40 pm I read the OP and some posts after but not the entire thread, so sorry if I'm repeating anything.

The issue I have with this, from a life cycle perspective, is you are buying only stocks for 10-20 years, depending when you start working and saving, and then after 40 you on average are buying mostly bonds. If stocks go up more than bonds, you get to the point where you buy only bonds to keep a target allocation. Now we are saying we go to 100% stocks to 60% in only 20 years. A 10-15 year buy in period is a fairly concentrated buy in period, increasing you risk of buying high and eventually selling low. Then after 40 you will have to more aggressively sell stocks from time to time, and to extent you have stocks in taxable it leads to more capital gains management.
I'm in this situation, and you should have plenty of tax sheltered stocks to convert into bonds. You should not plan on changing AA much within a taxable account, because of the large capital gains. You don't need to worry too much about selling stocks low, because low stocks values boosts your bond percentage of your portfolio. Plan ahead for what you need in your taxable accounts so that you don't have to buy or sell there before retiring.

Let's say you have 10x expenses at 40. You need to add 2% bonds per year, which is 0.2x expenses. That's doable with a 20% investment-to-annual-expenses ratio, so less than 16% of your income. The next year the market drops 25%. Now you have 7.5x expenses in stocks and 0.2x expenses in bonds, and you want 4% bonds. You only need 0.04*(7.7)x or .308x expense in bonds, so you buy .108x expenses in bonds and you have 0.092x expenses left to buy some more stocks.

Let's say you have 20x expenses at 50 and the recommended 20% in bonds, for a total of 16x earnings in stocks and 4x earnings in bonds. You need to add 2% bonds per year, which is 0.4x expenses. That's doable with a 40% investment-to-annual-expenses ratio, which is less than 28% of your income. The next year the market drops 25%. Now you have 12x expenses in stocks and 4x expenses in bonds, and you want 22% bonds. You only need 0.22*(16)x or 3.52x expense in bonds, so you can buy all stocks with your investments for the year and you would sell some bonds to purchase more stocks and get back to 78% stocks and 22% bonds.

What you see from these two examples is that it is easier to meet your bond goals when the market is down, and we are more likely to have to sell stocks when they are up. You can buy stocks when they are down, and even rebalance bonds into stocks.
User avatar
Topic Author
vineviz
Posts: 9837
Joined: Tue May 15, 2018 1:55 pm

Re: A bond rule of thumb better than "age in bonds"

Post by vineviz »

klaus14 wrote: Thu Dec 31, 2020 1:50 am How much bond do you think is reasonable for someone entering a hopefully long (50+ years) of retirement with adequate savings (~30x annual expenses) ?
“Reasonable” covers a lot of ground in that scenario, but IMHO the less allocated to bonds the better. Ideally? No bonds.

Fifty years is a long time to cover with 30x expenses, and portfolio returns become the limiting factor, not volatility.
"Far more money has been lost by investors preparing for corrections than has been lost in corrections themselves." ~~ Peter Lynch
klaus14
Posts: 706
Joined: Sun Nov 25, 2018 7:43 pm

Re: A bond rule of thumb better than "age in bonds"

Post by klaus14 »

vineviz wrote: Thu Dec 31, 2020 7:54 am
klaus14 wrote: Thu Dec 31, 2020 1:50 am How much bond do you think is reasonable for someone entering a hopefully long (50+ years) of retirement with adequate savings (~30x annual expenses) ?
“Reasonable” covers a lot of ground in that scenario, but IMHO the less allocated to bonds the better. Ideally? No bonds.

Fifty years is a long time to cover with 30x expenses, and portfolio returns become the limiting factor, not volatility.
Thanks. Let me revert the question then. What kind of asset allocation + savings should you aim to enter a long retirement given today's market conditions?

I guess 50x expenses with 40% bonds should work. But 50x is too difficult and you are working lots of extra years to achieve that. I am curious what your intuition says about the sweet spot here.
klaus14
Posts: 706
Joined: Sun Nov 25, 2018 7:43 pm

Re: A bond rule of thumb better than "age in bonds"

Post by klaus14 »

andrew99999 wrote: Thu May 07, 2020 9:26 pm
vineviz wrote: Thu May 07, 2020 9:58 am For those who prefer to think in terms of wealth, rather than age, there's a roughly equivalent wealth-based rule of thumb.

Compute your expected retirement withdrawals (in current dollars) as a multiple of your current retirement savings: this is (in percentage terms) your bond allocation.

So, for example, imagine that your current retirement savaging are $200,000 and you expect your annual portfolio withdrawals in retirement will be $20,000. 200,000 divided by 20,000 is 10. Your bond allocation is 10%.

If your current retirement savaging are $800,000 and you expect your annual portfolio withdrawals in retirement will be $20,000 then your bond allocation is 40% ( 800,000 divided by 20,000 is 40).

This will produce a slightly different glide path than the age-based rule, but the shape of the glide path will be roughly similar for most investors.
Trying to look at how this works in practice.
If I assume the 4% rule is being used, then it seems to be using a rule of

(Percent of way to goal) divided by 4

So if 20k/yr is required, and therefore target goal = 500k:

(100k / 500k) / 4 = 20% / 4 = 5% bonds
(200k / 500k) / 4 = 40% / 4 = 10% bonds
(300k / 500k) / 4 = 60% / 4 = 15% bonds
(400k / 500k) / 4 = 80% / 4 = 20% bonds
(500k / 500k) / 4 = 100% / 4 = 25% bonds

Have I got that right?

The reason I'm asking is that in your example, you only get to 40% and 50% bonds when you have an 800k and 1m portfolio, so with plans to withdraw 20k/yr it looks like you're basing the numbers on an idea of a 2 - 2.5% withdrawal rate?
I think you are correct except that this method doesn't assume anything about withdrawal rate. Even with less savings and high WR, you have a better chance with high equity allocation. Especially with today's low yields. And if you have high savings multiple then you can "afford" higher bond alloc. This feature is not present in any age based method.

I think this is a better method for people targeting early retirement.

I adopted this method for myself. I add a small tilt on top of it based on valuations.
Method suggests 28% bonds to me but I do 32% because i think stocks are overvalued (i estimate a fair value for SP500 and tilt according to distance).
I have SWR as 2.80% (weighted average of CAPE Yield and 3%)
BHawks87
Posts: 231
Joined: Mon May 07, 2012 2:03 am

Re: A bond rule of thumb better than "age in bonds"

Post by BHawks87 »

andrew99999 wrote: Thu May 07, 2020 9:26 pm
vineviz wrote: Thu May 07, 2020 9:58 am For those who prefer to think in terms of wealth, rather than age, there's a roughly equivalent wealth-based rule of thumb.

Compute your expected retirement withdrawals (in current dollars) as a multiple of your current retirement savings: this is (in percentage terms) your bond allocation.

So, for example, imagine that your current retirement savaging are $200,000 and you expect your annual portfolio withdrawals in retirement will be $20,000. 200,000 divided by 20,000 is 10. Your bond allocation is 10%.

If your current retirement savaging are $800,000 and you expect your annual portfolio withdrawals in retirement will be $20,000 then your bond allocation is 40% ( 800,000 divided by 20,000 is 40).

This will produce a slightly different glide path than the age-based rule, but the shape of the glide path will be roughly similar for most investors.
Trying to look at how this works in practice.
If I assume the 4% rule is being used, then it seems to be using a rule of

(Percent of way to goal) divided by 4

So if 20k/yr is required, and therefore target goal = 500k:

(100k / 500k) / 4 = 20% / 4 = 5% bonds
(200k / 500k) / 4 = 40% / 4 = 10% bonds
(300k / 500k) / 4 = 60% / 4 = 15% bonds
(400k / 500k) / 4 = 80% / 4 = 20% bonds
(500k / 500k) / 4 = 100% / 4 = 25% bonds

Have I got that right?

The reason I'm asking is that in your example, you only get to 40% and 50% bonds when you have an 800k and 1m portfolio, so with plans to withdraw 20k/yr it looks like you're basing the numbers on an idea of a 2 - 2.5% withdrawal rate?
Looks like this formula assumes you will be 75/25 at retirement. Another way to write it to change your desired bond allocation at retirement:

Current Portfolio / Retired Portfolio Goal * Desired Bond Allocation When Retired
hoops777
Posts: 3517
Joined: Sun Apr 10, 2011 12:23 pm

Re: A bond rule of thumb better than "age in bonds"

Post by hoops777 »

When one considers how totally different most people’s scenarios are from one another, I find it puzzling why people are always trying to come up with some “brilliant “ option that works for most people.
K.I.S.S........so easy to say so difficult to do.
UpperNwGuy
Posts: 5146
Joined: Sun Oct 08, 2017 7:16 pm

Re: A bond rule of thumb better than "age in bonds"

Post by UpperNwGuy »

hoops777 wrote: Tue May 04, 2021 7:55 pm When one considers how totally different most people’s scenarios are from one another, I find it puzzling why people are always trying to come up with some “brilliant “ option that works for most people.
Why do you find that to be puzzling? There are plenty of investors, many of whom are members of this forum, who want general guidance.
UpperNwGuy
Posts: 5146
Joined: Sun Oct 08, 2017 7:16 pm

Re: A bond rule of thumb better than "age in bonds"

Post by UpperNwGuy »

I agree with vineviz that an investor doesn't need bonds before age 40, and I like the results of his formula from age 40 to 65. I think the formula needs some refinement for the post-65 years, as it starts to produce very conservative results.
hoops777
Posts: 3517
Joined: Sun Apr 10, 2011 12:23 pm

Re: A bond rule of thumb better than "age in bonds"

Post by hoops777 »

UpperNwGuy wrote: Tue May 04, 2021 8:12 pm
hoops777 wrote: Tue May 04, 2021 7:55 pm When one considers how totally different most people’s scenarios are from one another, I find it puzzling why people are always trying to come up with some “brilliant “ option that works for most people.
Why do you find that to be puzzling? There are plenty of investors, many of whom are members of this forum, who want general guidance.
Just my opinion.
K.I.S.S........so easy to say so difficult to do.
Bridgebumbob
Posts: 15
Joined: Thu Aug 20, 2020 12:12 am

Re: A bond rule of thumb better than "age in bonds"

Post by Bridgebumbob »

When I was younger and in the accumulation phase, I did find the “age in bonds” advice useful. Now that I am well into retirement (age 78), and hoping to leave an inheritance to kids, I focus on actual dollar amounts instead of rules of thumb. Therefore, I plan to keep enough in treasuries (short term now) and cash to last for projected lifespan and the remainder (@30%) in equities (shades of rising equity glidepath) and a $900k home to leave to kids.
Post Reply