Can someone explain specifically how the "bucket" approach to retirement withdrawals is supposed to work?

Discuss all general (i.e. non-personal) investing questions and issues, investing news, and theory.
CULater
Posts: 1569
Joined: Sun Nov 13, 2016 10:59 am

Can someone explain specifically how the "bucket" approach to retirement withdrawals is supposed to work?

Post by CULater » Wed Nov 14, 2018 10:12 am

Generally speaking, the "bucket" approach for taking retirement withdrawals seems to involve the idea of dividing assets between stocks and safer assets such as bonds and cash, and then suspending withdrawals from the stock "bucket" during periods when stocks have negative returns while taking withdrawals from the safe bond or cash bucket instead. Beyond this general notion, I'm not aware of the actual details of how this is supposed to work in practice.

What are the triggers for which bucket withdrawals should come from? There are all kinds of "bucket" strategies, but the simplest version is having a stock and bond bucket: when stocks are "down" withdrawals should be taken from the bond "bucket", and when stocks are "up" then withdrawals should be taken from the "stock" bucket."

How is this "bucket" approach supposed to be implemented in actual practice? Does one determine which "bucket" to withdraw on a monthly distribution basis, or an annual distribution basis, or what? In other words, do I look at the monthly returns of stocks and if those returns are negative, then I take the next month's distribution from bonds and vice versa? Or do I do it on an annual basis; ie., if the annual return from stocks for a given year is negative, I take the next annual return from bonds and vice versa? Anybody have any actual specifics of how the rubber is supposed to hit the road or is the bucket approach just a mantra?

I'd like to look at some backtesting of the stock/bond bucket method but first I need to know if there is any consensus about how it should actually be put into practice. I sort of doubt there is a consensus, but maybe I'm wrong. If there is one, I"d like to do some backtesting to see if it has actually provided a benefit or is just a "feel good" strategy.
May you have the hindsight to know where you've been, The foresight to know where you're going, And the insight to know when you've gone too far. ~ Irish Blessing

AlohaJoe
Posts: 3916
Joined: Mon Nov 26, 2007 2:00 pm
Location: Saigon, Vietnam

Re: Can someone explain specifically how the "bucket" approach to retirement withdrawals is supposed to work?

Post by AlohaJoe » Wed Nov 14, 2018 10:21 am

There is no consensus.

The vagueness, and the inability of proponents to foresee and answer obvious questions, are the biggest weaknesses, in my opinion. It makes backtesting difficult and makes it clear the proponents never did a backtest of anything in their life.

marcopolo
Posts: 1270
Joined: Sat Dec 03, 2016 10:22 am

Re: Can someone explain specifically how the "bucket" approach to retirement withdrawals is supposed to work?

Post by marcopolo » Wed Nov 14, 2018 10:35 am

CULater wrote:
Wed Nov 14, 2018 10:12 am
Generally speaking, the "bucket" approach for taking retirement withdrawals seems to involve the idea of dividing assets between stocks and safer assets such as bonds and cash, and then suspending withdrawals from the stock "bucket" during periods when stocks have negative returns while taking withdrawals from the safe bond or cash bucket instead. Beyond this general notion, I'm not aware of the actual details of how this is supposed to work in practice.

What are the triggers for which bucket withdrawals should come from? There are all kinds of "bucket" strategies, but the simplest version is having a stock and bond bucket: when stocks are "down" withdrawals should be taken from the bond "bucket", and when stocks are "up" then withdrawals should be taken from the "stock" bucket."

How is this "bucket" approach supposed to be implemented in actual practice? Does one determine which "bucket" to withdraw on a monthly distribution basis, or an annual distribution basis, or what? In other words, do I look at the monthly returns of stocks and if those returns are negative, then I take the next month's distribution from bonds and vice versa? Or do I do it on an annual basis; ie., if the annual return from stocks for a given year is negative, I take the next annual return from bonds and vice versa? Anybody have any actual specifics of how the rubber is supposed to hit the road or is the bucket approach just a mantra?

I'd like to look at some backtesting of the stock/bond bucket method but first I need to know if there is any consensus about how it should actually be put into practice. I sort of doubt there is a consensus, but maybe I'm wrong. If there is one, I"d like to do some backtesting to see if it has actually provided a benefit or is just a "feel good" strategy.

I am no expert in this area, but that is very different than my understanding of the "bucket" approach.
What you are describing sounds more like an attempt to maintain a stable asses allocation (AA), where the withdrawals are used to re-balance towards target AA. Similar to what you would do during accumulation phase where contributions go into lagging asset class to nudge AA towards target.

My understanding of 'bucket" approach is that buckets are created for spending during different time periods in the future.
So, you would have one bucket (perhaps in cash equivalents) to carry you from early retirement to start of SS, and a second bucket (invested more aggressively) to cover the expenses beyond SS afterwards. Some use even more buckets to break put into finer granularity time periods.

Again, just my understanding. Agree that it often does seem vaguely defined.
Once in a while you get shown the light, in the strangest of places if you look at it right.

GrowthSeeker
Posts: 200
Joined: Tue May 15, 2018 10:14 pm

Re: Can someone explain specifically how the "bucket" approach to retirement withdrawals is supposed to work?

Post by GrowthSeeker » Wed Nov 14, 2018 10:36 am

CULater wrote:
Wed Nov 14, 2018 10:12 am
What are the triggers for which bucket withdrawals should come from? There are all kinds of "bucket" strategies, but the simplest version is having a stock and bond bucket: when stocks are "down" withdrawals should be taken from the bond "bucket", and when stocks are "up" then withdrawals should be taken from the "stock" bucket."

How is this "bucket" approach supposed to be implemented in actual practice? Does one determine which "bucket" to withdraw on a monthly distribution basis, or an annual distribution basis, or what? In other words, do I look at the monthly returns of stocks and if those returns are negative, then I take the next month's distribution from bonds and vice versa? Or do I do it on an annual basis; ie., if the annual return from stocks for a given year is negative, I take the next annual return from bonds and vice versa? Anybody have any actual specifics of how the rubber is supposed to hit the road or is the bucket approach just a mantra?

I'd like to look at some backtesting of the stock/bond bucket method but first I need to know if there is any consensus about how it should actually be put into practice. I sort of doubt there is a consensus, but maybe I'm wrong. If there is one, I"d like to do some backtesting to see if it has actually provided a benefit or is just a "feel good" strategy.
I don't have any answers. When I read about "buckets" it seemed to me to be a "way of thinking about" one's portfolio as a psychological crutch to help manage anxiety. I think many of the methodologies we discuss, not just on BH but throughout investment discussions tends to boil down to, at least partly, psychological management. I'm not downplaying the importance; one needs that peace of mind to keep it together during market crashes.
And it isn't just psychology, it leads to making real buy and sell choices (though perhaps not very often).

I, for one, would welcome some data of the form:
Way of thinking about it #1 --> these results,
Way of thinking about it #2 --> some other results

BH Wiki:
https://www.bogleheads.org/wiki/Buckets ... strategies

Benz:
https://www.morningstar.com/articles/70 ... estor.html
Just because you're paranoid doesn't mean they're NOT out to get you.

User avatar
Oicuryy
Posts: 1261
Joined: Thu Feb 22, 2007 10:29 pm

Re: Can someone explain specifically how the "bucket" approach to retirement withdrawals is supposed to work?

Post by Oicuryy » Wed Nov 14, 2018 10:37 am

That is a trade secret. You have to pay an investment manager an AUM fee to use their proprietary system to move money from bucket to bucket for you. None of them will tell you the details.

Ron
Money is fungible | Abbreviations and Acronyms

User avatar
goingup
Posts: 3295
Joined: Tue Jan 26, 2010 1:02 pm

Re: Can someone explain specifically how the "bucket" approach to retirement withdrawals is supposed to work?

Post by goingup » Wed Nov 14, 2018 10:40 am

CULater-
You might like reading Jane Bryant Quinn's book How to Make Your Money Last. It has a section that discusses bucketing. I recall the idea was to have a cash bucket, a bond bucket and a stock bucket.

I have some of the same questions you do, though. How do you decide which bucket to use? Most perplexing to me is how/when to replenish the buckets. I doubt we'll use formal buckets in retirement and rather just have a good mix of cash, bonds, and stock to draw from as needed.

Prudence
Posts: 392
Joined: Fri Mar 09, 2012 4:55 pm

Re: Can someone explain specifically how the "bucket" approach to retirement withdrawals is supposed to work?

Post by Prudence » Wed Nov 14, 2018 10:57 am

To me, the bucket approach is having a separate bucket for planned short term withdrawals (cash and cash equivalents and ST Treasuries for money needed for zero to three years); mid-term withdrawals (> 3 years to 6 to 10 years in a medium term bond fund); and long term (> 8 to 10 years in stock fund such as TSM). You can move funds from one bucket to the other once a year or whatever. You can include a TIPs fund if concerned about unexpected inflation. I don't think back testing is relevant or necessary.

nmclean
Posts: 9
Joined: Wed Nov 14, 2018 10:40 am

Re: Can someone explain specifically how the "bucket" approach to retirement withdrawals is supposed to work?

Post by nmclean » Wed Nov 14, 2018 11:30 am

marcopolo wrote:
Wed Nov 14, 2018 10:35 am
I am no expert in this area, but that is very different than my understanding of the "bucket" approach.
What you are describing sounds more like an attempt to maintain a stable asses allocation (AA), where the withdrawals are used to re-balance towards target AA. Similar to what you would do during accumulation phase where contributions go into lagging asset class to nudge AA towards target.

My understanding of 'bucket" approach is that buckets are created for spending during different time periods in the future.
So, you would have one bucket (perhaps in cash equivalents) to carry you from early retirement to start of SS, and a second bucket (invested more aggressively) to cover the expenses beyond SS afterwards. Some use even more buckets to break put into finer granularity time periods.

Again, just my understanding. Agree that it often does seem vaguely defined.
That's the thing, though. It could be described either way, because it's really just imaginary mental accounting.

If you have different buckets for "now" and "later", inescapably "later" becomes "now" and you'll obviously have to shift things to keep "now" replenished. In other words, you'll have to rebalance.

This is easy to see if you imagine rebalancing every single day. We have $x in a "short term" bucket, $y mid-term and $z long term, and we withdraw $x from the short-term bucket. At the end of the day, we would sell from some of the other two buckets and buy more investments for the short term. If we had withdrawn proportionately from each bucket, the result would have been the same.

A possible benefit of this thinking is that the usual approach focuses on percentage weighting. For the "safe" portion of your investments (like cash), it does make more sense to define it as a dollar amount based on your short-term needs, instead of a percentage of your entire savings. But that doesn't mean 100% of your withdrawals should be from that portion.

User avatar
Watty
Posts: 14569
Joined: Wed Oct 10, 2007 3:55 pm

Re: Can someone explain specifically how the "bucket" approach to retirement withdrawals is supposed to work?

Post by Watty » Wed Nov 14, 2018 11:36 am

CULater wrote:
Wed Nov 14, 2018 10:12 am
Generally speaking, the "bucket" approach for taking retirement withdrawals seems to involve the idea of dividing assets between stocks and safer assets such as bonds and cash, and then suspending withdrawals from the stock "bucket" during periods when stocks have negative returns while taking withdrawals from the safe bond or cash bucket instead. Beyond this general notion, I'm not aware of the actual details of how this is supposed to work in practice.
I don't think that is the correct definition of "bucket investing".

When I hear buckets for investing I think of something like a 40 year old might have the investments goals;
1) Kids college education
2) Buy a car in five years
3) emergency fund
4) high expenses in early retirement before they start Social Security and Medicare
5) Lower expenses in retirement after they start Social Security and Medicare
etc.

The money for each of these can be thought of as a separate bucket and invest with an appropriate asset allocation.

I don't know it people would always use it exactly the same way as "buckets" but another related more formal term is Liability Matching Portfolio (LMP). You can search the board or google that to find lots of information on that.

CULater
Posts: 1569
Joined: Sun Nov 13, 2016 10:59 am

Re: Can someone explain specifically how the "bucket" approach to retirement withdrawals is supposed to work?

Post by CULater » Wed Nov 14, 2018 11:55 am

Could be that I'm incorrectly calling the stock/bond approach a "bucket" strategy. What I'm trying to zero in on is what I've heard posters say about having enough of their nested in safe bonds or cash so that they can wait out a bear market in stocks without liquidating their stocks when they are losing value and locking in losses; instead, they would take all their distributions from the bond/cash portion. I know at least one person who told me this was his retirement spending strategy. Usually, the idea seems to be that if you have enough in safe assets to last at least 5 years, you won't have to sell any stocks during a downturn. I'd like to know how this is supposed to work in practice, rather than as a general idea.
May you have the hindsight to know where you've been, The foresight to know where you're going, And the insight to know when you've gone too far. ~ Irish Blessing

marcopolo
Posts: 1270
Joined: Sat Dec 03, 2016 10:22 am

Re: Can someone explain specifically how the "bucket" approach to retirement withdrawals is supposed to work?

Post by marcopolo » Wed Nov 14, 2018 12:02 pm

nmclean wrote:
Wed Nov 14, 2018 11:30 am
marcopolo wrote:
Wed Nov 14, 2018 10:35 am
I am no expert in this area, but that is very different than my understanding of the "bucket" approach.
What you are describing sounds more like an attempt to maintain a stable asses allocation (AA), where the withdrawals are used to re-balance towards target AA. Similar to what you would do during accumulation phase where contributions go into lagging asset class to nudge AA towards target.

My understanding of 'bucket" approach is that buckets are created for spending during different time periods in the future.
So, you would have one bucket (perhaps in cash equivalents) to carry you from early retirement to start of SS, and a second bucket (invested more aggressively) to cover the expenses beyond SS afterwards. Some use even more buckets to break put into finer granularity time periods.

Again, just my understanding. Agree that it often does seem vaguely defined.
That's the thing, though. It could be described either way, because it's really just imaginary mental accounting.

If you have different buckets for "now" and "later", inescapably "later" becomes "now" and you'll obviously have to shift things to keep "now" replenished. In other words, you'll have to rebalance.

This is easy to see if you imagine rebalancing every single day. We have $x in a "short term" bucket, $y mid-term and $z long term, and we withdraw $x from the short-term bucket. At the end of the day, we would sell from some of the other two buckets and buy more investments for the short term. If we had withdrawn proportionately from each bucket, the result would have been the same.

A possible benefit of this thinking is that the usual approach focuses on percentage weighting. For the "safe" portion of your investments (like cash), it does make more sense to define it as a dollar amount based on your short-term needs, instead of a percentage of your entire savings. But that doesn't mean 100% of your withdrawals should be from that portion.
I agree. I am not a fan of "bucket" approaches, for some the reason you mentioned.
I was just pointing out that my understanding of "bucketing" was quite different than what the OP described, not that i was in favor of it.
Once in a while you get shown the light, in the strangest of places if you look at it right.

marcopolo
Posts: 1270
Joined: Sat Dec 03, 2016 10:22 am

Re: Can someone explain specifically how the "bucket" approach to retirement withdrawals is supposed to work?

Post by marcopolo » Wed Nov 14, 2018 12:08 pm

CULater wrote:
Wed Nov 14, 2018 11:55 am
Could be that I'm incorrectly calling the stock/bond approach a "bucket" strategy. What I'm trying to zero in on is what I've heard posters say about having enough of their nested in safe bonds or cash so that they can wait out a bear market in stocks without liquidating their stocks when they are losing value and locking in losses; instead, they would take all their distributions from the bond/cash portion. I know at least one person who told me this was his retirement spending strategy. Usually, the idea seems to be that if you have enough in safe assets to last at least 5 years, you won't have to sell any stocks during a downturn. I'd like to know how this is supposed to work in practice, rather than as a general idea.
You could look at McClung's Prime Harvesting approach, I think that would largely accomplish what you are describing.

Similarly, even if you do not actively re-balance your portfolio, a simple rule such as "withdraw from whichever asset class is above your target" should accomplish what you are describing.

Let's say you decide you would like to be around 60/40 stock/bond. The market drops 50%, bonds stay flat, so now you are 43/57. So, you would withdraw from bonds, lowering their allocation a bit. As long as the market stays down, this rule would have you continue to withdraw from bonds until you got back to 60/40. Likewise, if the market goes up 20%, you would now be over-weighted in stocks, so that is where you would draw from.

My plan is to follow this "withdraw from whichever asset class is above your target", and then actively re-balance if i am still outside my target bands.
Once in a while you get shown the light, in the strangest of places if you look at it right.

Murgatroyd
Posts: 149
Joined: Sun Jan 21, 2018 8:23 pm

Re: Can someone explain specifically how the "bucket" approach to retirement withdrawals is supposed to work?

Post by Murgatroyd » Wed Nov 14, 2018 1:03 pm

This concise book by Larry Stein is very clear on how to create and adjust your buckets
https://www.amazon.com/Peace-Mind-Inves ... ence-ebook

The Wizard
Posts: 12432
Joined: Tue Mar 23, 2010 1:45 pm
Location: Reading, MA

Re: Can someone explain specifically how the "bucket" approach to retirement withdrawals is supposed to work?

Post by The Wizard » Wed Nov 14, 2018 1:15 pm

Watty wrote:
Wed Nov 14, 2018 11:36 am

I don't think that is the correct definition of "bucket investing".

When I hear buckets for investing I think of something like a 40 year old might have the investments goals;
1) Kids college education
2) Buy a car in five years
3) emergency fund
4) high expenses in early retirement before they start Social Security and Medicare
5) Lower expenses in retirement after they start Social Security and Medicare
etc.

The money for each of these can be thought of as a separate bucket and invest with an appropriate asset allocation.

I don't know it people would always use it exactly the same way as "buckets" but another related more formal term is Liability Matching Portfolio (LMP). You can search the board or google that to find lots of information on that.
OP is asking about bucket strategy for Retirement Withdrawals, not investing.
The two are different...
Attempted new signature...

John Z
Posts: 328
Joined: Sun Nov 14, 2010 5:42 pm

Re: Can someone explain specifically how the "bucket" approach to retirement withdrawals is supposed to work?

Post by John Z » Wed Nov 14, 2018 1:24 pm

marcopolo wrote:
Wed Nov 14, 2018 12:08 pm
CULater wrote:
Wed Nov 14, 2018 11:55 am
Could be that I'm incorrectly calling the stock/bond approach a "bucket" strategy. What I'm trying to zero in on is what I've heard posters say about having enough of their nested in safe bonds or cash so that they can wait out a bear market in stocks without liquidating their stocks when they are losing value and locking in losses; instead, they would take all their distributions from the bond/cash portion. I know at least one person who told me this was his retirement spending strategy. Usually, the idea seems to be that if you have enough in safe assets to last at least 5 years, you won't have to sell any stocks during a downturn. I'd like to know how this is supposed to work in practice, rather than as a general idea.
You could look at McClung's Prime Harvesting approach, I think that would largely accomplish what you are describing.
+1
You can download free the TofC and first 3 chapters of McClung's book at
http://livingoffyourmoney.com/
where he gives a simple example of the favored (by his research and backtesting) Prime Harvesting withdrawal method (along with about a dozen other withdrawal methods). I've been using Prime Harvesting since I retired nearly 7 years ago, before his book was published, and I didn't realize that my method had a name. The details of my method and his Prime Harvesting were minor but the concept was the same.
Years ago I was interested in the bucket approach but to me it was too short on details and backtesting and it seemed like I was moving money around based on time/balances/etc. Too complex for me.
Hope it works for you.

MrDrinkingWater
Posts: 98
Joined: Tue Dec 23, 2014 11:30 am

Re: Can someone explain specifically how the "bucket" approach to retirement withdrawals is supposed to work?

Post by MrDrinkingWater » Wed Nov 14, 2018 1:53 pm

A bucket strategy of taking withdrawals from certain buckets probably should not also have a rule of maintaining a constant fixed Asset Allocation (AA). Maybe that is what trips up the folks whose most important rule is maintaining a constant AA.

Making a flow chart that describes how a bucket strategy works is hard because the worst Use Case is spending down the bond bucket and cash bucket to zero over, say, five years, if the stock index funds in the stock bucket are not performing well enough. For some folks, the theoretical position on that last day of the fifth year of never selling any stocks would be an AA of 100%/0%, which might be hard to stick with and do.

I think bucket strategists should have a target AA, but they also need to define a range of tolerable AA ratios that fits into their withdrawal strategy and plan. When their AA reaches, say, 80%/20%, their flow chart should likely define that they need to sell some stocks, even if stocks are lower in value, and they must take a loss.

delamer
Posts: 6385
Joined: Tue Feb 08, 2011 6:13 pm

Re: Can someone explain specifically how the "bucket" approach to retirement withdrawals is supposed to work?

Post by delamer » Wed Nov 14, 2018 1:58 pm

Prudence wrote:
Wed Nov 14, 2018 10:57 am
To me, the bucket approach is having a separate bucket for planned short term withdrawals (cash and cash equivalents and ST Treasuries for money needed for zero to three years); mid-term withdrawals (> 3 years to 6 to 10 years in a medium term bond fund); and long term (> 8 to 10 years in stock fund such as TSM). You can move funds from one bucket to the other once a year or whatever. You can include a TIPs fund if concerned about unexpected inflation. I don't think back testing is relevant or necessary.
This is correct.

You replenish the cash bucket and bond buckets once a year by selling bonds and/or stocks if they have been on the rise.

If they have fallen, then you hold off on replenishing the cash bucket.

You can hold off, if necessary, because if you have 5 years (for example) in cash then you don’t need to sell bonds/stocks when they are down to meet expenses.

The whole concept is very similar to rebalancing to get to your preferred asset allocation. But it focuses on years of expenses rather than percentages.

I am think of Bill Bernstein’s recommendation of 25 years in cash equivalents and the rest of your portfolio in stocks as a radical version of buckets.

Sandi_k
Posts: 829
Joined: Sat May 16, 2015 11:55 am
Location: SF Bay Area

Re: Can someone explain specifically how the "bucket" approach to retirement withdrawals is supposed to work?

Post by Sandi_k » Wed Nov 14, 2018 2:06 pm

CULater wrote:
Wed Nov 14, 2018 11:55 am
Could be that I'm incorrectly calling the stock/bond approach a "bucket" strategy. What I'm trying to zero in on is what I've heard posters say about having enough of their nested in safe bonds or cash so that they can wait out a bear market in stocks without liquidating their stocks when they are losing value and locking in losses; instead, they would take all their distributions from the bond/cash portion. I know at least one person who told me this was his retirement spending strategy. Usually, the idea seems to be that if you have enough in safe assets to last at least 5 years, you won't have to sell any stocks during a downturn. I'd like to know how this is supposed to work in practice, rather than as a general idea.
I agree that this may be seen more as a "psychological crutch" method, AKA one of mental accounting. And frankly, I don't care about backtesting the results, as my focus will revolve around being able to SWAN once retired.

My plan:

For ages 60-64, I will have set aside $250k of my current "investment portfolio" into a 5 year, $50k per year pot of funds, held in bonds/CDs/TIPs. I then will have two buckets: the "investment portfolio" (invested at 40/60 to start) and the "withdrawal portfolio" of $250k.

Each month, we'll transfer $4k as income to the checking account from the "withdrawal portfolio" of $250k. The remaining "investment portfolio" will still be invested in the market, with an asset allocation of 40/60 in year one. Then I'll rebalance annually (which is what I do now). In year two, I move the investment portfolio to 42/58. In year three, 44/56. In year four, 46/54. In year five, 48/52%.

For ages 65-69, the house will be paid off, freeing up another $25k in cash flow annually. DH will also reach FRA in year six, which will increase our income/cash flow even more. We'll begin that year with a 50/50 asset allocation in the investment portfolio, and the "withdrawal portfolio" is now exhausted. This means the drawdown in years 6-10 would be from the regular investment portfolio, in whichever category did well that year. If the stock market is up, I sell stocks, and withdraw from that section of the portfolio. If stocks are down, I sell bonds instead. And I continue the "bond tent" method, moving the AA up 2% per year in stocks, so that I end at age 70 with a 60/40 portfolio, and RMDs and delayed (maximized!) SocSec for me beginning within the next 6 months.

The advantages, as I see it:

- SWAN factor in years 1-5.
- Easy to communicate to my husband.
- Rebalance once per year.
- Bond tent, which helps with the possibility of Sequence of Return risks.
- We allow the "investment portfolio" to ride for the first 5 years without substantial withdrawals, which also helps with SoRR.
- At 70, I'm likely to simply move everything into the Vanguard Wellington fund.

Monster99
Posts: 112
Joined: Tue Aug 29, 2017 10:28 am

Re: Can someone explain specifically how the "bucket" approach to retirement withdrawals is supposed to work?

Post by Monster99 » Wed Nov 14, 2018 2:24 pm

As was mentioned above,
Morningstar did a series on bucketing a few years ago - they did different levels of "aggressiveness" and used different fund families with a little bit of backtesting.
Last edited by Monster99 on Wed Nov 14, 2018 2:29 pm, edited 1 time in total.

User avatar
jazman12
Posts: 224
Joined: Tue Jan 26, 2016 4:38 pm

Re: Can someone explain specifically how the "bucket" approach to retirement withdrawals is supposed to work?

Post by jazman12 » Wed Nov 14, 2018 2:25 pm

Oicuryy wrote:
Wed Nov 14, 2018 10:37 am
That is a trade secret. You have to pay an investment manager an AUM fee to use their proprietary system to move money from bucket to bucket for you. None of them will tell you the details.

Ron
its VOODOO dont you know that!
Act soon... time is running out

Dottie57
Posts: 4773
Joined: Thu May 19, 2016 5:43 pm

Re: Can someone explain specifically how the "bucket" approach to retirement withdrawals is supposed to work?

Post by Dottie57 » Wed Nov 14, 2018 2:38 pm

Before SS I have 8 yearsin cash equivalents. Super safe. This heads more towards liability matching for this period of retirement.

After SS, I want to have 3 years of cash equivalents set up ... no need to decide immediately how to withdraw money from stock/bond funds and I can bide my time. Deciding. Truthfully, unless stocks are down horrifically, I will take distributions the same as my asset allocation.r

If people want to call it a crutch, fine by me.

@Sandy-k

I am trying to do similar.

JustinR
Posts: 775
Joined: Tue Apr 27, 2010 11:43 pm

Re: Can someone explain specifically how the "bucket" approach to retirement withdrawals is supposed to work?

Post by JustinR » Wed Nov 14, 2018 3:57 pm

It works by tricking yourself into thinking you're doing something special when really you're just mental accounting your asset allocation.

User avatar
munemaker
Posts: 3562
Joined: Sat Jan 18, 2014 6:14 pm

Re: Can someone explain specifically how the "bucket" approach to retirement withdrawals is supposed to work?

Post by munemaker » Wed Nov 14, 2018 4:15 pm

Christine Benz at Morningstar is an advocate of the bucket approach and has written a lot about it. Google Christine Benz and Bucket approach; you you will find a lot of information.

The Wizard
Posts: 12432
Joined: Tue Mar 23, 2010 1:45 pm
Location: Reading, MA

Re: Can someone explain specifically how the "bucket" approach to retirement withdrawals is supposed to work?

Post by The Wizard » Wed Nov 14, 2018 4:32 pm

munemaker wrote:
Wed Nov 14, 2018 4:15 pm
Christine Benz at Morningstar is an advocate of the bucket approach and has written a lot about it. Google Christine Benz and Bucket approach; you you will find a lot of information.
While this is correct, it certainly doesn't mean that many of us support a bucket strategy in retirement.
I don't.

What I recommend is having the bulk of your retirement income coming from pension/annuities + age 70 SS.
Portfolio withdrawals before/after age 70 can be supplemental income and should be pro rata monthly withdrawals from your portfolio at your preferred AA.
Simple...
Attempted new signature...

adam1712
Posts: 359
Joined: Fri Jun 01, 2007 5:21 pm

Re: Can someone explain specifically how the "bucket" approach to retirement withdrawals is supposed to work?

Post by adam1712 » Wed Nov 14, 2018 4:56 pm

Prudence wrote:
Wed Nov 14, 2018 10:57 am
To me, the bucket approach is having a separate bucket for planned short term withdrawals (cash and cash equivalents and ST Treasuries for money needed for zero to three years); mid-term withdrawals (> 3 years to 6 to 10 years in a medium term bond fund); and long term (> 8 to 10 years in stock fund such as TSM). You can move funds from one bucket to the other once a year or whatever. You can include a TIPs fund if concerned about unexpected inflation. I don't think back testing is relevant or necessary.
To me, the challenge with a bucket strategy is this: Say you plan for 30 years of retirement-
First year you should have: year 1 bucket, year 2 bucket, ..., year 28 bucket, year 29 bucket, year 30 bucket
Second year you should have: year 1 bucket, year 2 bucket, ..., year 28 bucket, year 29 bucket
Third year you should have: year 1 bucket, year 2 bucket, ..., year 28 bucket

So in reality, you're actually selling your later risky bucket each year to live off. Which seems not what one would want to be doing.

I think it's a useful exercise to determine what your expenses in future years would be and think what the asset allocation would look like for those buckets. But then I prefer just using those asset allocation percentages in a traditional way with your standard rebalancing.

Leesbro63
Posts: 5530
Joined: Mon Nov 08, 2010 4:36 pm

Re: Can someone explain specifically how the "bucket" approach to retirement withdrawals is supposed to work?

Post by Leesbro63 » Wed Nov 14, 2018 4:58 pm

What about for a portfolio 55/45 with a withdrawal rate of 2%? I am hoping to be able to do that in 7 years when I'm 65. I assume the income from the portfolio will fund the withdrawal rate (spending) without having to sell assets. I'm guessing that I'll need a 6 month cushion, at least for a few years, until I see how my income cash flow and outgo cash flow timing coordinate? Anyone else here do this?

CULater
Posts: 1569
Joined: Sun Nov 13, 2016 10:59 am

Re: Can someone explain specifically how the "bucket" approach to retirement withdrawals is supposed to work?

Post by CULater » Wed Nov 14, 2018 6:29 pm

So, I just did a little analysis. Let's assume the investor started with $20K allocated 50% in TSM and 50% in intermediate term Treasuries and exactly caught the stock market decline at the beginning of 2000. He was following the 4% rule, and began taking inflation-adjusted annual withdrawals of $800 from the bond side of his portfolio and not touching the stock side. The stock market did not recover to it's Jan 2000 price level for seven years, and the investor continued to take withdrawals from the bond side for that period of time. As it turns out, he would have ended up with $7,203 in real (inflation-adjusted) dollars remaining in Treasuries and $9,496 real dollars in TSM = a total of $16,699 real after accounting for the 4% real annual withdrawals (I can only look at yearly returns so not quite exact but close enough).

Let's compare that strategy to what would have happened had the investor started with $20K and simply taken the 4% real annual withdrawals from the total portfolio (both stocks and bonds proportionally) and rebalanced annually. Turns out he would have ended up with $16,789 in real (inflation-adjusted) dollars, virtually the same as the strategy of avoiding stock withdrawals in favor of bond withdrawals during this period.

So, it would appear that it wasn't worth the effort to use the stock/bond "buckets" strategy during the 2000-2007 period until stocks had largely recovered their bear market losses. What happened? Well, the strategy obviously allowed the investor to avoid selling stock shares at a loss during this period. However, in order to avoid selling stocks he was liquidating more Treasuries during a period when they had favorable returns, due in large part to the fact that Treasuries usually do well when stocks are doing poorly. There's the catch-22.

My conclusion: this "bucket" strategy doesn't appear to be worth the time and trouble, at least based on results during a time period where it clearly should have worked. Solution: rebalance to maintain your policy allocation and take withdrawals from the total portfolio. At the end of the day it won't make any difference.

All data courtesy of Portfolio Visualizer
May you have the hindsight to know where you've been, The foresight to know where you're going, And the insight to know when you've gone too far. ~ Irish Blessing

ThrustVectoring
Posts: 571
Joined: Wed Jul 12, 2017 2:51 pm

Re: Can someone explain specifically how the "bucket" approach to retirement withdrawals is supposed to work?

Post by ThrustVectoring » Wed Nov 14, 2018 6:51 pm

Leesbro63 wrote:
Wed Nov 14, 2018 4:58 pm
What about for a portfolio 55/45 with a withdrawal rate of 2%? I am hoping to be able to do that in 7 years when I'm 65. I assume the income from the portfolio will fund the withdrawal rate (spending) without having to sell assets. I'm guessing that I'll need a 6 month cushion, at least for a few years, until I see how my income cash flow and outgo cash flow timing coordinate? Anyone else here do this?
If your withdrawal rate is 2% at age 65, anything works. Literally stuffing cash under your mattress will work. 2% is an extraordinarily conservative withdrawal rate.

Anyhow, my personal "bucket" strategy is to use asset allocation at desired portfolio amount to determine breakpoints. So for me, my desired retirement portfolio is 30 years of expenses at 60% stocks / 40% bonds. The 12 years of bonds is to weather any short-term crisis, the 18 years of stocks is to drive long-term growth for later retirement years.

I've already decided to punt on extra crisis-y market returns. If global stock markets are down hard enough and long enough for it to wipe out my retirement, then I need to adjust things by going back to work. So the withdrawal strategy is to draw down from the fixed income bucket, and refill that bucket annually from stocks if I have more than 18x annual expenses in there. Otherwise, let the fixed income draw down and cover my expenses during the bear market. I'm even willing to borrow up to two years of annual expenses with my house and/or stock portfolio as collateral in order to avoid selling stocks in a downturn, so between that and receiving dividends + bond coupons as cash, that should handle any global stock market conditions that involve a negative real return over 16 years. Any worse than that and pretty much everyone needs to cut expenses and/or go back to work.
Current portfolio: 60% VTI / 40% VXUS

User avatar
Artful Dodger
Posts: 493
Joined: Thu Oct 20, 2016 12:56 pm

Re: Can someone explain specifically how the "bucket" approach to retirement withdrawals is supposed to work?

Post by Artful Dodger » Wed Nov 14, 2018 7:27 pm

MrDrinkingWater wrote:
Wed Nov 14, 2018 1:53 pm
A bucket strategy of taking withdrawals from certain buckets probably should not also have a rule of maintaining a constant fixed Asset Allocation (AA). Maybe that is what trips up the folks whose most important rule is maintaining a constant AA.

Making a flow chart that describes how a bucket strategy works is hard because the worst Use Case is spending down the bond bucket and cash bucket to zero over, say, five years, if the stock index funds in the stock bucket are not performing well enough. For some folks, the theoretical position on that last day of the fifth year of never selling any stocks would be an AA of 100%/0%, which might be hard to stick with and do.

I think bucket strategists should have a target AA, but they also need to define a range of tolerable AA ratios that fits into their withdrawal strategy and plan. When their AA reaches, say, 80%/20%, their flow chart should likely define that they need to sell some stocks, even if stocks are lower in value, and they must take a loss.
This is kind of what I've been thinking. I'm 64, spouse 65. Planning to retire in one or two years, delay SS to 70 for both. Current AA is 50 equity/40 bonds/10 cash. I use a mix (the Bucket) of cash, short duration bonds and CDs to to cover expenses until age 70. As I spend down the cash/bonds/CDs, I'll let my Equity percentage in the AA drift up. At age 70, our SS and pension will fund about 110% of necessary expenses. At that point, I'll reallocate, but probably still be comfortable with a 65/35 or so mix of equity to fixed.

RetiredAL
Posts: 167
Joined: Tue Jun 06, 2017 12:09 am
Location: SF Bay Area

Re: Can someone explain specifically how the "bucket" approach to retirement withdrawals is supposed to work?

Post by RetiredAL » Wed Nov 14, 2018 8:01 pm

CULater wrote:
Wed Nov 14, 2018 6:29 pm
So, I just did a little analysis.

My conclusion: this "bucket" strategy doesn't appear to be worth the time and trouble, at least based on results during a time period where it clearly should have worked. Solution: rebalance to maintain your policy allocation and take withdrawals from the total portfolio. At the end of the day it won't make any difference.

All data courtesy of Portfolio Visualizer
CULater,

When I retired, I looked into full blown buckets and I decided it would be an lot of extra work for little real gain.

So I chose a simplified mini bucket approach. The bulk of my withdrawals comes from my Pension Lump Sum derived IRA. That is separate from my 401K derived IRA. From within my Pension IRA, my withdrawal fund (bucket) is a short-term Treasury Fund that normally every 6 months is re-replenish to 18 months worth withdrawals. That Treasury Fund is part of the 55/45 equity ratio I use, along with other Funds/ETF's and CD's for my fixed income. My Investment Policy allows me skip a replenishment if BOTH equities and bonds are down at the same time, or very large drop in equities. I understand this actually does me little in the big picture, but it does give me piece of mind during downturns such as we currently are seeing. I would not do a skip unless there was major decline. This Treasury Fund at max, is only 20% of my Pension fixed income, thus I would still substantially get the bond gain one expects during a typical stock decline. Also, this Pension IRA is only 25% of all of our IRA's.

For the record, I chose at retirement to keep my Pension Lump Sum monies separate from my 401K monies. I did so for a more a simplified tracking for my Pension IRA $, as it actually is being used as a self built and managed account in place of a work provided annuity. This account is expected run out of money around year 20 to 24. Due to issues I have, it's very likely I be gone before that happens, thus my DW or our kids will benefit from the residual $.

randomguy
Posts: 6605
Joined: Wed Sep 17, 2014 9:00 am

Re: Can someone explain specifically how the "bucket" approach to retirement withdrawals is supposed to work?

Post by randomguy » Wed Nov 14, 2018 8:28 pm

delamer wrote:
Wed Nov 14, 2018 1:58 pm
Prudence wrote:
Wed Nov 14, 2018 10:57 am
To me, the bucket approach is having a separate bucket for planned short term withdrawals (cash and cash equivalents and ST Treasuries for money needed for zero to three years); mid-term withdrawals (> 3 years to 6 to 10 years in a medium term bond fund); and long term (> 8 to 10 years in stock fund such as TSM). You can move funds from one bucket to the other once a year or whatever. You can include a TIPs fund if concerned about unexpected inflation. I don't think back testing is relevant or necessary.
This is correct.

You replenish the cash bucket and bond buckets once a year by selling bonds and/or stocks if they have been on the rise.

If they have fallen, then you hold off on replenishing the cash bucket.

You can hold off, if necessary, because if you have 5 years (for example) in cash then you don’t need to sell bonds/stocks when they are down to meet expenses.

The whole concept is very similar to rebalancing to get to your preferred asset allocation. But it focuses on years of expenses rather than percentages.

I am think of Bill Bernstein’s recommendation of 25 years in cash equivalents and the rest of your portfolio in stocks as a radical version of buckets.
So what happens when I retire and the markets are down for 12+ years? That is the sequence of returns that causes problems. Not minor 3-5 blips in the market. You end up with slighlty different situations (buckets have a rising equity glide path, fixed AA are potentially rebalancing and buying low) but over a the difference between the approaches are pretty minimal.

Shadetreeinprocess
Posts: 3
Joined: Mon Aug 13, 2018 8:20 pm

Re: Can someone explain specifically how the "bucket" approach to retirement withdrawals is supposed to work?

Post by Shadetreeinprocess » Wed Nov 14, 2018 8:53 pm

I had always thought, possibly incorrectly after reading all these responses, that the buckets are more associated with tax strategies than asset allocation. I think of it like I have/will have four buckets of money that I can draw from in retirement; 401k, Roth IRA, aftertax savings and social security. Depending on my age, income requirements and legislation at the time, I can determine which is the best bucket to drain. If I can and want to retire at 55, I’ll most likely draw from my aftertax savings bucket. At 59.5; likely my 401k and at 70; ss. If this is a terrible plan; someone let me know!

randomguy
Posts: 6605
Joined: Wed Sep 17, 2014 9:00 am

Re: Can someone explain specifically how the "bucket" approach to retirement withdrawals is supposed to work?

Post by randomguy » Wed Nov 14, 2018 9:01 pm

Shadetreeinprocess wrote:
Wed Nov 14, 2018 8:53 pm
I had always thought, possibly incorrectly after reading all these responses, that the buckets are more associated with tax strategies than asset allocation. I think of it like I have/will have four buckets of money that I can draw from in retirement; 401k, Roth IRA, aftertax savings and social security. Depending on my age, income requirements and legislation at the time, I can determine which is the best bucket to drain. If I can and want to retire at 55, I’ll most likely draw from my aftertax savings bucket. At 59.5; likely my 401k and at 70; ss. If this is a terrible plan; someone let me know!
It is a terrible plan😁 You shoyld be using a strategy that maximizes your spendable income throughout retirement. That pretty much is always a blended approach.

CULater
Posts: 1569
Joined: Sun Nov 13, 2016 10:59 am

Re: Can someone explain specifically how the "bucket" approach to retirement withdrawals is supposed to work?

Post by CULater » Wed Nov 14, 2018 9:06 pm

Kitces compares the results of a "buckets" strategy to a total return rebalancing strategy for taking portfolio withdrawals. The portfolio allocates 60% to equities, 30% to bonds, and 10% to cash.

- The "buckets" are managed using the following decision rules:

1) If equities are up, take the retirement spending from equities

2) If equities are down but bonds are up, take the spending from bonds instead

3) If both equities and bonds are down in the same year, take the distribution from Treasury bills

- The total return rebalancing strategy simply takes withdrawals from the total portfolio and rebalances annually.

If the portfolio is rebalanced to the target bucket allocations every year,
the impact of decision rules is made null and void and the buckets are essentially just an asset allocation mirage, because the total amount of withdrawals is always the same (regardless of which asset classes it’s taken from) and the final allocation is always the same (due to the rebalancing)
In other words -- if you rebalance the portfolio annually, it doesn't make any difference. Your results are exactly the same. The only way there could be a difference is if you don't rebalance using the buckets strategy. In that case, you'll end up letting your portfolio allocations drift from their original targets and any return differences would be largely attributable to different risk portfolio levels.
Nonetheless, the point of all this discussion is not to make the case that decision-rules bucketing strategies are inferior. To the contrary, as long as they are implemented along with rebalancing, their results are exactly the same. And if clients are more comfortable with bucketing strategies – if only because they appeal more naturally to our tendency towards mental accounting – then so much the better.
https://www.kitces.com/blog/managing-se ... -approach/
Last edited by CULater on Wed Nov 14, 2018 9:12 pm, edited 1 time in total.
May you have the hindsight to know where you've been, The foresight to know where you're going, And the insight to know when you've gone too far. ~ Irish Blessing

randomguy
Posts: 6605
Joined: Wed Sep 17, 2014 9:00 am

Re: Can someone explain specifically how the "bucket" approach to retirement withdrawals is supposed to work?

Post by randomguy » Wed Nov 14, 2018 9:10 pm

CULater wrote:
Wed Nov 14, 2018 9:06 pm

In other words -- if you rebalance the portfolio annually, it doesn't make any difference. Your results are exactly the same. Sorta what I found when backtesting the 2-bucket strategy (stocks and Treasuries) but I
Sure but most bucket schemes dont rebalance every year. They have some rules (sticks up, above prior account vale,...) which result in more drift

CULater
Posts: 1569
Joined: Sun Nov 13, 2016 10:59 am

Re: Can someone explain specifically how the "bucket" approach to retirement withdrawals is supposed to work?

Post by CULater » Wed Nov 14, 2018 9:14 pm

randomguy wrote:
Wed Nov 14, 2018 9:10 pm
CULater wrote:
Wed Nov 14, 2018 9:06 pm

In other words -- if you rebalance the portfolio annually, it doesn't make any difference. Your results are exactly the same. Sorta what I found when backtesting the 2-bucket strategy (stocks and Treasuries) but I
Sure but most bucket schemes dont rebalance every year. They have some rules (sticks up, above prior account vale,...) which result in more drift
See above edit. If you don't rebalance then you're allowing the original target allocations to drift and any return difference vs. the total return/rebalancing strategy are trivially related to portfolio risk.
Last edited by CULater on Wed Nov 14, 2018 10:53 pm, edited 1 time in total.
May you have the hindsight to know where you've been, The foresight to know where you're going, And the insight to know when you've gone too far. ~ Irish Blessing

AlohaJoe
Posts: 3916
Joined: Mon Nov 26, 2007 2:00 pm
Location: Saigon, Vietnam

Re: Can someone explain specifically how the "bucket" approach to retirement withdrawals is supposed to work?

Post by AlohaJoe » Wed Nov 14, 2018 9:15 pm

delamer wrote:
Wed Nov 14, 2018 1:58 pm
Prudence wrote:
Wed Nov 14, 2018 10:57 am
To me, the bucket approach is having a separate bucket for planned short term withdrawals (cash and cash equivalents and ST Treasuries for money needed for zero to three years); mid-term withdrawals (> 3 years to 6 to 10 years in a medium term bond fund); and long term (> 8 to 10 years in stock fund such as TSM). You can move funds from one bucket to the other once a year or whatever. You can include a TIPs fund if concerned about unexpected inflation. I don't think back testing is relevant or necessary.
This is correct.

You replenish the cash bucket and bond buckets once a year by selling bonds and/or stocks if they have been on the rise.

If they have fallen, then you hold off on replenishing the cash bucket.
That's not right. Or, rather, sure, that's one possible way to do buckets. It isn't the only possible way. It certainly isn't the way that the guy who invented buckets told people to do it. In the book Buckets of Money, Ray Lucia tells you to:

1. Have 3 buckets: 1) a cash bucket holding 7 years of expenses; 2) a bond bucket holding another 7 years of expenses; 3) an equity bucket holding the rest
2. Spend from your cash bucket. Don't refill it every year. Once it is totally empty (i.e. after 7 years), the sell all of your bonds and put them in the cash bucket. Do not refill the bond bucket. At that point you will have only cash and equities, no more bonds.
3. When you finish the cash bucket again -- after 14 years -- then you sell from the equities bucket and refill both the cash and bond buckets with 7 years of expenses each.

It should be clear that that is very different from "refill the cash and bond buckets every year if stocks have been on the rise". And do we mean "on the rise on a nominal basis" or "on the rise on an inflation-adjusted basis"? In 1978, stocks were up by 5.8%...but inflation was 9% that year. Do you refill or not?

Or what about 2003. Stocks were up 28%. But they lost -22% the year before in 2002. So you still have less money than you did 2 years ago. Do you refill buckets? And actually the losing streak goes all the way back to 2000. $100 in equities in the year 2000 is only worth $80 in 2003 despite "being on the rise" over the past year. Do you refill buckets even though your stocks are still down 20%? Or do you wait until stocks get all the way back to their previous highwater mark? That takes until 2006 in nominal terms. But wait, didn't we decide we need to measure in real terms, not nominal? So even by 2007 you're still not back to where you were in 1999 in inflation-adjusted terms. And then...wouldn't you know it, 2008 happens. So it actually takes all the way until 2013 before you're back to where you were in 1999 in inflation-adjusted terms.

So which is it...do we refill the buckets in 2003? In 2006? Or 2013?

And then there's Paul Grangaard's bucket strategy, which is different from Christine Benz's bucket strategy and Ray Lucia's bucket strategy. He says not to replenish the bonds bucket unless at least 5 years have passed and the real return for stocks have been at least 7%.

It should be clear that all of those differences might lead to different outcomes. How do you know which one to pick? How do you begin to make that decision without doing back testing?

randomguy
Posts: 6605
Joined: Wed Sep 17, 2014 9:00 am

Re: Can someone explain specifically how the "bucket" approach to retirement withdrawals is supposed to work?

Post by randomguy » Wed Nov 14, 2018 9:44 pm

CULater wrote:
Wed Nov 14, 2018 9:14 pm
randomguy wrote:
Wed Nov 14, 2018 9:10 pm
CULater wrote:
Wed Nov 14, 2018 9:06 pm

In other words -- if you rebalance the portfolio annually, it doesn't make any difference. Your results are exactly the same. Sorta what I found when backtesting the 2-bucket strategy (stocks and Treasuries) but I
Sure but most bucket schemes dont rebalance every year. They have some rules (sticks up, above prior account vale,...) which result in more drift
See above edit. If you don't rebalance then you're allowing the original target allocations to drift and any return difference vs. the total return/rebalancing strategy are trivially related to risk.
Of course. The question is does the bucket approach result in either a higher SWR (you get to spend more money) or more stable (i.e. you sleep better at night) results than alternatives. A lot depends on the your system and preferences. Take the 1966 retiree. They would have had less volatility in most cases but they also would have arrived at 1981 with like 100% stocks. I am not sure if that encourages sleeping well at night:) Both end up with about the same total return.

Unfortunately there aren't any great ways of avoiding sequence of return risks. You can avoid market volatility by buying bonds but that does nothing as far as getting you returns. If we could just get half the expected long term returns, we would be north of 5% SWR. Unfortunately there is now good way to do that.

b0B
Posts: 339
Joined: Fri Oct 19, 2018 11:39 am

Re: Can someone explain specifically how the "bucket" approach to retirement withdrawals is supposed to work?

Post by b0B » Wed Nov 14, 2018 10:19 pm

CULater wrote:
Wed Nov 14, 2018 9:06 pm
Kitces compares the results of a "buckets" strategy to a total return rebalancing strategy for taking portfolio withdrawals. The portfolio allocates 60% to equities, 30% to bonds, and 10% to cash.

- The "buckets" are managed using the following decision rules:

1) If equities are up, take the retirement spending from equities

2) If equities are down but bonds are up, take the spending from bonds instead

3) If both equities and bonds are down in the same year, take the distribution from Treasury bills

- The total return rebalancing strategy simply takes withdrawals from the total portfolio and rebalances annually.

If the portfolio is rebalanced to the target bucket allocations every year,
the impact of decision rules is made null and void and the buckets are essentially just an asset allocation mirage, because the total amount of withdrawals is always the same (regardless of which asset classes it’s taken from) and the final allocation is always the same (due to the rebalancing)
In other words -- if you rebalance the portfolio annually, it doesn't make any difference. Your results are exactly the same. The only way there could be a difference is if you don't rebalance using the buckets strategy. In that case, you'll end up letting your portfolio allocations drift from their original targets and any return differences would be largely attributable to different risk portfolio levels.
Nonetheless, the point of all this discussion is not to make the case that decision-rules bucketing strategies are inferior. To the contrary, as long as they are implemented along with rebalancing, their results are exactly the same. And if clients are more comfortable with bucketing strategies – if only because they appeal more naturally to our tendency towards mental accounting – then so much the better.
https://www.kitces.com/blog/managing-se ... -approach/
That's obviously a prank post by Kitces.

dknightd
Posts: 1027
Joined: Wed Mar 07, 2018 11:57 am

Re: Can someone explain specifically how the "bucket" approach to retirement withdrawals is supposed to work?

Post by dknightd » Wed Nov 14, 2018 10:34 pm

Dottie57 wrote:
Wed Nov 14, 2018 2:38 pm
Before SS I have 8 yearsin cash equivalents. Super safe. This heads more towards liability matching for this period of retirement.

After SS, I want to have 3 years of cash equivalents set up ... no need to decide immediately how to withdraw money from stock/bond funds and I can bide my time. Deciding. Truthfully, unless stocks are down horrifically, I will take distributions the same as my asset allocation.r

If people want to call it a crutch, fine by me.

@Sandy-k

I am trying to do similar.
pretty much my plan. 2 buckets, one safe and covers necessary (and some fun, since we all need some fun) expenses. One bucket for pure fun, and who knows what. Define your buckets anyway you want. They are your buckets. Back test them all you want. Everybody has a different idea about buckets that will work for them. Yes it is an accounting tool. I hope you have been using accounting tools. Is it the best one, perhaps not.

User avatar
tadamsmar
Posts: 7878
Joined: Mon May 07, 2007 12:33 pm

Re: Can someone explain specifically how the "bucket" approach to retirement withdrawals is supposed to work?

Post by tadamsmar » Wed Nov 14, 2018 11:01 pm

The Boglehead emergency fund is a kind of bucket.

It's a fund that consists of a fixed amount of money invested in low-volatility liquid assets and it is not part of your AA.

But it gets depleted by an emergency and has to be restored at some point. There is suppose to be some advantage in allowing it to be depleted for a while so you don't have to dip into your more volatile allocation.

If you can get your head around an emergency fund, then I suppose there is some way to get your head around some other kind of bucket strategy.

randomguy
Posts: 6605
Joined: Wed Sep 17, 2014 9:00 am

Re: Can someone explain specifically how the "bucket" approach to retirement withdrawals is supposed to work?

Post by randomguy » Thu Nov 15, 2018 2:45 am

dknightd wrote:
Wed Nov 14, 2018 10:34 pm
Dottie57 wrote:
Wed Nov 14, 2018 2:38 pm
Before SS I have 8 yearsin cash equivalents. Super safe. This heads more towards liability matching for this period of retirement.

After SS, I want to have 3 years of cash equivalents set up ... no need to decide immediately how to withdraw money from stock/bond funds and I can bide my time. Deciding. Truthfully, unless stocks are down horrifically, I will take distributions the same as my asset allocation.r

If people want to call it a crutch, fine by me.

@Sandy-k

I am trying to do similar.
pretty much my plan. 2 buckets, one safe and covers necessary (and some fun, since we all need some fun) expenses. One bucket for pure fun, and who knows what. Define your buckets anyway you want. They are your buckets. Back test them all you want. Everybody has a different idea about buckets that will work for them. Yes it is an accounting tool. I hope you have been using accounting tools. Is it the best one, perhaps not.
The liability matching til SS is a slightly thing and you can make good arguments for it to match a liability that is going away in 8 years. On average you are giving up returns but you don't have time to wait out the market so you don't really have a choice.

The high level point is that stocks and bonds both return -1% real over the next 15 years, it doesn't matter if you use buckets, bond tents, rise equity glide paths or whatever crazy scheme you come up with. Your retirement will not be smooth and you will face some interesting decisions. At a high level holding 50/50 versus 50/30/20(5 years in cash) isn't going to vary much in results. Sure after 5 down years one person is still holding 50/50 while the other is holding 62/38/0 but that is a pretty minor difference in the grand scheme. To some extent it always feels like the bucket people don't understand that you are not going to be selling stocks in a downturn. You will be selling bonds for a long time before you get to the point where selling stocks is required for most retirement AA (i.e. talking the 60/40 to 30/70 range. Not the 100% stock folks:)).

User avatar
dogagility
Posts: 216
Joined: Fri Feb 24, 2017 6:41 am

Re: Can someone explain specifically how the "bucket" approach to retirement withdrawals is supposed to work?

Post by dogagility » Thu Nov 15, 2018 6:14 am

This is how I think of the bucket approach. In my opinion, this approach would be used so that one doesn't have to sell stocks when the market has declined drastically. The safe (bond or CD) bucket is a "bridge" that you spend from when stocks are "down".

One would need to define what "down" is for stocks. Let's say it's a 20% decline from its peak. One would need to have a "bridge" bucket for spending during this stock market trough. One spends equally from their stock/bond asset allocation when stocks are not "down". When the 20% threshold is reached, one spends only from the "safe" bucket. When the bridge is crossed, one replishes the bucket from the stock portion of the portfolio. This strategy by definition has a fluctuating asset allocation which might reach 100% equity depending upon how long the stock market is "down".

To implement this strategy, it seems one needs to predict one thing and know the other thing. One needs to predict how long the "down" (i.e. bridge) stock market will last and one needs to know how much money they will spend during the time the stock market is "down".

How long does the "bridge" need to be? According to this article (https://www.nytimes.com/2009/04/26/your ... 6stra.html),the bridge would have been (less than) 7 years during the Great Depression and (less than) 8 years during the 1970s stock market decline.

In the end, your "safe" bucket amount might be 8 years of spending multiplied by the amount you spend each year.

At least this is my perspective on the approach. YMMV
Taking "risk" since 1995.

User avatar
dogagility
Posts: 216
Joined: Fri Feb 24, 2017 6:41 am

Re: Can someone explain specifically how the "bucket" approach to retirement withdrawals is supposed to work?

Post by dogagility » Thu Nov 15, 2018 6:57 am

dogagility wrote:
Thu Nov 15, 2018 6:14 am
This is how I think of the bucket approach. In my opinion, this approach would be used so that one doesn't have to sell stocks when the market has declined drastically. The safe (bond or CD) bucket is a "bridge" that you spend from when stocks are "down".

One would need to define what "down" is for stocks. Let's say it's a 20% decline from its peak. One would need to have a "bridge" bucket for spending during this stock market trough. One spends equally from their stock/bond asset allocation when stocks are not "down". When the 20% threshold is reached, one spends only from the "safe" bucket. When the bridge is crossed, one replishes the bucket from the stock portion of the portfolio. This strategy by definition has a fluctuating asset allocation which might reach 100% equity depending upon how long the stock market is "down".

To implement this strategy, it seems one needs to predict one thing and know the other thing. One needs to predict how long the "down" (i.e. bridge) stock market will last and one needs to know how much money they will spend during the time the stock market is "down".

How long does the "bridge" need to be? According to this article (https://www.nytimes.com/2009/04/26/your ... 6stra.html),the bridge would have been (less than) 7 years during the Great Depression and (less than) 8 years during the 1970s stock market decline.

In the end, your "safe" bucket amount might be 8 years of spending multiplied by the amount you spend each year.

At least this is my perspective on the approach. YMMV
Thinking about this some more, it seems this strategy has no fixed asset allocation. The AA varies depending upon how the market performs.

A person would need a couple of attributes to implement this strategy:
1) the psychological ability to maintain a high equity exposure during market downturns (and the willingness to see 50 percent "losses" without selling)
2) faith that the stock market will rebound within the specified bridge period

However, it would seem to maximize exposure to the stock market, which in theory should maximize returns.
Taking "risk" since 1995.

CULater
Posts: 1569
Joined: Sun Nov 13, 2016 10:59 am

Re: Can someone explain specifically how the "bucket" approach to retirement withdrawals is supposed to work?

Post by CULater » Thu Nov 15, 2018 8:43 am

Well, first of all simple rebalancing will do most of what the "buckets" decision rules are designed to do, which is to shift more of your withdrawals to cash/bonds during stock market downturns and maintain your target asset allocation. But the bucket strategy can effectively end up being an over-rebalancing strategy that allows your original equity allocation to grow even larger because you are spending down safer assets while not touching stocks. At the end of a long bear market period, stocks can end up as a much larger percentage of your portfolio than at the start, and your portfolio has thereby become much riskier.

Yes, it makes sense to me during a bear market to deplete the safest assets I have while leaving the riskiest ones untouched, making my portfolio even more stock-heavy :oops:
May you have the hindsight to know where you've been, The foresight to know where you're going, And the insight to know when you've gone too far. ~ Irish Blessing

nmclean
Posts: 9
Joined: Wed Nov 14, 2018 10:40 am

Re: Can someone explain specifically how the "bucket" approach to retirement withdrawals is supposed to work?

Post by nmclean » Thu Nov 15, 2018 8:47 am

AlohaJoe wrote:
Wed Nov 14, 2018 9:15 pm
delamer wrote:
Wed Nov 14, 2018 1:58 pm
Prudence wrote:
Wed Nov 14, 2018 10:57 am
To me, the bucket approach is having a separate bucket for planned short term withdrawals (cash and cash equivalents and ST Treasuries for money needed for zero to three years); mid-term withdrawals (> 3 years to 6 to 10 years in a medium term bond fund); and long term (> 8 to 10 years in stock fund such as TSM). You can move funds from one bucket to the other once a year or whatever. You can include a TIPs fund if concerned about unexpected inflation. I don't think back testing is relevant or necessary.
This is correct.

You replenish the cash bucket and bond buckets once a year by selling bonds and/or stocks if they have been on the rise.

If they have fallen, then you hold off on replenishing the cash bucket.
That's not right. Or, rather, sure, that's one possible way to do buckets. It isn't the only possible way. It certainly isn't the way that the guy who invented buckets told people to do it. In the book Buckets of Money, Ray Lucia tells you to:

...

It should be clear that that is very different from "refill the cash and bond buckets every year if stocks have been on the rise".

...

It should be clear that all of those differences might lead to different outcomes. How do you know which one to pick? How do you begin to make that decision without doing back testing?
Classic asset allocation, the "withdraw from buckets and rebalance every year" strategy, Ray Lucia, etc. are all different, sure, but not substantially. The point being it's just rebalancing, which all of these are. The only real difference is rebalancing frequency. The rest is illusion.

As CULater pointed out, it doesn't seem to result in higher returns. Intuitively, there is no reason it should. How do you know which one to pick? I'm convinced the answer is whichever one makes you feel safer (regardless of returns, or actual safety), because that's what all this is really about.

The Wizard
Posts: 12432
Joined: Tue Mar 23, 2010 1:45 pm
Location: Reading, MA

Re: Can someone explain specifically how the "bucket" approach to retirement withdrawals is supposed to work?

Post by The Wizard » Thu Nov 15, 2018 9:17 am

So long as one has enough Millions in retirement for their needs, one can keep it ALL in a so called Cash Bucket and be totally insulated from stock market variations. That's fine.

As for me, a non-Bucketeer, I've been doing pro rata withdrawals from a tax deferred portfolio with 50% stocks and zero cash since I retired at age 63.
As I approach age 70 and the start of maximal SS, I'm letting my stock fund allocation drift upwards to 60% as the markets permit.

Withdrawals from portfolio constitute about 1/3 of my AGI in my pre-70 years and should decline to ~25% of AGI after age 70...
Attempted new signature...

Leesbro63
Posts: 5530
Joined: Mon Nov 08, 2010 4:36 pm

Re: Can someone explain specifically how the "bucket" approach to retirement withdrawals is supposed to work?

Post by Leesbro63 » Thu Nov 15, 2018 9:44 am

There have been multiple "bucket" threads here over the years. I believe that collectively the conclusion was that doing this is just basically mental accounting and that it's really, in the end, just an accounting gimmick that reduces your equity exposure to something below where you tell yourself it is.

The Wizard
Posts: 12432
Joined: Tue Mar 23, 2010 1:45 pm
Location: Reading, MA

Re: Can someone explain specifically how the "bucket" approach to retirement withdrawals is supposed to work?

Post by The Wizard » Thu Nov 15, 2018 9:54 am

Leesbro63 wrote:
Thu Nov 15, 2018 9:44 am
There have been multiple "bucket" threads here over the years. I believe that collectively the conclusion was that doing this is just basically mental accounting and that it's really, in the end, just an accounting gimmick that reduces your equity exposure to something below where you tell yourself it is.
Yep, that's a good short summary.
Plus the moving parts (refilling buckets and changing which bucket to withdraw from) can make it SEEM like it's a more functional plan.

In contrast, the generic Boglehead advice (Don't do something, just stand there) works fine with pro rata monthly withdrawals and occasional rebalancing to target AA...
Attempted new signature...

flee
Posts: 1
Joined: Thu Nov 15, 2018 10:29 am

Re: Can someone explain specifically how the "bucket" approach to retirement withdrawals is supposed to work?

Post by flee » Thu Nov 15, 2018 11:13 am

Seems everyone has a different withdrawal strategy. Basically, spending the higher gain investment. However, we do know rebalancing is always a good strategy that guarantees buying low selling high.

I read a piece from Ben Carlson about retirement investments outcomes for those that were unlucky enough to retire just before market crash. At the end of his evaluation he claimed to gain much new understanding on how important withdrawal strategy was. No, magic formula. Instead the differences we all must customise per our need for security and spending. Ben's most appreciated and powerful factor discovery was the power to have flexible spending. IOWs to lower ones cost of living in retirement. It's a huge advantage and more important than some ultimate withdrawal strategy. The ability to lower spending in bad market years.

Second takeaway from the article was the need for retirees to avoid mindlessly holding investment plan portfolio at all costs. This works for savers and the entire financial industry is more concerned with this saver asset class. It doesn't work so well to see your retirement spending account dwindle 50%. You can't afford the risk and it would be disaster to realize the recovery may be very long indeed. Your retirement income can't afford that time frame. But, nonetheless your long retirement period and ensuing market gains would get hammered if not taking risk and beating inflation. This safety at all cost thinking may be the most dangerous plan for your financial health. The article seem to allude to the idea that retirees should change investments and percentages based on their perceived risk of markets. To not "hold" a portfolio forever. To adjust if inflation is a problem, high p/e risk, raising interest risk, etc. To do so but know full well there will be low spending years ahead. To be invested in stock markets at something like 70/30 in all but the most dangerous time. Maybe dropping to 50/50 if fearful. Always a liquid (cash) hedge for whatever.

Post Reply