A different take on retirement income: time segmentation

Non-investing personal finance issues including insurance, credit, real estate, taxes, employment and legal issues such as trusts and wills
User avatar
willthrill81
Posts: 6436
Joined: Thu Jan 26, 2017 3:17 pm
Location: USA

Re: A different take on retirement income: time segmentation

Post by willthrill81 » Tue Jul 17, 2018 10:09 pm

petulant wrote:
Tue Jul 17, 2018 8:01 pm
willthrill81 wrote:
Tue Jul 17, 2018 7:35 pm
petulant wrote:
Tue Jul 17, 2018 6:39 pm
Well, my point is that once you consider the time impact, the essential portfolio should be more aggressive, while the discretionary portfolio should be more conservative, assuming one is actually at withdrawal. So any exercise in making the discretionary portfolio more aggressive because it’s discretionary would be counteracted by the timing need to keep it conservative, while the long-term needs for an essential portfolio counsel in favor of stocks and counteract your desire for complete stability. Not to be offensive, but it raises the question whether this is a worthwhile exercise.
Why would you want to be more conservative with your discretionary portfolio than with your essential portfolio? If you depleted most of your discretionary portfolio by the time you were in your mid-80s, that's not a real problem. But if you do that with your essential portfolio, you may be in real trouble.
I didn't say it should be more conservative overall. I just said there is a reason to be conservative with it that does not exist with the essential portfolio. The reason to be conservative is that if one retires in their mid-60s, they would reasonably expect a short time horizon to spend a lot of their discretionary portfolio--maybe 10 or 15 years. A shorter time horizon is a reason to stay conservative; on the other hand, the discretionary nature is a reason to stay aggressive. With the essential portfolio, there is a reason to be very conservative, but the time frame involved counsels toward being more aggressive. Overall, I would postulate that these factors are a wash, and there's no reason to create the segmented buckets in the first place. Under this postulate, retirees would be better off creating one portfolio that matches their overall risk tolerance.

Here's an example. Say with the essential portfolio, you suggest TIPS or other conservative investments. However, for a married couple between 60 and 65 looking to retire, there is a substantial possibility at least one will make it to age 90-95. They have 30 years left on their portfolio, which actually advocates for a stock allocation, perhaps a 30/70 stock-bond or 50/50 stock-bond portfolio. Then, with their discretionary assets, they would like to be more aggressive, but they also realize they plan to spend most of the money in the next 10 or so years. This counsels in favor of a 50/50 stock-bond allocation, or maybe a 70/30 at the most aggressive, for this couple. The result is that the couple could have a 50/50 stock-bond portfolio without creating segmented buckets. Do you see what I mean?
Keep in mind that just treating both your ESP and DSP as one portfolio with one withdrawal method suffers from the problems mentioned in the OP: either fixed or increasing withdrawals, when the data are clear that most retirees' spending declines throughout retirement.

If a retiree was fine with having no more than 10 years of discretionary spending, then a conservative allocation to something like TIPS would be fine. But for 15-20 year periods, I believe that most retirees would want some upside potential (e.g. some allocation to stocks).
“It's a dangerous business, Frodo, going out your door. You step onto the road, and if you don't keep your feet, there's no knowing where you might be swept off to.” J.R.R. Tolkien,The Lord of the Rings

User avatar
willthrill81
Posts: 6436
Joined: Thu Jan 26, 2017 3:17 pm
Location: USA

Re: A different take on retirement income: time segmentation

Post by willthrill81 » Tue Jul 17, 2018 10:46 pm

DC3509 wrote:
Tue Jul 17, 2018 8:38 pm
In Merton's interview -- which is well worth the full listen -- he basically says that not enough people look at annuities; everyone looks at what that total amount of the 401 (k) is -- not nearly enough people are thinking in terms of I make X income now, what can I do to get close to X income in retirement. Merton thinks that our 401 (K) statements should have a potential income annuity amount at the top -- not your total pot of money. He thinks the pot of money is meaningless to most people.
It sounds like Merton is in agreement with Wade Pfau regarding annuities. However, for better or worse, most investors are not at all a fan of full annuitization. Perhaps they would be more in favor of doing so for their ESP if they used the method outlined here. But again, the inflation risk of a SPIA without an inflation rider is not to be underestimated.
DC3509 wrote:
Tue Jul 17, 2018 8:38 pm
Yes, agree if inflation runs wild, annuities are risky. But if interest rates stay low, 4% withdrawal money is risky. Merton basically says -- let's say the 4% rule is right and has a 96% success rate -- what if you happen to be in the 4% that fails? Then what? At least with an income annuity -- you still have some cash coming in. You could theoretically have zero under the 4% method. I, for one, have not usually entertained the doomsday hypotheticals and preferred to play my chances, but if you are two middle or lower middle class retirees with no other assets -- the potential of being in the 4% perhaps is terrifying, or should be terrifying. And maybe the 4% failure rate will be a lot higher for this generation, given all of the various circumstances that we frequently discuss here.
To the extent that the remainder of this century resembles the 20th century, I'd take my chances with a 4% failure rate (which is a big misnomer) over having no inflation protection. But a 50/50 AA with the '4% rule' has never failed in U.S. history over a 30 year period. When people talk about it failing, like Early Retirement Now, it's usually because they're backtesting it higher stock allocations (ERN has used 80/20).

Regarding the 4% failure rate being a misnomer, failure doesn't mean 'go broke' because no sane retiree is going to voluntarily do that. It's far better referred to as the "risk of adjustment."

Don't forget that annuitized assets can never be recovered by the retiree. If they encounter a sudden need for some of those funds, even if accessing them would reduce their retirement income going forward, they cannot reverse the annuitization decision. It is irrevocable. That's a big drawback, both mentally and strategically, for a lot of investors.
DC3509 wrote:
Tue Jul 17, 2018 8:38 pm
If you take out the inflation rider, the SPIA pays out more than 4%. Not tons more, but more. For some people, eking out another $200-$300 bucks a month the difference between making it, and not making it. Again, not the crowd on this board, but a lot of regular people out there. In my prior example -- $300 is 6% of your total income for the month. Nothing to sniff at.
For a 66 year old opposite sex couple, a joint lifetime annuity pays about 5.75% right now. That additional 1.75% if nothing to sneeze out, but I wouldn't recommend that unless the retirees had literally no other choice. Even if inflation remains low, it is still very likely to take a toll on their income. A $1,000 monthly payout beginning in the year 2000 would now be worth $671, despite inflation being historically low over that period. That's a big hit, and if most of a retirees' spending is essential and roughly keeps pace with CPI, they could be in serious trouble right now, with another 11+ years to go in their 30 year retirement.
DC3509 wrote:
Tue Jul 17, 2018 8:38 pm
Yes, the Medicaid compliant annuity is a def option. One drawback -- and I have seen nothing written about this -- the annuity cannot run for longer than the anticipated lifespan of the annuitant. Wouldn't getting a regular life SPIA earlier be a better option? What if you outlive the life expectancy?
Your actuarial lifespan is what is used as the basis for Medicaid compliant annuities. And remember that the beneficiary of the annuity is the spouse who is not the recipient of the long-term care.

The other benefit of this approach is that the older you are when you begin the SPIA, the better the payout ratio. For someone in their 80s, the payout ratios are very high indeed.
“It's a dangerous business, Frodo, going out your door. You step onto the road, and if you don't keep your feet, there's no knowing where you might be swept off to.” J.R.R. Tolkien,The Lord of the Rings

DC3509
Posts: 280
Joined: Wed Jul 12, 2017 7:25 am

Re: A different take on retirement income: time segmentation

Post by DC3509 » Wed Jul 18, 2018 12:30 am

willthrill81 wrote:
Tue Jul 17, 2018 10:46 pm
DC3509 wrote:
Tue Jul 17, 2018 8:38 pm
In Merton's interview -- which is well worth the full listen -- he basically says that not enough people look at annuities; everyone looks at what that total amount of the 401 (k) is -- not nearly enough people are thinking in terms of I make X income now, what can I do to get close to X income in retirement. Merton thinks that our 401 (K) statements should have a potential income annuity amount at the top -- not your total pot of money. He thinks the pot of money is meaningless to most people.
It sounds like Merton is in agreement with Wade Pfau regarding annuities. However, for better or worse, most investors are not at all a fan of full annuitization. Perhaps they would be more in favor of doing so for their ESP if they used the method outlined here. But again, the inflation risk of a SPIA without an inflation rider is not to be underestimated.
DC3509 wrote:
Tue Jul 17, 2018 8:38 pm
Yes, agree if inflation runs wild, annuities are risky. But if interest rates stay low, 4% withdrawal money is risky. Merton basically says -- let's say the 4% rule is right and has a 96% success rate -- what if you happen to be in the 4% that fails? Then what? At least with an income annuity -- you still have some cash coming in. You could theoretically have zero under the 4% method. I, for one, have not usually entertained the doomsday hypotheticals and preferred to play my chances, but if you are two middle or lower middle class retirees with no other assets -- the potential of being in the 4% perhaps is terrifying, or should be terrifying. And maybe the 4% failure rate will be a lot higher for this generation, given all of the various circumstances that we frequently discuss here.
To the extent that the remainder of this century resembles the 20th century, I'd take my chances with a 4% failure rate (which is a big misnomer) over having no inflation protection. But a 50/50 AA with the '4% rule' has never failed in U.S. history over a 30 year period. When people talk about it failing, like Early Retirement Now, it's usually because they're backtesting it higher stock allocations (ERN has used 80/20).

Regarding the 4% failure rate being a misnomer, failure doesn't mean 'go broke' because no sane retiree is going to voluntarily do that. It's far better referred to as the "risk of adjustment."

Don't forget that annuitized assets can never be recovered by the retiree. If they encounter a sudden need for some of those funds, even if accessing them would reduce their retirement income going forward, they cannot reverse the annuitization decision. It is irrevocable. That's a big drawback, both mentally and strategically, for a lot of investors.
DC3509 wrote:
Tue Jul 17, 2018 8:38 pm
If you take out the inflation rider, the SPIA pays out more than 4%. Not tons more, but more. For some people, eking out another $200-$300 bucks a month the difference between making it, and not making it. Again, not the crowd on this board, but a lot of regular people out there. In my prior example -- $300 is 6% of your total income for the month. Nothing to sniff at.
For a 66 year old opposite sex couple, a joint lifetime annuity pays about 5.75% right now. That additional 1.75% if nothing to sneeze out, but I wouldn't recommend that unless the retirees had literally no other choice. Even if inflation remains low, it is still very likely to take a toll on their income. A $1,000 monthly payout beginning in the year 2000 would now be worth $671, despite inflation being historically low over that period. That's a big hit, and if most of a retirees' spending is essential and roughly keeps pace with CPI, they could be in serious trouble right now, with another 11+ years to go in their 30 year retirement.
DC3509 wrote:
Tue Jul 17, 2018 8:38 pm
Yes, the Medicaid compliant annuity is a def option. One drawback -- and I have seen nothing written about this -- the annuity cannot run for longer than the anticipated lifespan of the annuitant. Wouldn't getting a regular life SPIA earlier be a better option? What if you outlive the life expectancy?
Your actuarial lifespan is what is used as the basis for Medicaid compliant annuities. And remember that the beneficiary of the annuity is the spouse who is not the recipient of the long-term care.

The other benefit of this approach is that the older you are when you begin the SPIA, the better the payout ratio. For someone in their 80s, the payout ratios are very high indeed.
I don't know that full annuitized is the answer either, but some combination of a partial annuity with an investment portfolio being drawn down might be the answer.

There is no doubt that most people aren't interested in annuities. But most people aren't saving much for retirement either, and most people have some nostalgia for the days of "good-ole" defined benefit pension plans. Is it possible to change some of the very poor outcomes we are seeing now in retirement savings by redefining how people view retirement and looking at income as opposed to total assets? I think that is an interesting question worth exploring more; I don't have the answer.

As much as most people here enjoy this as a hobby, we are complete outliers in the general population. Most people don't "enjoy" this stuff -- it is viewed instead as a giant PIA which is why it gets farmed out to financial advisors, wealth advisors, etc. who charge those big AUM's that we rail against all the time. People would rather be doing something fun -- going to a sports event, the theatre, a concert, whatever you want -- than sitting around and digesting the debate over SPIA versus withdrawal rates. Yet, despite the fact that most people do not care and are not very good at this, we expect everyone to figure this out on their own and do a great job at it. It would be like giving someone all of the parts to a car and a manual on how to put together a car and telling that person to get to it for their next car. It's actually far worse than that because even if that happened, you might be able to ride a bike or take public transportation or call ubers all-day. You only get one shot at retirement and there are really no alternatives to it -- working forever I suppose.

My point is -- if there are ways to make this easier for average people to understand and better implement, we should be moving in that direction, I think. I even think semantically -- if you told people that when you retire, you can now buy a defined benefit pension plan and you are eligible for X income per month forever -- you might have a lot of people sign up for it, even if they don't know that they are really just buying an annuity.

User avatar
willthrill81
Posts: 6436
Joined: Thu Jan 26, 2017 3:17 pm
Location: USA

Re: A different take on retirement income: time segmentation

Post by willthrill81 » Wed Jul 18, 2018 12:52 am

DC3509 wrote:
Wed Jul 18, 2018 12:30 am
I don't know that full annuitized is the answer either, but some combination of a partial annuity with an investment portfolio being drawn down might be the answer.
I agree that it might be. It would be dependent on the retiree and their situation. Annuitizing up to 50% of an ESP might be worthwhile, retaining at least 25% in stocks and the remaining 25% in bonds of some flavor.
DC3509 wrote:
Wed Jul 18, 2018 12:30 am
There is no doubt that most people aren't interested in annuities. But most people aren't saving much for retirement either, and most people have some nostalgia for the days of "good-ole" defined benefit pension plans. Is it possible to change some of the very poor outcomes we are seeing now in retirement savings by redefining how people view retirement and looking at income as opposed to total assets? I think that is an interesting question worth exploring more; I don't have the answer.

As much as most people here enjoy this as a hobby, we are complete outliers in the general population. Most people don't "enjoy" this stuff -- it is viewed instead as a giant PIA which is why it gets farmed out to financial advisors, wealth advisors, etc. who charge those big AUM's that we rail against all the time. People would rather be doing something fun -- going to a sports event, the theatre, a concert, whatever you want -- than sitting around and digesting the debate over SPIA versus withdrawal rates. Yet, despite the fact that most people do not care and are not very good at this, we expect everyone to figure this out on their own and do a great job at it. It would be like giving someone all of the parts to a car and a manual on how to put together a car and telling that person to get to it for their next car. It's actually far worse than that because even if that happened, you might be able to ride a bike or take public transportation or call ubers all-day. You only get one shot at retirement and there are really no alternatives to it -- working forever I suppose.

My point is -- if there are ways to make this easier for average people to understand and better implement, we should be moving in that direction, I think. I even think semantically -- if you told people that when you retire, you can now buy a defined benefit pension plan and you are eligible for X income per month forever -- you might have a lot of people sign up for it, even if they don't know that they are really just buying an annuity.
I certainly agree that much of what is discussed on this forum is beyond what 90% of investors will ever consider. When I try to discuss much of this with my father who will retire very soon, he's just not interested in the process, even though he's very interested in the outcome. Discussions I've had with others with financial matters often seem to follow a similar direction, though some do not even seem concerned about the outcome, perhaps because they simply think that things will just work out somehow. A failure to plan...

And yes, I agree that annuities can certainly have their place and may be part of the answer for a lot of people, especially those who enter retirement with barely enough resources to scrap through it. I don't know much about them, but it seems that folks like those at Blueprint Income are trying to do precisely what you're proposing: reposition annuities as the pensions that many know and love. It will likely work for some. Still, I think that it will continue to be an uphill battle to convince even many of those who would benefit the most from an annuity to hand over a significant chunk of their relatively meager (but possibly large in their eyes) assets over to an insurance company in exchange for a promise to pay.
“It's a dangerous business, Frodo, going out your door. You step onto the road, and if you don't keep your feet, there's no knowing where you might be swept off to.” J.R.R. Tolkien,The Lord of the Rings

User avatar
willthrill81
Posts: 6436
Joined: Thu Jan 26, 2017 3:17 pm
Location: USA

Re: A different take on retirement income: time segmentation

Post by willthrill81 » Wed Jul 18, 2018 2:28 pm

Since several posters have indicated interest in this withdrawal strategy for retirees, I will post an example of how this could have worked for retirees in the past. The starting year for this analysis is the year 2000, the worst starting year for retirees in the last 40+ years.

Scenario
Let's assume that a retired couple, both aged 66 (Social Security's full retirement age) estimates that they will have $50k of essential spending needs, including taxes and some 'wiggle room' in case these needs increase faster than inflation. They will have $30k coming from Social Security, so they need an additional $20k of annual income. They have $1 million in their portfolio.

Essential Spending Portfolio (ESP)
Like most investors, they don't like annuities because they don't want to hand over $500k of their portfolio to an insurance company in return for promised future monthly payments adjusted for inflation, and they would also like the possibility of leaving legacy for others after their passing. They have some reservations about the '4% rule of thumb' for their essential spending needs, so they decide to use 3.5% instead (i.e. 3.5% of the starting portfolio balance, and that nominal dollar amount [$20k] adjusted for inflation every subsequent year). This means that the couple will need to set aside $571k for their ESP ($20k/3.5%). They decide to use a 50/50 AA, with the stocks being evenly split between U.S. and international equities, and total bond market for the rest.

Discretionary Spending Portfolio (ESP)
The couple then places the remaining $429k in their DSP. They are both in good health and hope that they will have twenty years to enjoy most of their discretionary spending (e.g. travel, dining out, nice vehicles). They are comfortable with some volatility in their withdrawals, so they decide to use the same 50/50 AA as their ESP, and they are fine if their discretionary withdrawals shrink over time. As such, they decide to withdraw 6% of their portfolio's balance annually when they rebalance.

ESP Performance
As of June, 2018, the couple's ESP would be worth an inflation-adjusted $470,089 ($703,847 in nominal dollars). Below is a graph of their ESP's performance over time. If their portfolio grew at 0% real, they still have 23.5 years of $20k withdrawals in their ESP. If they're still in good health and/or want to leave as large of a legacy as they comfortably can, they may want to maintain this AA. Otherwise, they may want to move to a more conservative AA.

Image

DSP Performance
As of June, 2018, the couple's DSP would be worth an inflation-adjusted $244,140 ($365,541 in nominal dollars). Below is a graph of their DSP's performance over time. Even with a real rate of growth of -3.00%, the couple still has over half of their starting DSP balance after adjusting for inflation.

Image

Below is a graph of their DSP withdrawals over time. In nominal terms, they peaked in 2007 at $26,250 and were lowest in 2008 at $20,293. Their 2017 withdrawal was $23,555, which was $15,732 in year 2000 dollars.

Image

Conclusion
After 18 years of withdrawals, the couple's ESP is in very good shape with 82% of their inflation-adjusted starting still intact. Even if they encountered a major financial need like $300k (nominal) for long-term care, they would still have 13.5 years of inflation-adjusted ES remaining.

The couple has depleted almost half of their DSP in year 2000 dollars. But they have had 18 years of (surprisingly to me at least) consistent withdrawals totaling about $421k in nominal dollars.

In sum, the couples' essential spending has been very well provided for, and they have depleted nearly half of their discretionary funds, leaving much for continued discretionary spending or any other use they wish.
“It's a dangerous business, Frodo, going out your door. You step onto the road, and if you don't keep your feet, there's no knowing where you might be swept off to.” J.R.R. Tolkien,The Lord of the Rings

The Wizard
Posts: 12433
Joined: Tue Mar 23, 2010 1:45 pm
Location: Reading, MA

Re: A different take on retirement income: time segmentation

Post by The Wizard » Wed Jul 18, 2018 3:15 pm

People are suspicious of insurance companies taking their money for single premium lifetime payout annuities and taking a hefty sales commission on the transaction.
This happens even though SPIA overhead/profit is said to be "low".

What I'd like to see is the Society of Actuaries come out with annually updated tables showing current fair SPIA payouts with ZERO expenses deducted. This would help people see what actual insurance companies are taking for expenses better.

I always preferred the monthly amount per $100k annuitized metric.
For example, at age 65, a theoretical lifetime annuity with ten year guarantee might pay $575 per month.
But immediate annuities dot com returns values from different companies ranging from $561 to $570 per month.

Data like this would help some people...
Attempted new signature...

User avatar
packer16
Posts: 1067
Joined: Sat Jan 04, 2014 2:28 pm

Re: A different take on retirement income: time segmentation

Post by packer16 » Wed Jul 18, 2018 5:07 pm

I know this is a different approach but I like the Kites upward ratchet approach. Where 4% is taken out every year then it is ratcheted upwards if the portfolio is greater than 50% of the inflation adjusted initial balance. The increases happen every 3 years as long as the portfolio is great than 50% of the initial portfolio.

This is discussed here: https://www.kitces.com/blog/the-ratchet ... he-4-rule/

This could be applied to your ESP & maybe a more liberal rule for the DSP portfolio. The advantage of this approach is that it spends the money & leaves little on the table beyond the initial amount invested. Thus maximizing spending but also preventing the portfolio from falling to the extent where it can longer support inflation adjusted income levels. This concept is also discussed in "How Much Can I Spend in Retirement" by Pfau & compares favorably with most other withdraw strategies (pg. 212 to 221).

Packer
Buy cheap and something good might happen

User avatar
willthrill81
Posts: 6436
Joined: Thu Jan 26, 2017 3:17 pm
Location: USA

Re: A different take on retirement income: time segmentation

Post by willthrill81 » Wed Jul 18, 2018 7:07 pm

packer16 wrote:
Wed Jul 18, 2018 5:07 pm
I know this is a different approach but I like the Kites upward ratchet approach. Where 4% is taken out every year then it is ratcheted upwards if the portfolio is greater than 50% of the inflation adjusted initial balance. The increases happen every 3 years as long as the portfolio is great than 50% of the initial portfolio.

This is discussed here: https://www.kitces.com/blog/the-ratchet ... he-4-rule/

This could be applied to your ESP & maybe a more liberal rule for the DSP portfolio. The advantage of this approach is that it spends the money & leaves little on the table beyond the initial amount invested. Thus maximizing spending but also preventing the portfolio from falling to the extent where it can longer support inflation adjusted income levels. This concept is also discussed in "How Much Can I Spend in Retirement" by Pfau & compares favorably with most other withdraw strategies (pg. 212 to 221).

Packer
I too like Kitces' ratcheting strategy. And if a retiree didn't want to start at 4% (I think it's perfectly fine but many are skittish about it), they could just start at a lower initial withdrawal rate and ratchet up later if possible. But remember that as far as the ESP is concerned, most of those expenses should be relatively 'fixed', so funds that exceeded 50% of the starting portfolio balance (adjusted for inflation) could be transferred to the DSP or possibly reserved for potential long-term care expenses later.

And in the scenario I ran above, it's clear that even though cumulative returns weren't great, a higher percentage of the portfolio (e.g. 7-8%) could have been withdrawn every year from the DSP so as to further 'front load' the DSP withdrawals, which declined in real dollars by only about one-third after the first 18 years of withdrawals.
“It's a dangerous business, Frodo, going out your door. You step onto the road, and if you don't keep your feet, there's no knowing where you might be swept off to.” J.R.R. Tolkien,The Lord of the Rings

User avatar
willthrill81
Posts: 6436
Joined: Thu Jan 26, 2017 3:17 pm
Location: USA

Re: A different take on retirement income: time segmentation

Post by willthrill81 » Wed Jul 18, 2018 7:12 pm

The Wizard wrote:
Wed Jul 18, 2018 3:15 pm
People are suspicious of insurance companies taking their money for single premium lifetime payout annuities and taking a hefty sales commission on the transaction.
This happens even though SPIA overhead/profit is said to be "low".

What I'd like to see is the Society of Actuaries come out with annually updated tables showing current fair SPIA payouts with ZERO expenses deducted. This would help people see what actual insurance companies are taking for expenses better.

I always preferred the monthly amount per $100k annuitized metric.
For example, at age 65, a theoretical lifetime annuity with ten year guarantee might pay $575 per month.
But immediate annuities dot com returns values from different companies ranging from $561 to $570 per month.

Data like this would help some people...
Such data would certainly be interesting.

However, I think that Kitces' hierarchy of retirement needs does a good job of explaining the 'annuity paradox'.
The significance of this “hierarchy of retirement needs” is that it helps to explain why some types of retirement income strategies like annuitization are very unpopular (despite the fact that retirees routinely state their biggest fear is outliving their retirement assets and annuitization can guarantee that will never happen), while others are used far more often even if their current guarantees are inferior to available alternatives (e.g., most guaranteed living benefit riders on today’s variable annuities).
“It's a dangerous business, Frodo, going out your door. You step onto the road, and if you don't keep your feet, there's no knowing where you might be swept off to.” J.R.R. Tolkien,The Lord of the Rings

dknightd
Posts: 1032
Joined: Wed Mar 07, 2018 11:57 am

Re: A different take on retirement income: time segmentation

Post by dknightd » Thu Jul 19, 2018 6:30 am

I think deciding to buy an SPIA is very much like the decision to cash out a pension. Except in reverse. The advantage we have is it does not have to be all or nothing. We can decide to use part of our money to "buy" a pension, and leave the rest invested. There is no right answer to this since we can't predict how long we will live, and how our investments might do. All we have are actuarial estimates, and prior market performance, on which to base our decision.

dknightd
Posts: 1032
Joined: Wed Mar 07, 2018 11:57 am

Re: A different take on retirement income: time segmentation

Post by dknightd » Thu Jul 19, 2018 7:46 am

willthrill81 wrote:
Wed Jul 18, 2018 7:12 pm

However, I think that Kitces' hierarchy of retirement needs does a good job of explaining the 'annuity paradox'.
It does. But he also says
Nonetheless, if the need for future income cannot be achieved until the need for current assets has been mentally satisfied first, retirees may continue to feel constrained by not having enough – even if they do – and/or to choose retirement income solutions that are mechanically inferior but psychologically more satisfying to our hierarchy of retirement needs!
When given the option of a "mechanically" superior plan, or a "psychologically more satisfying" plan. I might lean toward the second option as long both will likely work. I'd rather feel comfortable in retirement, than maximize what I have to spend.
When I ask my wife how much money she thinks we need in our retirement, her answer is always "enough". So my plan is to make sure I always feel like we'll have enough, even though that means we might have less total money to spend. I'm 60, seriously thinking about retiring in a year. I don't hate my job, I just don't having to work. I likely have the luxury of deciding when to retire. I don't want to keep working for money that we may not end up needing. I just want to feel comfortable when I pull the trigger. To me "feel comfortable" is the most important thing. YMMV

Edit: my the goal is to be happy happy happy. They say that once your expenses are covered more money does not buy more happiness. I agree with that. It all comes down to what your "expenses" are . . .

User avatar
siamond
Posts: 4305
Joined: Mon May 28, 2012 5:50 am

Re: A different take on retirement income: time segmentation

Post by siamond » Thu Jul 19, 2018 9:18 am

willthrill81 wrote:
Tue Jul 17, 2018 10:02 am
This strategy is an attempt to address one of the real world but often overlooked problem with constant or increased overall spending approaches like the '4% rule of thumb' and the VPW. Rather than remain flat or increase over time, the very clear trend is the opposite; retirees need/want to spend more early in retirement than later.
I didn't read all the posts, but the proposal seems to essentially be a direct derivative of the LMP/RP approach, which quite a few people are fond of on this board, while another large group (including myself) can't stand the idea... LOL

The biggest problem I see is the usual trap of mental accounting and segmentation. By separating the portfolio in two somewhat arbitrary goals (or more segmentation to address LTC, bequest, or whatever else), you lose optimality at the overall level (it's a bit like asset allocation where the sum of the parts is better than the parts acting in isolation). Case in point, the retiree will probably get overly conservative for the ESP investments (if there is a need beyond SS/Pension, that is), while there might be no need because the total portfolio would easily generate income significantly beyond the ESP needs. Also, the lines between ESP and DSP are quite fuzzy (where do you put what you could live without under severe duress, but really would like to keep otherwise?) and would be a constant headache to assess and adjust over time as life goes by. I actually tried to define a bare bone budget a while ago, scratched my head, and gave up at the end.

Please note that actuarial methods (PMT-based) as well as Guyton-Klinger decision rules can easily accommodate a higher spend in early years and a slow decline afterwards (at least statistically speaking), it's just a matter of increasing the base rate above the expected returns. VPW is unfortunately way too hardwired for its own sake, but it isn't hard to come back to the underlying PMT formula and do so. I do exactly that (in a rather cautious manner though), and I know that a few other folks on this forum do the same. With minimum care, this doesn't put my 'ESP' needs at risk, thanks to the intrinsic self-adjustment towards the lifetime goal. And this saves me from any segmentation line of thinking.

dknightd
Posts: 1032
Joined: Wed Mar 07, 2018 11:57 am

Re: A different take on retirement income: time segmentation

Post by dknightd » Thu Jul 19, 2018 9:58 am

Bottom line, is to spend less than you earn.
That is not so hard when you have a job.
But what happens when you have no job?
Saving money is part of the answer.
But saving money, and actually spending that money, are two different things.
When I retire I will likely not have money coming in. It will be hard to change from saver to spender.
:shock: :oops: :beer

User avatar
willthrill81
Posts: 6436
Joined: Thu Jan 26, 2017 3:17 pm
Location: USA

Re: A different take on retirement income: time segmentation

Post by willthrill81 » Thu Jul 19, 2018 10:06 am

dknightd wrote:
Thu Jul 19, 2018 7:46 am
willthrill81 wrote:
Wed Jul 18, 2018 7:12 pm

However, I think that Kitces' hierarchy of retirement needs does a good job of explaining the 'annuity paradox'.
It does. But he also says
Nonetheless, if the need for future income cannot be achieved until the need for current assets has been mentally satisfied first, retirees may continue to feel constrained by not having enough – even if they do – and/or to choose retirement income solutions that are mechanically inferior but psychologically more satisfying to our hierarchy of retirement needs!
When given the option of a "mechanically" superior plan, or a "psychologically more satisfying" plan. I might lean toward the second option as long both will likely work. I'd rather feel comfortable in retirement, than maximize what I have to spend.
When I ask my wife how much money she thinks we need in our retirement, her answer is always "enough". So my plan is to make sure I always feel like we'll have enough, even though that means we might have less total money to spend. I'm 60, seriously thinking about retiring in a year. I don't hate my job, I just don't having to work. I likely have the luxury of deciding when to retire. I don't want to keep working for money that we may not end up needing. I just want to feel comfortable when I pull the trigger. To me "feel comfortable" is the most important thing. YMMV

Edit: my the goal is to be happy happy happy. They say that once your expenses are covered more money does not buy more happiness. I agree with that. It all comes down to what your "expenses" are . . .
I'm agree with you. I've never said that annuities don't have any place in a retirees' toolbox; to the contrary, I said in the OP that they may be a good option for some of the essential spending portfolio laid out in this strategy. It's just that I don't think that they're the one-size-fits-all approach that some pundits believe that they are, even for essential spending needs. But if they are emotionally satisfying and allow a retiree to break free of the poverty mindset (i.e. "If I start spending anything from my portfolio, I might go broke one day!") that is all too common, then I think that they are great.
“It's a dangerous business, Frodo, going out your door. You step onto the road, and if you don't keep your feet, there's no knowing where you might be swept off to.” J.R.R. Tolkien,The Lord of the Rings

User avatar
willthrill81
Posts: 6436
Joined: Thu Jan 26, 2017 3:17 pm
Location: USA

Re: A different take on retirement income: time segmentation

Post by willthrill81 » Thu Jul 19, 2018 10:17 am

siamond wrote:
Thu Jul 19, 2018 9:18 am
willthrill81 wrote:
Tue Jul 17, 2018 10:02 am
This strategy is an attempt to address one of the real world but often overlooked problem with constant or increased overall spending approaches like the '4% rule of thumb' and the VPW. Rather than remain flat or increase over time, the very clear trend is the opposite; retirees need/want to spend more early in retirement than later.
I didn't read all the posts, but the proposal seems to essentially be a direct derivative of the LMP/RP approach, which quite a few people are fond of on this board, while another large group (including myself) can't stand the idea... LOL

The biggest problem I see is the usual trap of mental accounting and segmentation. By separating the portfolio in two somewhat arbitrary goals (or more segmentation to address LTC, bequest, or whatever else), you lose optimality at the overall level (it's a bit like asset allocation where the sum of the parts is better than the parts acting in isolation). Case in point, the retiree will probably get overly conservative for the ESP investments (if there is a need beyond SS/Pension, that is), while there might be no need because the total portfolio would easily generate income significantly beyond the ESP needs. Also, the lines between ESP and DSP are quite fuzzy (where do you put what you could live without under severe duress, but really would like to keep otherwise?) and would be a constant headache to assess and adjust over time as life goes by. I actually tried to define a bare bone budget a while ago, scratched my head, and gave up at the end.

Please note that actuarial methods (PMT-based) as well as Guyton-Klinger decision rules can easily accommodate a higher spend in early years and a slow decline afterwards (at least statistically speaking), it's just a matter of increasing the base rate above the expected returns. VPW is unfortunately way too hardwired for its own sake, but it isn't hard to come back to the underlying PMT formula and do so. I do exactly that (in a rather cautious manner though), and I know that a few other folks on this forum do the same. With minimum care, this doesn't put my 'ESP' needs at risk, thanks to the intrinsic self-adjustment towards the lifetime goal. And this saves me from any segmentation line of thinking.
I addressed earlier how this strategy is not just mental accounting. The basic idea of the approach is for retirees to separate their essential spending needs from their discretionary spending needs, both of which are defined by them, and to take a more conservative withdrawal strategy to cover their essential needs than their discretionary. There's no 'sub-optimal mental accounting' to making on a more liberal AA and/or withdrawals from funds that will cover discretionary spending.

Disaggregating essential and discretionary spending may actually help retirees to take on a less conservative approach to their cumulative withdrawals. I believe it's a fairly safe assumption to say that retirees are generally more willing to put their discretionary spending at risk than their essential spending, whatever they determine each to be comprised of. The folks around here that are in their 60s yet are fearful that 3% withdrawals might be too high could be well served by this type of strategy.

Further, the data are clear that the overwhelming majority of retirees' discretionary spending needs decline as they age. A withdrawal strategy that allows them to 'front load' their discretionary spending, then, will have greater economic utility than one that provides for flat or increasing total withdrawals over time.

If there's another specific method out there that can accomplish these goals in a more efficient manner, I'm all ears. I'm not saying that this method is the bees-knees and is perfect for everyone, but it's certainly not just mental accounting.
“It's a dangerous business, Frodo, going out your door. You step onto the road, and if you don't keep your feet, there's no knowing where you might be swept off to.” J.R.R. Tolkien,The Lord of the Rings

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

Re: A different take on retirement income: time segmentation

Post by marcopolo » Thu Jul 19, 2018 10:48 am

willthrill81 wrote:
Thu Jul 19, 2018 10:17 am

Disaggregating essential and discretionary spending may actually help retirees to take on a less conservative approach to their cumulative withdrawals. I believe it's a fairly safe assumption to say that retirees are generally more willing to put their discretionary spending at risk than their essential spending, whatever they determine each to be comprised of. The folks around here that are in their 60s yet are fearful that 3% withdrawals might be too high could be well served by this type of strategy.

If there's another specific method out there that can accomplish these goals in a more efficient manner, I'm all ears. I'm not saying that this method is the bees-knees and is perfect for everyone, but it's certainly not just mental accounting.
I tend to agree with Siamond on this one. I am not sure I agree with your assertion that the 3% withdrawal people will actually do much different by dis-aggregating. You might be right that they will be more willing to take risk with the discretionary portion, maybe they go to 4% and a higher equity allocation. But, they will also quite likely become more conservative with the essential portion, maybe going down to 2%, with a more FI allocation. Remember, they are only comfortable with the 3% because they know they have discretionary items they can cut back on if needed. So, in the end , they end up back close to 3%. Their overall AA probably does not change much either in aggregate for the same reasons. So, to me it still seems like a lot of complexity added to do the mental, or perhaps actual accounting in this case, to end up in the same place.
Once in a while you get shown the light, in the strangest of places if you look at it right.

dknightd
Posts: 1032
Joined: Wed Mar 07, 2018 11:57 am

Re: A different take on retirement income: time segmentation

Post by dknightd » Thu Jul 19, 2018 11:05 am

willthrill81 wrote:
Thu Jul 19, 2018 10:17 am
I addressed earlier how this strategy is not just mental accounting.
But the reality is it is a mental accounting. If I can't mentally account anymore, I suspect I'll be glad wife, or I, have an annuity (pension). Even if either one is us dead, or brain dead, we'll still have money coming in to help cover our expenses. I don't know how you attach a monetary value/expense to that.
Last edited by dknightd on Thu Jul 19, 2018 11:12 am, edited 1 time in total.

dknightd
Posts: 1032
Joined: Wed Mar 07, 2018 11:57 am

Re: A different take on retirement income: time segmentation

Post by dknightd » Thu Jul 19, 2018 11:07 am

I'm curious what your plan is. Can you share it with us?
Edit: if not, why not? My hunch is you are just collecting information, that you hope to monetize. Nothing wrong with that, but you should include a disclaimer.

User avatar
siamond
Posts: 4305
Joined: Mon May 28, 2012 5:50 am

Re: A different take on retirement income: time segmentation

Post by siamond » Thu Jul 19, 2018 12:20 pm

willthrill81 wrote:
Thu Jul 19, 2018 10:17 am
Further, the data are clear that the overwhelming majority of retirees' discretionary spending needs decline as they age. A withdrawal strategy that allows them to 'front load' their discretionary spending, then, will have greater economic utility than one that provides for flat or increasing total withdrawals over time.

If there's another specific method out there that can accomplish these goals in a more efficient manner, I'm all ears. I'm not saying that this method is the bees-knees and is perfect for everyone, but it's certainly not just mental accounting.
I agree with the top-level goal, to a large extent. Or at least, this is my perspective now, as an early retiree. I don't rule out changing my mind over time though (maybe my wife and I will develop a strong urge to donate to various causes and/or grandchildren while we are in our twilight years, or something like that, hence a higher spend/income pattern). I am an evolving individual, not a fixed average. Furthermore, I think the essential/discretionary lines are very blurry and somewhat arbitrary. So I'd much rather take a simpler approach that can easily be tweaked over time (in my case, by simply adjusting the base rate in the PMT formulas, and/or possibly the # of remaining periods).

Look, I didn't mean to say that your proposal is wrong. Those labels of 'mental accounting' (or 'market timing', etc) unfortunately convey a very negative connotation and are way too simplistic. Forget those labels. If what you suggest fits your personality/behavior well, then by all means, do it. Quite clearly, the large LMP/RP contingent on this board thinks in a such way and is happy with it. It seems suboptimal from a numbers standpoint to segregate (until you allow $$ to flow from one bucket to the other, but then you'd be undermining your own logic), but if it's optimal from a behavioral/comfort standpoint, or even enabling better individualized numbers-oriented decisions, then sure, go for it. No size fits all with decumulation strategies, it's all about balancing your own numbers and your own emotions.

All this being said, I do have in our IPS to consider the SPIA route to beef up our fixed income (SS/pension) at some point (e.g. when we turn 70), as a way to create a floor of sorts that would be fairly emotion-free. Sure, you can call it an ESP if you wish. We are NOT planning to actually do it, but my wife and I do acknowledge that by the time we turn 70-ish, after living through at least one more major crisis (my guess!), we might think differently, so we captured the possibility in writing. We'll see. Never say never.

dknightd
Posts: 1032
Joined: Wed Mar 07, 2018 11:57 am

Re: A different take on retirement income: time segmentation

Post by dknightd » Thu Jul 19, 2018 1:23 pm

dknightd wrote:
Thu Jul 19, 2018 11:07 am
I'm curious what your plan is. Can you share it with us?
I'll put this on ignore. Unless I bored.

User avatar
willthrill81
Posts: 6436
Joined: Thu Jan 26, 2017 3:17 pm
Location: USA

Re: A different take on retirement income: time segmentation

Post by willthrill81 » Thu Jul 19, 2018 2:19 pm

marcopolo wrote:
Thu Jul 19, 2018 10:48 am
Remember, they are only comfortable with the 3% because they know they have discretionary items they can cut back on if needed.
I'm not convinced that most of them are actually cognizant of this. HomerJ reminds people all the time that 'failure of the 4% rule' does not mean running out of money as much as it means making adjustments to one's discretionary spending.
“It's a dangerous business, Frodo, going out your door. You step onto the road, and if you don't keep your feet, there's no knowing where you might be swept off to.” J.R.R. Tolkien,The Lord of the Rings

User avatar
willthrill81
Posts: 6436
Joined: Thu Jan 26, 2017 3:17 pm
Location: USA

Re: A different take on retirement income: time segmentation

Post by willthrill81 » Thu Jul 19, 2018 2:26 pm

dknightd wrote:
Thu Jul 19, 2018 11:05 am
willthrill81 wrote:
Thu Jul 19, 2018 10:17 am
I addressed earlier how this strategy is not just mental accounting.
But the reality is it is a mental accounting.
Claiming it to be so doesn't make it so. Please provide your reasoning for this conclusion.

Would you claim that a LMP is nothing more than mental accounting? Do you believe that any withdrawal system that uses different withdrawal strategies to manage different types of risk is mental accounting? If so, you do not understand what mental accounting truly is.
dknightd wrote:
Thu Jul 19, 2018 11:07 am
I'm curious what your plan is. Can you share it with us?
Personal finance is a full-time hobby for me, and I'm mainly trying to improve on the existing withdrawal strategies that have obvious and documented problems noted by others.
dknightd wrote:
Thu Jul 19, 2018 11:07 am
Edit: if not, why not? My hunch is you are just collecting information, that you hope to monetize. Nothing wrong with that, but you should include a disclaimer.
No, I am not trying to peddle anything now or in the future.

Note that Larry Swedroe left this forum because people were making this precise accusation. Some of us really do have altruistic motivations for being here.
“It's a dangerous business, Frodo, going out your door. You step onto the road, and if you don't keep your feet, there's no knowing where you might be swept off to.” J.R.R. Tolkien,The Lord of the Rings

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

Re: A different take on retirement income: time segmentation

Post by marcopolo » Thu Jul 19, 2018 2:30 pm

willthrill81 wrote:
Thu Jul 19, 2018 2:19 pm
marcopolo wrote:
Thu Jul 19, 2018 10:48 am
Remember, they are only comfortable with the 3% because they know they have discretionary items they can cut back on if needed.
I'm not convinced that most of them are actually cognizant of this. HomerJ reminds people all the time that 'failure of the 4% rule' does not mean running out of money as much as it means making adjustments to one's discretionary spending.
That is a fair point. I guess i am projecting my biases onto others a bit here.
I prefer the simplicity of a single portfolio, but i can see where segregating can help some stay with their plans. In that sense, it is a helpful concept.
Once in a while you get shown the light, in the strangest of places if you look at it right.

User avatar
willthrill81
Posts: 6436
Joined: Thu Jan 26, 2017 3:17 pm
Location: USA

Re: A different take on retirement income: time segmentation

Post by willthrill81 » Thu Jul 19, 2018 2:55 pm

siamond wrote:
Thu Jul 19, 2018 12:20 pm
I agree with the top-level goal, to a large extent. Or at least, this is my perspective now, as an early retiree. I don't rule out changing my mind over time though (maybe my wife and I will develop a strong urge to donate to various causes and/or grandchildren while we are in our twilight years, or something like that, hence a higher spend/income pattern). I am an evolving individual, not a fixed average. Furthermore, I think the essential/discretionary lines are very blurry and somewhat arbitrary. So I'd much rather take a simpler approach that can easily be tweaked over time (in my case, by simply adjusting the base rate in the PMT formulas, and/or possibly the # of remaining periods).
Certainly this strategy isn't for everyone. While I agree that some of the lines between essential and discretionary spending can get blurry, I think we can all agree that globe trotting is more discretionary than paying for one's groceries or property taxes. Can there be discretionary elements to this latter group? Of course. Someone could eat nothing but rice and beans, and they could sell their home and live in a tent on public land (many actually do this). But those who have the ability to implement this strategy will likely have the view that certain expenses, while perhaps not strictly necessary in order to sustain life, are essential to them living the life they want. Other expenses may improve their quality of life, but they may be willing to forego or reduce these as they age and/or their portfolio suffers. It would be completely up to each retiree to determine how to do this categorization. The trade off they would need to be aware of is that it will take proportionately more of their portfolio to cover essential spending than discretionary spending (e.g. lower withdrawal rate).
siamond wrote:
Thu Jul 19, 2018 12:20 pm
Look, I didn't mean to say that your proposal is wrong. Those labels of 'mental accounting' (or 'market timing', etc) unfortunately convey a very negative connotation and are way too simplistic. Forget those labels. If what you suggest fits your personality/behavior well, then by all means, do it. Quite clearly, the large LMP/RP contingent on this board thinks in a such way and is happy with it. It seems suboptimal from a numbers standpoint to segregate (until you allow $$ to flow from one bucket to the other, but then you'd be undermining your own logic), but if it's optimal from a behavioral/comfort standpoint, or even enabling better individualized numbers-oriented decisions, then sure, go for it.
Aggregating all of one's expenses for the purposes of determining a needed portfolio size with one withdrawal strategy only makes sense if one first assumes that all of these expenses are created equal. For some retirees, this assumption is largely true; they only have enough of a portfolio to cover what they view as essential expenses. But for most Bogleheads, this assumption does not hold; they have enough to cover all of their essential spending as well as discretionary expenses (again, what is essential and discretionary is to be interpreted strictly by the retiree). Since discretionary spending can, by definition, be curbed in the event of poor market returns and/or as the retiree ages, retirees have the ability to take on greater risk with regard to portfolio construction and withdrawal strategy for the funds needed to provide for their discretionary spending than for their essential spending.

Recall that according to Larry Swedroe, risk tolerance should be based on need, willingness, and ability to take on risk. Logically, retirees are less able to take on risk when it comes to funding their essential spending than their discretionary spending.

As others have suggested, some may be implicitly using some form of this strategy already. We hear of many posters who state that while they would like to have X dollars of income during retirement, they could still make do with less than X. They then base their portfolio needs on X with the knowledge that if their strategy does not pan out as they hope, they can still get by with less.
siamond wrote:
Thu Jul 19, 2018 12:20 pm
No size fits all with decumulation strategies, it's all about balancing your own numbers and your own emotions.
I agree entirely. :beer
“It's a dangerous business, Frodo, going out your door. You step onto the road, and if you don't keep your feet, there's no knowing where you might be swept off to.” J.R.R. Tolkien,The Lord of the Rings

User avatar
willthrill81
Posts: 6436
Joined: Thu Jan 26, 2017 3:17 pm
Location: USA

Re: A different take on retirement income: time segmentation

Post by willthrill81 » Thu Jul 19, 2018 3:01 pm

marcopolo wrote:
Thu Jul 19, 2018 2:30 pm
willthrill81 wrote:
Thu Jul 19, 2018 2:19 pm
marcopolo wrote:
Thu Jul 19, 2018 10:48 am
Remember, they are only comfortable with the 3% because they know they have discretionary items they can cut back on if needed.
I'm not convinced that most of them are actually cognizant of this. HomerJ reminds people all the time that 'failure of the 4% rule' does not mean running out of money as much as it means making adjustments to one's discretionary spending.
That is a fair point. I guess i am projecting my biases onto others a bit here.
I prefer the simplicity of a single portfolio, but i can see where segregating can help some stay with their plans. In that sense, it is a helpful concept.
The people that may benefit the most from this strategy are those whose essential spending needs are very small in relation to their portfolio size. For instance, retirees whose annual essential spending is 2% of their portfolio have the ability to be more aggressive with their AA and/or withdrawal strategy than those whose annual essential spending is 4% of their portfolio. They may not desire to be more aggressive, but they certainly could be.
“It's a dangerous business, Frodo, going out your door. You step onto the road, and if you don't keep your feet, there's no knowing where you might be swept off to.” J.R.R. Tolkien,The Lord of the Rings

dknightd
Posts: 1032
Joined: Wed Mar 07, 2018 11:57 am

Re: A different take on retirement income: time segmentation

Post by dknightd » Thu Jul 19, 2018 6:16 pm

willthrill81 wrote:
Thu Jul 19, 2018 2:26 pm

Personal finance is a full-time hobby for me, and I'm mainly trying to improve on the existing withdrawal strategies that have obvious and documented problems noted by others.
Thanks

User avatar
siamond
Posts: 4305
Joined: Mon May 28, 2012 5:50 am

Re: A different take on retirement income: time segmentation

Post by siamond » Thu Jul 19, 2018 6:44 pm

Ok, let me ask a simple rhetorical question because a key point about the use of segmented buckets doesn't seem to come across.

Say the ESP portfolio is made of a base of fixed income (SS, Pension, SPIA, whatever) *and* some of the savings. As indicated in the OP, those savings should be invested in a fairly 'safe' manner (e.g. lots of bonds to dampen volatility, some stocks to keep some growth potential), with the '4% rule of thumb' withdrawal method. Then the DSP would use the rest of the savings in a more 'aggressive' manner.

Now for the great majority of starting years, the 4% rule of thumb actually totally underestimates the returns to come. And the ESP portfolio will probably keep growing. But then the '4% rule of thumb' is a fixed withdrawal method, period. Truth is in some rare cases, a giant crisis coming mid retirement might actually justify such conservatism. So you can't change the approach in foresight. But fact is, in hindsight, the excess returns for all other (rosier) scenarios will be left unused. Isn't that squarely sub-optimal? And a very direct consequence of the segmentation approach?

User avatar
willthrill81
Posts: 6436
Joined: Thu Jan 26, 2017 3:17 pm
Location: USA

Re: A different take on retirement income: time segmentation

Post by willthrill81 » Thu Jul 19, 2018 8:37 pm

siamond wrote:
Thu Jul 19, 2018 6:44 pm
Ok, let me ask a simple rhetorical question because a key point about the use of segmented buckets doesn't seem to come across.

Say the ESP portfolio is made of a base of fixed income (SS, Pension, SPIA, whatever) *and* some of the savings. As indicated in the OP, those savings should be invested in a fairly 'safe' manner (e.g. lots of bonds to dampen volatility, some stocks to keep some growth potential), with the '4% rule of thumb' withdrawal method. Then the DSP would use the rest of the savings in a more 'aggressive' manner.

Now for the great majority of starting years, the 4% rule of thumb actually totally underestimates the returns to come. And the ESP portfolio will probably keep growing. But then the '4% rule of thumb' is a fixed withdrawal method, period. Truth is in some rare cases, a giant crisis coming mid retirement might actually justify such conservatism. So you can't change the approach in foresight. But fact is, in hindsight, the excess returns for all other (rosier) scenarios will be left unused. Isn't that squarely sub-optimal? And a very direct consequence of the segmentation approach?
Good question. Here's my take.

First, retirees could use any withdrawal rate for either portfolio that they wanted, though I believe that 4% is a good starting role; some may want start even lower. But a key element of this strategy is that they should be more conservative with their ESP withdrawals than their DSP withdrawals for reasons already covered.

Second, as packer16 brought up, Kitces' 'ratcheting rule' can be used to either increase ESP withdrawals or, perhaps more logically, increase the DSP portfolio/withdrawals, leaving less money on the table than has been historically typical with implementation of the '4% rule of thumb'.

Keep in mind that any use of a conservative fixed withdrawal strategy like the '4% rule' is 'sub-optimal', regardless of whether a segmentation approach is used. You correctly point out that perfect foresight would be needed to safely exhaust the ESP, which is clearly impossible (with the exception of full annuitization). But the fact that there is a dearth of such foresight is a big part of the reason why this segmentation strategy may make sense. It is logical to take on a more conservative approach with one's essential spending needs than with one's discretionary spending needs.
“It's a dangerous business, Frodo, going out your door. You step onto the road, and if you don't keep your feet, there's no knowing where you might be swept off to.” J.R.R. Tolkien,The Lord of the Rings

User avatar
siamond
Posts: 4305
Joined: Mon May 28, 2012 5:50 am

Re: A different take on retirement income: time segmentation

Post by siamond » Thu Jul 19, 2018 9:15 pm

willthrill81 wrote:
Thu Jul 19, 2018 8:37 pm
siamond wrote:
Thu Jul 19, 2018 6:44 pm
[...] But fact is, in hindsight, the excess returns for all other (rosier) scenarios will be left unused. Isn't that squarely sub-optimal? And a very direct consequence of the segmentation approach?
[...] It is logical to take on a more conservative approach with one's essential spending needs than with one's discretionary spending needs.
I am afraid you're still missing my point. The root cause of the issue is the segmentation approach. If you don't segment, keeping the portfolio whole and using a decent variable withdrawal approach, you can easily manage your 'essential spending' as a soft floor while putting the excess returns to good use, i.e. variable discretionary spend over time, as any PMT backtesting would easily show (a modeling exercise I ran myself quite a few times to truly convince myself!). It is the precise introduction of the segmentation that makes you use a sub-optimal strategy on part of the portfolio, and reach this 'logical' inference that wasn't necessary in the first place. This is the fundamental issue with any LMP/RP bucketing scheme.

This being said, as I expressed before, this doesn't make the approach wrong. If behaviorally, this makes you feel good to use such segmenting (notably by creating a 'hard floor' instead of a 'soft floor'), and you don't mind the sub-optimality, then go for it. But you should understand the compromise you're making by doing so. Give it a good try, do the modeling, and you'll see the problem.

PS. I skipped the ratcheting tricks part of the answer, because they also suffer from significant issues (defeating the basic premise), but let's not derail the core point.

User avatar
willthrill81
Posts: 6436
Joined: Thu Jan 26, 2017 3:17 pm
Location: USA

Re: A different take on retirement income: time segmentation

Post by willthrill81 » Thu Jul 19, 2018 9:41 pm

siamond wrote:
Thu Jul 19, 2018 9:15 pm
willthrill81 wrote:
Thu Jul 19, 2018 8:37 pm
siamond wrote:
Thu Jul 19, 2018 6:44 pm
[...] But fact is, in hindsight, the excess returns for all other (rosier) scenarios will be left unused. Isn't that squarely sub-optimal? And a very direct consequence of the segmentation approach?
[...] It is logical to take on a more conservative approach with one's essential spending needs than with one's discretionary spending needs.
I am afraid you're still missing my point. The root cause of the issue is the segmentation approach. If you don't segment, keeping the portfolio whole and using a decent variable withdrawal approach, you can easily manage your 'essential spending' as a soft floor while putting the excess returns to good use, i.e. variable discretionary spend over time, as any PMT backtesting would easily show (a modeling exercise I ran myself quite a few times to truly convince myself!). It is the precise introduction of the segmentation that makes you use a sub-optimal strategy on part of the portfolio, and reach this 'logical' inference that wasn't necessary in the first place. This is the fundamental issue with any LMP/RP bucketing scheme.

This being said, as I expressed before, this doesn't make the approach wrong. If behaviorally, this makes you feel good to use such segmenting (notably by creating a 'hard floor' instead of a 'soft floor'), and you don't mind the sub-optimality, then go for it. But you should understand the compromise you're making by doing so. Give it a good try, do the modeling, and you'll see the problem.

PS. I skipped the ratcheting tricks part of the answer, because they also suffer from significant issues (defeating the basic premise), but let's not derail the core point.
Would you define what you mean by 'sub-optimal', preferably with some hard numbers? I'm having a hard time wrapping my mind around how using a "decent variable withdrawal approach" on the whole portfolio is mathematically preferable because you still have a withdrawal floor that must be accounted for.

Do you mean 'sub-optimal' to be a smaller ending or average portfolio, smaller withdrawals, 'smoother' withdrawals, etc. I'm not aware of a single method to define what is optimal with regard to withdrawal strategies.

Here's an example from an analysis done by Wade Pfau of a paper authored by David Blanchett, Maciej Kowara, and Peng Chen. It serves to illustrate that there is no consensus on what is 'optimal' with regard to withdrawal strategies.
The article creates a measure called “Withdrawal Efficiency Rate” (WER). The article is based on Monte Carlo simulations for asset returns and ages of death for a 65-year old couple. For each simulation, the authors first calculate the highest sustainable withdrawal rate for constant inflation-adjusted spending for someone with perfect foresight about the sequence of asset returns and the age of death that will be realized.

Then, for a variety of withdrawal strategies, the authors calculate the percentage of potential withdrawals the strategy allows with respect to the maximum sustainable withdrawal rate under perfect foresight. The higher the percentage, the more efficient is the withdrawal strategy. No importance is placed on leaving a bequest.
The paper suffers from the same problem I see in all of the variable withdrawal rate / safe withdrawal rate research. It tells you to spend as much as possible on a utility-adjusted basis. That’s great, but it isn’t otherwise connected to any sort of lifestyle spending goals or minimum needs. Though the utility function would discourage it to some extent, this approach would have you keep spending amounts well beyond your lifestyle goals. The utility function also does not incorporate a flooring level, which makes it unclear about where the punishing aspects of the utility function really start to kick in with respect to the minimum needs and desired lifestyle. I just mean that the spending paths coming from this will be somewhat arbitrary and not connected to a retiree’s needs. That is a potential problem.
emphasis added

I would argue that a withdrawal strategy that provides for higher withdrawals early in one's retirement and lower withdrawals later may be more 'optimal' in a way than another withdrawal strategy provides for lower starting withdrawals and higher withdrawals later on, even if the latter provided for more real dollars of withdrawals. This could be due to (1) the retiree has a higher likelihood of being alive to enjoy the higher starting withdrawals and (2) retirees tend to have higher spending needs when younger than older.

I'm reminded of a recent statement made by Paula Pant on the Afford Anything podcast. She said that most financial equations are only mathematical expressions of a particular way of thinking, which implies that there is no such thing as a universally optimal withdrawal strategy.
“It's a dangerous business, Frodo, going out your door. You step onto the road, and if you don't keep your feet, there's no knowing where you might be swept off to.” J.R.R. Tolkien,The Lord of the Rings

User avatar
siamond
Posts: 4305
Joined: Mon May 28, 2012 5:50 am

Re: A different take on retirement income: time segmentation

Post by siamond » Thu Jul 19, 2018 10:02 pm

willthrill81 wrote:
Thu Jul 19, 2018 9:41 pm
I would argue that a withdrawal strategy that provides for higher withdrawals early in one's retirement and lower withdrawals later may be more 'optimal' in a way than another withdrawal strategy provides for lower starting withdrawals and higher withdrawals later on, even if the latter provided for more real dollars of withdrawals. This could be due to (1) the retiree has a higher likelihood of being alive to enjoy the higher starting withdrawals and (2) retirees tend to have higher spending needs when younger than older.
I didn't express a disagreement with that view, did I? I am only debating the ways to get there... :wink:
willthrill81 wrote:
Thu Jul 19, 2018 9:41 pm
Do you mean 'sub-optimal' to be a smaller ending or average portfolio, smaller withdrawals, 'smoother' withdrawals, etc. I'm not aware of a single method to define what is optimal with regard to withdrawal strategies. [...]
I'm reminded of a recent statement made by Paula Pant on the Afford Anything podcast. She said that most financial equations are only mathematical expressions of a particular way of thinking, which implies that there is no such thing as a universally optimal withdrawal strategy.
Yes, yes, I wasn't trying to be precise about optimality, and you're correct this is subjective (I played with WER, HREFF, utility functions and all that stuff more than I should admit, none of that is terribly convincing). I like the quote, by the way.

I was just trying to make you understand that what you think you gained (income-wise) by being more aggressive on the DSP side might be more than counter-balanced by the decision-making on the ESP side, and that it is a direct consequence of hard partitioning in two buckets, which wasn't necessary in the first place. In such context, optimality is about income levels.

Beyond that, yes, agreed, optimality is subjective. The LMP/RP folks put a LOT of emphasis on having a hard floor, and optimality for them is mostly that. Personally, I am fine about having a soft floor, and I care more about maximizing income over the rest of my life. Yes, there is no universal optimal strategy, that part is clear.

User avatar
LadyGeek
Site Admin
Posts: 49146
Joined: Sat Dec 20, 2008 5:34 pm
Location: Philadelphia
Contact:

Re: A different take on retirement income: time segmentation

Post by LadyGeek » Thu Jul 19, 2018 10:04 pm

This thread is now in the Personal Finance (Not Investing) forum (withdrawal strategy).

Acronym help:

LMP: Liability Matching Portfolio
wiki wrote:A portfolio which produces enough guaranteed income to meet needs, such as an all-TIPS portfolio for retirement income. See: Matching strategy
VPW: Variable percentage withdrawal
Wiki To some, the glass is half full. To others, the glass is half empty. To an engineer, it's twice the size it needs to be.

User avatar
willthrill81
Posts: 6436
Joined: Thu Jan 26, 2017 3:17 pm
Location: USA

Re: A different take on retirement income: time segmentation

Post by willthrill81 » Thu Jul 19, 2018 10:06 pm

siamond wrote:
Thu Jul 19, 2018 10:02 pm
willthrill81 wrote:
Thu Jul 19, 2018 9:41 pm
I would argue that a withdrawal strategy that provides for higher withdrawals early in one's retirement and lower withdrawals later may be more 'optimal' in a way than another withdrawal strategy provides for lower starting withdrawals and higher withdrawals later on, even if the latter provided for more real dollars of withdrawals. This could be due to (1) the retiree has a higher likelihood of being alive to enjoy the higher starting withdrawals and (2) retirees tend to have higher spending needs when younger than older.
I didn't express a disagreement with that view, did I? I am only debating the ways to get there... :wink:
willthrill81 wrote:
Thu Jul 19, 2018 9:41 pm
Do you mean 'sub-optimal' to be a smaller ending or average portfolio, smaller withdrawals, 'smoother' withdrawals, etc. I'm not aware of a single method to define what is optimal with regard to withdrawal strategies. [...]
I'm reminded of a recent statement made by Paula Pant on the Afford Anything podcast. She said that most financial equations are only mathematical expressions of a particular way of thinking, which implies that there is no such thing as a universally optimal withdrawal strategy.
Yes, yes, I wasn't trying to be precise about optimality, and you're correct this is subjective (I played with WER, HREFF, utility functions and all that stuff more than I should admit, none of that is terribly convincing). I like the quote, by the way.

I was just trying to make you understand that what you think you gained (income-wise) by being more aggressive on the DSP side might be more than counter-balanced by the decision-making on the ESP side, and that it is a direct consequence of hard partitioning in two buckets, which wasn't necessary in the first place. In such context, optimality is about income levels.

Beyond that, yes, agreed, optimality is subjective. The LMP/RP folks put a LOT of emphasis on having a hard floor, and optimality for them is mostly that. Personally, I am fine about having a soft floor, and I care more about maximizing income over the rest of my life. Yes, there is no universal optimal strategy, that part is clear.
Cool. :beer

So are you more of a fan of something like the VPW for an entire portfolio, or do you have another method you personally prefer, and why?
“It's a dangerous business, Frodo, going out your door. You step onto the road, and if you don't keep your feet, there's no knowing where you might be swept off to.” J.R.R. Tolkien,The Lord of the Rings

User avatar
siamond
Posts: 4305
Joined: Mon May 28, 2012 5:50 am

Re: A different take on retirement income: time segmentation

Post by siamond » Thu Jul 19, 2018 10:19 pm

willthrill81 wrote:
Thu Jul 19, 2018 10:06 pm
So are you more of a fan of something like the VPW for an entire portfolio, or do you have another method you personally prefer, and why?
After long meandering and modeling/experimenting with numerous methods and metrics, I ended up settling on a use of the PMT function that shares numerous commonalities with VPW but also significantly departs from it. Yes, this is based on a non-partitioned portfolio with a fixed AA, aiming at fulfilling multiple goals of ours and addressing multiple fears of ours in one single swoop. This comes on top of the (future) fixed income my wife and I will get in a decade or so. Properly dealing with such future cash flows is actually a tricky part of the equation that few researchers ever discuss.

I'd rather not enter in more details in this thread, to avoid derailing the discussion. I'm happy to explain more details by PM if you're genuinely interested. And again, I wasn't trying to prove 'time segmentation' wrong. I was just trying to put it in some perspective.

User avatar
Sandtrap
Posts: 5454
Joined: Sat Nov 26, 2016 6:32 pm
Location: Hawaii😀 Northern AZ.😳 Retired.

Re: A different take on retirement income: time segmentation

Post by Sandtrap » Thu Jul 19, 2018 10:36 pm

willthrill81 wrote:
Mon Jul 16, 2018 11:38 am
goodenyou wrote:
Mon Jul 16, 2018 11:17 am
What do you think of the theory that the SPIA allows for a more aggressive equity portfolio that likely will lead to more growth and likely more legacy?
I don't buy it. Buying a SPIA is not likely to provide retirees with a meaningfully greater amount in their portfolio to allocate to equities. The payout ratio for a SPIA with an inflation-adjustment for a 65 year old couple is roughly 4% these days, which is close to the '4% rule of thumb'. Granted, the '4% rule' hasn't held with only nominal bonds in the past, but there's no need for a retiree to hold only nominal bonds any more. They can now buy TIPS, and with current real returns, they can guarantee about 27 years of inflation-adjusted spending, very close to the 30 year mark used with the '4% rule'. A small (e.g. 25%) allocation to stocks would historically be very prudent to add anyway to a ESP.

Yes, the SPIA is guaranteed for life, but that comes at the expense of giving up those assets forever, including any potential for growth and legacy that may come with retaining those assets.

So if a person who desires a 50/50 overall AA puts half of it in a SPIA instead of bonds, that doesn't enable them to be more 'aggressive' with the other half.
Excellent point!!!
j

User avatar
LadyGeek
Site Admin
Posts: 49146
Joined: Sat Dec 20, 2008 5:34 pm
Location: Philadelphia
Contact:

Re: A different take on retirement income: time segmentation

Post by LadyGeek » Fri Jul 20, 2018 9:22 am

siamond wrote:
Thu Jul 19, 2018 10:19 pm
willthrill81 wrote:
Thu Jul 19, 2018 10:06 pm
So are you more of a fan of something like the VPW for an entire portfolio, or do you have another method you personally prefer, and why?
After long meandering and modeling/experimenting with numerous methods and metrics, I ended up settling on a use of the PMT function that shares numerous commonalities with VPW but also significantly departs from it. Yes, this is based on a non-partitioned portfolio with a fixed AA, aiming at fulfilling multiple goals of ours and addressing multiple fears of ours in one single swoop. This comes on top of the (future) fixed income my wife and I will get in a decade or so. Properly dealing with such future cash flows is actually a tricky part of the equation that few researchers ever discuss.

I'd rather not enter in more details in this thread, to avoid derailing the discussion. I'm happy to explain more details by PM if you're genuinely interested. And again, I wasn't trying to prove 'time segmentation' wrong. I was just trying to put it in some perspective.
If it's worth a discussion, feel free to start a new thread.
Wiki To some, the glass is half full. To others, the glass is half empty. To an engineer, it's twice the size it needs to be.

smitcat
Posts: 2071
Joined: Mon Nov 07, 2016 10:51 am

Re: A different take on retirement income: time segmentation

Post by smitcat » Fri Jul 20, 2018 10:12 am

willthrill81 wrote:
Thu Jul 19, 2018 10:06 pm
siamond wrote:
Thu Jul 19, 2018 10:02 pm
willthrill81 wrote:
Thu Jul 19, 2018 9:41 pm
I would argue that a withdrawal strategy that provides for higher withdrawals early in one's retirement and lower withdrawals later may be more 'optimal' in a way than another withdrawal strategy provides for lower starting withdrawals and higher withdrawals later on, even if the latter provided for more real dollars of withdrawals. This could be due to (1) the retiree has a higher likelihood of being alive to enjoy the higher starting withdrawals and (2) retirees tend to have higher spending needs when younger than older.
I didn't express a disagreement with that view, did I? I am only debating the ways to get there... :wink:
willthrill81 wrote:
Thu Jul 19, 2018 9:41 pm
Do you mean 'sub-optimal' to be a smaller ending or average portfolio, smaller withdrawals, 'smoother' withdrawals, etc. I'm not aware of a single method to define what is optimal with regard to withdrawal strategies. [...]
I'm reminded of a recent statement made by Paula Pant on the Afford Anything podcast. She said that most financial equations are only mathematical expressions of a particular way of thinking, which implies that there is no such thing as a universally optimal withdrawal strategy.
Yes, yes, I wasn't trying to be precise about optimality, and you're correct this is subjective (I played with WER, HREFF, utility functions and all that stuff more than I should admit, none of that is terribly convincing). I like the quote, by the way.

I was just trying to make you understand that what you think you gained (income-wise) by being more aggressive on the DSP side might be more than counter-balanced by the decision-making on the ESP side, and that it is a direct consequence of hard partitioning in two buckets, which wasn't necessary in the first place. In such context, optimality is about income levels.

Beyond that, yes, agreed, optimality is subjective. The LMP/RP folks put a LOT of emphasis on having a hard floor, and optimality for them is mostly that. Personally, I am fine about having a soft floor, and I care more about maximizing income over the rest of my life. Yes, there is no universal optimal strategy, that part is clear.
Cool. :beer

So are you more of a fan of something like the VPW for an entire portfolio, or do you have another method you personally prefer, and why?

Similar to siamond our plan calls for all of our basic needs to be covered by 'fixed income' funds in our investments. The balnce of the portfolio is in stock or stock type investments. Our IPS calls for us to check each year that we have the balance of life needs coverewd in fixed income and that we never take more out of the balance of the portfolio that a VPW woudl indicate for that particlaur year. In the modeling that we have done we will be no where near the VPW rate so it is just there as a 'chack' to ensure we are correct and balcning if needed. There are no separate buckets in this plan for essentials & discretionary spending but we use bucket 'thinking' as a tool to make sure our AA is appropriate for our goals.
We also have a third 'and 4th 'bucket' in developing our plans rules but they are more personalized to us than anything else.

The Wizard
Posts: 12433
Joined: Tue Mar 23, 2010 1:45 pm
Location: Reading, MA

Re: A different take on retirement income: time segmentation

Post by The Wizard » Fri Jul 20, 2018 10:49 am

Sandtrap wrote:
Thu Jul 19, 2018 10:36 pm
willthrill81 wrote:
Mon Jul 16, 2018 11:38 am
goodenyou wrote:
Mon Jul 16, 2018 11:17 am
What do you think of the theory that the SPIA allows for a more aggressive equity portfolio that likely will lead to more growth and likely more legacy?
I don't buy it. Buying a SPIA is not likely to provide retirees with a meaningfully greater amount in their portfolio to allocate to equities. The payout ratio for a SPIA with an inflation-adjustment for a 65 year old couple is roughly 4% these days, which is close to the '4% rule of thumb'. Granted, the '4% rule' hasn't held with only nominal bonds in the past, but there's no need for a retiree to hold only nominal bonds any more. They can now buy TIPS, and with current real returns, they can guarantee about 27 years of inflation-adjusted spending, very close to the 30 year mark used with the '4% rule'. A small (e.g. 25%) allocation to stocks would historically be very prudent to add anyway to a ESP.

Yes, the SPIA is guaranteed for life, but that comes at the expense of giving up those assets forever, including any potential for growth and legacy that may come with retaining those assets.

So if a person who desires a 50/50 overall AA puts half of it in a SPIA instead of bonds, that doesn't enable them to be more 'aggressive' with the other half.
Excellent point!!!
j
Interesting point, but not totally correct.
As a single person, I did some Single Premium annuitization with TIAA at age 63 and got 6.43% initial payout with 10 year guarantee.
I did small additional annuitization at age 67 and got 7.04% initial annual payout, also with 10 year guarantee.

These lifetime annuities are with TIAA as I mentioned, and while some of them are "fixed" (based on TIAA Traditional) the majority are based on commercial real estate (TREA) and a lesser percentage on the broad stock market (CREF Stock) both of which will tend to increase over the years, as will my monthly income.

Unfortunately, decent annuitization options like these aren't available to the general public.

And to the point of this thread, annuity and SS income more than cover my basic/essential expenses, making my remaining portfolio available for discretionary expenses...
Attempted new signature...

User avatar
willthrill81
Posts: 6436
Joined: Thu Jan 26, 2017 3:17 pm
Location: USA

Re: A different take on retirement income: time segmentation

Post by willthrill81 » Fri Jul 20, 2018 11:30 am

smitcat wrote:
Fri Jul 20, 2018 10:12 am
willthrill81 wrote:
Thu Jul 19, 2018 10:06 pm
siamond wrote:
Thu Jul 19, 2018 10:02 pm
willthrill81 wrote:
Thu Jul 19, 2018 9:41 pm
I would argue that a withdrawal strategy that provides for higher withdrawals early in one's retirement and lower withdrawals later may be more 'optimal' in a way than another withdrawal strategy provides for lower starting withdrawals and higher withdrawals later on, even if the latter provided for more real dollars of withdrawals. This could be due to (1) the retiree has a higher likelihood of being alive to enjoy the higher starting withdrawals and (2) retirees tend to have higher spending needs when younger than older.
I didn't express a disagreement with that view, did I? I am only debating the ways to get there... :wink:
willthrill81 wrote:
Thu Jul 19, 2018 9:41 pm
Do you mean 'sub-optimal' to be a smaller ending or average portfolio, smaller withdrawals, 'smoother' withdrawals, etc. I'm not aware of a single method to define what is optimal with regard to withdrawal strategies. [...]
I'm reminded of a recent statement made by Paula Pant on the Afford Anything podcast. She said that most financial equations are only mathematical expressions of a particular way of thinking, which implies that there is no such thing as a universally optimal withdrawal strategy.
Yes, yes, I wasn't trying to be precise about optimality, and you're correct this is subjective (I played with WER, HREFF, utility functions and all that stuff more than I should admit, none of that is terribly convincing). I like the quote, by the way.

I was just trying to make you understand that what you think you gained (income-wise) by being more aggressive on the DSP side might be more than counter-balanced by the decision-making on the ESP side, and that it is a direct consequence of hard partitioning in two buckets, which wasn't necessary in the first place. In such context, optimality is about income levels.

Beyond that, yes, agreed, optimality is subjective. The LMP/RP folks put a LOT of emphasis on having a hard floor, and optimality for them is mostly that. Personally, I am fine about having a soft floor, and I care more about maximizing income over the rest of my life. Yes, there is no universal optimal strategy, that part is clear.
Cool. :beer

So are you more of a fan of something like the VPW for an entire portfolio, or do you have another method you personally prefer, and why?

Similar to siamond our plan calls for all of our basic needs to be covered by 'fixed income' funds in our investments. The balnce of the portfolio is in stock or stock type investments. Our IPS calls for us to check each year that we have the balance of life needs coverewd in fixed income and that we never take more out of the balance of the portfolio that a VPW woudl indicate for that particlaur year. In the modeling that we have done we will be no where near the VPW rate so it is just there as a 'chack' to ensure we are correct and balcning if needed. There are no separate buckets in this plan for essentials & discretionary spending but we use bucket 'thinking' as a tool to make sure our AA is appropriate for our goals.
We also have a third 'and 4th 'bucket' in developing our plans rules but they are more personalized to us than anything else.
Very interesting. Like marcopolo pointed out early in the thread, it seems that you might be using some kind of version of this essential-discretionary segmentation strategy. While you don't explicitly have separate buckets for these categories, you're keeping your basic needs covered by fixed income (conservative) and placing the remainder (discretionary) in stocks (aggressive).
“It's a dangerous business, Frodo, going out your door. You step onto the road, and if you don't keep your feet, there's no knowing where you might be swept off to.” J.R.R. Tolkien,The Lord of the Rings

User avatar
willthrill81
Posts: 6436
Joined: Thu Jan 26, 2017 3:17 pm
Location: USA

Re: A different take on retirement income: time segmentation

Post by willthrill81 » Fri Jul 20, 2018 11:33 am

The Wizard wrote:
Fri Jul 20, 2018 10:49 am
Sandtrap wrote:
Thu Jul 19, 2018 10:36 pm
willthrill81 wrote:
Mon Jul 16, 2018 11:38 am
goodenyou wrote:
Mon Jul 16, 2018 11:17 am
What do you think of the theory that the SPIA allows for a more aggressive equity portfolio that likely will lead to more growth and likely more legacy?
I don't buy it. Buying a SPIA is not likely to provide retirees with a meaningfully greater amount in their portfolio to allocate to equities. The payout ratio for a SPIA with an inflation-adjustment for a 65 year old couple is roughly 4% these days, which is close to the '4% rule of thumb'. Granted, the '4% rule' hasn't held with only nominal bonds in the past, but there's no need for a retiree to hold only nominal bonds any more. They can now buy TIPS, and with current real returns, they can guarantee about 27 years of inflation-adjusted spending, very close to the 30 year mark used with the '4% rule'. A small (e.g. 25%) allocation to stocks would historically be very prudent to add anyway to a ESP.

Yes, the SPIA is guaranteed for life, but that comes at the expense of giving up those assets forever, including any potential for growth and legacy that may come with retaining those assets.

So if a person who desires a 50/50 overall AA puts half of it in a SPIA instead of bonds, that doesn't enable them to be more 'aggressive' with the other half.
Excellent point!!!
j
Interesting point, but not totally correct.
As a single person, I did some Single Premium annuitization with TIAA at age 63 and got 6.43% initial payout with 10 year guarantee.
I did small additional annuitization at age 67 and got 7.04% initial annual payout, also with 10 year guarantee.

These lifetime annuities are with TIAA as I mentioned, and while some of them are "fixed" (based on TIAA Traditional) the majority are based on commercial real estate (TREA) and a lesser percentage on the broad stock market (CREF Stock) both of which will tend to increase over the years, as will my monthly income.

Unfortunately, decent annuitization options like these aren't available to the general public.

And to the point of this thread, annuity and SS income more than cover my basic/essential expenses, making my remaining portfolio available for discretionary expenses...
Since your annuity payouts did not seem to include an inflation rider, that, along with you being a single male, explains the higher payout ratio. But over the long-term, you'll likely at least keep pace with inflation and probably do better.

I'm glad that I'll have access to TIAA's annuities; they definitely seem to be the 'good guys' in the annuity arena. I still have serious reservations about annuities, but I may feel differently when I'm staring down the barrel of a 40+ year retirement.

Are your annuity payouts guaranteed to never go down?
“It's a dangerous business, Frodo, going out your door. You step onto the road, and if you don't keep your feet, there's no knowing where you might be swept off to.” J.R.R. Tolkien,The Lord of the Rings

User avatar
siamond
Posts: 4305
Joined: Mon May 28, 2012 5:50 am

Re: A different take on retirement income: time segmentation

Post by siamond » Fri Jul 20, 2018 11:51 am

smitcat wrote:
Fri Jul 20, 2018 10:12 am
Similar to siamond our plan calls for all of our basic needs to be covered by 'fixed income' funds in our investments. The balance of the portfolio is in stock or stock type investments. Our IPS calls for us to check each year that we have the balance of life needs covered in fixed income and that we never take more out of the balance of the portfolio that a VPW would indicate for that particular year.
Actually, our essential needs are not fully covered by fixed income, and our portfolio (via a VPW/PMT derivative method) will have to take care of a chunk of that, plus discretionary expenses. We *may* decide by the time we turn 70 to increase fixed income (creating a hard floor for essentials) by turning some of the portfolio in an SPIA, but I would be quite reluctant to do so (I'd rather give a bequest to my heirs than to insurers!).

Smitcat's approach makes sense to me though. I would tend to think that if somebody is truly intent on having a hard floor for essential expenses, it has better be truly 'hard' (i.e. no 4% rule of thumb or things like that) and address longevity risk (i.e. no time-limited TIPS ladder). Which probably spells out SPIA if SS/Pension are not enough. I am not quite ready to follow such track, but I appreciate the fact that at an older age, our thinking might shift towards it. Anyhoo, we have more than a decade to reflect upon this! :wink:

User avatar
Sandtrap
Posts: 5454
Joined: Sat Nov 26, 2016 6:32 pm
Location: Hawaii😀 Northern AZ.😳 Retired.

Re: A different take on retirement income: time segmentation

Post by Sandtrap » Fri Jul 20, 2018 12:06 pm

The Wizard wrote:
Fri Jul 20, 2018 10:49 am
Sandtrap wrote:
Thu Jul 19, 2018 10:36 pm
willthrill81 wrote:
Mon Jul 16, 2018 11:38 am
goodenyou wrote:
Mon Jul 16, 2018 11:17 am
What do you think of the theory that the SPIA allows for a more aggressive equity portfolio that likely will lead to more growth and likely more legacy?
I don't buy it. Buying a SPIA is not likely to provide retirees with a meaningfully greater amount in their portfolio to allocate to equities. The payout ratio for a SPIA with an inflation-adjustment for a 65 year old couple is roughly 4% these days, which is close to the '4% rule of thumb'. Granted, the '4% rule' hasn't held with only nominal bonds in the past, but there's no need for a retiree to hold only nominal bonds any more. They can now buy TIPS, and with current real returns, they can guarantee about 27 years of inflation-adjusted spending, very close to the 30 year mark used with the '4% rule'. A small (e.g. 25%) allocation to stocks would historically be very prudent to add anyway to a ESP.

Yes, the SPIA is guaranteed for life, but that comes at the expense of giving up those assets forever, including any potential for growth and legacy that may come with retaining those assets.

So if a person who desires a 50/50 overall AA puts half of it in a SPIA instead of bonds, that doesn't enable them to be more 'aggressive' with the other half.
Excellent point!!!
j
Interesting point, but not totally correct.
As a single person, I did some Single Premium annuitization with TIAA at age 63 and got 6.43% initial payout with 10 year guarantee.
I did small additional annuitization at age 67 and got 7.04% initial annual payout, also with 10 year guarantee.

These lifetime annuities are with TIAA as I mentioned, and while some of them are "fixed" (based on TIAA Traditional) the majority are based on commercial real estate (TREA) and a lesser percentage on the broad stock market (CREF Stock) both of which will tend to increase over the years, as will my monthly income.

Unfortunately, decent annuitization options like these aren't available to the general public.

And to the point of this thread, annuity and SS income more than cover my basic/essential expenses, making my remaining portfolio available for discretionary expenses...
If these decent annuitization options are not available to most of the general public,
which decent ones are available to most of us?

mahalo,
jim

The Wizard
Posts: 12433
Joined: Tue Mar 23, 2010 1:45 pm
Location: Reading, MA

Re: A different take on retirement income: time segmentation

Post by The Wizard » Fri Jul 20, 2018 2:11 pm

willthrill81 wrote:
Fri Jul 20, 2018 11:33 am

Since your annuity payouts did not seem to include an inflation rider, that, along with you being a single male, explains the higher payout ratio. But over the long-term, you'll likely at least keep pace with inflation and probably do better.

I'm glad that I'll have access to TIAA's annuities; they definitely seem to be the 'good guys' in the annuity arena. I still have serious reservations about annuities, but I may feel differently when I'm staring down the barrel of a 40+ year retirement.

Are your annuity payouts guaranteed to never go down?
Definitely no guarantee on my annuity payments not going down.
To do so would require some sort of creative foolishness, the sort that tends to give variable annuities a bad name.

As a practical matter, my TIAA Traditional annuities are quite unlikely to have a decrease in monthly payout. And unlike SPIAs, Trad annuities have had unscheduled 1% to 2% increases in some recent years due to TIAA's stronger than expected reserves.

But the TREA and CREF Stock payout annuities are pretty much tied to the 4% Assumed Investment Return algorithm, meaning that if the NAV of those funds increases 4% each year, my monthly payout stays the same. Changes up or down from +4.0% cause proportional changes in my monthly payout. Really simple to compute yourself, with no caps on either the upside or the downside.
I say "pretty much" because they tweak the payout adjustments slightly based on actual annuitant mortality experience over the past year vs actuarial expectations...
Attempted new signature...

The Wizard
Posts: 12433
Joined: Tue Mar 23, 2010 1:45 pm
Location: Reading, MA

Re: A different take on retirement income: time segmentation

Post by The Wizard » Fri Jul 20, 2018 2:15 pm

Sandtrap wrote:
Fri Jul 20, 2018 12:06 pm
The Wizard wrote:
Fri Jul 20, 2018 10:49 am

...Unfortunately, decent annuitization options like these aren't available to the general public.

And to the point of this thread, annuity and SS income more than cover my basic/essential expenses, making my remaining portfolio available for discretionary expenses...
If these decent annuitization options are not available to most of the general public,
which decent ones are available to most of us?

mahalo,
jim
Aside from standard fixed monthly payment SPIAs, I don't really know...
:(

Edit: I'm aware that Vanguard offers variable annuities, but I've made no effort to understand them or compare them to TIAA.
With TIAA, I don't have to be in variable annuities during the accumulation phase, I can be in their S&P 500 index fund TISPX, with an ER of 0.06%, not as low as Vanguard VFIAX, but decent.
But I can't "annuitize" TISPX. But I can transfer $100k from TISPX into CREF Stock a week before I retire and annuitize that for lifetime income.
No idea if Vanguard offers similar flexibility...
Last edited by The Wizard on Fri Jul 20, 2018 2:50 pm, edited 1 time in total.
Attempted new signature...

User avatar
willthrill81
Posts: 6436
Joined: Thu Jan 26, 2017 3:17 pm
Location: USA

Re: A different take on retirement income: time segmentation

Post by willthrill81 » Fri Jul 20, 2018 2:22 pm

The Wizard wrote:
Fri Jul 20, 2018 2:11 pm
willthrill81 wrote:
Fri Jul 20, 2018 11:33 am

Since your annuity payouts did not seem to include an inflation rider, that, along with you being a single male, explains the higher payout ratio. But over the long-term, you'll likely at least keep pace with inflation and probably do better.

I'm glad that I'll have access to TIAA's annuities; they definitely seem to be the 'good guys' in the annuity arena. I still have serious reservations about annuities, but I may feel differently when I'm staring down the barrel of a 40+ year retirement.

Are your annuity payouts guaranteed to never go down?
Definitely no guarantee on my annuity payments not going down.
To do so would require some sort of creative foolishness, the sort that tends to give variable annuities a bad name.

As a practical matter, my TIAA Traditional annuities are quite unlikely to have a decrease in monthly payout. And unlike SPIAs, Trad annuities have had unscheduled 1% to 2% increases in some recent years due to TIAA's stronger than expected reserves.

But the TREA and CREF Stock payout annuities are pretty much tied to the 4% Assumed Investment Return algorithm, meaning that if the NAV of those funds increases 4% each year, my monthly payout stays the same. Changes up or down from +4.0% cause proportional changes in my monthly payout. Really simple to compute yourself, with no caps on either the upside or the downside.
I say "pretty much" because they tweak the payout adjustments slightly based on actual annuitant mortality experience over the past year vs actuarial expectations...
So for the TREA and CREF stock annuities, those payouts would have taken a big hit after 2008-2009? That seems to partially defeat the purpose of an annuity, besides the mortality credits you seem to get regardless.
“It's a dangerous business, Frodo, going out your door. You step onto the road, and if you don't keep your feet, there's no knowing where you might be swept off to.” J.R.R. Tolkien,The Lord of the Rings

The Wizard
Posts: 12433
Joined: Tue Mar 23, 2010 1:45 pm
Location: Reading, MA

Re: A different take on retirement income: time segmentation

Post by The Wizard » Fri Jul 20, 2018 3:01 pm

willthrill81 wrote:
Fri Jul 20, 2018 2:22 pm
The Wizard wrote:
Fri Jul 20, 2018 2:11 pm

Definitely no guarantee on my annuity payments not going down.
To do so would require some sort of creative foolishness, the sort that tends to give variable annuities a bad name.

As a practical matter, my TIAA Traditional annuities are quite unlikely to have a decrease in monthly payout. And unlike SPIAs, Trad annuities have had unscheduled 1% to 2% increases in some recent years due to TIAA's stronger than expected reserves.

But the TREA and CREF Stock payout annuities are pretty much tied to the 4% Assumed Investment Return algorithm, meaning that if the NAV of those funds increases 4% each year, my monthly payout stays the same. Changes up or down from +4.0% cause proportional changes in my monthly payout. Really simple to compute yourself, with no caps on either the upside or the downside.
I say "pretty much" because they tweak the payout adjustments slightly based on actual annuitant mortality experience over the past year vs actuarial expectations...
So for the TREA and CREF stock annuities, those payouts would have taken a big hit after 2008-2009? That seems to partially defeat the purpose of an annuity, besides the mortality credits you seem to get regardless.
A big income hit in 08-09, indeed, which is why one should probably not have income tied too heavily to the stock market or CRE.
But consider a retiree with $500k in TSM or S&P 500 index fund going into 2008. 4% swr gives them $20k per year from that segment of their portfolio. But then that segment drops 40% (?) by March 2009.
So, yeah, stock market volatility is an issue we need to contend with either way...
Attempted new signature...

longinvest
Posts: 3108
Joined: Sat Aug 11, 2012 8:44 am

Re: A different take on retirement income: time segmentation

Post by longinvest » Fri Jul 20, 2018 10:50 pm

Willthrill81,
willthrill81 wrote:
Sun Jul 15, 2018 5:36 pm
Another problem not mentioned nearly as often is that while most of the traditional withdrawal strategies assume relatively fixed income needs from the point of retirement until death, this is not supported by real world data. Approximately two-thirds of retirees report that their spending declines over their retirement (half of the remainder report no change in spending, and the other half report spending increases). Retirees tend to reduce their annual spending in real dollars by 1-2% annually, even during their latter years when increased medical costs are offset by reduced transportation, food, and discretionary spending; there are clearly exceptions to this, but they are indeed exceptions to the widespread trend.
I would be very careful with such studies. Averages are just that: averages. Give me $1,000,000 to spend every year (wouldn't that be nice?!), and I'll sure find a way to spend or give it. There might exist people who wouldn't succeed to spend and give as much as that, but, I feel pretty confident I would be able to do it, even when older.

Also, these are just statistics of the past. The future might differ. Improvements in healthcare might increase longevity but require more money. It's really difficult to predict the future.
willthrill81 wrote:
Sun Jul 15, 2018 5:36 pm
This means that retirees' are likely to need/want to spend more in the early years of retirement and less in the later years of retirement.
Personally, I don't care about spending less later; I care about spending as much as reasonably possible as soon as possible. That's what VPW does. It doesn't, in any way, predict stable withdrawals (that would require a stable and predictable portfolio).
willthrill81 wrote:
Sun Jul 15, 2018 5:36 pm
To date, the focus has been primarily on sacrificing retirees' need/want for more income in the early stages of retirement for the security of reducing the likelihood of running out of money.
I disagree. The whole idea of variable percentage withdrawal (VPW) is exactly to spend as much as is reasonably possible, given the current state of things at at every point in time! It's actually quite aggressive at it when compared with "safe withdrawal rates".
willthrill81 wrote:
Sun Jul 15, 2018 5:36 pm
The basic gist of what I propose is this: split retirees' portfolio into two, one for essential spending (ES portfolio) and another for discretionary spending (DS portfolio). Then, apply different withdrawal methods for the two portfolios, based on your need, willingness, and ability to take on risk for both ES and DS. Income strategies involving use of the ESP should logically be more conservative than those involving the DSP. And while the primary purpose of the DSP would be to fund discretionary spending, which typically declines as retirees' age, a secondary purpose could be to at least partially fund significant essential spending needs that might arise, such as long-term care.
The problem is the cost of the ES portfolio, especially for those retiring early. One can buy an inflation-indexed Single Premium Immediate Annuity (inflation-indexed SPIA) at age 50; it won't be cheap, though!

I won't repeat what I've written in the VPW thread (see posts like this one).

There are countless approaches to planning retirement. There are simpler approaches and more complex approaches.

The more difficult challenge is to to design a simple, yet robust approach that could easily be used by common people. My personal goal is for my wife (who gets bored in less than 10 minutes, when discussing money) to be able to manage our portfolio and retirement.

That's where the whole VPW approach comes in (including base income, Social Security delayed to 70 and a CD ladder to fill the gap until then, and annuitization at age 80, as explained in our wiki). It is very simple to both explain and use.

A more comprehensive and somewhat simpler* approach is presented in the thread Variable Percentage Withdrawal (VPW-adv) for Advanced Users. It simplifies portfolio management using a single portfolio, with no need for a CD ladder to fill the gap in pension payments, and it includes provisions to keep lifelong liquidity (something that isn't specifically discussed in the VPW wiki page).

* It's simpler when using a spreadsheet, but it involves more complex calculations.

One can easily develop lots of variations on these ideas. But basically, the main trade-off is likely to remain between guaranteed stable lifelong income (such as pensions, SPIAs, Social Security, and non-rolling ladders to cover the gap of delayed pensions) and keeping liquidity (e.g. a portfolio of fluctuating assets) along with flexible portfolio withdrawals. One caveat: guaranteed stable lifelong income is expensive early in life.
Bogleheads investment philosophy | Lifelong Portfolio: 25% each of (domestic / international) stocks / domestic (nominal / inflation-indexed) long-term bonds | VCN/VXC/VLB/ZRR

User avatar
willthrill81
Posts: 6436
Joined: Thu Jan 26, 2017 3:17 pm
Location: USA

Re: A different take on retirement income: time segmentation

Post by willthrill81 » Fri Jul 20, 2018 11:50 pm

longinvest wrote:
Fri Jul 20, 2018 10:50 pm
willthrill81 wrote:
Sun Jul 15, 2018 5:36 pm
Another problem not mentioned nearly as often is that while most of the traditional withdrawal strategies assume relatively fixed income needs from the point of retirement until death, this is not supported by real world data. Approximately two-thirds of retirees report that their spending declines over their retirement (half of the remainder report no change in spending, and the other half report spending increases). Retirees tend to reduce their annual spending in real dollars by 1-2% annually, even during their latter years when increased medical costs are offset by reduced transportation, food, and discretionary spending; there are clearly exceptions to this, but they are indeed exceptions to the widespread trend.
I would be very careful with such studies. Averages are just that: averages. Give me $1,000,000 to spend every year (wouldn't that be nice?!), and I'll sure find a way to spend or give it. There might exist people who wouldn't succeed to spend and give as much as that, but, I feel pretty confident I would be able to do it, even when older.

Also, these are just statistics of the past. The future might differ. Improvements in healthcare might increase longevity but require more money. It's really difficult to predict the future.
That's true. It could also be said of much of our knowledge of finance, which is largely derived from the knowledge of the past.
longinvest wrote:
Fri Jul 20, 2018 10:50 pm
willthrill81 wrote:
Sun Jul 15, 2018 5:36 pm
To date, the focus has been primarily on sacrificing retirees' need/want for more income in the early stages of retirement for the security of reducing the likelihood of running out of money.
I disagree. The whole idea of variable percentage withdrawal (VPW) is exactly to spend as much as is reasonably possible, given the current state of things at at every point in time! It's actually quite aggressive at it when compared with "safe withdrawal rates".
You'll note that I said "primarily." But with regard to VPW, the percentage withdrawals increase as retirees' age. Depending on how their portfolio performs, this could easily (likely in younger years) lead to larger withdrawals in real dollars over time. But the research is clear that two-thirds of retirees need more money when they're younger rather than older.

So again, there are trade offs at work. If retirees are trying to squeeze as many dollars as they can out of their portfolio at every given point along the way, then VPW might be appropriate for them. But many retirees may want to front-load their withdrawals so as to allow for greater discretionary spending while they're younger. I know that my parents certainly do.
longinvest wrote:
Fri Jul 20, 2018 10:50 pm
willthrill81 wrote:
Sun Jul 15, 2018 5:36 pm
The basic gist of what I propose is this: split retirees' portfolio into two, one for essential spending (ES portfolio) and another for discretionary spending (DS portfolio). Then, apply different withdrawal methods for the two portfolios, based on your need, willingness, and ability to take on risk for both ES and DS. Income strategies involving use of the ESP should logically be more conservative than those involving the DSP. And while the primary purpose of the DSP would be to fund discretionary spending, which typically declines as retirees' age, a secondary purpose could be to at least partially fund significant essential spending needs that might arise, such as long-term care.
The problem is the cost of the ES portfolio, especially for those retiring early. One can buy an inflation-indexed Single Premium Immediate Annuity (inflation-indexed SPIA) at age 50; it won't be cheap, though!
This 'problem' is a drawback of every withdrawal/income strategy: it costs more to provide income for younger retirees. And I don't personally think that SPIAs are a one-size-fits-all solution for retirees at any stage of life.

And while I'm not a huge fan of fixed-dollar withdrawal strategies, history has shown that it may not be necessary for retirees to go below about 3% for starting withdrawals, regardless of the presumed length of their retirement. So the 'cost' of a withdrawal strategy for essential spending may not be greater than 33X annual spending. I emphasize "may."
longinvest wrote:
Fri Jul 20, 2018 10:50 pm
I won't repeat what I've written in length in the VPW thread (see posts like this one).

There are countless approaches to planning retirement. There are simpler approaches and more complex approaches.

The more difficult challenge is to to design a simple, yet robust approach that could be used not only by people who eat spreadsheets for breakfast, but also by common people. My personal goal is for my wife (who gets bored in less than 10 minutes, when discussing money) to be able to manage our portfolio and retirement.

That's where the whole VPW approach comes in (including base income, Social Security delayed to 70 and a CD ladder to fill the gap until then, and annuitization at age 80, as explained in our wiki). It is very simple to both explain and use.

A more comprehensive and somewhat simpler* approach is presented in the thread Variable Percentage Withdrawal (VPW-adv) for Advanced Users. It simplifies portfolio management using a single portfolio, with no need for a CD ladder to fill the gap in pension payments, and it includes provisions to keep lifelong liquidity (something that isn't specifically discussed in the VPW wiki page).

* It is simpler when using a spreadsheet, but it involves more complex calculations.

One can easily develop lots of variations on these ideas. But basically, the main trade-off is likely to remain between guaranteed stable lifelong income (such as pensions, SPIAs, Social Security, and non-rolling ladders to cover the gap of delayed pensions) and keeping liquidity (e.g. a portfolio of fluctuating assets) along with flexible portfolio withdrawals. One caveat: guaranteed stable lifelong income is expensive early in life.
I'm glad that VPW works for you, and it clearly works well for some others. But it's not a one-size-fits-all approach, and I've never claimed that this 'essential-discretionary-time' segmentation approach is either. Beyond the numbers, much of a retiree's choice of withdrawal strategies depends on what they wish to accomplish. There is no 'optimum' approach for everyone because 'optimum' is subjective.
“It's a dangerous business, Frodo, going out your door. You step onto the road, and if you don't keep your feet, there's no knowing where you might be swept off to.” J.R.R. Tolkien,The Lord of the Rings

longinvest
Posts: 3108
Joined: Sat Aug 11, 2012 8:44 am

Re: A different take on retirement income: time segmentation

Post by longinvest » Sat Jul 21, 2018 1:05 am

willthrill81 wrote:
Fri Jul 20, 2018 11:50 pm
You'll note that I said "primarily." But with regard to VPW, the percentage withdrawals increase as retirees' age. Depending on how their portfolio performs, this could easily (likely in younger years) lead to larger withdrawals in real dollars over time. But the research is clear that two-thirds of retirees need more money when they're younger rather than older.
Increasing VPW percentages could as easily deliver decreasing withdrawal amounts! Withdrawal amounts will be dictated by (unfortunately unknown) future market returns.
willthrill81 wrote:
Fri Jul 20, 2018 11:50 pm
So again, there are trade offs at work. If retirees are trying to squeeze as many dollars as they can out of their portfolio at every given point along the way, then VPW might be appropriate for them. But many retirees may want to front-load their withdrawals so as to allow for greater discretionary spending while they're younger. I know that my parents certainly do.
First, two notes:

1- The VPW spreadsheet already has parameters to allow all sorts of customization (like increasing the likelihood of a decreasing withdrawal amount, even with good market returns, at the cost of worsening the outcome with bad market returns). I invite you to download the backtesting spreadsheet and play with it. While I don't recommend changing the default settings in the "Table" sheet for a real life use of VPW, playing with these settings to look at various backtests can be very instructive.

2- Inflation-indexed SPIAs can be bought with a CPI minus 1% or minus 2% indexing. This guarantees a controlled decrease in spending power of SPIA payments over time. Many work pensions aren't indexed to inflation and, as a result, have a decreasing purchase power over time.

Second, there's a deeper issue, here, that deserves discussing and that has nothing to do with withdrawal methods.

Who wouldn't dream of always having as much money as one needs at the exact time it is needed? Young parents in their 20s would love to have as much money as retirees. They often don't and have to borrow money to buy a house, for example.

Predicting the future, including future money needs, is difficult. Naturally, most retirees want to spend more on travel and things like that in the earlier days of their retirement. But, there are other money needs that appear in older age. One might not be able to care for one's house anymore, and the proceeds of the sale of the house might not be sufficient to provide for equivalent housing (and the eventual necessary help).

I would think twice before planning for a miserable end of life... But, that's me. :wink:
willthrill81 wrote:
Fri Jul 20, 2018 11:50 pm
And while I'm not a huge fan of fixed-dollar withdrawal strategies, history has shown that it may not be necessary for retirees to go below about 3% for starting withdrawals, regardless of the presumed length of their retirement. So the 'cost' of a withdrawal strategy for essential spending may not be greater than 33X annual spending. I emphasize "may."
I don't see the need to repeat old discussions about fixed-dollar methods, like this one: Michael Kitces 4% rule podcast on Madfientist.
Bogleheads investment philosophy | Lifelong Portfolio: 25% each of (domestic / international) stocks / domestic (nominal / inflation-indexed) long-term bonds | VCN/VXC/VLB/ZRR

longinvest
Posts: 3108
Joined: Sat Aug 11, 2012 8:44 am

Re: A different take on retirement income: time segmentation

Post by longinvest » Sat Jul 21, 2018 1:32 am

Willthrill81,

I've provided a link to a thread presenting a spreadsheet that combines various calculations (and pensions) into a cohesive and robust retirement plan.

Here's the quote:
longinvest wrote:
Fri Jul 20, 2018 10:50 pm
A more comprehensive and somewhat simpler* approach is presented in the thread Variable Percentage Withdrawal (VPW-adv) for Advanced Users. It simplifies portfolio management using a single portfolio, with no need for a CD ladder to fill the gap in pension payments, and it includes provisions to keep lifelong liquidity (something that isn't specifically discussed in the VPW wiki page).

* It's simpler when using a spreadsheet, but it involves more complex calculations.
This spreadsheet can be easily customized to include provisions for travel spending during a number of years, while managing the whole as a single portfolio. The thing is that markets could be uncooperative, in which case the retiree will have to change his plans according to the reality of the moment.

One could alternately buy a term-certain CPI-U-indexed annuity to provide special travel funding for a predetermined number of years. It could be more expensive to do so, but annual payments wouldn't be affected by market returns. I like this example, because it exposes the weakness of seeking too much income certainty: there's no guarantee that the cost of plane tickets and hotels won't increase faster than CPI-U!

Anyway, I invite you again to play with this spreadsheet for "advanced users". Maybe you'll find a way to tweak it to your liking. It's open and free. It's meant to be adapted to one's particular needs.
Bogleheads investment philosophy | Lifelong Portfolio: 25% each of (domestic / international) stocks / domestic (nominal / inflation-indexed) long-term bonds | VCN/VXC/VLB/ZRR

Post Reply