The 4% Rule Just Became a Whole Lot Easier - Allan Roth

Discuss all general (i.e. non-personal) investing questions and issues, investing news, and theory.
User avatar
vineviz
Posts: 14456
Joined: Tue May 15, 2018 1:55 pm
Location: Baltimore, MD

Re: The 4% Rule Just Became a Whole Lot Easier - Allan Roth

Post by vineviz »

Prokofiev wrote: Sun Nov 20, 2022 11:38 am
I don't think this is correct. Getting a graded 2, 3 or 5% SPIA will mean the break-even age MUST be higher than a 0% SPIA. You are effectively
moving payments from the present into the future and your starting point is much lower with the 5% annuity. They cannot both give you the same
dollars up to "average lifespan" since the 5% annuity would then be so much more valuable for the 50% of the pool that lives beyond that age.
It is correct that the amount of fixed increase ( whether 0% or more) is actuarially neutral.

It’s true that IF you could be SURE when you bought the annuity that’s you’d live longer than the insurance company expects you to, the 5% annuity would be very valuable to you.

But none of us know that, and the insurance companies know that none of us know that.
"Far more money has been lost by investors preparing for corrections than has been lost in corrections themselves." ~~ Peter Lynch
ScubaHogg
Posts: 1758
Joined: Sun Nov 06, 2011 3:02 pm

Re: The 4% Rule Just Became a Whole Lot Easier - Allan Roth

Post by ScubaHogg »

nedsaid wrote: Sun Nov 20, 2022 11:49 am
ScubaHogg wrote: Thu Nov 17, 2022 10:51 am
dcabler wrote: Thu Nov 17, 2022 8:06 am
nedsaid wrote: Wed Nov 16, 2022 8:17 pm
ScubaHogg wrote: Wed Nov 16, 2022 6:36 am

Really? This is curiously at odds with what he has said and written about planning as if you are going to have a long retirement



Safety First, Wade Pfau
No. What I said was not at odds with Dr. Wade Pfau. He actually did favor the nominal SPIAs over the CPI adjusted SPIAs because the breakeven point took so long. From memory, it was 12-15 years. There was a thread on this topic some time ago, in any case the CPI adjusted Single Premium Annuities are no longer available. Things have changed, including inflation and interest rates, I don't know his most very recent views.
I looked at this several months ago for the current choices of fixed COLA adjustments available with SPIAs today and it was still around 14 years or so. The surprising thing to me (and, yeah, this is from my memory) was that the breakeven period seemed to be largely independent on the amount of COLA chosen. And with a higher COLA resulting in lower initial payouts, yet another thing to think about when considering one's range of remaining life expectancy.

Cheers.
It’s probably not surprising since they are looking to pay roughly the same amount for a given cohort. I’m guessing the break even is roughly the life expectancy of the cohort.

Nedsaid: Ding! Ding! Ding! Ding! Ding! Ding! Ding! For an average lifespan, a CPI adjusted annuity and a nominal annuity are actuarily neutral. This is the way it has to be. An annuity is a bet with the insurance company that you will live a longer than average lifespan, the actuaries have it all figured out so that the insurance company wins over the many annuitants that it insures.

A big issue here is that people who buy Single Premium Immediate Annuities know this, so you get the healthiest folks with the best prospects of the longer lifespans who buy these. Of course, the insurance companies know this very well and this is priced in. So if your prospects are for an average to less than average lifespan, you would want to think hard about buying an SPIA as the odds are not in your favor. Still, even in this case purchasing an SPIA can be a rational choice even if only to achieve a certain peace of mind.


I think it’s important to realize a COLA is just trading a lower initial payout for a larger later payout. And it’s totally different than a cpi adjusted annuity.
My only quibble is that I think a COLA adjusted annuity is actuarially neutral. I don’t know that a CPI adjusted annuity actually would be since we don’t know the future inflation rate

But my intuition might be incorrect about that
“… the fact remains that buying a nominal bond ladder to defease future living expenses can prove disastrous.” - Bill Bernstein | | “…something unusual happens—usually.” - Nassim Taleb
User avatar
nedsaid
Posts: 16938
Joined: Fri Nov 23, 2012 12:33 pm

Re: The 4% Rule Just Became a Whole Lot Easier - Allan Roth

Post by nedsaid »

Allan Roth wrote: Thu Nov 17, 2022 3:09 pm
vineviz wrote: Thu Nov 17, 2022 2:55 pm
dcabler wrote: Thu Nov 17, 2022 8:06 am I looked at this several months ago for the current choices of fixed COLA adjustments available with SPIAs today and it was still around 14 years or so. The surprising thing to me (and, yeah, this is from my memory) was that the breakeven period seemed to be largely independent on the amount of COLA chosen. And with a higher COLA resulting in lower initial payouts, yet another thing to think about when considering one's range of remaining life expectancy.
The fixed COLA is set to be actuarially neutral, which means the "breakeven point" (I hate that phrase) will always be near the life expectancy value used by the insurance company for its annuitant pool.
I think Vineviz deserves the medal of honor for financial bravery. I'm too chicken to buy a long term nominal bond (yielding less than intermediate-term bonds) much less a SPIA which is an indirect bond with a duration for the rest of my life that also carries credit risk. The only thing riskier I can think of is a SPIA with a COLA since that increases the duration (larger payments later) and thus the inflation risk.
Hi Allan, one reason I got a kick out of this thread is seeing a Nuerosurgeon telling an Economist that he doesn't know what he is talking about and watching a Financial Advisor telling another Financial Advisor that he doesn't know what he is talking about either. Vince has an MBA and a CFA and that isn't shabby. You, Bill Bernstein, Bob Kaminsky, and Vince Vizachero all have had good careers and credentials as well as the respect of this forum. Having both a CPA and a CFP designation speaks well of you, I know how difficult those credentials are to get and what they mean.

It just shows how human we all are, that we all have our opinions, and that we all have our unique life experiences and perspective. The exchanges have been fun as well as very informative.

I will say here what I said in response to Bill Bernstein. This is a vastly imperfect world with no perfect solutions, often it is just trade-offs. You are correct to point out there are risks with annuities but still perfectly rational people buy them for their own reasons and to fit their own life situations. Higher rates of inflation that we are seeing now have thrown a monkey wrench into the assumptions many of us held, and many are having to rethink some things. But it doesn't mean that choices that others make are reckless or irrational.
A fool and his money are good for business.
billaster
Posts: 814
Joined: Wed Apr 27, 2022 2:21 pm

Re: The 4% Rule Just Became a Whole Lot Easier - Allan Roth

Post by billaster »

I wish people would stop using COLA to describe a fixed rate of increase for a SPIA. COLA is a cost of living adjustment and is indexed to the change in cost of living. The fixed increase annuity does not do that. There is no link to cost of living. It is simply an agreement to take less money now in exchange for more money later.

And in that sense, the fixed increase annuity is like standard annuity mortality credits on steroids. The person surviving less than average receives even less money with this type of annuity which is a bonus transferred to those surviving longer. That's why you often see prices on these fixed increase annuities that are slightly better than standard annuities. You are leveraging your longevity bet.
User avatar
nedsaid
Posts: 16938
Joined: Fri Nov 23, 2012 12:33 pm

Re: The 4% Rule Just Became a Whole Lot Easier - Allan Roth

Post by nedsaid »

ScubaHogg wrote: Sun Nov 20, 2022 12:45 pm
nedsaid wrote: Sun Nov 20, 2022 11:49 am
ScubaHogg wrote: Thu Nov 17, 2022 10:51 am
dcabler wrote: Thu Nov 17, 2022 8:06 am
nedsaid wrote: Wed Nov 16, 2022 8:17 pm

No. What I said was not at odds with Dr. Wade Pfau. He actually did favor the nominal SPIAs over the CPI adjusted SPIAs because the breakeven point took so long. From memory, it was 12-15 years. There was a thread on this topic some time ago, in any case the CPI adjusted Single Premium Annuities are no longer available. Things have changed, including inflation and interest rates, I don't know his most very recent views.
I looked at this several months ago for the current choices of fixed COLA adjustments available with SPIAs today and it was still around 14 years or so. The surprising thing to me (and, yeah, this is from my memory) was that the breakeven period seemed to be largely independent on the amount of COLA chosen. And with a higher COLA resulting in lower initial payouts, yet another thing to think about when considering one's range of remaining life expectancy.

Cheers.
It’s probably not surprising since they are looking to pay roughly the same amount for a given cohort. I’m guessing the break even is roughly the life expectancy of the cohort.

Nedsaid: Ding! Ding! Ding! Ding! Ding! Ding! Ding! For an average lifespan, a CPI adjusted annuity and a nominal annuity are actuarily neutral. This is the way it has to be. An annuity is a bet with the insurance company that you will live a longer than average lifespan, the actuaries have it all figured out so that the insurance company wins over the many annuitants that it insures.

A big issue here is that people who buy Single Premium Immediate Annuities know this, so you get the healthiest folks with the best prospects of the longer lifespans who buy these. Of course, the insurance companies know this very well and this is priced in. So if your prospects are for an average to less than average lifespan, you would want to think hard about buying an SPIA as the odds are not in your favor. Still, even in this case purchasing an SPIA can be a rational choice even if only to achieve a certain peace of mind.


I think it’s important to realize a COLA is just trading a lower initial payout for a larger later payout. And it’s totally different than a cpi adjusted annuity.
My only quibble is that I think a COLA adjusted annuity is actuarially neutral. I don’t know that a CPI adjusted annuity actually would be since we don’t know the future inflation rate

But my intuition might be incorrect about that
It took me a while to understand how to guarantee an inflation adjusted income with duration matching with TIPS funds. Taking into account the variable pricing of TIPS, from positive real yields to negative back to positive, is something that I haven't thought through yet. How do you guarantee inflation adjusted income when TIPS are priced at negative yields? The answer is that you can't, you throw up your hands and reason that some inflation protection is better than none.

To your point, a COLA annuity has to be actuarially neutral to a nominal annuity in order to stay in business. It might be that insurance companies were finding this hard to do in practice and this might be one reason they stopped offering these. I don't know, I am not an actuary but we can have our suspicions. I think what killed off COLA adjusted annuities is that customers sacrificed too much up front income compared to a nominal annuity and thus did not want to buy them. A great product that very few wanted to buy.
A fool and his money are good for business.
User avatar
nedsaid
Posts: 16938
Joined: Fri Nov 23, 2012 12:33 pm

Re: The 4% Rule Just Became a Whole Lot Easier - Allan Roth

Post by nedsaid »

billaster wrote: Sun Nov 20, 2022 1:09 pm I wish people would stop using COLA to describe a fixed rate of increase for a SPIA. COLA is a cost of living adjustment and is indexed to the change in cost of living. The fixed increase annuity does not do that. There is no link to cost of living. It is simply an agreement to take less money now in exchange for more money later.

And in that sense, the fixed increase annuity is like standard annuity mortality credits on steroids. The person surviving less than average receives even less money with this type of annuity which is a bonus transferred to those surviving longer. That's why you often see prices on these fixed increase annuities that are slightly better than standard annuities. You are leveraging your longevity bet.
Not sure there is confusion here. It is understood that insurance companies no longer offer COLA adjusted SPIA's but you can buy annual fixed increases anywhere from 1% to 5%. Folks here have been careful to make the distinction.
A fool and his money are good for business.
StillGoing
Posts: 98
Joined: Mon Nov 04, 2019 4:43 am
Location: U.K.

Re: The 4% Rule Just Became a Whole Lot Easier - Allan Roth

Post by StillGoing »

billaster wrote: Sun Nov 20, 2022 1:09 pm I wish people would stop using COLA to describe a fixed rate of increase for a SPIA. COLA is a cost of living adjustment and is indexed to the change in cost of living. The fixed increase annuity does not do that. There is no link to cost of living. It is simply an agreement to take less money now in exchange for more money later.
In the UK, the fixed increases are commonly known as 'escalations' and the inflation-linked ones as 'RPI' which (at least to me) is a bit less ambiguous.

cheers
StillGoing
protagonist
Posts: 7989
Joined: Sun Dec 26, 2010 12:47 pm

Re: The 4% Rule Just Became a Whole Lot Easier - Allan Roth

Post by protagonist »

Bill Bernstein wrote: Tue Nov 15, 2022 6:48 pm
I'd hope people in this thread are about facts, and the fact is that in the worst historical period for inflation a retiree would have been better off with a nominal annuity than with nominal bonds.
I don't disagree at all with the above statement, for the simple reason that long nominal bonds are just as risky in terms of inflation risk as annuities, and in addition don't carry the longevity feature of annuities. To repeat, the comparison isn't annuities versus long bonds; it's annuities versus a TIPS ladder.

I don't consider it scaremongering to point out that for four decades after 1940, long bonds lost more than 60% of total return, and an annuity bought back then would have done nearly as poorly in terms of real consumption, since a $1,000 annuity payout in, say, 1951 bought $290 of real consumption by the end of 1980.

Perhaps inflation will indeed be muted over the next 30 years, but I don't want to unnecessarily bet my retirement that it will be when I can avoid that risk by substituting a TIPS ladder for a nominal annuity, let alone for long nominal bonds

Bill
+1
I think the point Bill was making (correct me if I am wrong, Bill), is that if you have "won the game" (defined by having enough saved to fund your retirement), any investment that is not inflation-protected using the money that you need for your retirement carries unnecessary risk. There is no statistically valid way to even estimate the amount of risk involved. That is a critical point. Past performance data is false security. There are no guarantees that future inflation will remain within a specific limit and no way to determine the probability that it will. So why would you want to put the money that you need for your retirement at risk unnecessarily when you could almost certainly eliminate that risk with a TIPS ladder and live out the rest of your days without financial worry?
Last edited by protagonist on Sun Nov 20, 2022 2:16 pm, edited 2 times in total.
billaster
Posts: 814
Joined: Wed Apr 27, 2022 2:21 pm

Re: The 4% Rule Just Became a Whole Lot Easier - Allan Roth

Post by billaster »

nedsaid wrote: Sun Nov 20, 2022 1:34 pm Not sure there is confusion here. It is understood that insurance companies no longer offer COLA adjusted SPIA's but you can buy annual fixed increases anywhere from 1% to 5%. Folks here have been careful to make the distinction.
Obviously there is some confusion in this thread because some are using CPI adjusted annuity to mean inflation adjusted which they distinguish from COLA annuity to mean fixed increases.
StillGoing wrote: Sun Nov 20, 2022 1:53 pm In the UK, the fixed increases are commonly known as 'escalations' and the inflation-linked ones as 'RPI' which (at least to me) is a bit less ambiguous.
Yes, I think escalated annuity is more accurate than "COLA" for fixed increases.
User avatar
bobcat2
Posts: 5820
Joined: Tue Feb 20, 2007 3:27 pm
Location: just barely Outside the Beltway

Re: The 4% Rule Just Became a Whole Lot Easier - Allan Roth

Post by bobcat2 »

billaster wrote: Sun Nov 20, 2022 2:05 pmYes, I think escalated annuity is more accurate than "COLA" for fixed increases.
I believe graded life annuity is accurate for life annuities that have fixed increases. Cost of living adjustment (COLA) is inaccurate. Social Security is an example of a COLA adjusted life annuity.

BobK
In finance risk is defined as uncertainty that is consequential (nontrivial). | The two main methods of dealing with financial risk are the matching of assets to goals & diversifying.
User avatar
vineviz
Posts: 14456
Joined: Tue May 15, 2018 1:55 pm
Location: Baltimore, MD

Re: The 4% Rule Just Became a Whole Lot Easier - Allan Roth

Post by vineviz »

billaster wrote: Sun Nov 20, 2022 2:05 pm
nedsaid wrote: Sun Nov 20, 2022 1:34 pm Not sure there is confusion here. It is understood that insurance companies no longer offer COLA adjusted SPIA's but you can buy annual fixed increases anywhere from 1% to 5%. Folks here have been careful to make the distinction.
Obviously there is some confusion in this thread because some are using CPI adjusted annuity to mean inflation adjusted which they distinguish from COLA annuity to mean fixed increases.
StillGoing wrote: Sun Nov 20, 2022 1:53 pm In the UK, the fixed increases are commonly known as 'escalations' and the inflation-linked ones as 'RPI' which (at least to me) is a bit less ambiguous.
Yes, I think escalated annuity is more accurate than "COLA" for fixed increases.
In the US, at least, the term "cost of living adjustment (COLA)" can refer to any form of payment escalation whether indexed directly to prices (e.g. to CPI) in part or in full, discretionary (as with many defined benefit plans), or fixed (e.g. a constant 2% or 3%). It is standard usage to refer to ALL of these as a COLA.

If the payment increase isn't actually indexed to inflation, the COLA should not be called an "inflation adjustment" IMHO but it is still a COLA..
"Far more money has been lost by investors preparing for corrections than has been lost in corrections themselves." ~~ Peter Lynch
billaster
Posts: 814
Joined: Wed Apr 27, 2022 2:21 pm

Re: The 4% Rule Just Became a Whole Lot Easier - Allan Roth

Post by billaster »

vineviz wrote: Sun Nov 20, 2022 2:38 pm In the US, at least, the term "cost of living adjustment (COLA)" can refer to any form of payment escalation whether indexed directly to prices (e.g. to CPI) in part or in full, discretionary (as with many defined benefit plans), or fixed (e.g. a constant 2% or 3%). It is standard usage to refer to ALL of these as a COLA.
I don't think that is true. At least as commonly understood. Since at least the 1950s a COLA in a union contract has meant a CPI adjustment. Fixed increases in labor contracts are not called COLAs. The same is true for Social Security.

The Bureau of Labor statistics defines a COLA as "contractual provisions in collective bargaining that automatically tie wage changes to changes in the Consumer Price Index." A COLA always implies increases tied to an index.

Another example is a COLA for child support. It ties increases in required child support payment to a price index.

The same is true in contract law. Service contracts often include a COLA clause tying multi-year service charges to a price index.

But I'm sure you can find some exceptions that are just as misleading as using COLA for an annuity with fixed increases. This misleading terminology should be discouraged
User avatar
vineviz
Posts: 14456
Joined: Tue May 15, 2018 1:55 pm
Location: Baltimore, MD

Re: The 4% Rule Just Became a Whole Lot Easier - Allan Roth

Post by vineviz »

billaster wrote: Sun Nov 20, 2022 4:12 pm
vineviz wrote: Sun Nov 20, 2022 2:38 pm In the US, at least, the term "cost of living adjustment (COLA)" can refer to any form of payment escalation whether indexed directly to prices (e.g. to CPI) in part or in full, discretionary (as with many defined benefit plans), or fixed (e.g. a constant 2% or 3%). It is standard usage to refer to ALL of these as a COLA.
I don't think that is true. At least as commonly understood. Since at least the 1950s a COLA in a union contract has meant a CPI adjustment. Fixed increases in labor contracts are not called COLAs. The same is true for Social Security.

The Bureau of Labor statistics defines a COLA as "contractual provisions in collective bargaining that automatically tie wage changes to changes in the Consumer Price Index." A COLA always implies increases tied to an index.

Another example is a COLA for child support. It ties increases in required child support payment to a price index.

The same is true in contract law. Service contracts often include a COLA clause tying multi-year service charges to a price index.

But I'm sure you can find some exceptions that are just as misleading as using COLA for an annuity with fixed increases. This misleading terminology should be discouraged
As I said, the term "COLA" can refer to any type of payment escalation. That certainly includes inflation-linked COLAs, but it includes other types as well.

Every insurance company I know uses the umbrella term "Cost-Of-Living Rider" to refer to the attachment to the income annuity which specifies the amount of annual increase. They use this term in their product descriptions, their state regulators refer to this term on their websites and registration exams, etc.

Earlier this year the Ohio STRS voted to approve a 3% cost-of-living adjustment after its actuary decided they could afford more than the 2% originally proposed. I can think of several other state pension boards that likewise have to set the COLA (if any) by decision, not by automatic indexing.
"Far more money has been lost by investors preparing for corrections than has been lost in corrections themselves." ~~ Peter Lynch
GAAP
Posts: 1801
Joined: Fri Apr 08, 2016 12:41 pm

Re: The 4% Rule Just Became a Whole Lot Easier - Allan Roth

Post by GAAP »

vineviz wrote: Sun Nov 20, 2022 4:34 pm As I said, the term "COLA" can refer to any type of payment escalation. That certainly includes inflation-linked COLAs, but it includes other types as well.

Every insurance company I know uses the umbrella term "Cost-Of-Living Rider" to refer to the attachment to the income annuity which specifies the amount of annual increase. They use this term in their product descriptions, their state regulators refer to this term on their websites and registration exams, etc.

Earlier this year the Ohio STRS voted to approve a 3% cost-of-living adjustment after its actuary decided they could afford more than the 2% originally proposed. I can think of several other state pension boards that likewise have to set the COLA (if any) by decision, not by automatic indexing.
And many union contracts have a fixed COLA that is not indexed to inflation.

COLA is a loose, lousy name for "some sort of adjustment that may or may not match, keep up with, or exceed inflation".

Theory and practice, etc.
“Adapt what is useful, reject what is useless, and add what is specifically your own.” ― Bruce Lee
billaster
Posts: 814
Joined: Wed Apr 27, 2022 2:21 pm

Re: The 4% Rule Just Became a Whole Lot Easier - Allan Roth

Post by billaster »

You are talking about a one-time or special increase due to cost of living. They look at the CPI and make an adjustment based on cost of living. That is different from referring to annual fixed increases as COLAs that have absolutely nothing whatsoever to do with cost of living. It's misleading terminology.

I certainly wouldn't use the questionable practices of the insurance industry as my model for clarity.
User avatar
vineviz
Posts: 14456
Joined: Tue May 15, 2018 1:55 pm
Location: Baltimore, MD

Re: The 4% Rule Just Became a Whole Lot Easier - Allan Roth

Post by vineviz »

billaster wrote: Sun Nov 20, 2022 5:12 pm You are talking about a one-time or special increase due to cost of living. They look at the CPI and make an adjustment based on cost of living. That is different from referring to annual fixed increases as COLAs that have absolutely nothing whatsoever to do with cost of living. It's misleading terminology.

I certainly wouldn't use the questionable practices of the insurance industry as my model for clarity.
I'm merely pointing out that the term "COLA" has a widely accepted meaning which does not REQUIRE the adjustment to be directly indexed to inflation.

No one is forcing you to use any words you don't want to use, but I don't think it's productive to suggest that an industry stop using a term merely because some people who think they know what means actually don't.
"Far more money has been lost by investors preparing for corrections than has been lost in corrections themselves." ~~ Peter Lynch
billaster
Posts: 814
Joined: Wed Apr 27, 2022 2:21 pm

Re: The 4% Rule Just Became a Whole Lot Easier - Allan Roth

Post by billaster »

vineviz wrote: Sun Nov 20, 2022 5:32 pm
billaster wrote: Sun Nov 20, 2022 5:12 pm You are talking about a one-time or special increase due to cost of living. They look at the CPI and make an adjustment based on cost of living. That is different from referring to annual fixed increases as COLAs that have absolutely nothing whatsoever to do with cost of living. It's misleading terminology.

I certainly wouldn't use the questionable practices of the insurance industry as my model for clarity.
I'm merely pointing out that the term "COLA" has a widely accepted meaning which does not REQUIRE the adjustment to be directly indexed to inflation.

No one is forcing you to use any words you don't want to use, but I don't think it's productive to suggest that an industry stop using a term merely because some people who think they know what means actually don't.
If widely accepted means the insurance industry, yeah, I'll take a pass. They are explicitly the ones who try to mislead people with questionable terminology.
User avatar
Prokofiev
Posts: 1196
Joined: Mon Feb 19, 2007 9:45 pm
Location: New Orleans

Re: The 4% Rule Just Became a Whole Lot Easier - Allan Roth

Post by Prokofiev »

vineviz wrote: Sun Nov 20, 2022 12:01 pm
Prokofiev wrote: Sun Nov 20, 2022 11:38 am
I don't think this is correct. Getting a graded 2, 3 or 5% SPIA will mean the break-even age MUST be higher than a 0% SPIA. You are effectively
moving payments from the present into the fu}ture and your starting point is much lower with the 5% annuity. They cannot both give you the same
dollars up to "average lifespan" since the 5% annuity would then be so much more valuable for the 50% of the pool that lives beyond that age.
It is correct that the amount of fixed increase ( whether 0% or more) is actuarially neutral.

It’s true that IF you could be SURE when you bought the annuity that’s you’d live longer than the insurance company expects you to, the 5% annuity would be very valuable to you.

But none of us know that, and the insurance companies know that none of us know that.
Well, I hate to press the point. but I believe it depends on what you think "actuarially neutral" means. Let me start by saying I have no deep knowledge of the insurance or annuity business. I did risk analysis for a major oil firm, and we did not need to use mortality functions. This is more of a math/statistics approach.

Consider two groups of annuitants. All are exactly 70yo, male and want to buy a SPIA. To make matters easier, we live in a zero-discount rate world where $1 today equals $1 30 years from now. No time value of money. Group A wants a simple 0% SPIA, while Group B prefers a 5% graded
life annuity. The insurance company needs to make a profit. Say 2% to the salesman, 2% to write checks for 30+ years and 6% net profit. So, lets
say 10% (I have no idea of the correct number). Each group buys $100 million in contracts and the insurance company wants to pay-out exactly
$90 million to each group over the next 50 years at which time everyone is gone.

There is a mortality distribution function known to the company and assume it is perfect. Given this function and the 10% profitability goal, can we calculate exactly how much each person in Group A should receive each year or month? Sure. Easy. Let's say everyone gets 6.67% a year of their initial premium (completely made-up number). They all can expect to break-even in exactly 15 years. For Group B using the same mortality function (more on this later) can they calculate what each member should receive to make their break-even exactly 15 years? Yup. 4.63%. But the problem now is that the insurance company will NOT pay-out exactly $90 million for their 10% profitability. It will probably be more. So can they
calculate what the initial pay-out should be to make 10% profit? Sure. But in general, these numbers are not the same. For a given mortality distribution function and 2 different pay-out structures, you cannot match BOTH the total payout (profitability) and the break-even age with a single initial pay-out number.

So here is my question. Does being actuarially fair or neutral mean both groups get the same total pay-out or the same break-even age? I would vote for the insurance company making exactly the same profit for both groups. That is their perspective. But many reading this thread think it
means "if I live exactly my expected lifespan, then all SPIA annuities will be exactly the same, be it 0%,1%,2%,5% or even true CPI". I don't believe that is the case. I haven't checked in quite a while, but I believe that adding a larger increase will also increase your break-even age. Now it may be a rather small increase, but that has more to do with the mortality function shape than any need for them to be equal. The difference between an age 85 or 86 break-even seems rather small, barely 1%. But the expected lifespan of an 85 yo is about 6 years. So that is a big % of your remaining life. A small increase of only a few months can mean a large difference in profitability to the policy writer. Maybe this is a very small point. Kind of like saying SS is actuarially neutral. For men or for women? For a single person or married? For a 60 yo or a 70 yo. It may be a small difference, but they cannot ALL be fair.

I would not be surprised that companies use a different mortality table for different pay-out escalators. Does a 70 yo with a 5% annuity live a little longer than those with 2%? And 2% live longer that 0% who surely will live longer than the general public. Is that slicing things too thin? If I ran the company, I would certainly look at that and I'm pretty sure they do. Perhaps there is someone doing this work who could chime it. I am too lazy to ask for different graded quotes and I don't want to send them my personal info. The calculators seem to only provide a simple 0% answer.

My other point is that it is more than semantics to say COLA doesn't equal CPI. We can agree that a SPIA with an escalator of x% has nothing directly to do with inflation or your personal cost of living. It just changes the pay-out structure of the annuity for the annuitant. The insurance company is accepting no risk nor changing their expected profit.
Everything should be made as simple as possible, but not simpler - Einstein
User avatar
nedsaid
Posts: 16938
Joined: Fri Nov 23, 2012 12:33 pm

Re: The 4% Rule Just Became a Whole Lot Easier - Allan Roth

Post by nedsaid »

billaster wrote: Sun Nov 20, 2022 6:10 pm
vineviz wrote: Sun Nov 20, 2022 5:32 pm
billaster wrote: Sun Nov 20, 2022 5:12 pm You are talking about a one-time or special increase due to cost of living. They look at the CPI and make an adjustment based on cost of living. That is different from referring to annual fixed increases as COLAs that have absolutely nothing whatsoever to do with cost of living. It's misleading terminology.

I certainly wouldn't use the questionable practices of the insurance industry as my model for clarity.
I'm merely pointing out that the term "COLA" has a widely accepted meaning which does not REQUIRE the adjustment to be directly indexed to inflation.

No one is forcing you to use any words you don't want to use, but I don't think it's productive to suggest that an industry stop using a term merely because some people who think they know what means actually don't.
If widely accepted means the insurance industry, yeah, I'll take a pass. They are explicitly the ones who try to mislead people with questionable terminology.
Billaster, you did a good service by raising this point. It is important that language be precise and commonly understood. Sloppy definitions and sloppy phrasing can muddy a discussion and we can wind up talking past each other because one side doesn't fully understand what the other side means.

My understanding is that a COLA refers to an adjustment tied to an inflation index like the CPI. Yes, a looser definition of a COLA could mean an annual 2% adjustment to a pension, like what the Public Retirement system does in my state. But as said above, imprecise definitions does muddy discussions. Words do need to mean actual things.
A fool and his money are good for business.
User avatar
vineviz
Posts: 14456
Joined: Tue May 15, 2018 1:55 pm
Location: Baltimore, MD

Re: The 4% Rule Just Became a Whole Lot Easier - Allan Roth

Post by vineviz »

billaster wrote: Sun Nov 20, 2022 6:10 pm If widely accepted means the insurance industry, yeah, I'll take a pass. They are explicitly the ones who try to mislead people with questionable terminology.
"Widely accepted" means that the definition I gave you is the standard one used in all areas of finance and economics.

Perhaps you were merely confused before, but now you know.
"Far more money has been lost by investors preparing for corrections than has been lost in corrections themselves." ~~ Peter Lynch
User avatar
vineviz
Posts: 14456
Joined: Tue May 15, 2018 1:55 pm
Location: Baltimore, MD

Re: The 4% Rule Just Became a Whole Lot Easier - Allan Roth

Post by vineviz »

Prokofiev wrote: Sun Nov 20, 2022 6:33 pm

So here is my question. Does being actuarially fair or neutral mean both groups get the same total pay-out or the same break-even age? I would vote for the insurance company making exactly the same profit for both groups. That is their perspective. But many reading this thread think it
means "if I live exactly my expected lifespan, then all SPIA annuities will be exactly the same, be it 0%,1%,2%,5% or even true CPI".
No CPI-linked annuities are currently available for purchase, and because inflation can't be known in advance it's complete neutrality can't be known a priori.

Apart from that quibble, though, the two interpretations you provide are synonymous. And both are true.
Last edited by vineviz on Sun Nov 20, 2022 7:52 pm, edited 1 time in total.
"Far more money has been lost by investors preparing for corrections than has been lost in corrections themselves." ~~ Peter Lynch
User avatar
vineviz
Posts: 14456
Joined: Tue May 15, 2018 1:55 pm
Location: Baltimore, MD

Re: The 4% Rule Just Became a Whole Lot Easier - Allan Roth

Post by vineviz »

nedsaid wrote: Sun Nov 20, 2022 7:10 pm My understanding is that a COLA refers to an adjustment tied to an inflation index like the CPI.
Apparently you are not the only one who assumed that this was the only interpretation, but it has always had a broader meaning.
"Far more money has been lost by investors preparing for corrections than has been lost in corrections themselves." ~~ Peter Lynch
User avatar
nedsaid
Posts: 16938
Joined: Fri Nov 23, 2012 12:33 pm

Re: The 4% Rule Just Became a Whole Lot Easier - Allan Roth

Post by nedsaid »

vineviz wrote: Sun Nov 20, 2022 7:52 pm
nedsaid wrote: Sun Nov 20, 2022 7:10 pm My understanding is that a COLA refers to an adjustment tied to an inflation index like the CPI.
Apparently you are not the only one who assumed that this was the only interpretation, but it has always had a broader meaning.
Its okay. Sometimes words have both a specific and a broader meaning and this is a good example. I think from now on we will have to be specific how the COLA adjustment is calculated, tied to an inflation index like the CPI or a fixed amount like the 2% each year for PERS pensions in my state. I am not a grammar, spelling, or dictionary fanatic. Just saying we need to be clear on what we are saying. Billaster brought up a point, things got clarified, and I am satisfied with that.
A fool and his money are good for business.
User avatar
bobcat2
Posts: 5820
Joined: Tue Feb 20, 2007 3:27 pm
Location: just barely Outside the Beltway

Re: The 4% Rule Just Became a Whole Lot Easier - Allan Roth

Post by bobcat2 »

To avoid confusion I suggest the following.

Graded life annuities should be called - graded life annuities.

CPI inflation adjusted life annuities should be called - CPI inflation adjusted life annuities.

In this way we can put the COLA kerfuffle to bed. :wink:

BobK
In finance risk is defined as uncertainty that is consequential (nontrivial). | The two main methods of dealing with financial risk are the matching of assets to goals & diversifying.
billaster
Posts: 814
Joined: Wed Apr 27, 2022 2:21 pm

Re: The 4% Rule Just Became a Whole Lot Easier - Allan Roth

Post by billaster »

vineviz wrote: Sun Nov 20, 2022 7:46 pm
billaster wrote: Sun Nov 20, 2022 6:10 pm If widely accepted means the insurance industry, yeah, I'll take a pass. They are explicitly the ones who try to mislead people with questionable terminology.
"Widely accepted" means that the definition I gave you is the standard one used in all areas of finance and economics.
That simply is not true, as much as you want to believe it so.
StillGoing
Posts: 98
Joined: Mon Nov 04, 2019 4:43 am
Location: U.K.

Re: The 4% Rule Just Became a Whole Lot Easier - Allan Roth

Post by StillGoing »

vineviz wrote: Sun Nov 20, 2022 7:50 pm
Prokofiev wrote: Sun Nov 20, 2022 6:33 pm

So here is my question. Does being actuarially fair or neutral mean both groups get the same total pay-out or the same break-even age? I would vote for the insurance company making exactly the same profit for both groups. That is their perspective. But many reading this thread think it
means "if I live exactly my expected lifespan, then all SPIA annuities will be exactly the same, be it 0%,1%,2%,5% or even true CPI".
No CPI-linked annuities are currently available for purchase, and because inflation can't be known in advance it's complete neutrality can't be known a priori.

Apart from that quibble, though, the two interpretations you provide are synonymous. And both are true.
You are correct that the neutrality cannot be known because future inflation is unknowable. But, it is interesting to note that (at least in the UK), RPI annuities are more expensive than the value of implied inflation would suggest.

For example, a single life annuity with 5 year guarantee (period certain) for a 65 year old is currently (https://www.hl.co.uk/retirement/annuiti ... -buy-rates) 7.3% for level, 5.1% for a 3% escalation, and 4.1% for the RPI-linked version. Very roughly, an escalation reduces the annuity rate by percentage points equivalent to somewhere between half and two thirds of the escalation (so approximately 2 percentage points for a 3% escalation). On the same basis, the 3.2 percentage point difference in the rate between the level and RPI annuities would suggest an implied inflation of about 4.8%. The current spot rate of implied inflation for the UK (see https://www.bankofengland.co.uk/statistics/yield-curves) is around 3.5% so is less than that.

If future inflation turns out to be higher than implied inflation, the RPI annuity will have paid out more than the equivalent nominal annuity and vice versa. It will have been actuarially neutral if, and only if, the actual inflation turns out to have been the same as the implied inflation at the time of purchase. I also note that an extended period of deflation would be unpleasant for the holders of RPI linked annuities.

So why are the RPI annuities more expensive than might be expected? My guess is
1) There are few commercial inflation linked bonds, so the insurance company is more reliant on government gilts (unlike nominal annuities which will be supported by both gilts and AAA rated commercial bonds with a higher yield).
2) There are fewer RPI annuities sold (although recent inflation may change that), and therefore the mortality rates are a bit less certain and the insurance companies may be a bit more conservative and, I'm not sure whether all UK insurance companies offer them (hence price competition is less).
3) RPI gilts, and hence annuities, are likely to be moving to CPI in 2030 or so.

cheers
StillGoing
User avatar
vineviz
Posts: 14456
Joined: Tue May 15, 2018 1:55 pm
Location: Baltimore, MD

Re: The 4% Rule Just Became a Whole Lot Easier - Allan Roth

Post by vineviz »

billaster wrote: Sun Nov 20, 2022 9:08 pm
vineviz wrote: Sun Nov 20, 2022 7:46 pm
billaster wrote: Sun Nov 20, 2022 6:10 pm If widely accepted means the insurance industry, yeah, I'll take a pass. They are explicitly the ones who try to mislead people with questionable terminology.
"Widely accepted" means that the definition I gave you is the standard one used in all areas of finance and economics.
That simply is not true, as much as you want to believe it so.
I've got enough years of experience and training in finance and economics to know what terms like this mean or don't mean. You've been provided with multiple examples of the usage of "COLA" which are consistent with the broader meaning I've supplied but inconsistent with the narrower meaning you've supplied.

You made an incorrect assumption about what the term meant. I'm sorry, but that's how it is.
"Far more money has been lost by investors preparing for corrections than has been lost in corrections themselves." ~~ Peter Lynch
User avatar
vineviz
Posts: 14456
Joined: Tue May 15, 2018 1:55 pm
Location: Baltimore, MD

Re: The 4% Rule Just Became a Whole Lot Easier - Allan Roth

Post by vineviz »

bobcat2 wrote: Sun Nov 20, 2022 9:00 pm To avoid confusion I suggest the following.

Graded life annuities should be called - graded life annuities.

CPI inflation adjusted life annuities should be called - CPI inflation adjusted life annuities.

In this way we can put the COLA kerfuffle to bed. :wink:

BobK
It'd be nice if it were that easy, but the term "graded" has its own set of various meanings in the life insurance industry that aren't always clearly applicable to the concept of an income annuity. I'm not sure it would be helpful.

Besides, with most contracts anyone who wants to add a fixed 2% or 3% increase to their income annuity payments is going to have to sign a form that is literally called a "Cost-of-Living Rider".

Despite whatever Quixotic objections some people seem to have to this term, it's usage isn't going to go away. In my experience most professionals will use a term like the one you suggested (e.g. "inflation-linked" or "inflation-indexed" or "CPI-indexed") when that's what they mean.
"Far more money has been lost by investors preparing for corrections than has been lost in corrections themselves." ~~ Peter Lynch
GAAP
Posts: 1801
Joined: Fri Apr 08, 2016 12:41 pm

Re: The 4% Rule Just Became a Whole Lot Easier - Allan Roth

Post by GAAP »

The COLA term is an acronym for Cost of Living Adjustment. How the adjustment is calculated is not specified --for that matter, neither is the definition of "Cost of Living".

Any method that provides some degree of extra income over the base qualifies as an adjustment. Even if you assume CPI for the metric, there are multiple versions, not to mention multiple regional differences. It is unlikely that any "official" definition of an inflation-based adjustment will actually match your experience with inflation. That difference between your income -- "adjusted" or not -- and your reality is what is important.

It would be nice if we can get past circular arguments about terms and return to discussion of concepts, strategies, etc.

One example -- should you or should you not build a TIPS ladder beyond age 85? What are the risks and benefits from such a strategy? Are there alternatives with different sets of risks and benefits?

Or, what about planning for an expected potential (uncertain) loss of funding in an inflation-linked defined benefit like Social Security, some pensions, or an annuity?

Surely we can find better things to talk about. :annoyed
“Adapt what is useful, reject what is useless, and add what is specifically your own.” ― Bruce Lee
User avatar
vineviz
Posts: 14456
Joined: Tue May 15, 2018 1:55 pm
Location: Baltimore, MD

Re: The 4% Rule Just Became a Whole Lot Easier - Allan Roth

Post by vineviz »

GAAP wrote: Mon Nov 21, 2022 1:08 pm
One example -- should you or should you not build a TIPS ladder beyond age 85? What are the risks and benefits from such a strategy? Are there alternatives with different sets of risks and benefits?
If a client (let's say a health couple) was setting up a TIPS ladder I can't imagine on what basis I'd suggest ending it at age 85.

There's a >50% chance that at least one of them will still be alive at age 90 and a >25% chance that one of them will still be alive at age 95.

It seems contraindicated to me to make a plan that involves making additional decisions and/or taking on more risk in a future period when it is likely to be unnecessary AND could very well be marked by some degree of cognitive decline.

Of course all such decisions depend on the preferences, resources, and constraints of the investor but in general I'd set the ladder up at least to age 90 and possibly age 95. Even if they both pass away earlier than that, having a handful of TIPS in a portfolio is not likely to be significant hardship for the estate and/or heirs. Certainly no more difficult than dealing with the alternative, which would mean selling an ETF/mutual fund or two.

It's not like we're building a portofio involving dozens or hundreds of individual municipal bonds or something.
"Far more money has been lost by investors preparing for corrections than has been lost in corrections themselves." ~~ Peter Lynch
retireIn2020
Posts: 320
Joined: Sat Jan 04, 2020 6:13 pm

Re: The 4% Rule Just Became a Whole Lot Easier - Allan Roth

Post by retireIn2020 »

Prokofiev wrote: Sun Nov 20, 2022 11:38 am
dcabler wrote: Thu Nov 17, 2022 3:35 pm
GaryA505 wrote: Thu Nov 17, 2022 2:58 pm
vineviz wrote: Thu Nov 17, 2022 2:55 pm
dcabler wrote: Thu Nov 17, 2022 8:06 am I looked at this several months ago for the current choices of fixed COLA adjustments available with SPIAs today and it was still around 14 years or so. The surprising thing to me (and, yeah, this is from my memory) was that the breakeven period seemed to be largely independent on the amount of COLA chosen. And with a higher COLA resulting in lower initial payouts, yet another thing to think about when considering one's range of remaining life expectancy.
The fixed COLA is set to be actuarially neutral, which means the "breakeven point" (I hate that phrase) will always be near the life expectancy value used by the insurance company for its annuitant pool.
So, assuming you have a roughly average lifespan, it doesn't really matter much which one you use. It just depends on whether you would like to have bigger payments earlier or later.
Yep, that's what I concluded after looking at it a few months back.
I don't think this is correct. Getting a graded 2, 3 or 5% SPIA will mean the break-even age MUST be higher than a 0% SPIA. You are effectively
moving payments from the present into the future and your starting point is much lower with the 5% annuity. They cannot both give you the same
dollars up to "average lifespan" since the 5% annuity would then be so much more valuable for the 50% of the pool that lives beyond that age.

There are 2 risks we are discussing. Longevity and Inflation. Today, any SPIA will only protect against the first. Life insurance companies
can handle longevity or mortality. It's what they do for a living, and they do it very well. A graded 5% SPIA is neither CPI protected or COLAed.
The insurance company doesn't care what the CPI is or what your "cost of living" is. They know in advance and with NO RISK what the payout
schedule will be. $1000/mo for first year. $1050/mo for second, $1103/mo for third, then $1158 etc. They can calculate PV and future value
of this cash flow and adjust it using their mortality tables to give the annuitant the same AVERAGE pay-out. You can think of the 5% increase
as a COLA or inflation adjustment, but the insurance company does not. They are just modifying the cash flows which they know in advance.

I would prefer to think of this as doubling-down on longevity. It will be that much more important to live a long life. Of course, in a way this certainly can compensate for cost of living. A 70-year-old who only lives 5 more years to 75 is not as exposed to inflation as someone who
lives to 100. But these increases are not connected to either CPI or a COLA. The companies could cover both these risks if they wanted and
until several years ago, did offer CPI adjusted SPIAs. Many have said that they were not popular, so they are no longer offered. But they were
not popular because most annuitants underestimate the value or cost of inflation protection. No one wanted to take a 15-20% or more haircut
on the first year's payout. Or move the break-even age up 2-3 years or even more.
I ran quotes for myself (almost 63) and the breakeven for 0 COLA adjustment is age 76, the breakeven for a 2% COLA is 77. The plus for the COLA SPIA is after age 72-73 the yearly income starts to aggressively pull away from the "no COLA".

Here are the quotes do what you will with the data!

Image
Image
Retired as of July 2020
StillGoing
Posts: 98
Joined: Mon Nov 04, 2019 4:43 am
Location: U.K.

Re: The 4% Rule Just Became a Whole Lot Easier - Allan Roth

Post by StillGoing »

vineviz wrote: Mon Nov 21, 2022 3:34 pm
GAAP wrote: Mon Nov 21, 2022 1:08 pm
One example -- should you or should you not build a TIPS ladder beyond age 85? What are the risks and benefits from such a strategy? Are there alternatives with different sets of risks and benefits?
If a client (let's say a health couple) was setting up a TIPS ladder I can't imagine on what basis I'd suggest ending it at age 85.

There's a >50% chance that at least one of them will still be alive at age 90 and a >25% chance that one of them will still be alive at age 95.

It seems contraindicated to me to make a plan that involves making additional decisions and/or taking on more risk in a future period when it is likely to be unnecessary AND could very well be marked by some degree of cognitive decline.

Of course all such decisions depend on the preferences, resources, and constraints of the investor but in general I'd set the ladder up at least to age 90 and possibly age 95. Even if they both pass away earlier than that, having a handful of TIPS in a portfolio is not likely to be significant hardship for the estate and/or heirs. Certainly no more difficult than dealing with the alternative, which would mean selling an ETF/mutual fund or two.

It's not like we're building a portofio involving dozens or hundreds of individual municipal bonds or something.
I posted the following graph which shows the number of years before the income falls below the target (the colour scale) as a function of the ladder duration and ladder spend (i.e., the proportion of the initial portfolio used to construct the ladder). upthread, but it may be worth revisiting...

Image

The graph assumes a target CPI linked income rate of 4% of the initial portfolio, such that if the income from the ladder exceeded the target, the excess was returned to the portfolio, while if the income from the ladder fell below the target, the shortfall was made up from the residual portfolio. A yield to maturity of 1.7% has been used for the bonds in the ladder and the residual portfolio consisted of 80% stocks and 20% nominal bonds, rebalanced annually.

Without a ladder, the income dropped below target (i.e. the portfolio was exhausted) after about 25 years. Ladder durations of just under 20 to 30 years and spends between 40% and 80% resulted in income failure after 35 years or more with longer ladders tending to require a greater spend to achieve the same longevity. The results will be different with a different yield to maturity or a different target income (obviously, reducing the target income increases the longevity of the income - even dropping to 3.8%, improves it by 4 years or more).

Fully agree with vineviz that an obvious problem with a shorter duration ladder is that a 20 year ladder would expire at 85 (for a 65 year old retiree), leaving them with a choice of continuing with portfolio withdrawals only (as assumed in the above graph), the construction of a further ladder (with the possibility that interest rates were much less than those at retirement), or annuitization of some or all of the remaining portfolio (with subsequent inflation risks for nominal annuities). However, even a 30 year ladder will expire, and as vineviz, says, for a couple that means there is a 25% probability that one member of the couple may have to deal with the consequences aged 95.

cheers
StillGoing
User avatar
Prokofiev
Posts: 1196
Joined: Mon Feb 19, 2007 9:45 pm
Location: New Orleans

Re: The 4% Rule Just Became a Whole Lot Easier - Allan Roth

Post by Prokofiev »

retireIn2020 wrote: Tue Nov 22, 2022 12:53 am
Prokofiev wrote: Sun Nov 20, 2022 11:38 am
dcabler wrote: Thu Nov 17, 2022 3:35 pm
GaryA505 wrote: Thu Nov 17, 2022 2:58 pm
vineviz wrote: Thu Nov 17, 2022 2:55 pm

The fixed COLA is set to be actuarially neutral, which means the "breakeven point" (I hate that phrase) will always be near the life expectancy value used by the insurance company for its annuitant pool.
So, assuming you have a roughly average lifespan, it doesn't really matter much which one you use. It just depends on whether you would like to have bigger payments earlier or later.
Yep, that's what I concluded after looking at it a few months back.
I don't think this is correct. Getting a graded 2, 3 or 5% SPIA will mean the break-even age MUST be higher than a 0% SPIA. You are effectively
moving payments from the present into the future and your starting point is much lower with the 5% annuity. They cannot both give you the same
dollars up to "average lifespan" since the 5% annuity would then be so much more valuable for the 50% of the pool that lives beyond that age.

There are 2 risks we are discussing. Longevity and Inflation. Today, any SPIA will only protect against the first. Life insurance companies
can handle longevity or mortality. It's what they do for a living, and they do it very well. A graded 5% SPIA is neither CPI protected or COLAed.
The insurance company doesn't care what the CPI is or what your "cost of living" is. They know in advance and with NO RISK what the payout
schedule will be. $1000/mo for first year. $1050/mo for second, $1103/mo for third, then $1158 etc. They can calculate PV and future value
of this cash flow and adjust it using their mortality tables to give the annuitant the same AVERAGE pay-out. You can think of the 5% increase
as a COLA or inflation adjustment, but the insurance company does not. They are just modifying the cash flows which they know in advance.

I would prefer to think of this as doubling-down on longevity. It will be that much more important to live a long life. Of course, in a way this certainly can compensate for cost of living. A 70-year-old who only lives 5 more years to 75 is not as exposed to inflation as someone who
lives to 100. But these increases are not connected to either CPI or a COLA. The companies could cover both these risks if they wanted and
until several years ago, did offer CPI adjusted SPIAs. Many have said that they were not popular, so they are no longer offered. But they were
not popular because most annuitants underestimate the value or cost of inflation protection. No one wanted to take a 15-20% or more haircut
on the first year's payout. Or move the break-even age up 2-3 years or even more.
I ran quotes for myself (almost 63) and the breakeven for 0 COLA adjustment is age 76, the breakeven for a 2% COLA is 77. The plus for the COLA SPIA is after age 72-73 the yearly income starts to aggressively pull away from the "no COLA".

Here are the quotes do what you will with the data!
Thanks for posting this. It is consistent with what I remember from quotes several years ago. Of course, these real-world numbers are
going to be different from my example since we are not in a 0% discount rate world. True break-even will push the 2% COLA annuity slightly higher still if one reinvests the larger early payments received from the 0% SPIA. But it illustrates the basic point.
Everything should be made as simple as possible, but not simpler - Einstein
GaryA505
Posts: 1741
Joined: Wed Feb 08, 2017 2:59 pm
Location: New Mexico

Re: The 4% Rule Just Became a Whole Lot Easier - Allan Roth

Post by GaryA505 »

On the SPIA COLA vs no-COLA question ...

It seems the usual advice advice is to only use a SPIA at higher ages (70s or 80s).

It seems that when looking at the break-even point, using the 2% COLA option might be advantageous for an early retiree who lives a long time. At least that's what my very crude Excel modeling seems to indicate.
Get most of it right and don't make any big mistakes. Other things being equal (or close enough), simpler is better.
GAAP
Posts: 1801
Joined: Fri Apr 08, 2016 12:41 pm

Re: The 4% Rule Just Became a Whole Lot Easier - Allan Roth

Post by GAAP »

StillGoing wrote: Tue Nov 22, 2022 4:41 amFully agree with vineviz that an obvious problem with a shorter duration ladder is that a 20 year ladder would expire at 85 (for a 65 year old retiree), leaving them with a choice of continuing with portfolio withdrawals only (as assumed in the above graph), the construction of a further ladder (with the possibility that interest rates were much less than those at retirement), or annuitization of some or all of the remaining portfolio (with subsequent inflation risks for nominal annuities). However, even a 30 year ladder will expire, and as vineviz, says, for a couple that means there is a 25% probability that one member of the couple may have to deal with the consequences aged 95.

cheers
StillGoing
We're really discussing how to handle fortuitous longevity that exceeds the original planned expectation. The 4% Rule is generally based around a 30-year expectation -- at some point, one can recognize that the original plan is not sufficient.

Successfully waking up each morning increases the likelihood that the interval will increase over the original plan -- the IRS RMD tables don't get down to 1-year life expectancy until age 120. Those same tables would allow someone to recognize a potential issue. For someone who retires at age 65 with a 30-year drawdown expectation, IRS table I recognizes a problem at age 89. Unfortunately, that is pretty late to do much.

Table III doesn't start until age 72, but recognizes a problem immediately. Building a ladder based upon Table III mortality, with the assumed need not starting until age 95 would provide 27 years advance notice, while only needing to cover about 4 years of expenses at first. This would require taking funds out of the portfolio -- but that is no different than buying an annuity. I would probably build this with ETFs for simplicity. The duration needed would transition from a long/intermediate mix to intermediate/short around age 89, and would drop to entirely short term at around age 106. An annual AA calculation plus a plan to withdraw in descending term order would be sufficient to manage duration changes.

You could actually start this process before retirement, using an assumption of the age 72 mortality to calculate duration. Since the upper end need would be age 99, starting as early as age 59 would be possible with currently available TIPS funds.
“Adapt what is useful, reject what is useless, and add what is specifically your own.” ― Bruce Lee
SteveinVanvcouverWA
Posts: 47
Joined: Sun Feb 17, 2008 12:26 pm

Re: The 4% Rule Just Became a Whole Lot Easier - Allan Roth

Post by SteveinVanvcouverWA »

vineviz wrote: Mon Nov 21, 2022 3:34 pm
GAAP wrote: Mon Nov 21, 2022 1:08 pm
One example -- should you or should you not build a TIPS ladder beyond age 85? What are the risks and benefits from such a strategy? Are there alternatives with different sets of risks and benefits?
If a client (let's say a health couple) was setting up a TIPS ladder I can't imagine on what basis I'd suggest ending it at age 85.

There's a >50% chance that at least one of them will still be alive at age 90 and a >25% chance that one of them will still be alive at age 95.

It seems contraindicated to me to make a plan that involves making additional decisions and/or taking on more risk in a future period when it is likely to be unnecessary AND could very well be marked by some degree of cognitive decline.

Of course all such decisions depend on the preferences, resources, and constraints of the investor but in general I'd set the ladder up at least to age 90 and possibly age 95. Even if they both pass away earlier than that, having a handful of TIPS in a portfolio is not likely to be significant hardship for the estate and/or heirs. Certainly no more difficult than dealing with the alternative, which would mean selling an ETF/mutual fund or two.

It's not like we're building a portofio involving dozens or hundreds of individual municipal bonds or something.
Very helpful, vineviz , thank you! :happy
StillGoing
Posts: 98
Joined: Mon Nov 04, 2019 4:43 am
Location: U.K.

Re: The 4% Rule Just Became a Whole Lot Easier - Allan Roth

Post by StillGoing »

GAAP wrote: Tue Nov 22, 2022 5:22 pm
StillGoing wrote: Tue Nov 22, 2022 4:41 amFully agree with vineviz that an obvious problem with a shorter duration ladder is that a 20 year ladder would expire at 85 (for a 65 year old retiree), leaving them with a choice of continuing with portfolio withdrawals only (as assumed in the above graph), the construction of a further ladder (with the possibility that interest rates were much less than those at retirement), or annuitization of some or all of the remaining portfolio (with subsequent inflation risks for nominal annuities). However, even a 30 year ladder will expire, and as vineviz, says, for a couple that means there is a 25% probability that one member of the couple may have to deal with the consequences aged 95.

cheers
StillGoing
We're really discussing how to handle fortuitous longevity that exceeds the original planned expectation. The 4% Rule is generally based around a 30-year expectation -- at some point, one can recognize that the original plan is not sufficient.

Successfully waking up each morning increases the likelihood that the interval will increase over the original plan -- the IRS RMD tables don't get down to 1-year life expectancy until age 120. Those same tables would allow someone to recognize a potential issue. For someone who retires at age 65 with a 30-year drawdown expectation, IRS table I recognizes a problem at age 89. Unfortunately, that is pretty late to do much.

Table III doesn't start until age 72, but recognizes a problem immediately. Building a ladder based upon Table III mortality, with the assumed need not starting until age 95 would provide 27 years advance notice, while only needing to cover about 4 years of expenses at first. This would require taking funds out of the portfolio -- but that is no different than buying an annuity. I would probably build this with ETFs for simplicity. The duration needed would transition from a long/intermediate mix to intermediate/short around age 89, and would drop to entirely short term at around age 106. An annual AA calculation plus a plan to withdraw in descending term order would be sufficient to manage duration changes.

You could actually start this process before retirement, using an assumption of the age 72 mortality to calculate duration. Since the upper end need would be age 99, starting as early as age 59 would be possible with currently available TIPS funds.
Thank you for your reply. I was partially answering your earlier question...
One example -- should you or should you not build a TIPS ladder beyond age 85? What are the risks and benefits from such a strategy? Are there alternatives with different sets of risks and benefits?
However, I may have misinterpreted your question since I assumed you meant building the ladder at 65 to last until 85 or beyond, but you may have meant, relying on portfolio withdrawals until 85, then building the ladder*. I think I've partially answered the first, but not the second. Both cases are interesting!

Agreed, the 4% rule is based around a 30 year expectation - for a single person at 65, there is some chance of exceeding that age (it is about 10% for the general population in the UK, and according to the SOA 2000 tables (I only currently have those coded up), it is a bit higher for the US at around 15% or more). If i understand correctly, your comments about Table 1 are then mixing probabilities - at 65 you initially plan for a (say) 15% event, i.e. living to 95, but then update the plan using the mean (and, hence, very roughly, median) life expectancy. A more consistent approach would be to use the same probability for both planning and updating and this would provide an earlier warning of unexpected longevity. For example, using Table 1, at 65, the mean longevity would be to aged 88 (and, rounded to the nearest integer age, this would be exceeded 4 years later at 69).

* I'm probably being dense, but I'm not sure that I understand your example using Table 3 (but it does make me question whether I have understood what you meant in your original question), so please could you explain further?

cheers
StillGoing
User avatar
WoodSpinner
Posts: 2808
Joined: Mon Feb 27, 2017 1:15 pm

Re: The 4% Rule Just Became a Whole Lot Easier - Allan Roth

Post by WoodSpinner »

StillGoing wrote: Tue Nov 22, 2022 4:41 am I posted the following graph which shows the number of years before the income falls below the target (the colour scale) as a function of the ladder duration and ladder spend (i.e., the proportion of the initial portfolio used to construct the ladder). upthread, but it may be worth revisiting...

Image

The graph assumes a target CPI linked income rate of 4% of the initial portfolio, such that if the income from the ladder exceeded the target, the excess was returned to the portfolio, while if the income from the ladder fell below the target, the shortfall was made up from the residual portfolio. A yield to maturity of 1.7% has been used for the bonds in the ladder and the residual portfolio consisted of 80% stocks and 20% nominal bonds, rebalanced annually.

Without a ladder, the income dropped below target (i.e. the portfolio was exhausted) after about 25 years. Ladder durations of just under 20 to 30 years and spends between 40% and 80% resulted in income failure after 35 years or more with longer ladders tending to require a greater spend to achieve the same longevity. The results will be different with a different yield to maturity or a different target income (obviously, reducing the target income increases the longevity of the income - even dropping to 3.8%, improves it by 4 years or more).

Fully agree with vineviz that an obvious problem with a shorter duration ladder is that a 20 year ladder would expire at 85 (for a 65 year old retiree), leaving them with a choice of continuing with portfolio withdrawals only (as assumed in the above graph), the construction of a further ladder (with the possibility that interest rates were much less than those at retirement), or annuitization of some or all of the remaining portfolio (with subsequent inflation risks for nominal annuities). However, even a 30 year ladder will expire, and as vineviz, says, for a couple that means there is a 25% probability that one member of the couple may have to deal with the consequences aged 95.

cheers
StillGoing
StillGoing,

I need some remedial help in understanding this graph. It looks like it is insightful but I just don’t get it.

Can we start with what the y-axis represents?

What do the colors in the heat chart indicate?

How can I apply this in a real world example? For instance, I have projected Cashflow requirements through retirement but they are very lumpy due to changes in income and expenses over time.

Apologies for being a bit slow….

WoodSpinner
StillGoing
Posts: 98
Joined: Mon Nov 04, 2019 4:43 am
Location: U.K.

Re: The 4% Rule Just Became a Whole Lot Easier - Allan Roth

Post by StillGoing »

WoodSpinner wrote: Wed Nov 23, 2022 10:53 am
StillGoing wrote: Tue Nov 22, 2022 4:41 am I posted the following graph which shows the number of years before the income falls below the target (the colour scale) as a function of the ladder duration and ladder spend (i.e., the proportion of the initial portfolio used to construct the ladder). upthread, but it may be worth revisiting...

Image

The graph assumes a target CPI linked income rate of 4% of the initial portfolio, such that if the income from the ladder exceeded the target, the excess was returned to the portfolio, while if the income from the ladder fell below the target, the shortfall was made up from the residual portfolio. A yield to maturity of 1.7% has been used for the bonds in the ladder and the residual portfolio consisted of 80% stocks and 20% nominal bonds, rebalanced annually.

Without a ladder, the income dropped below target (i.e. the portfolio was exhausted) after about 25 years. Ladder durations of just under 20 to 30 years and spends between 40% and 80% resulted in income failure after 35 years or more with longer ladders tending to require a greater spend to achieve the same longevity. The results will be different with a different yield to maturity or a different target income (obviously, reducing the target income increases the longevity of the income - even dropping to 3.8%, improves it by 4 years or more).

Fully agree with vineviz that an obvious problem with a shorter duration ladder is that a 20 year ladder would expire at 85 (for a 65 year old retiree), leaving them with a choice of continuing with portfolio withdrawals only (as assumed in the above graph), the construction of a further ladder (with the possibility that interest rates were much less than those at retirement), or annuitization of some or all of the remaining portfolio (with subsequent inflation risks for nominal annuities). However, even a 30 year ladder will expire, and as vineviz, says, for a couple that means there is a 25% probability that one member of the couple may have to deal with the consequences aged 95.

cheers
StillGoing
StillGoing,

I need some remedial help in understanding this graph. It looks like it is insightful but I just don’t get it.

Can we start with what the y-axis represents?

What do the colors in the heat chart indicate?

How can I apply this in a real world example? For instance, I have projected Cashflow requirements through retirement but they are very lumpy due to changes in income and expenses over time.

Apologies for being a bit slow….

WoodSpinner
No worries - this is my fault for separating the explanation and the graph!

Assuming a ladder is constructed at the beginning of retirement
1) The y axis is the percentage of the portfolio spent on the ladder (ranging from 10% to 100%)
2) The x axis is the time the ladder will provide income for (ranging from 5 to 30 years)
3) The colour scale is the number of years before the income falls below the inflation-adjusted target of 4% (for comparison, in the portfolio only case, the money runs out in year 25) after historical backtesting.
4) The asset allocation immediately after constructing the ladder is 80% stocks and 20% bonds, while after the ladder expired it was 60% stocks and 40% bonds.
5) Fees and taxes are not considered.

In other words, the method is a version of the '4% rule' (i.e. where income is level in real terms), but one where an inflation linked bond ladder is constructed instead of holding solely nominal bonds. The yield to maturity of the inflation linked bonds is 1.7% (i.e., the same as the example given in the OP) resulting in an income of 4.2% for a 30 year ladder. In common with many such backtesting experiments it ignores other income (e.g. social security and DB pensions) and lumpy expenditure. The method also suffers from the same weaknesses as other historical backtesting.

Bearing in mind that the income would be level in real terms, here are a couple of possibilities for someone with a portfolio of $1m.
1) Use the portfolio only with 60% stocks and 40% bonds and withdraw $40k per year and adjust for inflation. Roughly 5% of historical cases failed (i.e. the portfolio went to zero) before year 30 and the earliest (i.e., worst case) failure occurred after about 25 years.
2) Use 40% of the portfolio (i.e., $400k) to construct a 20 year bond ladder. Use an asset allocation of 80% stocks and 20% bonds for the residual portfolio. While the ladder exists (i.e., for the first 20 years) it will provide an inflation adjusted $23.4k per year (pmt(1.7%,20,-400000,0,1)). For the first 20 years, the residual portfolio of $600k will then need to provide an inflation adjusted $16.6k per year, while after 20 years, it will need to provide $40k. From the graph, the first failure was after more than 35 years and, hence, there were no failures in the first 30 years.

My apologies for not explaining this better in the first place.

cheers
StillGoing
GAAP
Posts: 1801
Joined: Fri Apr 08, 2016 12:41 pm

Re: The 4% Rule Just Became a Whole Lot Easier - Allan Roth

Post by GAAP »

StillGoing wrote: Wed Nov 23, 2022 6:51 am
However, I may have misinterpreted your question since I assumed you meant building the ladder at 65 to last until 85 or beyond, but you may have meant, relying on portfolio withdrawals until 85, then building the ladder*. I think I've partially answered the first, but not the second. Both cases are interesting!

Agreed, the 4% rule is based around a 30 year expectation - for a single person at 65, there is some chance of exceeding that age (it is about 10% for the general population in the UK, and according to the SOA 2000 tables (I only currently have those coded up), it is a bit higher for the US at around 15% or more). If i understand correctly, your comments about Table 1 are then mixing probabilities - at 65 you initially plan for a (say) 15% event, i.e. living to 95, but then update the plan using the mean (and, hence, very roughly, median) life expectancy. A more consistent approach would be to use the same probability for both planning and updating and this would provide an earlier warning of unexpected longevity. For example, using Table 1, at 65, the mean longevity would be to aged 88 (and, rounded to the nearest integer age, this would be exceeded 4 years later at 69).

* I'm probably being dense, but I'm not sure that I understand your example using Table 3 (but it does make me question whether I have understood what you meant in your original question), so please could you explain further?

cheers
StillGoing
Actually, I was primarily trying to get back to a topic raised up-thread that rapidly devolved into an argument about annuities -- and I was just using that topic as one example of how to get away from a pointless discussion about a term definition and on to something potentially useful...

I should probably note at this time that I am not a fan of SWR methods, and longevity risk is just one reason why. If I were to use such a method, I would also need to plan to deal with potential "excess longevity".

The basic idea is to use a longevity table to estimate expected longevity, set aside a portion of the portfolio to deal with the potential income risks associated with living longer, and then adjust on an annual basis going forward. If someone entered retirement at age 65 with a 30-year TIPS ladder, anything beyond age 95 is at risk -- especially if they are relying upon a method that may very well run out of money in 30 years. That corner case is much easier to deal with if you start early.

The IRS RMD tables are actually quite conservative for this sort of thing since they hit the upper range of the distribution -- nearly double what the Social Security Administration uses. How much longer they will publish Table 1 is an open question given the SECURE act changes, but table 3 is sufficient for this purpose. At age 72, that hypothetical 30-year ladder has 23 years remaining to age 95, while the table says there is potential need out to age 99. Ten years later, the ladder has 13 years remaining while the table says there is potential for 18 more years of need. The table goes out to age 120, a quarter century past the end of the TIPS ladder.

My proposal is to start building a fund at age 72 to manage this risk. The retiree would calculate the NPV of the at-risk years income each year, and set aside money in a combination of TIPS funds with the necessary duration, with the mix changing as the need gets closer. This would act somewhat as the purchase/creation of a deferred real annuity over time -- and one that has potential/probable residual value for heirs. Since the risk and potential duration changes each year, I think a blend of funds would be easier to deal with than trying to buy individual TIPS -- even thought the duration match would not be perfectly accurate at first.

At the moment, this method would require about 3.5 years of expected income to be set aside at age 72. The retiree could start sooner with smaller amounts -- or ideally start building that contingency fund before retirement. If they retire in a good year, they could just set aside the money up-front rather than calculating their withdrawal rate based upon the total portfolio. Choosing the age 72 NPV while retiring at 65 should provide plenty of cushion.
“Adapt what is useful, reject what is useless, and add what is specifically your own.” ― Bruce Lee
bugday
Posts: 5
Joined: Mon Oct 24, 2022 6:22 am

Re: The 4% Rule Just Became a Whole Lot Easier - Allan Roth

Post by bugday »

My understanding of the "normal" way of building a TIPS ladder in one fell swoop is you do the following: you first purchase a 30 years TIPS which gives lump sum in 30 years, and a coupon each year along the way. You then buy a 29 year TIPS, though that purchase is smaller than it would otherwise be because you take into account the coupon you receive from the 30 year TIPS. You then buy a 28 year TIPS, which is even smaller due to the coupons from the 29 and 30 year, etc.

Another method would be to slowly build a ladder over time (ie buy a 30 year and 15 year every year for the 15 years preceding retirement). If one took this approach, how would you account for the coupon payments coming from the bonds you haven't purchased yet?

For example, in 2022 I buy bonds maturing in 2037 and 2052. When all is said and done, in 2037 I will get the face value of the 2037 bond and coupons from bonds maturing in 2048-2067 (starting to feel like a science fiction book!), but in 2022 when I'm buying the bond maturing in 2037, I only know what the coupon would be from the 2052 bond.
ollistu
Posts: 21
Joined: Fri Jan 10, 2020 12:35 pm

Re: The 4% Rule Just Became a Whole Lot Easier - Allan Roth

Post by ollistu »

If one were to purchase the 30 year TIPS ladder in a non-retirement brokerage account, I understand the income tax due on the"phantom income". Would this be a worse-case scenario?: Year 1 the inflation index increases your bond by 5%, and you owe the tax on the phantom income. Year 2 there is deflation by 5%. And so on.....on alternate years, the inflation number goes up, and then goes down by the same amount. So when the bond matures, the value will even out to the face value (equal number of inflation and deflation years), but you will have paid tax every other year for no income at the end.
I know this is highly unlikely, but just checking to see if I understand this correctly.
Leesbro63
Posts: 8788
Joined: Mon Nov 08, 2010 4:36 pm

Re: The 4% Rule Just Became a Whole Lot Easier - Allan Roth

Post by Leesbro63 »

ollistu wrote: Fri Dec 02, 2022 1:42 pm If one were to purchase the 30 year TIPS ladder in a non-retirement brokerage account, I understand the income tax due on the"phantom income". Would this be a worse-case scenario?: Year 1 the inflation index increases your bond by 5%, and you owe the tax on the phantom income. Year 2 there is deflation by 5%. And so on.....on alternate years, the inflation number goes up, and then goes down by the same amount. So when the bond matures, the value will even out to the face value (equal number of inflation and deflation years), but you will have paid tax every other year for no income at the end.
I know this is highly unlikely, but just checking to see if I understand this correctly.
I'm honestly not sure I understand your question, but you mentioned "worst-case scenario". The worst case scenario is big inflation, like 2022, for an extended period of time, like the late 1960s through early 1980s period. The very reason you buy TIPS...for inflation protection...gets negated as much of your "phantom income" gets "taxflated" away. For this very reason, the TIPS solution doesn't work well for large taxable portfolios.
User avatar
vineviz
Posts: 14456
Joined: Tue May 15, 2018 1:55 pm
Location: Baltimore, MD

Re: The 4% Rule Just Became a Whole Lot Easier - Allan Roth

Post by vineviz »

Leesbro63 wrote: Fri Dec 02, 2022 4:12 pm
ollistu wrote: Fri Dec 02, 2022 1:42 pm If one were to purchase the 30 year TIPS ladder in a non-retirement brokerage account, I understand the income tax due on the"phantom income". Would this be a worse-case scenario?: Year 1 the inflation index increases your bond by 5%, and you owe the tax on the phantom income. Year 2 there is deflation by 5%. And so on.....on alternate years, the inflation number goes up, and then goes down by the same amount. So when the bond matures, the value will even out to the face value (equal number of inflation and deflation years), but you will have paid tax every other year for no income at the end.
I know this is highly unlikely, but just checking to see if I understand this correctly.
I'm honestly not sure I understand your question, but you mentioned "worst-case scenario". The worst case scenario is big inflation, like 2022, for an extended period of time, like the late 1960s through early 1980s period. The very reason you buy TIPS...for inflation protection...gets negated as much of your "phantom income" gets "taxflated" away. For this very reason, the TIPS solution doesn't work well for large taxable portfolios.
This myth has been repeatedly debunked.
"Far more money has been lost by investors preparing for corrections than has been lost in corrections themselves." ~~ Peter Lynch
Leesbro63
Posts: 8788
Joined: Mon Nov 08, 2010 4:36 pm

Re: The 4% Rule Just Became a Whole Lot Easier - Allan Roth

Post by Leesbro63 »

vineviz wrote: Fri Dec 02, 2022 4:43 pm
Leesbro63 wrote: Fri Dec 02, 2022 4:12 pm
ollistu wrote: Fri Dec 02, 2022 1:42 pm If one were to purchase the 30 year TIPS ladder in a non-retirement brokerage account, I understand the income tax due on the"phantom income". Would this be a worse-case scenario?: Year 1 the inflation index increases your bond by 5%, and you owe the tax on the phantom income. Year 2 there is deflation by 5%. And so on.....on alternate years, the inflation number goes up, and then goes down by the same amount. So when the bond matures, the value will even out to the face value (equal number of inflation and deflation years), but you will have paid tax every other year for no income at the end.
I know this is highly unlikely, but just checking to see if I understand this correctly.
I'm honestly not sure I understand your question, but you mentioned "worst-case scenario". The worst case scenario is big inflation, like 2022, for an extended period of time, like the late 1960s through early 1980s period. The very reason you buy TIPS...for inflation protection...gets negated as much of your "phantom income" gets "taxflated" away. For this very reason, the TIPS solution doesn't work well for large taxable portfolios.
This myth has been repeatedly debunked.
I do remember some “soft” points that conflict with the assertion that TIPS don’t work well for large taxable portfolios. But I don’t remember any outright debunking.
User avatar
nedsaid
Posts: 16938
Joined: Fri Nov 23, 2012 12:33 pm

Re: The 4% Rule Just Became a Whole Lot Easier - Allan Roth

Post by nedsaid »

vineviz wrote: Fri Dec 02, 2022 4:43 pm
Leesbro63 wrote: Fri Dec 02, 2022 4:12 pm
ollistu wrote: Fri Dec 02, 2022 1:42 pm If one were to purchase the 30 year TIPS ladder in a non-retirement brokerage account, I understand the income tax due on the"phantom income". Would this be a worse-case scenario?: Year 1 the inflation index increases your bond by 5%, and you owe the tax on the phantom income. Year 2 there is deflation by 5%. And so on.....on alternate years, the inflation number goes up, and then goes down by the same amount. So when the bond matures, the value will even out to the face value (equal number of inflation and deflation years), but you will have paid tax every other year for no income at the end.
I know this is highly unlikely, but just checking to see if I understand this correctly.
I'm honestly not sure I understand your question, but you mentioned "worst-case scenario". The worst case scenario is big inflation, like 2022, for an extended period of time, like the late 1960s through early 1980s period. The very reason you buy TIPS...for inflation protection...gets negated as much of your "phantom income" gets "taxflated" away. For this very reason, the TIPS solution doesn't work well for large taxable portfolios.
This myth has been repeatedly debunked.
For one thing, the tax brackets get adjusted for inflation each year. Other things in the tax code like IRA and 401(k) contribution limits are also inflation adjusted. Taxflation might have been a big issue in the 1970's but the issue would be pretty minor today as many things but not everything in the tax code is inflation adjusted.
A fool and his money are good for business.
Leesbro63
Posts: 8788
Joined: Mon Nov 08, 2010 4:36 pm

Re: The 4% Rule Just Became a Whole Lot Easier - Allan Roth

Post by Leesbro63 »

nedsaid wrote: Fri Dec 02, 2022 5:44 pm
vineviz wrote: Fri Dec 02, 2022 4:43 pm
Leesbro63 wrote: Fri Dec 02, 2022 4:12 pm
ollistu wrote: Fri Dec 02, 2022 1:42 pm If one were to purchase the 30 year TIPS ladder in a non-retirement brokerage account, I understand the income tax due on the"phantom income". Would this be a worse-case scenario?: Year 1 the inflation index increases your bond by 5%, and you owe the tax on the phantom income. Year 2 there is deflation by 5%. And so on.....on alternate years, the inflation number goes up, and then goes down by the same amount. So when the bond matures, the value will even out to the face value (equal number of inflation and deflation years), but you will have paid tax every other year for no income at the end.
I know this is highly unlikely, but just checking to see if I understand this correctly.
I'm honestly not sure I understand your question, but you mentioned "worst-case scenario". The worst case scenario is big inflation, like 2022, for an extended period of time, like the late 1960s through early 1980s period. The very reason you buy TIPS...for inflation protection...gets negated as much of your "phantom income" gets "taxflated" away. For this very reason, the TIPS solution doesn't work well for large taxable portfolios.
This myth has been repeatedly debunked.
For one thing, the tax brackets get adjusted for inflation each year. Other things in the tax code like IRA and 401(k) contribution limits are also inflation adjusted. Taxflation might have been a big issue in the 1970's but the issue would be pretty minor today as many things but not everything in the tax code is inflation adjusted.
If one has $10M (my definition of “a large taxable account”) invested in TIPS and we have a 9% inflation year, that’s $900,000 of phantom income that gets taxed. Adjusting tax brackets helps a little, but over a 1966-82 period, a lotta wealth will erode.
User avatar
nedsaid
Posts: 16938
Joined: Fri Nov 23, 2012 12:33 pm

Re: The 4% Rule Just Became a Whole Lot Easier - Allan Roth

Post by nedsaid »

Leesbro63 wrote: Fri Dec 02, 2022 6:57 pm
nedsaid wrote: Fri Dec 02, 2022 5:44 pm
vineviz wrote: Fri Dec 02, 2022 4:43 pm
Leesbro63 wrote: Fri Dec 02, 2022 4:12 pm
ollistu wrote: Fri Dec 02, 2022 1:42 pm If one were to purchase the 30 year TIPS ladder in a non-retirement brokerage account, I understand the income tax due on the"phantom income". Would this be a worse-case scenario?: Year 1 the inflation index increases your bond by 5%, and you owe the tax on the phantom income. Year 2 there is deflation by 5%. And so on.....on alternate years, the inflation number goes up, and then goes down by the same amount. So when the bond matures, the value will even out to the face value (equal number of inflation and deflation years), but you will have paid tax every other year for no income at the end.
I know this is highly unlikely, but just checking to see if I understand this correctly.
I'm honestly not sure I understand your question, but you mentioned "worst-case scenario". The worst case scenario is big inflation, like 2022, for an extended period of time, like the late 1960s through early 1980s period. The very reason you buy TIPS...for inflation protection...gets negated as much of your "phantom income" gets "taxflated" away. For this very reason, the TIPS solution doesn't work well for large taxable portfolios.
This myth has been repeatedly debunked.
For one thing, the tax brackets get adjusted for inflation each year. Other things in the tax code like IRA and 401(k) contribution limits are also inflation adjusted. Taxflation might have been a big issue in the 1970's but the issue would be pretty minor today as many things but not everything in the tax code is inflation adjusted.
If one has $10M (my definition of “a large taxable account”) invested in TIPS and we have a 9% inflation year, that’s $900,000 of phantom income that gets taxed. Adjusting tax brackets helps a little, but over a 1966-82 period, a lotta wealth will erode.
Sort of like the commercial, you can pay me now or pay me later. The relative tax inefficiency of TIPS is one reason that it is recommended that these be placed in tax deferred retirement accounts or ROTH accounts. I saw an article on tipswatch regarding taxes on TIPS in taxable accounts. The author gave an example and said the tax burden isn't always as bad as you think. During 1966-82, tax brackets were not indexed for inflation, this didn't happen until 1985.
A fool and his money are good for business.
Leesbro63
Posts: 8788
Joined: Mon Nov 08, 2010 4:36 pm

Re: The 4% Rule Just Became a Whole Lot Easier - Allan Roth

Post by Leesbro63 »

So I ran some numbers to see if my point was correct or not about TIPS not working well for large taxable portfolios. Say you accumulated a (non tax sheltered) $10M stash; my definition of a large taxable portfolio (sold a business, inheritance, lottery). And you wanted to "lock in the real value forever" and bought TIPS. You won the game and want to totally quit "playing". If inflation was 9% as it almost is this year, you'd owe $317,000 in Federal income tax if you had no other income besides the $900,000 (9% of $10M) phantom TIPS income. That's a 3% drag. If you still had a 30 year life expectancy (or joint life expectancy), wouldn't a 50/50 portfolio of Vanguard Total Market/Intermediate Term Municipal Bond Fund make more sense than a guaranteed annual tax drag...through, say, a 1966-1982 period? Yeah, I get it that inflation won't be 9% every year. But it could be even more. And yeah, I get it that you might have other income. And yeah, I also get it that even the 50/50 portfolio would have had a tough time from 1966-1982. But it just seems like signing up for a guaranteed tax-loss-drag doesn't fit my definition of TIPS working for a large taxable portfolio. And that the idea of TIPS being some sort of magical way to lock in value just does not work for large taxable portfolios. It's a good solution for the upper middle class and moderately wealthy, only.
User avatar
Svensk Anga
Posts: 1184
Joined: Sun Dec 23, 2012 5:16 pm

Re: The 4% Rule Just Became a Whole Lot Easier - Allan Roth

Post by Svensk Anga »

Leesbro63 wrote: Sat Dec 03, 2022 10:30 am So I ran some numbers to see if my point was correct or not about TIPS not working well for large taxable portfolios. Say you accumulated a (non tax sheltered) $10M stash; my definition of a large taxable portfolio (sold a business, inheritance, lottery). And you wanted to "lock in the real value forever" and bought TIPS. You won the game and want to totally quit "playing". If inflation was 9% as it almost is this year, you'd owe $317,000 in Federal income tax if you had no other income besides the $900,000 (9% of $10M) phantom TIPS income. That's a 3% drag. If you still had a 30 year life expectancy (or joint life expectancy), wouldn't a 50/50 portfolio of Vanguard Total Market/Intermediate Term Municipal Bond Fund make more sense than a guaranteed annual tax drag...through, say, a 1966-1982 period? Yeah, I get it that inflation won't be 9% every year. But it could be even more. And yeah, I get it that you might have other income. And yeah, I also get it that even the 50/50 portfolio would have had a tough time from 1966-1982. But it just seems like signing up for a guaranteed tax-loss-drag doesn't fit my definition of TIPS working for a large taxable portfolio. And that the idea of TIPS being some sort of magical way to lock in value just does not work for large taxable portfolios. It's a good solution for the upper middle class and moderately wealthy, only.
No, it's not so good for $10 MM, but that kind of stash is rare. It looks bad because of the progressive tax system. At lower rates, the real return on the TIPS can pay the taxes on a good bit of inflation. It looks pretty good if you can stay in the 12% bracket. Your premise was to lock in the real value. At present you can do about 1.5%/year better than that. That 1.5% real can pay your taxes and still have a shot at preserving the real value of the stash. True, you may not have 1.5% of coupons, but you could sell an issue or two to get down to preserving and not growing real value.
Post Reply