22 multiple rule

Discuss all general (i.e. non-personal) investing questions and issues, investing news, and theory.
Post Reply
Topic Author
investment rube
Posts: 20
Joined: Fri Oct 04, 2013 1:04 pm

22 multiple rule

Post by investment rube »

I read an article by Jason Zweig in the WSJ today that stated, "To be assured of having enough money to fund a comfortable retirement, you should save a total of 22 times the annual income you want to earn when you retire."
First, wondering the Boglehead view of this statement.
Also, would this be impacted by age when you retire? 65 vs, say, 50?
Thanks.
User avatar
JoMoney
Posts: 16260
Joined: Tue Jul 23, 2013 5:31 am

Re: 22 multiple rule

Post by JoMoney »

My view is that the 4% SWR (equivalent to 25x income desired) is a "rule of thumb: a principle with broad application that is not intended to be strictly accurate or reliable for every situation" not a hard "rule: statement that tells you what is allowed or what will happen within a particular system"
Age does make a difference. You need a higher growth rate (or a bigger pot of money) to have a portfolio last a longer period of time.

I like the figure of:
((Life expectancy)-(Current Age, adjusting every year)) / (Portfolio Balance)
or as a percentage
100 / ((Life expectancy)-(Current Age, adjusting every year))

Most Single Life Expectancy tables estimate around 84 years (until you get closer to that age, then it goes up), so for a 65 YO:
100 / (84-65) = 5.26 % , or around 19x desired income
for a 50 YO:
100 / (84-50) = 2.94 % , or around 34x desired income

What I like about this method is that it doesn't really rely on future growth. It just tells you what may be prudent given what you know today. It may hamper spending early on, but this is when you have the most potential from your human capital to earn more if you want to spend more. If you have good growth in the portfolio, It may be a little heavy on how much you can withdraw in the later years, but this is when you may not have any other choice but to live on your portfolio, and I'd rather be in a position where I may have too much than too little. If the portfolio is invested in assets that fluctuate in value a lot (i.e. stock market) it could make the actual dollar amount withdrawn fluctuate wildly.
"To achieve satisfactory investment results is easier than most people realize; to achieve superior results is harder than it looks." - Benjamin Graham
MooreBonds
Posts: 153
Joined: Sat Apr 27, 2013 8:54 pm
Location: Missouri

Re: 22 multiple rule

Post by MooreBonds »

investment rube wrote:I read an article by Jason Zweig in the WSJ today that stated, "To be assured of having enough money to fund a comfortable retirement, you should save a total of 22 times the annual income you want to earn when you retire."
First, wondering the Boglehead view of this statement.
Also, would this be impacted by age when you retire? 65 vs, say, 50?
Thanks.
The obvious first answer is that it all depends on your annual expenses. But since most people have no clue what they spend each year (they likely would underestimate their annual budget), it's easier for the press to deal in multiples of annual salary, because nearly everyone knows what their annual salary is.

As many on the forum would say, retiring at 50 would warrant a more conservative WR for security (perhaps 3% withdrawal rate), rather than closer to a 4% withdrawal rate if you're 65 and don't have much expectations of living beyond 95.

Using strictly back-of-the-envelope calculations, the 22x income it would be a somewhat close approximation for someone saving roughly 15%/year.

While working:
1.0 Annual income
-0.0765 SS/Medicare
-0.7735 Taxes/Spending
0.15 Left for saving

If you retire with the same spending level as working, and if you retire with 22x annual income, it leaves you with a portfolio that's 28x expenses. A 4% WR from a 28x expenses portfolio pulls out 1.14x times expenses each year, which is more than needed if you retire at 67 and can live with a slight chance of running out of money at 97.

If you start work at 22, retire at 67, that's a 45 year career. Saving 15% each year gives you total savings of 6.75x income over 45 years. To get to 22x annual income from 6.75x income saved up, you need 3.25 'doubling periods'. Assuming 5% average annualized returns AFTER TAXES, you'd need roughly 14 years to double. So 3.25 doubling periods @ 5% after-taxes would take 45 years. (yes, I know it's not purely mathematically correct due to compounding, but I'm only doing back of the envelope calculations here)

This does ignore Social Security, but it also ignores the impacts of inflation, as a 5% annualized return after taxes and after inflation would be hard to come by. Also, if you're an 82 year old still spending the same as you did when you were working, you either have one heck of a retirement, or an overly conservative budget. And finally, this simple result gives you 28x EXPENSES at 65, which, with a 4% WR, gives you 14% more than your budget. So you could reduce your annual return a little for inflation, and add in SS, and arrive at an approximately close realistic number.

If you retire at 50, you'd need to save a far higher % of your income - but then you would need to accumulate less in order to live off of a lower withdrawal rate.
YDNAL
Posts: 13774
Joined: Tue Apr 10, 2007 4:04 pm
Location: Biscayne Bay

Re: 22 multiple rule

Post by YDNAL »

investment rube wrote:I read an article by Jason Zweig in the WSJ today that stated, "To be assured of having enough money to fund a comfortable retirement, you should save a total of 22 times the annual income you want to earn when you retire."
The often-used *guideline* is 25 x Expenses for normal retirement [say 65].
  • This is equivalent to the 4% withdrawal guideline.
  • Want 3% withdrawal?... 33.3 x Expenses.
I have zero clue where 22 comes from, but there must be some funky calculation behind it - one that I also have zero interest in. :)

My motto: save the most and consume the less while enjoying life!
Landy | Be yourself, everyone else is already taken -- Oscar Wilde
Call_Me_Op
Posts: 9881
Joined: Mon Sep 07, 2009 2:57 pm
Location: Milky Way

Re: 22 multiple rule

Post by Call_Me_Op »

YDNAL wrote: My motto: save the most and consume the less while enjoying life!
Less or least? :)
Best regards, -Op | | "In the middle of difficulty lies opportunity." Einstein
cherijoh
Posts: 6591
Joined: Tue Feb 20, 2007 3:49 pm
Location: Charlotte NC

Re: 22 multiple rule

Post by cherijoh »

JoMoney wrote:My view is that the 4% SWR (equivalent to 25x income desired) is a "rule of thumb: a principle with broad application that is not intended to be strictly accurate or reliable for every situation" not a hard "rule: statement that tells you what is allowed or what will happen within a particular system"
Age does make a difference. You need a higher growth rate (or a bigger pot of money) to have a portfolio last a longer period of time.

I like the figure of:
((Life expectancy)-(Current Age, adjusting every year)) / (Portfolio Balance)
or as a percentage
100 / ((Life expectancy)-(Current Age, adjusting every year))

Most Single Life Expectancy tables estimate around 84 years (until you get closer to that age, then it goes up), so for a 65 YO:
100 / (84-65) = 5.26 % , or around 19x desired income
for a 50 YO:
100 / (84-50) = 2.94 % , or around 34x desired income

What I like about this method is that it doesn't really rely on future growth. It just tells you what may be prudent given what you know today. It may hamper spending early on, but this is when you have the most potential from your human capital to earn more if you want to spend more. If you have good growth in the portfolio, It may be a little heavy on how much you can withdraw in the later years, but this is when you may not have any other choice but to live on your portfolio, and I'd rather be in a position where I may have too much than too little. If the portfolio is invested in assets that fluctuate in value a lot (i.e. stock market) it could make the actual dollar amount withdrawn fluctuate wildly.
I'm not sure what you were getting at with your first formula, but it doesn't make sense to me - did you leave something out? The numerator will be a number in the range of something like 20 to 40, then you are dividing it by your portfolio balance which will hopefully be quite large. It bears no relationship as written to the second formula, which is how the IRS determines RMDs for your IRA.

IMO the weak point of this approach is that the life expectancy tables are based on your median life expectancy, so 50% of the people will live longer - some, a lot longer! So I would hesitate to use it to calculate how much I need to fund my portfolio to have an adequate retirement nest egg. And unfortunately, whether you run out of money or not will depend on the future growth of your portfolio. There isn't any way around it unless you have a very large multiple of your spending needs.
User avatar
midareff
Posts: 7711
Joined: Mon Nov 29, 2010 9:43 am
Location: Biscayne Bay, South Florida

Re: 22 multiple rule

Post by midareff »

Bengen, the father of the 4% rule (25X) used some interesting data in his calculations of portfolio life. His model portfolio was 50% stock returning 10.3% and 50% IT Treasury notes returning 5.1%. While expectations are nothing more than expectations, it seems the prospects of receiving 10.3% average return from stocks over the next decade don't seem very high. Treasuries as measured by the VG IT Government Bond Index (all treasuries) currently has a prior 12 month yield of 1.33%.

Bengen's projection data, upon which he calculated portfolio survivability, is not what we expect moving forward subjecting today's withdrawal rate to sequence of return risks. Stock returns are generally estimated to be 50-60% of what Bengen used for projections, and current treasury yields are roughly 30% of his data used for the projection calculations.

AFAIK, you can follow the Pied Piper's and math wizards of 4% and higher inflation indexed withdrawals, but do so at your own risk understanding you are making the later golden year's cat food and life under the bridge bet.
User avatar
matjen
Posts: 2189
Joined: Sun Nov 20, 2011 10:30 pm

Re: 22 multiple rule

Post by matjen »

YDNAL wrote:I have zero clue where 22 comes from, but there must be some funky calculation behind it - one that I also have zero interest in. :)
OP's original question arises out of a WSJ article I posted and commented on here. We need BobK to chime in on these threads. This is his sweet spot!

http://www.bogleheads.org/forum/viewtop ... 0&t=132044
A man is rich in proportion to the number of things he can afford to let alone.
User avatar
gatorking
Posts: 1498
Joined: Tue Feb 20, 2007 5:15 pm
Location: MA

Re: 22 multiple rule

Post by gatorking »

MooreBonds wrote: The obvious first answer is that it all depends on your annual expenses. But since most people have no clue what they spend each year (they likely would underestimate their annual budget), it's easier for the press to deal in multiples of annual salary, because nearly everyone knows what their annual salary is.
The original post says "you should save a total of 22 times the annual income you want to earn when you retire." I read this as 22 times your annual expenses in retirement.
User avatar
TheTimeLord
Posts: 12130
Joined: Fri Jul 26, 2013 2:05 pm

Re: 22 multiple rule

Post by TheTimeLord »

matjen wrote:
YDNAL wrote:I have zero clue where 22 comes from, but there must be some funky calculation behind it - one that I also have zero interest in. :)
OP's original question arises out of a WSJ article I posted and commented on here. We need BobK to chime in on these threads. This is his sweet spot!

http://www.bogleheads.org/forum/viewtop ... 0&t=132044
It seems years ago it was 20 then with the 4% rule it became 25 now you see folks pushing towards 33. It is hard to use the amount of money you earn because you maybe be saving 20% of that plus people also have benefits like pension and SS that change the equation. I don't know if 22 is the right answer but it probably isn't a bad goal. Personally because of small pensions, SS and the amount we save the target number I have for my wife and I is south of 22 times our combined earnings but each situation is unique and I assume 22 is a rule of thumb not an absolute. Our target number is based on SS, pensions and the 4% rule to put us within a range of income annually. From experience I know we can live comfortably at the bottom of the range and the top of the would enable pretty much anything I can think of us ever needing in a year. The range itself is only $35K wide.
IMHO, Investing should be about living the life you want, not avoiding the life you fear. | Run, You Clever Boy! [9085]
Rodc
Posts: 13601
Joined: Tue Jun 26, 2007 9:46 am

Re: 22 multiple rule

Post by Rodc »

25x, 22x whatever. This is a very rough guestimate and so there is little difference.

FWIW: 22x just means a 4.5% withdrawal rate vs 4%. The usually worked historically for 30 years, but obviously not as often as 4%. Take your pick: it is just a rough rule of thumb. If you want to be even more safe, pick 30x.

For example if you want to buy an inflation indexed annuity, I might guess more often than not you can do better than 4% real return, and so 4.5% might be a better guestimate for someone willing to buy an annuity. FWIW-II: recently while bonds were providing such meager returns the rate I think was down around 4% indicating that in more typical bond climates you could do a little better.
We live a world with knowledge of the future markets has less than one significant figure. And people will still and always demand answers to three significant digits.
User avatar
Cb
Posts: 275
Joined: Tue Feb 20, 2007 11:39 pm

Re: 22 multiple rule

Post by Cb »

midareff wrote:...AFAIK, you can follow the Pied Piper's and math wizards of 4% and higher inflation indexed withdrawals, but do so at your own risk understanding you are making the later golden year's cat food and life under the bridge bet...
This sort of thing turns up in most SWR threads. I wonder what % of the retired population lives under bridges, subsisting on a diet of cat food, and of that population, what % had entered their retirements with an investment portfolio of ~22 times their annual expenses.
Topic Author
investment rube
Posts: 20
Joined: Fri Oct 04, 2013 1:04 pm

Re: 22 multiple rule

Post by investment rube »

Thanks everyone for the replies. All very helpful.
I thought 22x was too hopeful. Oh well, back to the coal mine...
basspond
Posts: 1500
Joined: Wed Nov 27, 2013 3:01 am

Re: 22 multiple rule

Post by basspond »

Like in any of these retirement assets calculations, your are dealing with three unknowns, actual retirement expenses, returns, and your drop dead date. It is impossible for anyone to tell you what that number is. If they could then they would be richer than Carlos Slim. You will have to run your own numbers and decide how much risk you are willing to take. In my case it is at least 30* guesstimated expenses.
User avatar
Offshore
Posts: 510
Joined: Wed Apr 09, 2008 9:43 am
Location: Michigan's West Coast

Re: 22 multiple rule

Post by Offshore »

Mr. Zweig makes the following statement:
The closest thing to a riskless asset is Treasury inflation-protected securities, U.S. government bonds that promise to preserve your purchasing power. Conservatively assuming a rate of return of zero, a $2 million portfolio of TIPS will enable you to withdraw $100,000 a year for 20 years.
May I presume he is refering to holding TIPS directly and not via a fund? He never tells the reader if a TIPS fund can substitute for direct bond ownership. I have sent him this question via email, but haven't heard back yet.

My longtime concern has been dealing with Treasury Direct. This board has several threads on the difficulties investors face with TD.
User avatar
bobcat2
Posts: 6076
Joined: Tue Feb 20, 2007 2:27 pm
Location: just barely Outside the Beltway

Re: 22 multiple rule

Post by bobcat2 »

matjen wrote:
YDNAL wrote:I have zero clue where 22 comes from, but there must be some funky calculation behind it - one that I also have zero interest in. :)
OP's original question arises out of a WSJ article I posted and commented on here. We need BobK to chime in on these threads. This is his sweet spot!

http://www.bogleheads.org/forum/viewtop ... 0&t=132044
IMO once you are within 20 years of your expected retirement year you should target your retirement income goals by pricing the cost of a real life annuity at your retirement age that generates your retirement income goals. This needs to be adjusted each year as the cost changes primarily due to changes in real interest rates and longevity. This does not mean you are necessarily purchasing a real life annuity at retirement, but instead you are using its cost to price your income goals, both floor and aspirational. You do not want to price the amount needed by 25x or any other (x times) simplistic rule of thumb.

For example, you are a 50 year old single male and you intend to retire in 15 years at age 65. You want your portfolio to be able to generate $11,000 real in safe retirement income above your safe SS income and $31,000 real in total aspirational income. Today's price for a real life annuity producing $11,000 in income for a 65 year old male is $220,000. You aim to hit your floor income by saving and investing safely between now and retirement in order to have $220,000 to purchase a real life annuity at age 65 that generates $11,000 in safe income. The rest of your portfolio needs to grow to $400,000, again by pricing a real life annuity, by age 65 to generate the additional $20,000 in retirement income to get you to $31,000. You do not intend to purchase a life annuity at age 65 to generate the additional $20,000 in retirement income, but you use the life annuity to price how much assets are needed to reach the aspirational income goal.

In addition you want a reserve fund of at least $100,000 at retirement. At retirement you need a portfolio in today's dollars of $720,000 ($220k+$400k+$100k). This gets adjusted regularly over time between now and retirement primarily because of changes in real interest rates and longevity at age 65, which the pricing of real annuities takes into account explicitly, as well as some other factors.

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.
Rodc
Posts: 13601
Joined: Tue Jun 26, 2007 9:46 am

Re: 22 multiple rule

Post by Rodc »

Today's price for a real life annuity producing $11,000 in income for a 65 year old male is $220,000.
Which in other words, says today this person wants a multiplier of 20x, and this will go up or down depending on real rates.

20x, 22x, seems both are basically the same given the level of uncertainty.

I do like the idea of also planning on a portion set aside for a reserve fund.

So, to get $31K, and have a $100K reserve, you show a portfolio of $720K, which is a multiplier of

23x desired income

Seems the WSJ rule of thumb at least today is right on the money.
We live a world with knowledge of the future markets has less than one significant figure. And people will still and always demand answers to three significant digits.
User avatar
wade
Posts: 612
Joined: Fri Sep 17, 2010 12:38 am
Location: Main Line
Contact:

Re: 22 multiple rule

Post by wade »

I believe that the 22x number is based on the specific strategy of retiring at 65, buying a 20-year TIPS bond ladder, and buying a deferred income annuity which begins payments at age 85. This allows for a higher than 4% withdrawal rate, though with a great deal of inflation risk related to the deferred income annuity.
User avatar
bobcat2
Posts: 6076
Joined: Tue Feb 20, 2007 2:27 pm
Location: just barely Outside the Beltway

Re: 22 multiple rule

Post by bobcat2 »

Rodc wrote:
Today's price for a real life annuity producing $11,000 in income for a 65 year old male is $220,000.
Which in other words, says today this person wants a multiplier of 20x, and this will go up or down depending on real rates.

20x, 22x, seems both are basically the same given the level of uncertainty.

I do like the idea of also planning on a portion set aside for a reserve fund.

So, to get $31K, and have a $100K reserve, you show a portfolio of $720K, which is a multiplier of

23x desired income

Seems the WSJ rule of thumb at least today is right on the money.
I made up the cost of a real life annuity today for a 65 year old male. I used a reasonable number that made the math easy. My guess is today it would cost more than the $720k I came up with in the example for the single male. It would be even more expensive for a single woman or a couple in any event.

The level of real interest rates makes a "real" difference both when planning for retirement and for having a pleasant retirement lifestyle. :wink: It's unfortunate that many people have trouble grasping this simple fact. :(

Any simply multiplier cannot possibly be generally correct for a single male, a single female, and a couple. Each of these households will have a different correct multiplier at any point in time. The correct multiplier will also be different for every level of current real LT interest rates. In addition it will be different as longevity evolves over time and obviously for every age of retirement. Any time you see a fixed multiplier such as 23x what you know for sure is that it is wrong for most people. What you don't know for sure is whether it is wrong for nearly everyone. :D

BobK
Last edited by bobcat2 on Tue Feb 04, 2014 11:38 am, edited 2 times in total.
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.
TareNeko
Posts: 585
Joined: Wed Aug 14, 2013 1:27 pm

Re: 22 multiple rule

Post by TareNeko »

22 seems aggressive to me. I would (and do) target for 30x of retirement spending where the house is paid off.

The problem is, I do not count on SS, especially if I can retire early. So I need to be self sufficient.
User avatar
bobcat2
Posts: 6076
Joined: Tue Feb 20, 2007 2:27 pm
Location: just barely Outside the Beltway

Re: 22 multiple rule

Post by bobcat2 »

Erhan wrote:22 seems aggressive to me. I would (and do) target for 30x of retirement spending where the house is paid off.

The problem is, I do not count on SS, especially if I can retire early. So I need to be self sufficient.
Medicare is in much worse financial shape than SS. So you had better not count on Medicare. If you are married and want to be truly self-sufficient add another $800,000 for self insuring retirement medical costs. If you are single you can cut that almost in half. :happy

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.
richard
Posts: 7961
Joined: Tue Feb 20, 2007 2:38 pm
Contact:

Re: 22 multiple rule

Post by richard »

bobcat2 wrote:
Erhan wrote:22 seems aggressive to me. I would (and do) target for 30x of retirement spending where the house is paid off.

The problem is, I do not count on SS, especially if I can retire early. So I need to be self sufficient.
Medicare is in much worse financial shape than SS. So you had better not count on Medicare. If you are married and want to be truly self-sufficient add another $800,000 for self insuring retirement medical costs. If you are single you can cut that almost in half. :happy

BobK
Just to show how bad things are, the market is willing to lend to the US at approximately the lowest rate in over 50 years. Consider the risk premium that implies.
Image
User avatar
bobcat2
Posts: 6076
Joined: Tue Feb 20, 2007 2:27 pm
Location: just barely Outside the Beltway

Re: 22 multiple rule

Post by bobcat2 »

Here is yet another problem with any simple multiplier. SS replaces a higher percentage of pre-retirement income the lower your level of pre-retirement income. If the goal of saving for retirement is to have the same standard of living as you had in your latter working years then it follows that higher earners must have higher multipliers, because a lower proportion of their pre-retirement income is replaced by SS.

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.
richard
Posts: 7961
Joined: Tue Feb 20, 2007 2:38 pm
Contact:

Re: 22 multiple rule

Post by richard »

bobcat2 wrote:Here is yet another problem with any simple multiplier. SS replaces a higher percentage of pre-retirement income the lower your level of pre-retirement income. If the goal of saving for retirement is to have the same standard of living as you had in your latter working years then it follows that higher earners must have higher multipliers, because a lower proportion of their pre-retirement income is replaced by SS.

BobK
Social Security covers a portion of income and income from the portfolio covers the rest. Use a 22x multiplier (or whatever) for portfolio withdrawals, whatever the amount of income to be filled in. For $10,000/year above SS save $220,000, for $200,000 above SS save $4.4 million.

What am I missing?
TareNeko
Posts: 585
Joined: Wed Aug 14, 2013 1:27 pm

Re: 22 multiple rule

Post by TareNeko »

@Richard: If you retire early (say in your 50s), you won't have any SS. So you need to take that into account.
Rodc
Posts: 13601
Joined: Tue Jun 26, 2007 9:46 am

Re: 22 multiple rule

Post by Rodc »

bobcat2 wrote:Here is yet another problem with any simple multiplier. SS replaces a higher percentage of pre-retirement income the lower your level of pre-retirement income. If the goal of saving for retirement is to have the same standard of living as you had in your latter working years then it follows that higher earners must have higher multipliers, because a lower proportion of their pre-retirement income is replaced by SS.

BobK
I always presumed the multiplier was for the income you want to generate from your portfolio and thus naturally is independent of SS income or pension income. It does not matter if you use a generic multiplier (e.g. 25x based on 4% SWR rule) or a better real annuity rate multiplier, to estimate the desired portfolio size. I'm not sure I have ever seen someone use the multiplier as you are suggesting.
We live a world with knowledge of the future markets has less than one significant figure. And people will still and always demand answers to three significant digits.
User avatar
bobcat2
Posts: 6076
Joined: Tue Feb 20, 2007 2:27 pm
Location: just barely Outside the Beltway

Re: 22 multiple rule

Post by bobcat2 »

Different income levels generate different multipliers.

Fred & Ethel earn $45,000/year. To keep the same living standard in retirement they need 2/3's of their pre-retirement income, or $30,000. SS replaces 40% of their pre-retirement income, so they need to self-finance only 27% of their pre-retirement income for retirement.

Desi & Lucy earn $240,000/year. To keep their pre-retirement standard of living they also need 2/3's of their pre-retirement income, or $160,000. SS replaces 29% of their pre-retirement income. They need to self-finance 38%, not 27%, of their pre-retirement income. Their multiplier will be higher than that of Fred & Ethel.

The idea that you only save a multiple above what you receive in SS and DB pension benefits makes no sense. How much of your pre-retirement income is replaced by SS and DB pension income makes a large difference in how much you need to provide for any particular retirement living standard.

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.
Rodc
Posts: 13601
Joined: Tue Jun 26, 2007 9:46 am

Re: 22 multiple rule

Post by Rodc »

The correct multiplier will also be different for every level of current real LT interest rates.
Right. And so if one is many years out from retirement they are going to be wrong with a fixed rule of thumb multiplier and wrong with a more detailed real rate estimated multiplier.

If I am say 20 years out, single male, what is the range of real rates I might expect for a real life annuity?

If I am say 10 years out, single male, what is the range of real rates I might expect for a real life annuity?

Assume I'll be a healthy 65 year old.

Say in those two cases I want to generate a real $20k annual income referenced to 1 January 2014. What is the range of portfolio values that I will need on retirement day to buy a real annuity equal to $20K today? That is, what is the range I better plan on now to achieve my goal.

Is the range all that different from the range given in generic rules of thumb which seem to range from about an optimistic 20x (real rate annuity conversion of 5% or a SWR rule of 5%) to a pessimistic 30x (3.33%) with a more central rate of 25x (the 4% SWR "rule", which I agree completely does not a sound plan make, but does give a written in pencil income approximation).
We live a world with knowledge of the future markets has less than one significant figure. And people will still and always demand answers to three significant digits.
Rodc
Posts: 13601
Joined: Tue Jun 26, 2007 9:46 am

Re: 22 multiple rule

Post by Rodc »

The idea that you only save a multiple above what you receive in SS and DB pension benefits makes no sense.
Of course it makes sense.

If I want an income of $20k above pensions and SS I need the same amount of money regardless of how much I am otherwise going to add to it.

Now if I am used to a grand life style I might decide I want $40K, but that is an entirely different issue. And at any rate unless wanting $40K vs wanting $20K makes me a better or worse investor I get the same multiplier either way, because the multiplier to get the portfolio is applied only to the income I want from the portfolio.

Why would anybody ever multiply times total income from pensions, ss, etc?
We live a world with knowledge of the future markets has less than one significant figure. And people will still and always demand answers to three significant digits.
freebeer
Posts: 2014
Joined: Wed May 02, 2007 8:30 am
Location: Seattle area USA

Re: 22 multiple rule

Post by freebeer »

bobcat2 wrote:Different income levels generate different multipliers.

Fred & Ethel earn $45,000/year. To keep the same living standard in retirement they need 2/3's of their pre-retirement income, or $30,000. SS replaces 40% of their pre-retirement income, so they need to self-finance only 27% of their pre-retirement income for retirement.

Desi & Lucy earn $240,000/year. To keep their pre-retirement standard of living they also need 2/3's of their pre-retirement income, or $160,000. SS replaces 29% of their pre-retirement income. They need to self-finance 38%, not 27%, of their pre-retirement income. Their multiplier will be higher than that of Fred & Ethel.

The idea that you only save a multiple above what you receive in SS and DB pension benefits makes no sense. How much of your pre-retirement income is replaced by SS and DB pension income makes a large difference in how much you need to provide for any particular retirement living standard.

BobK
Bob & Ethel will get $18K/year from SS so need an extra $12K/year, which they can achieve by saving 25x that amount = $300K portfolio (4% withdrawals assumed). Desi & Lucy will get $69,600 from SS so need an extra $90,400/year, which they can achieve by saving 25x that amount = $2,260,000 (4% withdrawals assumed). What makes no sense about that (as a rough SWAG)?
User avatar
bobcat2
Posts: 6076
Joined: Tue Feb 20, 2007 2:27 pm
Location: just barely Outside the Beltway

Re: 22 multiple rule

Post by bobcat2 »

Rodc wrote:
bobcat2 wrote:Here is yet another problem with any simple multiplier. SS replaces a higher percentage of pre-retirement income the lower your level of pre-retirement income. If the goal of saving for retirement is to have the same standard of living as you had in your latter working years then it follows that higher earners must have higher multipliers, because a lower proportion of their pre-retirement income is replaced by SS.

BobK
I always presumed the multiplier was for the income you want to generate from your portfolio and thus naturally is independent of SS income or pension income. It does not matter if you use a generic multiplier (e.g. 25x based on 4% SWR rule) or a better real annuity rate multiplier, to estimate the desired portfolio size. I'm not sure I have ever seen someone use the multiplier as you are suggesting.
If you a retirement goal, such as having the same living standard in retirement as in your latter working years. then how much income you receive from pensions and SS has a huge impact on how much you need to save to reach that goal. Why wouldn't it. Someone earns $120,000/year and in saving and investing for retirement it makes no difference to them whether SS and pension income together are $66,000 or $33,000. Why would that be?

The goal is how much income you want in retirement. How much of that you have to come up depends on how much you receive from SS and pensions. To do it any other way says that it makes no difference when I take SS, because I only care in the amount I provide above SS income. So whether I take SS at age 62 or wait to age 70 for a higher benefit makes no difference because I only save for the amount above SS, regardless of what the SS amount is.

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.
Rodc
Posts: 13601
Joined: Tue Jun 26, 2007 9:46 am

Re: 22 multiple rule

Post by Rodc »

If you a retirement goal, such as having the same living standard in retirement as in your latter working years. then how much income you receive from pensions and SS has a huge impact on how much you need to save to reach that goal.
You are not reading what anyone (myself, richard or freebeer) is saying to you. You are not addressing the issue anyone else is asking about.

You have simply made up a different issue. It is a valid issue, it is just not what we are talking about in this thread which is how to go from desiring $X/year to how large the portfolio needs to be to generate that $X/year.

You are arguing some people will want a bigger or smaller X which is not germane to this thread.
We live a world with knowledge of the future markets has less than one significant figure. And people will still and always demand answers to three significant digits.
freebeer
Posts: 2014
Joined: Wed May 02, 2007 8:30 am
Location: Seattle area USA

Re: 22 multiple rule

Post by freebeer »

Rodc wrote:...
You have simply made up a different issue. It is a valid issue, it is just not what we are talking about in this thread which is how to go from desiring $X/year to how large the portfolio needs to be to generate that $X/year...
Rodc, in fairness I think Bob''s germane point is that it hasn't been clear throughout the thread that desiring $X/year means **excluding SS/pension income** which is necessary if the idea to save a multiplier of $X is going to make any sense... some have talked about $X goal in terms of a say a % of working salary ... if that's your $X you almost certainly need to calculate a $Y (income replaceent goal) and then you can swag a portfolio amount based on that $Y.
Rodc
Posts: 13601
Joined: Tue Jun 26, 2007 9:46 am

Re: 22 multiple rule

Post by Rodc »

freebeer wrote:
Rodc wrote:...
You have simply made up a different issue. It is a valid issue, it is just not what we are talking about in this thread which is how to go from desiring $X/year to how large the portfolio needs to be to generate that $X/year...
Rodc, in fairness I think Bob''s germane point is that it hasn't been clear throughout the thread that desiring $X/year means **excluding SS/pension income** which is necessary if the idea to save a multiplier of $X is going to make any sense... some have talked about $X goal in terms of a say a % of working salary ... if that's your $X you almost certainly need to calculate a $Y (income replaceent goal) and then you can swag a portfolio amount based on that $Y.
Ok, let's be clear then. :)

A multiplier (fixed, rule of thumb, real annuity based, or otherwise) of income desired to estimate portfolio size required only makes sense if we are talking about a portfolio and the income to be generated from that portfolio.

Better? :)
We live a world with knowledge of the future markets has less than one significant figure. And people will still and always demand answers to three significant digits.
YDNAL
Posts: 13774
Joined: Tue Apr 10, 2007 4:04 pm
Location: Biscayne Bay

Re: 22 multiple rule

Post by YDNAL »

bobcat2 wrote:Fred & Ethel earn $45,000/year. To keep the same living standard in retirement they need 2/3's of their pre-retirement income, or $30,000. SS replaces 40% of their pre-retirement income, so they need to self-finance only 27% of their pre-retirement income for retirement.

Desi & Lucy earn $240,000/year. To keep their pre-retirement standard of living they also need 2/3's of their pre-retirement income, or $160,000. SS replaces 29% of their pre-retirement income. They need to self-finance 38%, not 27%, of their pre-retirement income. Their multiplier will be higher than that of Fred & Ethel.
There is a glaring definition (nothing new) problem here.

Code: Select all

                F&E	      D&L
Income	      45,000	  240,000 
2/3 Income	  30,000	  160,000 
SS	   (40%) (18,000)	 (69,600) (29%)
Net	         12,000 	  90,400 
Multiplier	      22 	      22
Savings	    264,000   1,988,800  <-- OP's question
I say to-may-to, you say to-mah-to. :D
Last edited by YDNAL on Tue Feb 04, 2014 2:30 pm, edited 1 time in total.
Landy | Be yourself, everyone else is already taken -- Oscar Wilde
User avatar
bobcat2
Posts: 6076
Joined: Tue Feb 20, 2007 2:27 pm
Location: just barely Outside the Beltway

Re: 22 multiple rule

Post by bobcat2 »

It is certainly true that once you factor in SS you apply rules for the amount above that. But a rule like 22x is very quiet about the SS part - mentioned only passing in parens by Zweig. It sounds to many people that they need to save 22x of their targeted retirement income. What they need to save for is an amount above SS and DB pension income. This can be a relativity small amount if they need 64% of their pre-retirement income and SS supplies 40% and DB pension 15%.

How much you will need to save and invest depends on a host of factors including your gender, marital status, intended retirement age, SS and DB pension replacement rates, longevity at retirement, real interest rates, and other factors. It is imprudent to pick a fixed multiplier and apply it to all people. A well off couple retiring at age 63 will have a much different multiplier than a single less affluent male retiring in the same year at age 67 and his multiplier will be applied to a smaller percentage of income to be replaced. How these multipliers play out will also be affected by longevity in the retirement year as well as real interest rate levels.

The multiplier in any individual situation is going to change over time as well as the amount to be covered changing. It makes little sense to stick to a fixed number that represents some sort of average household. A household that is apparently a combination of single male, single female, and married couple. Not only that but all households retire at the same age, real interest rates are always average, and longevity and income levels never change. Instead update the number over time as the inevitable changes occur to your particular situation. It seems to me the easiest way to update the plan over time is by pricing real annuities which explicitly take into account changes in such things as longevity, real interest rates, and marital status.

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
midareff
Posts: 7711
Joined: Mon Nov 29, 2010 9:43 am
Location: Biscayne Bay, South Florida

Re: 22 multiple rule

Post by midareff »

Cb wrote:
midareff wrote:...AFAIK, you can follow the Pied Piper's and math wizards of 4% and higher inflation indexed withdrawals, but do so at your own risk understanding you are making the later golden year's cat food and life under the bridge bet...
This sort of thing turns up in most SWR threads. I wonder what % of the retired population lives under bridges, subsisting on a diet of cat food, and of that population, what % had entered their retirements with an investment portfolio of ~22 times their annual expenses.
If you want to go in with an inflation indexed WR of 4.55% healthy, with a lineage of long life, and a 30 year retirement in mind for you and your other half, .. go for it. I'm hoping it works out for you, it didn't for Bengen, Otar, Phau and a ton of others. .. but I'm sure you know more than they did.
User avatar
bobcat2
Posts: 6076
Joined: Tue Feb 20, 2007 2:27 pm
Location: just barely Outside the Beltway

Re: 22 multiple rule

Post by bobcat2 »

IMO you need a way to calculate directly the level of assets you need to fund retirement given your unique situation. You don't want to get there indirectly by calculating some average ratio of assets to income regardless of your individual circumstances concerning gender, marital status, age at retirement, nor do you want to ignore states of the world such as longevity and real interest rates. Costing out retirement income by pricing a real annuity gets directly to the level of assets taking into account longevity, real interest rates, age at retirement, gender, and marital status explicitly.

Yes you can take 22x or some other x ratio and fudge it for gender, fudge it for marital status, fudge it for age at retirement, fudge it for longevity, etc etc, but at the end of the day you are going to have a highly fudged estimate, which may be far away in either direction from the number you need to reach your retirement goal.

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.
Rodc
Posts: 13601
Joined: Tue Jun 26, 2007 9:46 am

Re: 22 multiple rule

Post by Rodc »

midareff wrote:
Cb wrote:
midareff wrote:...AFAIK, you can follow the Pied Piper's and math wizards of 4% and higher inflation indexed withdrawals, but do so at your own risk understanding you are making the later golden year's cat food and life under the bridge bet...
This sort of thing turns up in most SWR threads. I wonder what % of the retired population lives under bridges, subsisting on a diet of cat food, and of that population, what % had entered their retirements with an investment portfolio of ~22 times their annual expenses.
If you want to go in with an inflation indexed WR of 4.55% healthy, with a lineage of long life, and a 30 year retirement in mind for you and your other half, .. go for it. I'm hoping it works out for you, it didn't for Bengen, Otar, Phau and a ton of others. .. but I'm sure you know more than they did.
I don't think any real person has ever actually used a 4% plus inflation type rule for anything other than estimating potential income. That it is not a plan anyone should implement, and likely extremely few ever have. Certainly Bengen, Otar, Phau have never actually used it other than as a backtesting example (so they are not eating cat food as a result).

So it is unlikely it has ever led anyone to eat cat food.

Real people (virtually) always react to a serious down market or portfolio balance dropping more than planned by cutting back on expenses, unless already living in deep poverty.
We live a world with knowledge of the future markets has less than one significant figure. And people will still and always demand answers to three significant digits.
Rodc
Posts: 13601
Joined: Tue Jun 26, 2007 9:46 am

Re: 22 multiple rule

Post by Rodc »

but at the end of the day you are going to have a highly fudged estimate
I guess my question still stands.

It seems to me that the game of guessing how big a portfolio you "need" is largely a game for people a long way from retirement. Close to retirement or in retirement, you know pretty well what you have and you don't have too much you can do about it (other than put off retirement), so you live with what you have.

So I'll comment on what I see as the situation for people a long way from retirement, say 20 years or more.

Trying to guess future, possibly deep in the future, real rates or real annuity costs is just a different fudge.

How much practical difference is there? After all you also have to fudge future investment returns to get the amount you have to spend on an annuity, which may be such a big fudge that it swamps any attempt at precision. Not to mention if you are 20 or more years from retirement even the desire income is often little more than a wild guess.

If you are 20, 30, very likely 40 the uncertainty in the future is such that the best you can do is a rough estimate for any of these factors, so unclear to me that you can do a lot better than pick your favorite multiplier, recognizing you could easily be off by +/-5% in the multiplier, could easily be off by way more than that in the desired income.

(FWIW: I'm not all that far from retirement, so things are sort of starting to come into focus as far as income desires and expected portfolio size, and I have switched to estimating potential portfolio income using real annuities so I am not at all opposed to the idea. Of course I might not choose to buy such a thing, and not sure if I do when I would do so.)
We live a world with knowledge of the future markets has less than one significant figure. And people will still and always demand answers to three significant digits.
User avatar
bobcat2
Posts: 6076
Joined: Tue Feb 20, 2007 2:27 pm
Location: just barely Outside the Beltway

Re: 22 multiple rule

Post by bobcat2 »

I don't believe looking at the size of assets needed at retirement is very useful much more than 20 years from retirement. The income goal itself is too fuzzy 30 years out. Between 20-5 years it seems very useful as you can adjust savings/year, retirement age, and risk level in order to keep on track to reaching your retirement income goals even as the goals themselves may evolve.

Less than five years from expected retirement you want to be looking at your AA and whether it is appropriate in the tradeoff between reaching goals and risk level. You want to have all of your assets dedicated to your safe floor income invested safely, a combination of safe assets and dedicated future contributions, and you want to have your risk level for aspirational income appropriate for reaching your goal. The temptation to raise the aspirational level of income by taking on more risk has to be balanced with the lower probability of having your targeted safe floor of income met, as well as falling further from your aspirational income level if the risk shows up.

OTOH if you are falling short of your aspirational income goal within five years of retirement you have hard choices to make regarding savings per year, retirement date, and higher risk/return tradeoff. In addition you would need to perhaps reconsider your SS strategy. downsizing the house or reverse mortgage, and how much and when to annuitize. Or, do you simply accept lower retirement income than you had planned for?

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
Kevin M
Posts: 15789
Joined: Mon Jun 29, 2009 3:24 pm
Contact:

Re: 22 multiple rule

Post by Kevin M »

investment rube wrote:I read an article by Jason Zweig in the WSJ today that stated, "To be assured of having enough money to fund a comfortable retirement, you should save a total of 22 times the annual income you want to earn when you retire."
First, wondering the Boglehead view of this statement..
Reading the linked thread in post above by matjen explains the idea behind the rule, which Wade has summarized succinctly:
wade wrote:I believe that the 22x number is based on the specific strategy of retiring at 65, buying a 20-year TIPS bond ladder, and buying a deferred income annuity which begins payments at age 85. This allows for a higher than 4% withdrawal rate, though with a great deal of inflation risk related to the deferred income annuity.
So 20X for the TIPS ladder and 2X for the deferred annuity.

Prominent BH author William Bernstein also recommends a TIPS ladder as a core of one's Liability Matching Portfolio (LMP), along with delaying SS payments until age 70. He is concerned about the systemic risk of annuities. Aside from systemic risk, there is huge inflation risk with a non-inflation-adjusted annuity, as Wade mentions.
investment rube wrote:Also, would this be impacted by age when you retire? 65 vs, say, 50?
Clearly the answer is yes. 20X in TIPS gives you 20 years, and at age 70 (50+20) you are subject to much more longevity/inflation risk than at age 85 (65+20), so counting on a non-inflation-adjusted deferred annuity for the remaining years is much riskier if your expected lifetime is much longer.

Kevin
If I make a calculation error, #Cruncher probably will let me know.
freebeer
Posts: 2014
Joined: Wed May 02, 2007 8:30 am
Location: Seattle area USA

Re: 22 multiple rule

Post by freebeer »

Rodc wrote:...
It seems to me that the game of guessing how big a portfolio you "need" is largely a game for people a long way from retirement. Close to retirement or in retirement, you know pretty well what you have and you don't have too much you can do about it (other than put off retirement), so you live with what you have.

So I'll comment on what I see as the situation for people a long way from retirement, say 20 years or more...
I disagree (as someone who's anywhere from 0 to infinity years from retirement). First of all, if you're behind, close to retirement as in 5 to 10 years is still plenty of time for "catch up" and compounding effect. More importantly, if you start saving 25% of your gross salary you are perforce not going to be spending more than say 55% so you will necessarily start getting some LBYM discipline, maybe even LBYM "religion". And then you can consider a possible further reduction in retirement spending from that base of while-working spending. Lastly, many people have highly variable end-of-career earnings. Some keep climbing that BigCo ladder, others stay where they are, still others downsize. It was a hard look at our status towards "FIRE readiness" that gave me the comfort to change to a position at half the compensation I previously had (but twice the intangible rewards and 4x the flexibility). OTOH if we had been behind I would have definitely not made that move.
Post Reply