## Safe Withdrawal Rate from SPIA???

Discuss all general (i.e. non-personal) investing questions and issues, investing news, and theory.
pezblanco
Posts: 477
Joined: Thu Sep 12, 2013 8:02 pm

### Safe Withdrawal Rate from SPIA???

From time to time BH's consider the purchase of an SPIA as a do it yourself pension. You can buy these with consumer price index increases built in but they are quite a bit more expensive. The other strategy is to purchase one without such a CPI rider and then when inflation has eaten away it's value too much, purchase another one. Another strategy might be to purchase one and spend part of it each year and invest the rest in a stock fund. This last strategy is what I'm considering here.

I have seen a couple of threads of where people have asked what is the "sustainable expenditure" that one can expect from an SPIA. The idea is that you purchase an SPIA that supplies you with say 1000 dollars per year. At year one, you spend an initial amount of say 800 dollars. You take the excess 200 dollars and you invest it in the S&P. This is your SPIA excess investment fund.

At year two you spend 800*(1+CPI rate of inflation). If that amount exceeds 1000 dollars plus whatever you have in the investment fund, you have a FAIL. If we don't have a FAIL in the first year, we now continue this process for the term period. I assumed 30 years. What what be the probability of FAIL over this period?

To get a handle on this, I took the last 51 years of the CPI inflation rate and the last 51 years of returns from the S&P from the DFA Data Matrix. To simulate each years investment and CPI, I sampled randomly from those data.

The results are:

Starting Inital Amount Probability of Failure in 30 years

800................................................96
700................................................58
600................................................13.5
500................................................1

I hope this is of some interest. I would also be interested in other approaches to how to go about evaluating the long term value of an SPIA as part of a retirement portfolio. I was surprised that in order to maintain a safe withdrawal rate, you had to invest such a large portion of the SPIA income. Perhaps those CPI riders are not as expensive as is commonly believed. Also, it might be that my definition of FAIL is too stringent .... what we might have is a sequence of return risk going on ... if you get large inflation and bad stock returns early on before you build up the "investment fund", then you FAIL. Once you build up the fund, your rate of expending could very possibly go up .... I didn't investigate this.

ResearchMed
Posts: 7473
Joined: Fri Dec 26, 2008 11:25 pm

### Re: Safe Withdrawal Rate from SPIA???

pezblanco wrote:
Wed Dec 06, 2017 10:36 am
From time to time BH's consider the purchase of an SPIA as a do it yourself pension. You can buy these with consumer price index increases built in but they are quite a bit more expensive. The other strategy is to purchase one without such a CPI rider and then when inflation has eaten away it's value too much, purchase another one. Another strategy might be to purchase one and spend part of it each year and invest the rest in a stock fund. This last strategy is what I'm considering here.

I have seen a couple of threads of where people have asked what is the "sustainable expenditure" that one can expect from an SPIA. The idea is that you purchase an SPIA that supplies you with say 1000 dollars per year. At year one, you spend an initial amount of say 800 dollars. You take the excess 200 dollars and you invest it in the S&P. This is your SPIA excess investment fund.

At year two you spend 800*(1+CPI rate of inflation). If that amount exceeds 1000 dollars plus whatever you have in the investment fund, you have a FAIL. If we don't have a FAIL in the first year, we now continue this process for the term period. I assumed 30 years. What what be the probability of FAIL over this period?

To get a handle on this, I took the last 51 years of the CPI inflation rate and the last 51 years of returns from the S&P from the DFA Data Matrix. To simulate each years investment and CPI, I sampled randomly from those data.

The results are:

Starting Inital Amount Probability of Failure in 30 years

800................................................96
700................................................58
600................................................13.5
500................................................1

I hope this is of some interest. I would also be interested in other approaches to how to go about evaluating the long term value of an SPIA as part of a retirement portfolio. I was surprised that in order to maintain a safe withdrawal rate, you had to invest such a large portion of the SPIA income. Perhaps those CPI riders are not as expensive as is commonly believed. Also, it might be that my definition of FAIL is too stringent .... what we might have is a sequence of return risk going on ... if you get large inflation and bad stock returns early on before you build up the "investment fund", then you FAIL. Once you build up the fund, your rate of expending could very possibly go up .... I didn't investigate this.
I think most people think of a long term SPIA to be used as a "lifetime annuity".
There's no "safe withdrawal rate" for an SPIA lifetime annuity.
It would pay out for the entire lifetime, and the entire point is to avoid the juggling you are creating.

I'm not sure why an "annuity" would be needed in your case, instead of, say, laddered bonds or some other fixed income, as you know the time frame exactly.

Have you considered how a life annuity would work for your needs?

RM
This signature is a placebo. You are in the control group.

moghopper
Posts: 134
Joined: Fri Oct 18, 2013 10:07 am

### Re: Safe Withdrawal Rate from SPIA???

Many studies indicate that spending decreases over time.

(ref: https://www.forbes.com/sites/wadepfau/2 ... -over-time)

Why assume a need for a CPI adjustment on the SPIA, or a "man-made" one in your scenario? I realize that is the thing to do when planning, but it appears not to jive with reality.

Sandtrap
Posts: 5379
Joined: Sat Nov 26, 2016 6:32 pm
Location: Hawaii😀 Northern AZ.😳 Retired.

### Re: Safe Withdrawal Rate from SPIA???

IMHO an SPIA is an insurance product (lifetime annuity/or fixed term) that can supplement one's income stream through retirement.
One of its positives is providing a consistent return where one's portfolio, and other income streams, are not enough to provide for anticipated expenses. To that end, one might ladder in each policy perhaps from age 65 or 70 and then progressively evaluate and purchase as needed every 5 or more years thereafter. There are other strategies.

Your approach is interesting and very unlike most past threads on SPIA's. I don't understand the advantage of it as it doesn't seem to fit into the primary purpose of an annuity. A large part of the initial proceeds are a return on one's money. Please explain.
IMHO purchasing a larger policy than is needed and attempting to invest the unused income seems counterproductive.

Immediate annuity.com for estimates
https://www.immediateannuities.com/a/an ... gL5L_D_BwE

mahalo
j

pezblanco
Posts: 477
Joined: Thu Sep 12, 2013 8:02 pm

### Re: Safe Withdrawal Rate from SPIA???

Thank you for your comments!

Yes, agreed .... the usual way to deal with inflation risk is to ladder SPIAs ... this has the advantage of "doing it when you need it". Yes, I would guess that spending goes down with age (except perhaps at the very end with end-of-life-care and terminal care medical expenses).

However, this isn't just a SPIA problem is it? There are many "defined benefit" pensions that have no CPI increase built into them. (I actually have one.) Say, that I'm being given 1000 dollars per year out of my pension, what sort of continuous spending in todays dollars can I expect from that cash flow? The answer that I'm being given from this particular (probably too conservative) approach is something like 500 to 600 today's dollars per year can be expended for the rest of my life.

pezblanco
Posts: 477
Joined: Thu Sep 12, 2013 8:02 pm

### Re: Safe Withdrawal Rate from SPIA???

ResearchMed wrote:
Wed Dec 06, 2017 10:45 am
pezblanco wrote:
Wed Dec 06, 2017 10:36 am
From time to time BH's consider the purchase of an SPIA as a do it yourself pension. You can buy these with consumer price index increases built in but they are quite a bit more expensive. The other strategy is to purchase one without such a CPI rider and then when inflation has eaten away it's value too much, purchase another one. Another strategy might be to purchase one and spend part of it each year and invest the rest in a stock fund. This last strategy is what I'm considering here.

I have seen a couple of threads of where people have asked what is the "sustainable expenditure" that one can expect from an SPIA. The idea is that you purchase an SPIA that supplies you with say 1000 dollars per year. At year one, you spend an initial amount of say 800 dollars. You take the excess 200 dollars and you invest it in the S&P. This is your SPIA excess investment fund.

At year two you spend 800*(1+CPI rate of inflation). If that amount exceeds 1000 dollars plus whatever you have in the investment fund, you have a FAIL. If we don't have a FAIL in the first year, we now continue this process for the term period. I assumed 30 years. What what be the probability of FAIL over this period?

To get a handle on this, I took the last 51 years of the CPI inflation rate and the last 51 years of returns from the S&P from the DFA Data Matrix. To simulate each years investment and CPI, I sampled randomly from those data.

The results are:

Starting Inital Amount Probability of Failure in 30 years

800................................................96
700................................................58
600................................................13.5
500................................................1

I hope this is of some interest. I would also be interested in other approaches to how to go about evaluating the long term value of an SPIA as part of a retirement portfolio. I was surprised that in order to maintain a safe withdrawal rate, you had to invest such a large portion of the SPIA income. Perhaps those CPI riders are not as expensive as is commonly believed. Also, it might be that my definition of FAIL is too stringent .... what we might have is a sequence of return risk going on ... if you get large inflation and bad stock returns early on before you build up the "investment fund", then you FAIL. Once you build up the fund, your rate of expending could very possibly go up .... I didn't investigate this.
I think most people think of a long term SPIA to be used as a "lifetime annuity".
There's no "safe withdrawal rate" for an SPIA lifetime annuity.
It would pay out for the entire lifetime, and the entire point is to avoid the juggling you are creating.

I'm not sure why an "annuity" would be needed in your case, instead of, say, laddered bonds or some other fixed income, as you know the time frame exactly.

Have you considered how a life annuity would work for your needs?

RM
RM, agreed it's just a cash stream. I'm not sure I'm in complete agreement that the whole point is to avoid the juggling. It just seemed to me that there might be interest in how one goes about protecting that cash stream against inflation. The typical time frame for evaluation of retirement strategies is 30 years ... for example when developing the 4 percent rule. Bond ladders are typically (unless they are TIPS) are not completely inflation protected either, are they?

As I indicated, some BHs already have life annuities that are not inflation protected (myself!). How does one go about evaluating what the real cash flow to be expected from that pension/life annuity really might be.

ResearchMed
Posts: 7473
Joined: Fri Dec 26, 2008 11:25 pm

### Re: Safe Withdrawal Rate from SPIA???

pezblanco wrote:
Wed Dec 06, 2017 11:33 am
ResearchMed wrote:
Wed Dec 06, 2017 10:45 am
pezblanco wrote:
Wed Dec 06, 2017 10:36 am
From time to time BH's consider the purchase of an SPIA as a do it yourself pension. You can buy these with consumer price index increases built in but they are quite a bit more expensive. The other strategy is to purchase one without such a CPI rider and then when inflation has eaten away it's value too much, purchase another one. Another strategy might be to purchase one and spend part of it each year and invest the rest in a stock fund. This last strategy is what I'm considering here.

I have seen a couple of threads of where people have asked what is the "sustainable expenditure" that one can expect from an SPIA. The idea is that you purchase an SPIA that supplies you with say 1000 dollars per year. At year one, you spend an initial amount of say 800 dollars. You take the excess 200 dollars and you invest it in the S&P. This is your SPIA excess investment fund.

At year two you spend 800*(1+CPI rate of inflation). If that amount exceeds 1000 dollars plus whatever you have in the investment fund, you have a FAIL. If we don't have a FAIL in the first year, we now continue this process for the term period. I assumed 30 years. What what be the probability of FAIL over this period?

To get a handle on this, I took the last 51 years of the CPI inflation rate and the last 51 years of returns from the S&P from the DFA Data Matrix. To simulate each years investment and CPI, I sampled randomly from those data.

The results are:

Starting Inital Amount Probability of Failure in 30 years

800................................................96
700................................................58
600................................................13.5
500................................................1

I hope this is of some interest. I would also be interested in other approaches to how to go about evaluating the long term value of an SPIA as part of a retirement portfolio. I was surprised that in order to maintain a safe withdrawal rate, you had to invest such a large portion of the SPIA income. Perhaps those CPI riders are not as expensive as is commonly believed. Also, it might be that my definition of FAIL is too stringent .... what we might have is a sequence of return risk going on ... if you get large inflation and bad stock returns early on before you build up the "investment fund", then you FAIL. Once you build up the fund, your rate of expending could very possibly go up .... I didn't investigate this.
I think most people think of a long term SPIA to be used as a "lifetime annuity".
There's no "safe withdrawal rate" for an SPIA lifetime annuity.
It would pay out for the entire lifetime, and the entire point is to avoid the juggling you are creating.

I'm not sure why an "annuity" would be needed in your case, instead of, say, laddered bonds or some other fixed income, as you know the time frame exactly.

Have you considered how a life annuity would work for your needs?

RM
RM, agreed it's just a cash stream. I'm not sure I'm in complete agreement that the whole point is to avoid the juggling. It just seemed to me that there might be interest in how one goes about protecting that cash stream against inflation. The typical time frame for evaluation of retirement strategies is 30 years ... for example when developing the 4 percent rule. Bond ladders are typically (unless they are TIPS) are not completely inflation protected either, are they?

As I indicated, some BHs already have life annuities that are not inflation protected (myself!). How does one go about evaluating what the real cash flow to be expected from that pension/life annuity really might be.
First, I wasn't sure if you were using SPIA for "lifetime" income or not, so I may have misunderstood.

But assuming you are, yes there is the inflation "problem". Not much now, be we sure remember times when inflation was a real concern.
Using a hypothetical length of time isn't a good reflection of the realities of a true life annuity. The entire point of a lifetime SPIA is that it is for a "lifetime", however long or short.

When you write "How does one go about evaluating what the real cash flow to be expected from that pension/life annuity really might be." wouldn't that be the monthly/annual payments times the average life expectancy?

We are approaching the time to start some of this.
We'll probably do a bit of laddering of SPIA's. One larger one, then perhaps a few much smaller over time, for the inflation coverage.
The money for those might come from what we "used to spend" on things like travel, that get cut back as we get much older.
We might also get some small Deferred Single Payment Annuities, given that they'd be so much less expensive. Problem is any inflation prior to their start, so we might just do one of those, as a sort of diversification.

It's tricky, of course, given that one wouldn't want to annuitize *everything*, and most people aren't considering an inflation adjusted SPIA.

We are also speaking with a CCRC type place, where we'd probably go after selling our house. (MIL is there now, and we are incredibly impressed, with a good "inside" view.) We are making sure that we don't screw it up if they need a lump sum vs annuity income. Turns out, they'll consider either, as equivalents, with at least a minimum required lump, but the rest could be lump or annuity income.

RM
This signature is a placebo. You are in the control group.

Swelfie
Posts: 200
Joined: Mon Mar 14, 2016 12:54 am

### Re: Safe Withdrawal Rate from SPIA???

ResearchMed wrote:
Wed Dec 06, 2017 11:47 am
When you write "How does one go about evaluating what the real cash flow to be expected from that pension/life annuity really might be." wouldn't that be the monthly/annual payments times the average life expectancy?
No, because mortality isn't a normal distribution. To get this number you need a mortality table then adjust each future years chance of perishing by forward looking rates of increased life expectancy. There is a professional association that publishes these. It's a pretty hairy spreadsheet to build though, as I've done it. I use it to measure the future value of social security which started out as an experimental curiosity and ended up a cool toy. Then multiply each year's return by the odds of being alive to receive it. Most mortality tables stop at 120 years old and call it 0% odds since the future value of those payments becomes minute.

ResearchMed
Posts: 7473
Joined: Fri Dec 26, 2008 11:25 pm

### Re: Safe Withdrawal Rate from SPIA???

Swelfie wrote:
Wed Dec 06, 2017 12:15 pm
ResearchMed wrote:
Wed Dec 06, 2017 11:47 am
When you write "How does one go about evaluating what the real cash flow to be expected from that pension/life annuity really might be." wouldn't that be the monthly/annual payments times the average life expectancy?
No, because mortality isn't a normal distribution. To get this number you need a mortality table then adjust each future years chance of perishing by forward looking rates of increased life expectancy. There is a professional association that publishes these. It's a pretty hairy spreadsheet to build though, as I've done it. I use it to measure the future value of social security which started out as an experimental curiosity and ended up a cool toy. Then multiply each year's return by the odds of being alive to receive it. Most mortality tables stop at 120 years old and call it 0% odds since the future value of those payments becomes minute.
Right.
And at any given age (could be adjusted per month or week, or however finely those actuarial tables are done), one would use these to get an average, like a weighted average (weighted by life expectancy for each age interval/point).

IF it were to be a normal distribution, which is obviously not the case at all, then it would be easy... just using the simple mean age times payout.
[For simplicity, I didn't get into why the arithmetic mean should *not* be used. But "expected values", and similar calculations, are certainly used in non-normal distributions...]

Presumably that's what the insurers are doing to get estimates of what the expected (aka "average") payout would be at each starting age, and then add in their desired profit plus some fudge/safety factor, etc.

This is very simplified, as you no doubt understand given your background.

My understanding of the original question about "how could some sort of value be assigned" is that this would provide the best estimate, even though it is obviously simplifying a huge amount of actuarial research (and constant updating, as mortality changes, for better or worse, etc., for each age....).

I was trying to give some sense of "how it could be figured", without getting into actuarial details.
Apologies if I am not explaining this generality well for either experts or non-experts.

RM
This signature is a placebo. You are in the control group.

randomguy
Posts: 6504
Joined: Wed Sep 17, 2014 9:00 am

### Re: Safe Withdrawal Rate from SPIA???

pezblanco wrote:
Wed Dec 06, 2017 11:24 am
Thank you for your comments!

Yes, agreed .... the usual way to deal with inflation risk is to ladder SPIAs ... this has the advantage of "doing it when you need it". Yes, I would guess that spending goes down with age (except perhaps at the very end with end-of-life-care and terminal care medical expenses).

And it has the disadvantage of exposing you to market risk. In year 1 you can take your million dollars and buy an inflation adjusted SPIA that will give you 4% for the rest of your life. If you sink half into a nonInflation adjusted SPIA and the other half in the market with plans to buy another SPIA in 10 years, what happens if at that time you only have enough money to give you say a 3% SWR? Most of the time things work out. But those are also the cases where you didn't need to buy a SPIA to begin with.

It always seemed like to me that for a 65 year old buying the inflation adjusted SPIA works out. The lower payouts seem brutual at first but since you are worried about the risk to living 90+, you might as well have protection for when you get there. the 80 year old on the other hand is facing a lot fewer inflation years and might want to gamble on the nominal one.

If you are willing to take on market risk, you would want to compare buying stocks with annuity payouts to buying annuity payouts with stocks (i.e. imagine every year you buy an annuity with the dividends). I imagine you get roughly the same results but with different worse years.

ResearchMed
Posts: 7473
Joined: Fri Dec 26, 2008 11:25 pm

### Re: Safe Withdrawal Rate from SPIA???

randomguy wrote:
Wed Dec 06, 2017 12:58 pm
pezblanco wrote:
Wed Dec 06, 2017 11:24 am
Thank you for your comments!

Yes, agreed .... the usual way to deal with inflation risk is to ladder SPIAs ... this has the advantage of "doing it when you need it". Yes, I would guess that spending goes down with age (except perhaps at the very end with end-of-life-care and terminal care medical expenses).

And it has the disadvantage of exposing you to market risk. In year 1 you can take your million dollars and buy an inflation adjusted SPIA that will give you 4% for the rest of your life. If you sink half into a nonInflation adjusted SPIA and the other half in the market with plans to buy another SPIA in 10 years, what happens if at that time you only have enough money to give you say a 3% SWR? Most of the time things work out. But those are also the cases where you didn't need to buy a SPIA to begin with.

It always seemed like to me that for a 65 year old buying the inflation adjusted SPIA works out. The lower payouts seem brutual at first but since you are worried about the risk to living 90+, you might as well have protection for when you get there. the 80 year old on the other hand is facing a lot fewer inflation years and might want to gamble on the nominal one.

If you are willing to take on market risk, you would want to compare buying stocks with annuity payouts to buying annuity payouts with stocks (i.e. imagine every year you buy an annuity with the dividends). I imagine you get roughly the same results but with different worse years.
Are there any comparisons from the past (future may/will differ, etc.) about how an inflation adjusted SPIA really did/did not keep up with the actual inflation as it occurred going forward?
It would be interesting to know if it *ever* came even close, etc.

RM
This signature is a placebo. You are in the control group.

randomguy
Posts: 6504
Joined: Wed Sep 17, 2014 9:00 am

### Re: Safe Withdrawal Rate from SPIA???

ResearchMed wrote:
Wed Dec 06, 2017 1:02 pm

Are there any comparisons from the past (future may/will differ, etc.) about how an inflation adjusted SPIA really did/did not keep up with the actual inflation as it occurred going forward?
It would be interesting to know if it *ever* came even close, etc.

RM
Obviously it depends on what you buy. Buy one that matches CPI, that is what you get. Inflation can be pretty personal as we all have different exposures to housing, gas, medical, food and so on.

ResearchMed
Posts: 7473
Joined: Fri Dec 26, 2008 11:25 pm

### Re: Safe Withdrawal Rate from SPIA???

randomguy wrote:
Wed Dec 06, 2017 1:13 pm
ResearchMed wrote:
Wed Dec 06, 2017 1:02 pm

Are there any comparisons from the past (future may/will differ, etc.) about how an inflation adjusted SPIA really did/did not keep up with the actual inflation as it occurred going forward?
It would be interesting to know if it *ever* came even close, etc.

RM
Obviously it depends on what you buy. Buy one that matches CPI, that is what you get. Inflation can be pretty personal as we all have different exposures to housing, gas, medical, food and so on.
I'm asking if the inflation adjusted SPIA's "do what they are supposed to do", in real life, going forward, at least for past years.

I've never known anyone who had one, other than SS, and I guess part of my question might include how commercial inflation adjusted SPIA's actually compared with SS benefits, counting both starting the same year.
(But with many years of data, over many past year starting points.)

For example, are there unexpected "gotchas", like with some "equity indexed annuities" where there are caps on returns that are not usually understood/expected, etc. (I understand these are entirely different, but it's the "unexpected gotchas" that is part of what I'm wondering about.)

How one needs/uses the money isn't really different from the actual living costs each of us chooses or needs, regardless of income source.

RM
This signature is a placebo. You are in the control group.

pezblanco
Posts: 477
Joined: Thu Sep 12, 2013 8:02 pm

### Re: Safe Withdrawal Rate from SPIA???

Interesting discussion. Just a data point that I got from immediateannuties.com For a 65 year old male living in my state:

For a \$1,000,000 payment, they will give \$3759.00 per month a 4.51% payout rate with COLA.
they will give \$5512.00 per month a 6.61% payout rate without COLA.

So a 68% discoount for providing a COLA. This sort of makes sense to me .... I was finding from simulation that 50 to 60 percent would be the discount but I thought that my method of deciding a FAIL was too conservative.

The Wizard
Posts: 12358
Joined: Tue Mar 23, 2010 1:45 pm
Location: Reading, MA

### Re: Safe Withdrawal Rate from SPIA???

Just ladder your immediate annuities if you need to.
I got a bit over 6.5% payout at age 63 from TIAA.
Then a 7.0% payout this year for a small increment at age 67.
My TIAA immediate annuities are both fixed (Trad) and variable (TREA and CREF Stock).
I'm sorry that TIAA is not available to the general public...
Attempted new signature...

technovelist
Posts: 2917
Joined: Wed Dec 30, 2009 9:02 pm
Contact:

### Re: Safe Withdrawal Rate from SPIA???

pezblanco wrote:
Wed Dec 06, 2017 3:44 pm
Interesting discussion. Just a data point that I got from immediateannuties.com For a 65 year old male living in my state:

For a \$1,000,000 payment, they will give \$3759.00 per month a 4.51% payout rate with COLA.
they will give \$5512.00 per month a 6.61% payout rate without COLA.

So a 68% discoount for providing a COLA. This sort of makes sense to me .... I was finding from simulation that 50 to 60 percent would be the discount but I thought that my method of deciding a FAIL was too conservative.
That's a 32% discount, not a 68% discount. I think you mean the COLA'd annuity pays 68% of what the non-COLA'd one pays.
In theory, theory and practice are identical. In practice, they often differ.

heyyou
Posts: 3175
Joined: Tue Feb 20, 2007 4:58 pm

### Re: Safe Withdrawal Rate from SPIA???

Henry "Bud" Hebeler advocated saving a portion of any type of no-COLA annuity/pension.
http://www.analyzenow.com/Articles/Annu ... pdf%20.pdf
The third page shows calculations for how much to save and how much to spend of the annual annuity income.

technovelist
Posts: 2917
Joined: Wed Dec 30, 2009 9:02 pm
Contact:

### Re: Safe Withdrawal Rate from SPIA???

ResearchMed wrote:
Wed Dec 06, 2017 1:27 pm
randomguy wrote:
Wed Dec 06, 2017 1:13 pm
ResearchMed wrote:
Wed Dec 06, 2017 1:02 pm

Are there any comparisons from the past (future may/will differ, etc.) about how an inflation adjusted SPIA really did/did not keep up with the actual inflation as it occurred going forward?
It would be interesting to know if it *ever* came even close, etc.

RM
Obviously it depends on what you buy. Buy one that matches CPI, that is what you get. Inflation can be pretty personal as we all have different exposures to housing, gas, medical, food and so on.
I'm asking if the inflation adjusted SPIA's "do what they are supposed to do", in real life, going forward, at least for past years.

I've never known anyone who had one, other than SS, and I guess part of my question might include how commercial inflation adjusted SPIA's actually compared with SS benefits, counting both starting the same year.
(But with many years of data, over many past year starting points.)

For example, are there unexpected "gotchas", like with some "equity indexed annuities" where there are caps on returns that are not usually understood/expected, etc. (I understand these are entirely different, but it's the "unexpected gotchas" that is part of what I'm wondering about.)

How one needs/uses the money isn't really different from the actual living costs each of us chooses or needs, regardless of income source.

RM
If you buy a "CPI-adjusted" annuity and it doesn't adjust payments according to the CPI, then the insurer is not living up to their side of the contract. I assume that any licensed insurer would have to do that to stay out of trouble.

Now whether CPI is a good measure of inflation is another matter...
In theory, theory and practice are identical. In practice, they often differ.

technovelist
Posts: 2917
Joined: Wed Dec 30, 2009 9:02 pm
Contact:

### Re: Safe Withdrawal Rate from SPIA???

heyyou wrote:
Wed Dec 06, 2017 6:45 pm
Henry "Bud" Hebeler advocated saving a portion of any type of no-COLA annuity/pension.
http://www.analyzenow.com/Articles/Annu ... pdf%20.pdf
The third page shows calculations for how much to save and how much to spend of the annual annuity income.
My retirement income analyzer includes that as part of its "level real after-tax spending" calculations.
In theory, theory and practice are identical. In practice, they often differ.

ResearchMed
Posts: 7473
Joined: Fri Dec 26, 2008 11:25 pm

### Re: Safe Withdrawal Rate from SPIA???

heyyou wrote:
Wed Dec 06, 2017 6:45 pm
Henry "Bud" Hebeler advocated saving a portion of any type of no-COLA annuity/pension.
http://www.analyzenow.com/Articles/Annu ... pdf%20.pdf
The third page shows calculations for how much to save and how much to spend of the annual annuity income.
Oh dear.... Mr. Hebeler passed away this August... what a shame.
He was 84.
He did so much to help others, at no cost to them.

R.I.P.

RM
This signature is a placebo. You are in the control group.

ResearchMed
Posts: 7473
Joined: Fri Dec 26, 2008 11:25 pm

### Re: Safe Withdrawal Rate from SPIA???

technovelist wrote:
Wed Dec 06, 2017 6:47 pm
ResearchMed wrote:
Wed Dec 06, 2017 1:27 pm
randomguy wrote:
Wed Dec 06, 2017 1:13 pm
ResearchMed wrote:
Wed Dec 06, 2017 1:02 pm

Are there any comparisons from the past (future may/will differ, etc.) about how an inflation adjusted SPIA really did/did not keep up with the actual inflation as it occurred going forward?
It would be interesting to know if it *ever* came even close, etc.

RM
Obviously it depends on what you buy. Buy one that matches CPI, that is what you get. Inflation can be pretty personal as we all have different exposures to housing, gas, medical, food and so on.
I'm asking if the inflation adjusted SPIA's "do what they are supposed to do", in real life, going forward, at least for past years.

I've never known anyone who had one, other than SS, and I guess part of my question might include how commercial inflation adjusted SPIA's actually compared with SS benefits, counting both starting the same year.
(But with many years of data, over many past year starting points.)

For example, are there unexpected "gotchas", like with some "equity indexed annuities" where there are caps on returns that are not usually understood/expected, etc. (I understand these are entirely different, but it's the "unexpected gotchas" that is part of what I'm wondering about.)

How one needs/uses the money isn't really different from the actual living costs each of us chooses or needs, regardless of income source.

RM
If you buy a "CPI-adjusted" annuity and it doesn't adjust payments according to the CPI, then the insurer is not living up to their side of the contract. I assume that any licensed insurer would have to do that to stay out of trouble.

Now whether CPI is a good measure of inflation is another matter...
Good.
Is this the same inflation adjustment that is used by SS?

Thanks.

RM
This signature is a placebo. You are in the control group.

technovelist
Posts: 2917
Joined: Wed Dec 30, 2009 9:02 pm
Contact:

### Re: Safe Withdrawal Rate from SPIA???

ResearchMed wrote:
Wed Dec 06, 2017 6:57 pm
technovelist wrote:
Wed Dec 06, 2017 6:47 pm
ResearchMed wrote:
Wed Dec 06, 2017 1:27 pm
randomguy wrote:
Wed Dec 06, 2017 1:13 pm
ResearchMed wrote:
Wed Dec 06, 2017 1:02 pm

Are there any comparisons from the past (future may/will differ, etc.) about how an inflation adjusted SPIA really did/did not keep up with the actual inflation as it occurred going forward?
It would be interesting to know if it *ever* came even close, etc.

RM
Obviously it depends on what you buy. Buy one that matches CPI, that is what you get. Inflation can be pretty personal as we all have different exposures to housing, gas, medical, food and so on.
I'm asking if the inflation adjusted SPIA's "do what they are supposed to do", in real life, going forward, at least for past years.

I've never known anyone who had one, other than SS, and I guess part of my question might include how commercial inflation adjusted SPIA's actually compared with SS benefits, counting both starting the same year.
(But with many years of data, over many past year starting points.)

For example, are there unexpected "gotchas", like with some "equity indexed annuities" where there are caps on returns that are not usually understood/expected, etc. (I understand these are entirely different, but it's the "unexpected gotchas" that is part of what I'm wondering about.)

How one needs/uses the money isn't really different from the actual living costs each of us chooses or needs, regardless of income source.

RM
If you buy a "CPI-adjusted" annuity and it doesn't adjust payments according to the CPI, then the insurer is not living up to their side of the contract. I assume that any licensed insurer would have to do that to stay out of trouble.

Now whether CPI is a good measure of inflation is another matter...
Good.
Is this the same inflation adjustment that is used by SS?

Thanks.

RM
"The Social Security Act specifies a formula for determining each COLA. According to the formula, COLAs are based on increases in the Consumer Price Index for Urban Wage Earners and Clerical Workers (CPI-W). CPI-Ws are calculated on a monthly basis by the Bureau of Labor Statistics."

(https://www.ssa.gov/OACT/COLA/latestCOLA.html)

Of course it's not as simple as just multiplying the previous payment by 1 + the CPI-W. After all, this is the Social Security Administration! Nothing they do is simple. The rest of the factors for the calculation are on that same page.

But now the question is what "CPI" do the "CPI-adjusted" annuities use? That may be different for different insurance companies, and even possibly by what state you live in (because different states have different insurance regulations). May I ask what state you live in? I can probably find out how your state's regulations might affect CPI-adjusted annuities.
In theory, theory and practice are identical. In practice, they often differ.

pezblanco
Posts: 477
Joined: Thu Sep 12, 2013 8:02 pm

### Re: Safe Withdrawal Rate from SPIA???

heyyou wrote:
Wed Dec 06, 2017 6:45 pm
Henry "Bud" Hebeler advocated saving a portion of any type of no-COLA annuity/pension.
http://www.analyzenow.com/Articles/Annu ... pdf%20.pdf
The third page shows calculations for how much to save and how much to spend of the annual annuity income.
Thank you so much for this reference. The article states:

you can get a reasonable idea which might be best for you by multiplying the quote
for the fixed payment annuity times your current age / 110 as above. If the resulting payment is
significantly better than the inflation-adjusted quote, go for the fixed payment IF you feel you can
maintain a separate reserve with secure investments that will have returns that are close to, or better
than, inflation. Such a reserve might be dominated by Savings I Bonds or TIPS. The former is better for a
non-qualified account (taxable) and the latter for a qualified account (IRA).

So for a 65 year old, the ratio that he gives is 59% of the payout is the true long term inflation adjusted cash flow. This again is very much in line with what we've been seeing from other arguments. I'm not sure exactly how his caveat about "secure investments ... returns .. close to or better than inflation" would work out .... we've just gone through a stage where secure interest rates were less than inflation. My analysis assumed putting the money in the stock market and going for greater average returns than that. I could do some simulations with returns based on T-Bills instead of the S&P 500 but I suspect that the FAIL rate would go up quite a bit.

heyyou
Posts: 3175
Joined: Tue Feb 20, 2007 4:58 pm

### Re: Safe Withdrawal Rate from SPIA???

My analysis assumed putting the money in the stock market and going for greater average returns than that.
Not a problem if you can wait long enough before starting to spend from it, or only need to withdraw a little of it annually to supplement the annuity income.

At my house, we expect to adapt some to our income in retirement, and we have 30+ years of experience at living within our means.

itstoomuch
Posts: 5343
Joined: Mon Dec 15, 2014 12:17 pm
Location: midValley OR

### Re: Safe Withdrawal Rate from SPIA???

op, your idea won't work.

You are assuming that SWR is sacrosanct, its not.
You are assuming that a SPIA is perfectly priced at any time at purchase, it's not.
Both may appear to you as stationary but to the vendors they are highly variable; which is how the make their money and hedge their positions.

It can get complicated.
But you can beat the percentages
Ymmv
Rev012718; 4 Incm stream buckets: SS+pension; dfr'd GLWB VA & FI anntys, by time & \$\$ laddered; Discretionary; Rentals. LTCi. Own, not asset. Tax TBT%. Early SS. FundRatio (FR) >1.1 67/70yo

pezblanco
Posts: 477
Joined: Thu Sep 12, 2013 8:02 pm

### Re: Safe Withdrawal Rate from SPIA???

itstoomuch wrote:
Wed Dec 06, 2017 11:06 pm
op, your idea won't work.

You are assuming that SWR is sacrosanct, its not.
You are assuming that a SPIA is perfectly priced at any time at purchase, it's not.
Both may appear to you as stationary but to the vendors they are highly variable; which is how the make their money and hedge their positions.

It can get complicated.
But you can beat the percentages
Ymmv
Ummmmm ..... what exactly is my idea that won't work? It's not an original idea by any means. There is a referenced article to a previous author addressing this scheme.

I'm not assuming anything other than what I originally posted. I don't know what you mean by your assertion that I'm assuming the SWR is sacrosanct. I don't think anyone believes that (including myself). I'm not assuming (nor do I need to) that SPIAs are "perfectly priced" either .... ??????

pezblanco
Posts: 477
Joined: Thu Sep 12, 2013 8:02 pm

### Re: Safe Withdrawal Rate from SPIA???

Just to finish off this sequence of simulations:

One BH by private message asked what would be the affect of taking the whole first year payout and just putting it into the investment fund thereby minimizing a bit the sequence of returns risk: That strategy does improve things:

800...........71
700...........30
600............6
500............(.5)

Just as a matter of interest. Instead of holding the investment account in the S&P500, I decided to use a super safe investment .... One month T-bills. Again, I used 51 past years of data from the DFA Matrix Book. The results are (as expected) decidedly worse:

700..............98
600..............47
500...............2.7

All of these simulations were done with Matlab .... I would be happy to give the code to anyone that wants to play with it for themselves. You can run the code on-line via an open source Octave compiler if you don't have Matlab or Octave installed on your own computer.

willthrill81
Posts: 6146
Joined: Thu Jan 26, 2017 3:17 pm
Location: USA

### Re: Safe Withdrawal Rate from SPIA???

moghopper wrote:
Wed Dec 06, 2017 10:50 am
Many studies indicate that spending decreases over time.

(ref: https://www.forbes.com/sites/wadepfau/2 ... -over-time)

Why assume a need for a CPI adjustment on the SPIA, or a "man-made" one in your scenario? I realize that is the thing to do when planning, but it appears not to jive with reality.
Yes, it's true and often overlooked that retirees aged 65 and older tend to spend 1-2% less in real dollars every year throughout their retirement, even well into their 80s.

So if inflation was no more than the Fed's target of 2%, a CPI adjustment or COLA is probably unnecessary if your spending is going down at approximately the same rate as inflation, particularly if the SPIA is not your primary source of income (which it usually isn't).

However, if retirees experienced the kind of inflation seen in the 1970s and 1980s, the lack of a CPI adjustment could literally destroy the value of their SPIA payouts. It's already happened, and it could happen again. How likely it is is anyone's guess.

What some of us discussed a while back was the idea of getting a SPIA without a COLA adjustment and then saving a portion of the monthly payout, possibly the difference in monthly payouts between a SPIA with and without a COLA adjustment, and investing it in TIPS. This would provide somewhat of an inflation hedge for later years, but one problem with this strategy is that it foregoes current spending (when retirees typically need more money) for later spending (when retirees typically need less money).

A SPIA without a CPI adjustment for an opposite-sex couple aged 65 pays about 5.5% annually right now. Given that the historical safe withdrawal rate has been around 4-4.5% (depending on the portfolio), many retirees aren't willing to hand over too much of their assets to purchase a SPIA. There are other reasons as well why retirees tend to steer away from SPIAs, a prominent one being that they tend to not want to part with their currently owned assets to provide a future income; they have a strong tendency to literally have their assets and eat them too.
“It's a dangerous business, Frodo, going out your door. You step onto the road, and if you don't keep your feet, there's no knowing where you might be swept off to.” J.R.R. Tolkien,The Lord of the Rings

jalbert
Posts: 3828
Joined: Fri Apr 10, 2015 12:29 am

### Re: Safe Withdrawal Rate from SPIA???

The idea is that you purchase an SPIA that supplies you with say 1000 dollars per year. At year one, you spend an initial amount of say 800 dollars. You take the excess 200 dollars and you invest it in the S&P. This is your SPIA excess investment fund.
Why not just start with the original principal and instead of using all of it to buy a SPIA, use 80% of it to buy a SPIA and put the other 20% in a stock fund?
Risk is not a guarantor of return.

technovelist
Posts: 2917
Joined: Wed Dec 30, 2009 9:02 pm
Contact:

### Re: Safe Withdrawal Rate from SPIA???

jalbert wrote:
Thu Dec 07, 2017 12:12 am
The idea is that you purchase an SPIA that supplies you with say 1000 dollars per year. At year one, you spend an initial amount of say 800 dollars. You take the excess 200 dollars and you invest it in the S&P. This is your SPIA excess investment fund.
Why not just start with the original principal and instead of using all of it to buy a SPIA, use 80% of it to buy a SPIA and put the other 20% in a stock fund?
Good point, and that would avoid the annuity sales charges on the 20% as well.
In theory, theory and practice are identical. In practice, they often differ.

castlemodesto
Posts: 144
Joined: Sun Sep 15, 2013 12:31 pm

### Re: Safe Withdrawal Rate from SPIA???

i Have earned/ purchased nominal annuities to provide an income floor. In order to ( at least partially), hedge inflation ,I am currently purchasing a ladder of TIPS so that the annuities plus the TIPS will give an exact (more or less) 2% pa COLA. This will not match high inflation, but it should help somewhat. I am using TIPS for liability matching: the liability being that the income is not COLA or inflation adjusted.
the TIPs ladder I am purchasing starts at \$4,000 and goes up each year by \$500 to give the 2% COLA. So a TIPS of 4k for year 1, 4.5k for year 2 etc. Not a perfect plan, but it does give a reliable "paycheck" ( if that is what someone prefers) unlike investing in stocks or the market with plans to "supersize" the nominal annuity in the future ( and hope the markets co-operate at the correct time) .

pezblanco
Posts: 477
Joined: Thu Sep 12, 2013 8:02 pm

### Re: Safe Withdrawal Rate from SPIA???

jalbert wrote:
Thu Dec 07, 2017 12:12 am
The idea is that you purchase an SPIA that supplies you with say 1000 dollars per year. At year one, you spend an initial amount of say 800 dollars. You take the excess 200 dollars and you invest it in the S&P. This is your SPIA excess investment fund.
Why not just start with the original principal and instead of using all of it to buy a SPIA, use 80% of it to buy a SPIA and put the other 20% in a stock fund?
You could do that. But that isn't the point of what we're trying to get at.

The point of the exercise is to determine how many real dollars can you spend from an SPIA or a non-COLA adjusted pension during the time of your retirement. I.e. what constant dollar stream can be extracted from a GIVEN SPIA or non-COLA adjusted pension? The SPIA has already been bought or the pension has already been given to you ... you don't get to change the parameters of its existence.

Based on historical data for the last 50 years and taking part of your money and investing it in the S&P500, you can spend about 60% of your first year payout in real dollars for the time period of your retirement. For me, this is a useful rule of thumb ....

longinvest
Posts: 3053
Joined: Sat Aug 11, 2012 8:44 am

### Re: Safe Withdrawal Rate from SPIA???

Pezblanco,
pezblanco wrote:
Thu Dec 07, 2017 1:26 am
The point of the exercise is to determine how many real dollars can you spend from an SPIA or a non-COLA adjusted pension during the time of your retirement. I.e. what constant dollar stream can be extracted from a GIVEN SPIA or non-COLA adjusted pension? The SPIA has already been bought or the pension has already been given to you ... you don't get to change the parameters of its existence.
The same question was discussed a few months ago:
Any way to "translate" a fixed pension/annuity to an inflation adjusted one?
Last edited by longinvest on Thu Dec 07, 2017 2:43 pm, edited 2 times in total.
Bogleheads investment philosophy | Lifelong Portfolio: 25% each of (domestic/international) stocks / (nominal/inflation-indexed) long-term bonds | VCN/VXC/VLB/ZRR

Stormbringer
Posts: 609
Joined: Sun Jun 14, 2015 7:07 am

### Re: Safe Withdrawal Rate from SPIA???

moghopper wrote:
Wed Dec 06, 2017 10:50 am
Why assume a need for a CPI adjustment on the SPIA, or a "man-made" one in your scenario?
Because if you bought an annuity in the early 1970s, you lost half of your purchasing power in about eight years. As plenty of other countries have demonstrated, next time could be much worse.
"Compound interest is the most powerful force in the universe." - Albert Einstein

itstoomuch
Posts: 5343
Joined: Mon Dec 15, 2014 12:17 pm
Location: midValley OR

### Re: Safe Withdrawal Rate from SPIA???

The other alternative we did: We invested (buy) a rental. After all we are only looking for Income that tracks inflation in the intermediate time and long-term asset growth tracking for our heir.
Ymmv
Rev012718; 4 Incm stream buckets: SS+pension; dfr'd GLWB VA & FI anntys, by time & \$\$ laddered; Discretionary; Rentals. LTCi. Own, not asset. Tax TBT%. Early SS. FundRatio (FR) >1.1 67/70yo

pezblanco
Posts: 477
Joined: Thu Sep 12, 2013 8:02 pm

### Re: Safe Withdrawal Rate from SPIA???

longinvest wrote:
Thu Dec 07, 2017 12:43 pm
Pezblanco,
pezblanco wrote:
Thu Dec 07, 2017 1:26 am
The point of the exercise is to determine how many real dollars can you spend from an SPIA or a non-COLA adjusted pension during the time of your retirement. I.e. what constant dollar stream can be extracted from a GIVEN SPIA or non-COLA adjusted pension? The SPIA has already been bought or the pension has already been given to you ... you don't get to change the parameters of its existence.
The same question was discussed a few months ago:
Any way to "translate" a fixed pension/annuity to an inflation adjusted one?
Thank you. I have put a post into that thread so that this thread can be found. In that thread longinvest (and another BH numbercruncher) perform a deterministic analysis assuming a CONSTANT inflation rate and a CONSTANT investment return which allows you to get a deterministic answer. It turns out that answer is slightly more optimistic than that produced by stochastic simulation using the actual return and inflation data from the past 50 years.

longinvest
Posts: 3053
Joined: Sat Aug 11, 2012 8:44 am

### Re: Safe Withdrawal Rate from SPIA???

pezblanco wrote:
Thu Dec 07, 2017 5:39 pm
longinvest wrote:
Thu Dec 07, 2017 12:43 pm
Pezblanco,
pezblanco wrote:
Thu Dec 07, 2017 1:26 am
The point of the exercise is to determine how many real dollars can you spend from an SPIA or a non-COLA adjusted pension during the time of your retirement. I.e. what constant dollar stream can be extracted from a GIVEN SPIA or non-COLA adjusted pension? The SPIA has already been bought or the pension has already been given to you ... you don't get to change the parameters of its existence.
The same question was discussed a few months ago:
Any way to "translate" a fixed pension/annuity to an inflation adjusted one?
Thank you. I have put a post into that thread so that this thread can be found. In that thread longinvest (and another BH numbercruncher) perform a deterministic analysis assuming a CONSTANT inflation rate and a CONSTANT investment return which allows you to get a deterministic answer. It turns out that answer is slightly more optimistic than that produced by stochastic simulation using the actual return and inflation data from the past 50 years.
You're welcome.

Let me clarify that, in the linked thread, I used a deterministic analysis to find the ratio of pension payment to put aside, but I also proposed to use a fixed percentage to determine the fluctuating amount to withdraw from the inflation-increment portfolio. So, instead of delivering constant inflation-adjusted total income, the method delivers a slightly varying inflation-adjusted amount in real life (depending on actual inflation and portfolio returns).

In other words, I think that it is important to adapt withdrawals to portfolio returns, so as to eliminate the possibility of premature portfolio depletion. This leads to an imperfect solution.

It is always preferable to select an inflation-adjusted income stream (COLA pension, inflation-indexed SPIA) over a fixed nominal income stream, when one has the opportunity to choose between the two. Unfortunately, some of us are vested into pension plans with no cost of living adjustment provisions.
Bogleheads investment philosophy | Lifelong Portfolio: 25% each of (domestic/international) stocks / (nominal/inflation-indexed) long-term bonds | VCN/VXC/VLB/ZRR

jalbert
Posts: 3828
Joined: Fri Apr 10, 2015 12:29 am

### Re: Safe Withdrawal Rate from SPIA???

pezblanco wrote:
Thu Dec 07, 2017 1:26 am
jalbert wrote:
Thu Dec 07, 2017 12:12 am
Why not just start with the original principal and instead of using all of it to buy a SPIA, use 80% of it to buy a SPIA and put the other 20% in a stock fund?
You could do that. But that isn't the point of what we're trying to get at.

The point of the exercise is to determine how many real dollars can you spend from an SPIA or a non-COLA adjusted pension during the time of your retirement. I.e. what constant dollar stream can be extracted from a GIVEN SPIA or non-COLA adjusted pension? The SPIA has already been bought or the pension has already been given to you ... you don't get to change the parameters of its existence.
I’m confused. This was your original problem statement.
From time to time BH's consider the purchase of an SPIA as a do it yourself pension. You can buy these with consumer price index increases built in but they are quite a bit more expensive. The other strategy is to purchase one without such a CPI rider and then when inflation has eaten away it's value too much, purchase another one. Another strategy might be to purchase one and spend part of it each year and invest the rest in a stock fund. This last strategy is what I'm considering here.
Risk is not a guarantor of return.

pezblanco
Posts: 477
Joined: Thu Sep 12, 2013 8:02 pm

### Re: Safe Withdrawal Rate from SPIA???

jalbert wrote:
Thu Dec 07, 2017 8:20 pm
pezblanco wrote:
Thu Dec 07, 2017 1:26 am
jalbert wrote:
Thu Dec 07, 2017 12:12 am
Why not just start with the original principal and instead of using all of it to buy a SPIA, use 80% of it to buy a SPIA and put the other 20% in a stock fund?
You could do that. But that isn't the point of what we're trying to get at.

The point of the exercise is to determine how many real dollars can you spend from an SPIA or a non-COLA adjusted pension during the time of your retirement. I.e. what constant dollar stream can be extracted from a GIVEN SPIA or non-COLA adjusted pension? The SPIA has already been bought or the pension has already been given to you ... you don't get to change the parameters of its existence.
I’m confused. This was your original problem statement.
From time to time BH's consider the purchase of an SPIA as a do it yourself pension. You can buy these with consumer price index increases built in but they are quite a bit more expensive. The other strategy is to purchase one without such a CPI rider and then when inflation has eaten away it's value too much, purchase another one. Another strategy might be to purchase one and spend part of it each year and invest the rest in a stock fund. This last strategy is what I'm considering here.
My original statement was incomplete. Please accept my apologies.

pezblanco
Posts: 477
Joined: Thu Sep 12, 2013 8:02 pm

### Re: Safe Withdrawal Rate from SPIA???

longinvest wrote:
Thu Dec 07, 2017 5:55 pm
pezblanco wrote:
Thu Dec 07, 2017 5:39 pm
longinvest wrote:
Thu Dec 07, 2017 12:43 pm
Pezblanco,
pezblanco wrote:
Thu Dec 07, 2017 1:26 am
The point of the exercise is to determine how many real dollars can you spend from an SPIA or a non-COLA adjusted pension during the time of your retirement. I.e. what constant dollar stream can be extracted from a GIVEN SPIA or non-COLA adjusted pension? The SPIA has already been bought or the pension has already been given to you ... you don't get to change the parameters of its existence.
The same question was discussed a few months ago:
Any way to "translate" a fixed pension/annuity to an inflation adjusted one?
Thank you. I have put a post into that thread so that this thread can be found. In that thread longinvest (and another BH numbercruncher) perform a deterministic analysis assuming a CONSTANT inflation rate and a CONSTANT investment return which allows you to get a deterministic answer. It turns out that answer is slightly more optimistic than that produced by stochastic simulation using the actual return and inflation data from the past 50 years.
You're welcome.

Let me clarify that, in the linked thread, I used a deterministic analysis to find the ratio of pension payment to put aside, but I also proposed to use a fixed percentage to determine the fluctuating amount to withdraw from the inflation-increment portfolio. So, instead of delivering constant inflation-adjusted total income, the method delivers a slightly varying inflation-adjusted amount in real life (depending on actual inflation and portfolio returns).

In other words, I think that it is important to adapt withdrawals to portfolio returns, so as to eliminate the possibility of premature portfolio depletion. This leads to an imperfect solution.

It is always preferable to select an inflation-adjusted income stream (COLA pension, inflation-indexed SPIA) over a fixed nominal income stream, when one has the opportunity to choose between the two. Unfortunately, some of us are vested into pension plans with no cost of living adjustment provisions.
Yes. I would also agree that these sorts of things are not "set and forget". I was thinking that with my Monte Carlo approach, it should be re-evaluated every year depending on what the value is in the "investment" account. We should always take into account our present circumstances. Just as the 4% rule is just a rule of thumb to get you going ... you should reevaluate it every once in a while as your retirement accounts will typically grow ....

I liked your analysis ... it's nice that all of these approaches are coming up with the same ballpark figure ..... 55-65 % should be socked away at the start to allow a constant real income stream.

randomguy
Posts: 6504
Joined: Wed Sep 17, 2014 9:00 am

### Re: Safe Withdrawal Rate from SPIA???

technovelist wrote:
Thu Dec 07, 2017 1:12 am
jalbert wrote:
Thu Dec 07, 2017 12:12 am
The idea is that you purchase an SPIA that supplies you with say 1000 dollars per year. At year one, you spend an initial amount of say 800 dollars. You take the excess 200 dollars and you invest it in the S&P. This is your SPIA excess investment fund.
Why not just start with the original principal and instead of using all of it to buy a SPIA, use 80% of it to buy a SPIA and put the other 20% in a stock fund?
Good point, and that would avoid the annuity sales charges on the 20% as well.
And exposure yourself to sequence of return issues. Buying the annuity largely avoids that in that you are DCA for all those years till inflation catches up. You have to pick what risks you want to deal with.

jalbert
Posts: 3828
Joined: Fri Apr 10, 2015 12:29 am

### Re: Safe Withdrawal Rate from SPIA???

Based on historical data for the last 50 years and taking part of your money and investing it in the S&P500, you can spend about 60% of your first year payout in real dollars for the time period of your retirement
The extra cost for having a SPIA be COLA'd is typically 30% of the payout rate of a nominal one of the same credit quality. So buying a COLA'd SPIA still gives you a higher real payout rate than using 60% of a nominal one plus stock to simulate a real payout.
Risk is not a guarantor of return.