SWR - No Heirs

[Independent]

Would this work.

Let say you decide you need money for another 30 years. At that point you will be 95 and with your health and family background that is ten years past what you really think you will live. Firecalc says 3.5% (Im making up percentage because I'm too lazy to enter it) So you take your nest egg and live on 3.5% and the world is good.

Five years later, the market has been good to you and you now have 50% more than you had, even with the withdrawals. You run Firecalc again this time for 25 years and it say 4%. You reset and start your retirement over using the new amount, knowing that Firecalc says you will not run out of money before you expect to run out of time.

If the market goes down you just stay the course. You should not have not have to lower your withdrawal from a previous high point, unless you forecast living beyond 95. In my example, you figured your increase on only living 25 years and increased funds. If you now believe you will live to be 100, you would have readjust your SWR. However, that is the same for all of us that decide we are going to live longer than the what original SWR was figured on, and our nest egg is lower.

I will admit that I have not gone through ever scenario, but it seems like anytime your nest egg goes up beyond it's set point, you could reset, using a realistic expected life plus pad. As you get older, the projected end of life date becomes a little easier to grasp.

Now before people jump all over me, I don't do this! As I have posted before, we do not use our nest egg for necessary retirement spending. Currently money from our IRA is rolled over to savings, and will most likely continue to be that way. I have just always pondered why you could not reset your SWR based on expected age and nest egg.

I'll look at it this way. The 3.5% comes from some back-testing or Monte Carlo analysis that had a 95% confidence level. That is, you think the 3.5% will fail 5% of the time.

Looking at the individual scenarios in the analysis, you see that none of the failures started with strong returns. All the failures had average or below average (usually below) in the first five years.

If the analysis is reliable, a portfolio growth of 50% indicates that you are in the 100% success group of scenarios. If you continue with your original withdrawal level (which is now 2.33% of your current portfolio) you have a 100% chance of success.

Buy upping your SWR to 4.0%, you put yourself back into the 5% risk of failure world.

I'll try an analogy. Let's suppose someone pays me $$$ to let him shoot at me once from a distance. I'm fairly confident he only has a 5% chance of hitting me, so I take the risk.
He misses on the first shot. Should I say "I'll play again. I'll get a second $$$, and the chance of him hitting me is still just 5%"?
If he misses on the second shot, should I play a third round?

Somehow, ratcheting seems like it introduces more risk.

I might try this rule instead: I'll take the greater of 3.5% of my (inflation adjusted) original portfolio, or 2.33% of this year's portfolio.
Without the ratchet, I will drop my withdrawals when the market goes down.

Another approach, is that I ratchet up, but only to the 100% success SWR.

(Oh, and since I know I've lived the first 5 years, if I'm still in good health, I should change my end point to 26 years or whatever.)

 

[ERD50]

...

Buy upping your SWR to 4.0%, you put yourself back into the 5% risk of failure world.

I'll try an analogy. Let's suppose someone pays me $$$ to let him shoot at me once from a distance. I'm fairly confident he only has a 5% chance of hitting me, so I take the risk.
He misses on the first shot. Should I say "I'll play again. I'll get a second $$$, and the chance of him hitting me is still just 5%"?
If he misses on the second shot, should I play a third round?

Somehow, ratcheting seems like it introduces more risk.
...

I think there is a flaw in your analogy.

The historical reporters like FIRECalc are already running all the available scenarios. That 'second shot' is in all of them.

Again, we have to stick to the historical data itself, the future is unknown territory. But re-running FIRECalc isn't a 'second shot'. All the shots were included, so the second run can't be worse than the first, it was already considered.

It's easier to see if you use a 100% success rate for the time period. Obviously, if there are no failures in the first run, there can't be any in the second either. One caveat - as your subsequent runs are for shorter and shorter time periods, FIRECalc can include more and more complete sets of data (ie, a 50 year run does not include 1965 and later, as those are less than complete, inclusive 50 year runs). *(see edit/add note)

In your analogy, it would be more like you spin the wheel and FIRECalc spits out data for one specific randomly chosen starting year - 95% of the time you win (with defaults), yeah! Then your 'second shot' would be like another spin, and another random start year. Etc. But that is not how it works.

* edit/add: I think the killer year is 1966, so for now, as long as you are using less than about a 46 year period (47-48? - gotta watch the 'inclusive math)), you should not run into this apparent anomaly.

-ERD50

 

[dallas27]

Even rerunning firecalc, you are "adversely selecting" the a probability worse (lower) line in firecalc each time you adjust.

In essence, if the future mirrored the past in firecalc, over enough adjustments(could take a long while), you would end up at the lowest possible line, meaning you would always be "close" to not having enough.


Sent from my iPhone using Early Retirement Forum

 

[Rustic23]

Carrying it to extreme my thought process was this.

When I'm 95 my portfolio has performed at the high end. When I started it said 4% of one million $40,000, plus annual inflation. So here I am with an expected life of two years, an and $8,000,000 in the bank. Not sure what I would spend it on, but I think I would be taking out a heck of a lot more than what I started at. i.e. I would reset long before that point. Ucretch's first post was more in line with my thoughts. Firecalc does not know when you start. As long as you adjust remaining life it should work.

 

[photoguy]

Somehow, ratcheting seems like it introduces more risk.

Ratcheting up definitely increases the risk. The question is whether it's worth the tradeoff of more spending money.

 

[Independent]

Ratcheting up definitely increases the risk. The question is whether it's worth the tradeoff of more spending money.

I'd agree with this. Growing my portfolio by 50% in the first five years is a Good Thing. I've got an important piece of uncertainty behind me.

Now, the question is how I respond to that good news.
A) I could use it entirely to increase my spending - ratchet up, or even spend 1/3 of my new portfolio in one year.
B) I could use it entirely to sleep better at night - keep my original spending plan, but note that I have a lot more margin than when I started.

Or, I could use some of A and some of B. Maybe spend some extra in only one year, but still leave more in the portfolio than I started with.

 

[Independent]

I think there is a flaw in your analogy.

The historical reporters like FIRECalc are already running all the available scenarios. That 'second shot' is in all of them.

Again, we have to stick to the historical data itself, the future is unknown territory. But re-running FIRECalc isn't a 'second shot'. All the shots were included, so the second run can't be worse than the first, it was already considered.

It's easier to see if you use a 100% success rate for the time period. Obviously, if there are no failures in the first run, there can't be any in the second either. One caveat - as your subsequent runs are for shorter and shorter time periods, FIRECalc can include more and more complete sets of data (ie, a 50 year run does not include 1965 and later, as those are less than complete, inclusive 50 year runs). *(see edit/add note)

In your analogy, it would be more like you spin the wheel and FIRECalc spits out data for one specific randomly chosen starting year - 95% of the time you win (with defaults), yeah! Then your 'second shot' would be like another spin, and another random start year. Etc. But that is not how it works.

* edit/add: I think the killer year is 1966, so for now, as long as you are using less than about a 46 year period (47-48? - gotta watch the 'inclusive math)), you should not run into this apparent anomaly.

-ERD50

I guess no analogy is perfect. I'll just drop it and go to the FireCalc back-testing.

If I just load FireCalc, and don't change any assumptions, I'll get 6 failures in 115 runs - 1965, 66, 67, 68, 69, and 73.

If I change the "years" to 25, and change the payout from $30,000 to $32,250 (i.e. 4.3%) I also get 6 failures in the same 6 years. (I pick up an extra 5 start years, but none of them is a failure.)
So, it looks like 4.3% is a reasonable SWR for 25 years.

Now I change FireCalc* so it withdraws $30,000 for the first 5 years. Then, if the 5th year ending balance is more than 150% of the (inflation adjusted) original $750,000, I will calculate a new withdrawal amount of 4.3% of the end of 5th year balance, and carry it forward with inflation.

It turns out that only 14 start years hit the 150%, and they are all successes with this ratcheted payout.
We could say "that's no additional risk". Or we could note that since we're looking for events that only occur 5% the time, we just didn't get the additional one failure due to statistical variation.

I dialed back the 150% to 120% and tried again. This time, I had 40 start years hit the 120%, and two of them became failures 1960 and 1961. I'll say that the ratchet created failures where there had been none before.

The original FireCalc run didn't have this ratchet up withdrawal plan, so introducing changes the failure rate.

As I mentioned above, I'd rather use my strong early years to do a combination of a somewhat better payout, and also reduce my risk.

* Yes, "change FireCalc" means to go offline and code a new payout formula.
Since FireCalc will produce detailed Excel worksheets that we can download, this is possible, though a little time consuming.

 

[dallas27]

A

 

[dallas27]

For the record - if you, or anybody else here needs an heir, my good wife and I will be you heir's. We'll even execute your will in good faith, assuming we got a little.

Certainly there are worse options than using a forum member.







Sent from my iPhone using Early Retirement Forum

 

[eta2020]

(investable_assets + 401k + IRA)/(70 - AGE) = withdrawal rate

At 70 with no money get SS and live rest of life off of it.

If you were a high earner you are looking at 3k plus a months of SS at time when you spending goes down. Good enough IMO. Enjoy healthy years . If you have spouse high earner you are looking at 7k a month.

You do not need to worry about any MEANS tests since you will be broke by the time you get SS.

If any means testing ever passes I would right away institute plan how to spend money or give it to my kids by the time I am 70. Who are those morons that come up with those ideas ?

 

[NW-Bound]

They are the same morons that have passed all the laws we are living under.

 

[dallas27]

If they ever do means testing, there will be loopholes, there always are.


Sent from my iPhone using Early Retirement Forum

 

[ERD50]

...
...
I dialed back the 150% to 120% and tried again. This time, I had 40 start years hit the 120%, and two of them became failures 1960 and 1961. I'll say that the ratchet created failures where there had been none before.

The original FireCalc run didn't have this ratchet up withdrawal plan, so introducing changes the failure rate. ...

I'm tempted to say that something in your coding accounts for this. Unless I'm missing something, it doesn't seem to add up.

Those years that peaked at 120%, were also part of the runs in FIRECalc that passed (they were the starting year when they were at 120% of some earlier reference), so if they passed at 4.3% as part of that run, they must also pass when you bring the rate to 4.3% at 120%.

It doesn't make any difference how you got there, if year X can survive 25 years at 4.3%. The 5 years previous could be low or high withdraws, but if you take 4.3% of a value at year X, that's all there is. History doesn't matter.

Did I not follow something in your explanation?

-ERD50

 

[Independent]

I'm tempted to say that something in your coding accounts for this. Unless I'm missing something, it doesn't seem to add up.

Those years that peaked at 120%, were also part of the runs in FIRECalc that passed (they were the starting year when they were at 120% of some earlier reference), so if they passed at 4.3% as part of that run, they must also pass when you bring the rate to 4.3% at 120%.

It doesn't make any difference how you got there, if year X can survive 25 years at 4.3%. The 5 years previous could be low or high withdraws, but if you take 4.3% of a value at year X, that's all there is. History doesn't matter.

Did I not follow something in your explanation?

-ERD50

Regarding the coding, just download the detail for 1960 and you can check my numbers. I'll be happy to walk through my approach with you.

The issue is that year X did not survive 25 years at 4.3%. If I run Firecalc at 4.3% for 25 years, I'll get 6 failures. One of them is X = 1965.

If I start in 1960 and ratchet up 5 years later, I'm putting myself into that 1965 failure situation.

OTOH, if I run Firecalc at 3.33% for 25 years, I'll get zero failures. Even 1965 survives.

The person who starts in 1960, collecting $30,000 on a $750,000 portfolio, has good news after 5 years. His portfolio grew in spite of his withdrawals. He is in a stronger position than when he started, $30,000 is only 3.33% of his $900,000 portfolio. Even though the years after 1965 are tough, he has built up enough cushion to survive. Firecalc has no cases of 3.33% withdrawal rates failing in 25 years.

But, by ratcheting up, he gives away his cushion. He switches out of a 100% success rate into a 95% success rate. Of course, we should expect that 95% of the time he simply enjoys his higher payout. But, 5% of the time he will trade into a failure.

 

[ERD50]

Regarding the coding, just download the detail for 1960 and you can check my numbers. I'll be happy to walk through my approach with you. ...

OK thanks, I will run through this later, don't have time right now.

-ERD50

 

[PaddyMac]

We also have no heirs or nearby relatives. The problem with planning to die broke is you just don't know how expensive the last year of care might be for the survivor. The first spouse to pass will have a spouse (hopefully in sound mind) to take care of things, but the second spouse may need lots of help. I would not want to drain the accounts down close to nothing when I'm 90, in case I live to 100 and need lots of help. At the very least, we'll have a nice home paid for and the idea is to sell that when the time comes and hopefully it will pay for a few years of care for one, plus whatever is left in the portfolio.

 

[rayvt]

I wholeheartedly agree with your idea on delaying SPIA's until age 70 or so.

But I also notice that online quotes seem to stop at age 80, there are no quotes for age 85, 90 etc. So I've wondered for quite a while if there's an upper age limit at which an annuity provider will even sell an annuity (IOW, there's an age window to buy) OR does the cost start to increase at some age relative for a given payout. I'm asking...

Christopher Lee just died. Age 93. Darned few people live to 93.

SPIA's that I've looked at seem to take 10-15 years until you've gotten back all the money you paid in. How many people would buy an annuity that will almost certainly they'll never get to the break-even point of?

And why would insurance companies bother with quotes for a product that has no customers wanting to buy it?

Buy an annuity at 80, get to the crossover 5 years after your death? Can't be many people wanting that.

 

[rayvt]

Assuming this assumption is accurate, what happens when you have no heirs and have no intention to leaving it to anyone.

I went through this exercise a few years before we retired, using a calculator similar to Firecalc that my employer provided. My wife wanted to know what it said--would we be leaving any money to the kids?

I showed her the output. We'd be somewhere between living our last few years in a refrigerator box, or leaving $8M for the kids to split.

Problem is, you can't cut it that fine so as to run out of money with your last breath. Look at the detailed output from Firecalc. I ran it with $1M portfolio and standard 4% SWR.
6 of 115 runs you ran out of money before 30 years.
33 of 115 runs you had more than $5M at the 30 year mark.

95 of 115 runs, you ended with more than the $1M you started with.

It's quite a bang-bang toggle at the threshold at 30 years.
Only 5 of the 115 runs had between $0 and $500K. Those would be what I'd call "not leaving any money behind."

$10 difference in your withdrawal amount can have a $1M differerence at the final outcome -- there is just no way to plan it so that you leave no money when you die. Either you go broke first, or you die with more (usually much more) than you started wth.

 

[ERD50]

Christopher Lee just died. Age 93. Darned few people live to 93.

SPIA's that I've looked at seem to take 10-15 years until you've gotten back all the money you paid in. How many people would buy an annuity that will almost certainly they'll never get to the break-even point of?

And why would insurance companies bother with quotes for a product that has no customers wanting to buy it?

Buy an annuity at 80, get to the crossover 5 years after your death? Can't be many people wanting that.

Darned few?

https://personal.vanguard.com/us/insights/retirement/plan-for-a-long-retirement-tool

For a couple, each 60 years old, there is a 27% chance that one of them will make it to 93. Is that 'darned few'?

I agree that the ins cos may not be selling them at age 80, the market and their pool may be too small. But 70 makes a lot of sense to me, at least to consider.

Most people considering these products are not looking at 'break even', they are looking at it for what it is - insurance. Do you 'expect' to get in a serious accident on the way to the store? No, and it very likely will not happen. But you buckle your seat belt anyhow, just in case.

-ERD50

 

[ERD50]

I went through this exercise a few years before we retired, using a calculator similar to Firecalc that my employer provided. My wife wanted to know what it said--would we be leaving any money to the kids?

I showed her the output. We'd be somewhere between living our last few years in a refrigerator box, or leaving $8M for the kids to split.

Problem is, you can't cut it that fine so as to run out of money with your last breath. Look at the detailed output from Firecalc. I ran it with $1M portfolio and standard 4% SWR.
6 of 115 runs you ran out of money before 30 years.
33 of 115 runs you had more than $5M at the 30 year mark.

95 of 115 runs, you ended with more than the $1M you started with. ...

And that is exactly why someone might consider annuities at age 70 - it smooths out those variations.

-ERD50

 

[pixelville]

(investable_assets + 401k + IRA)/(70 - AGE) = withdrawal rate

At 70 with no money get SS and live rest of life off of it.

If you were a high earner you are looking at 3k plus a months of SS at time when you spending goes down. Good enough IMO. Enjoy healthy years . If you have spouse high earner you are looking at 7k a month.

You do not need to worry about any MEANS tests since you will be broke by the time you get SS.

If any means testing ever passes I would right away institute plan how to spend money or give it to my kids by the time I am 70. Who are those morons that come up with those ideas ?

That is kind of my plan. Except also have a small amount to annutize around 70.
Might as well spend during the healthy years.

So many seniors are currently living on SS alone.

 

[eta2020]

That is kind of my plan. Except also have a small amount to annutize around 70.
Might as well spend during the healthy years.

So many seniors are currently living on SS alone.

If you and your spouse were high earners and you take SS at 70 you are looking at about 70k to 75k a year. (Even if you stopped working at around 50-55)

That is plenty to live on.

 

[Davis65]

There seem to be a lot of retired people in this forum who report using a 3.5-4% SWR, and their assets are growing. There could be a number if reasons for this or combinations of reasons: (1) people with growing assets like to brag about it (and who wouldn't?), while people with declining assets aren't so eager to write about it; (2) people here are better investors than most; (3) the markets have been rising the last couple of years; and (4) the 3.5-4% SWR is safe for worst case scenarios, which happen only rarely. I kind of think that #4 is a big reason. I plan to use a higher withdrawal rate, and keep a cash cushion of term deposits so that I don't have to sell off holdings in a down market.

 

[eta2020]

There seem to be a lot of retired people in this forum who report using a 3.5-4% SWR, and their assets are growing. There could be a number if reasons for this or combinations of reasons: (1) people with growing assets like to brag about it (and who wouldn't?), while people with declining assets aren't so eager to write about it; (2) people here are better investors than most; (3) the markets have been rising the last couple of years; and (4) the 3.5-4% SWR is safe for worst case scenarios, which happen only rarely. I kind of think that #4 is a big reason. I plan to use a higher withdrawal rate, and keep a cash cushion of term deposits so that I don't have to sell off holdings in a down market.

I think once you cross 65 there is more value in bragging about perfect health then about portfolio growth.... As long as you have at least 60k to live on.

 

[Jack_Pine]

There seem to be a lot of retired people in this forum who report using a 3.5-4% SWR, and their assets are growing. There could be a number if reasons for this or combinations of reasons: (1) people with growing assets like to brag about it (and who wouldn't?), while people with declining assets aren't so eager to write about it; (2) people here are better investors than most; (3) the markets have been rising the last couple of years; and (4) the 3.5-4% SWR is safe for worst case scenarios, which happen only rarely. I kind of think that #4 is a big reason. I plan to use a higher withdrawal rate, and keep a cash cushion of term deposits so that I don't have to sell off holdings in a down market.

I agree with this in general. Many of those things can/could be true but I think number 4 is very possible. I also plan on a relatively high SWR but am willing to adjust for down return years.

I hate to say it but I am not sold on waiting until 70 for SS. This subject has been beat to death for sure and there are a ton of threads on it.....