Clarification on 25x annual spending rule

Non-investing personal finance issues including insurance, credit, real estate, taxes, employment and legal issues such as trusts and wills
Topic Author
coldweather
Posts: 11
Joined: Thu Oct 15, 2020 6:28 pm

Clarification on 25x annual spending rule

Post by coldweather »

Hello, I'm just looking for clarification on the 25x annual spending rule in order to fund retirement. I understand this is just a rough guideline, but I'm unclear about how inflation plays into this. For example, if I need $40,000 per year to live on, and I reach a point where I have $1,000,000 in my taxable account, can I safely retire, or would I need more that that to account for inflation? 20 years into this theoretical retirement, that 40k needed per year might become 50k or who knows how much. Am I on the right track here and if so, how do I calculate a FI number that includes inflation in it? Thank you!
Marseille07
Posts: 3994
Joined: Fri Nov 06, 2020 1:41 pm

Re: Clarification on 25x annual spending rule

Post by Marseille07 »

It's 40K on 1M to start with, but each year you add CPI to it. We just saw CPI of 4.2%; so if you use this number then you're looking at 41.68K next year instead of 40K.
User avatar
Kenkat
Posts: 7326
Joined: Thu Mar 01, 2007 11:18 am
Location: Cincinnati, OH

Re: Clarification on 25x annual spending rule

Post by Kenkat »

The 4% guideline takes inflation into account; it is an inflation adjusted withdraw rate.
User avatar
Nicolas
Posts: 2855
Joined: Wed Aug 22, 2012 7:41 am
Location: Ashtabula, 56th and Wabasha
Contact:

Re: Clarification on 25x annual spending rule

Post by Nicolas »

Your portfolio, unless it’s all in cash, should keep up with inflation so that’s implicit in the “rule”. Most projections are built upon a 60/40 stocks/bonds asset allocation, though yours may differ. Of course if you hold only cash you could be in trouble.
Topic Author
coldweather
Posts: 11
Joined: Thu Oct 15, 2020 6:28 pm

Re: Clarification on 25x annual spending rule

Post by coldweather »

Got it. Thanks everybody!
stuper1
Posts: 318
Joined: Tue Apr 03, 2018 9:30 am

Re: Clarification on 25x annual spending rule

Post by stuper1 »

You probably need at least 30% in stocks in your portfolio, preferably a bit more, in order for your portfolio to keep up with inflation.
26% US large cap, 26% US small-cap value, 15% long-term Treasuries, 15% short-term Treasuries, 18% gold
User avatar
bengal22
Posts: 1982
Joined: Sat Dec 03, 2011 6:20 pm
Location: Ohio

Re: Clarification on 25x annual spending rule

Post by bengal22 »

For me it's projected income minus projected expense.
That gives me an idea how long my nest egg will last.
Not sure why we forget the income part: ssa, pension, etc.
"Earn All You Can; Give All You Can; Save All You Can." .... John Wesley
delamer
Posts: 11486
Joined: Tue Feb 08, 2011 6:13 pm

Re: Clarification on 25x annual spending rule

Post by delamer »

bengal22 wrote: Sun May 16, 2021 6:15 pm For me it's projected income minus projected expense.
That gives me an idea how long my nest egg will last.
Not sure why we forget the income part: ssa, pension, etc.
I think you mean projected expenses minus projected income (Social Security, pensions, etc.) equals projected withdrawals from portfolio.
One thing that humbles me deeply is to see that human genius has its limits while human stupidity does not. | | Alexandre Dumas, fils
User avatar
bengal22
Posts: 1982
Joined: Sat Dec 03, 2011 6:20 pm
Location: Ohio

Re: Clarification on 25x annual spending rule

Post by bengal22 »

delamer wrote: Sun May 16, 2021 6:35 pm
bengal22 wrote: Sun May 16, 2021 6:15 pm For me it's projected income minus projected expense.
That gives me an idea how long my nest egg will last.
Not sure why we forget the income part: ssa, pension, etc.
I think you mean projected expenses minus projected income (Social Security, pensions, etc.) equals projected withdrawals from portfolio.
No I meant exactly what I said. That's how you get projected withdrawals.
"Earn All You Can; Give All You Can; Save All You Can." .... John Wesley
azanon
Posts: 3026
Joined: Mon Nov 07, 2011 10:34 am

Re: Clarification on 25x annual spending rule

Post by azanon »

coldweather wrote: Sun May 16, 2021 5:50 pm Hello, I'm just looking for clarification on the 25x annual spending rule in order to fund retirement. I understand this is just a rough guideline, but I'm unclear about how inflation plays into this. For example, if I need $40,000 per year to live on, and I reach a point where I have $1,000,000 in my taxable account, can I safely retire, or would I need more that that to account for inflation? 20 years into this theoretical retirement, that 40k needed per year might become 50k or who knows how much. Am I on the right track here and if so, how do I calculate a FI number that includes inflation in it? Thank you!
As others have said, the 4% takes into account inflation. There is a catch though - and that it's based off of the trinity study (Bill Bengen's work), and it assumes a portfolio of at least 50% stocks, which for me would cause quite a bit of uncomfortable feelings with no human capital left. Bengen said, and recently confirmed, that any percentage of stock under 50% is suboptimal, and would begin to lower the optimal percentage to something less than that.

I won't go into the details, but there are ways to fine-tune those aforementioned issues; For instance, I plan to use a variable withdrawal method that intentionally, but gradually spends into principle, which will allow me to safely lower my equity percent to the ~ 30% range while still withdrawing near 4%. I prefer that over a perpetual method (such as constant percent method) having more volatility of a 50% or even higher equity percentage. To each his own, I guess. A couple of solid variable methods that do that are Boglehead VPW, or even just the vanilla RMD method.
tibbitts
Posts: 13846
Joined: Tue Feb 27, 2007 6:50 pm

Re: Clarification on 25x annual spending rule

Post by tibbitts »

azanon wrote: Mon May 17, 2021 8:22 am I won't go into the details, but there are ways to fine-tune those aforementioned issues; For instance, I plan to use a variable withdrawal method that intentionally, but gradually spends into principle, which will allow me to safely lower my equity percent to the ~ 30% range while still withdrawing near 4%. I prefer that over a perpetual method (such as constant percent method) having more volatility of a 50% or even higher equity percentage. To each his own, I guess. A couple of solid variable methods that do that are Boglehead VPW, or even just the vanilla RMD method.
Just to clarify for the OP, the "but gradually spends into principle" part might be confusing, since the traditional 4% SWR method does include the possibility of depleting or nearly depleting the original amount.
azanon
Posts: 3026
Joined: Mon Nov 07, 2011 10:34 am

Re: Clarification on 25x annual spending rule

Post by azanon »

tibbitts wrote: Mon May 17, 2021 8:39 am
azanon wrote: Mon May 17, 2021 8:22 am I won't go into the details, but there are ways to fine-tune those aforementioned issues; For instance, I plan to use a variable withdrawal method that intentionally, but gradually spends into principle, which will allow me to safely lower my equity percent to the ~ 30% range while still withdrawing near 4%. I prefer that over a perpetual method (such as constant percent method) having more volatility of a 50% or even higher equity percentage. To each his own, I guess. A couple of solid variable methods that do that are Boglehead VPW, or even just the vanilla RMD method.
Just to clarify for the OP, the "but gradually spends into principle" part might be confusing, since the traditional 4% SWR method does include the possibility of depleting or nearly depleting the original amount.
Thanks for that clarification. Used properly (my opinion again), the 4% rule is only appropriate for getting an initial idea of how much you can withdraw through your retirement, given the assumptions of the rule. But actually using it as a system for retirement income, in the strictest sense, is pretty reckless. It's also known as the "constant dollar" method. So absolutely, it has a chance of running out prematurely, and if your stock percentage is high enough and with a bad enough market, that could be way sooner than you planned.
AnEngineer
Posts: 554
Joined: Sat Jun 27, 2020 4:05 pm

Re: Clarification on 25x annual spending rule

Post by AnEngineer »

azanon wrote: Mon May 17, 2021 9:12 amBut actually using it as a system for retirement income, in the strictest sense, is pretty reckless.
Can you explain this?
Jags4186
Posts: 5785
Joined: Wed Jun 18, 2014 7:12 pm

Re: Clarification on 25x annual spending rule

Post by Jags4186 »

AnEngineer wrote: Mon May 17, 2021 9:16 am
azanon wrote: Mon May 17, 2021 9:12 amBut actually using it as a system for retirement income, in the strictest sense, is pretty reckless.
Can you explain this?
While the 4% rule has a very high probability of success, it is not a 100% probability of success. Since you are 1 person living 1 life during 1 time period, it would be reckless to continue to increase your spending if your portfolio is getting destroyed. You may never recover and portfolio failure could lead to personal catastrophe.

Of course this is why someone like Paul Merriman recommends saving more than you need. If you need $40k/yr from your portfolio, have a $1.5mm portfolio. Then if the market does perform poorly, you can reduce your withdrawals while maintaining lifestyle, and in good times either ratchet up the fun or give money away. Of course, most people barely save “enough” let alone well more than enough.
AnEngineer
Posts: 554
Joined: Sat Jun 27, 2020 4:05 pm

Re: Clarification on 25x annual spending rule

Post by AnEngineer »

Jags4186 wrote: Mon May 17, 2021 9:24 am
AnEngineer wrote: Mon May 17, 2021 9:16 am
azanon wrote: Mon May 17, 2021 9:12 amBut actually using it as a system for retirement income, in the strictest sense, is pretty reckless.
Can you explain this?
While the 4% rule has a very high probability of success, it is not a 100% probability of success. Since you are 1 person living 1 life during 1 time period, it would be reckless to continue to increase your spending if your portfolio is getting destroyed. You may never recover and portfolio failure could lead to personal catastrophe.
If 100% is your standard then I suppose you consider driving to be reckless too? If your standard for "reckless" if out of step with everyone, fine, but it sounded like you had some complaint about how the 4% rule works in practice that doesn't align with the success rates reported in the studies.
dbr
Posts: 35886
Joined: Sun Mar 04, 2007 9:50 am

Re: Clarification on 25x annual spending rule

Post by dbr »

tibbitts wrote: Mon May 17, 2021 8:39 am
azanon wrote: Mon May 17, 2021 8:22 am I won't go into the details, but there are ways to fine-tune those aforementioned issues; For instance, I plan to use a variable withdrawal method that intentionally, but gradually spends into principle, which will allow me to safely lower my equity percent to the ~ 30% range while still withdrawing near 4%. I prefer that over a perpetual method (such as constant percent method) having more volatility of a 50% or even higher equity percentage. To each his own, I guess. A couple of solid variable methods that do that are Boglehead VPW, or even just the vanilla RMD method.
Just to clarify for the OP, the "but gradually spends into principle" part might be confusing, since the traditional 4% SWR method does include the possibility of depleting or nearly depleting the original amount.
A more generalized thought process for this that I think more clearly represents the dynamics of holding portfolios with contributions or withdrawals is that there is no useful concept of principal. Rather the parameters are the portfolio value at any time and how it varies over time. The arithmetic is very simple as it is just that the evolution to the next portfolio value is the old portfolio value plus the gain from return plus the contributions minus the withdrawals. Then you compound that over time. Part of the importance of this model is that it naturally incorporates the standard definition of return without getting confused over principal vs earnings vs capital gains vs dividends, etc. The data sets on investment performance, past or estimated are in terms of return and that is for a reason, namely simplicity, completeness, and clarity of thought process.

If a person is concerned whether not the past or estimated future value of the portfolio ever becomes less than the starting value, that result drops right out of the analysis immediately. In that regard it is rational to compute all of this in real dollars. The idea that one has "not invaded principal" when the nominal value but not the real value is maintained seems not useful for most cases. Another fallacy is to value the holding in "shares" while deciding to ignore the value of those shares. While there can be some psychological advantage to that there are lots of ways that doing so is not useful where valuing the portfolio would be useful in being more complete and more relevant to most purposes.
Jags4186
Posts: 5785
Joined: Wed Jun 18, 2014 7:12 pm

Re: Clarification on 25x annual spending rule

Post by Jags4186 »

AnEngineer wrote: Mon May 17, 2021 9:29 am
Jags4186 wrote: Mon May 17, 2021 9:24 am
AnEngineer wrote: Mon May 17, 2021 9:16 am
azanon wrote: Mon May 17, 2021 9:12 amBut actually using it as a system for retirement income, in the strictest sense, is pretty reckless.
Can you explain this?
While the 4% rule has a very high probability of success, it is not a 100% probability of success. Since you are 1 person living 1 life during 1 time period, it would be reckless to continue to increase your spending if your portfolio is getting destroyed. You may never recover and portfolio failure could lead to personal catastrophe.
If 100% is your standard then I suppose you consider driving to be reckless too? If your standard for "reckless" if out of step with everyone, fine, but it sounded like you had some complaint about how the 4% rule works in practice that doesn't align with the success rates reported in the studies.
If the speed limit on a road is 65 MPH you might drive 65 MPH. If you’re an aggressive driver you might drive 80 MPH. If you’re a conservative driver you might drive 55 MPH. If it’s raining hard, regardless of how you normally drive, you slow down because conditions have changed. If it’s raining so hard you can’t see, you might pull over. You don’t just driving the same speed because there’s a sign on the side of the road that says 65 MPH. If you don’t slow down you have a high likelihood of ending up in a ditch.

Likewise, a retired person may be 60/40, 70/30, or 50/50 AA. They may all be withdrawing 4%, or maybe aggressive ones are withdrawing 5% or conservative ones are withdrawing 3%. But if the market crashes 50%, its reckless to continue going at the same speed until conditions improve.
User avatar
David Jay
Posts: 10989
Joined: Mon Mar 30, 2015 5:54 am
Location: Michigan

Re: Clarification on 25x annual spending rule

Post by David Jay »

AnEngineer wrote: Mon May 17, 2021 9:29 am
Jags4186 wrote: Mon May 17, 2021 9:24 am
AnEngineer wrote: Mon May 17, 2021 9:16 am
azanon wrote: Mon May 17, 2021 9:12 amBut actually using it as a system for retirement income, in the strictest sense, is pretty reckless.
Can you explain this?
While the 4% rule has a very high probability of success, it is not a 100% probability of success. Since you are 1 person living 1 life during 1 time period, it would be reckless to continue to increase your spending if your portfolio is getting destroyed. You may never recover and portfolio failure could lead to personal catastrophe.
If 100% is your standard then I suppose you consider driving to be reckless too? If your standard for "reckless" if out of step with everyone, fine, but it sounded like you had some complaint about how the 4% rule works in practice that doesn't align with the success rates reported in the studies.
Note that Jags is talking about continuing to spend the inflation-adjusted amount after a "black swan" type event in one's retirement years that has devastated one's portfolio. Because it is the only portfolio one has, it would be necessary to find a way to adjust expenses.

It feels like you read his comment differently, i.e. that the risk of retiring with 25x is too great a risk to take.

Or to use your example, he is not saying to give up driving due to risk. I believe he is saying to go ahead and take the road trip but if the "low tire pressure" light comes on while driving down the highway at 70MPH, it would be prudent to do something about it before losing control of the vehicle.
Prediction is very difficult, especially about the future - Niels Bohr | To get the "risk premium", you really do have to take the risk - nisiprius
wolf359
Posts: 2399
Joined: Sun Mar 15, 2015 8:47 am

Re: Clarification on 25x annual spending rule

Post by wolf359 »

coldweather wrote: Sun May 16, 2021 5:50 pm Hello, I'm just looking for clarification on the 25x annual spending rule in order to fund retirement. I understand this is just a rough guideline, but I'm unclear about how inflation plays into this. For example, if I need $40,000 per year to live on, and I reach a point where I have $1,000,000 in my taxable account, can I safely retire, or would I need more that that to account for inflation? 20 years into this theoretical retirement, that 40k needed per year might become 50k or who knows how much. Am I on the right track here and if so, how do I calculate a FI number that includes inflation in it? Thank you!
William Bengen's original article that proposed the 4% withdrawal rate isn't a very long read, and might help clarify what he meant. There have been a lot of interpretations over the years.

The paper is located here: https://www.retailinvestor.org/pdf/Bengen1.pdf

Some takeaways that may surprise you:

1) The guidance was originally aimed at financial planners, not individuals. It was to provide guidance to advisors as to how to best manage client portfolios and withdrawal rates. That doesn't mean that you can't use the guidance yourself, but it was aimed at people who provided an extra safeguard to keep their clients from panicking during a downturn. If you need that help, you might need to see it out separately.

2) The 4% rate was a guideline, based on what worked in the past. He stresses flexibility, and increasing or decreasing that rate depending upon market conditions. It wasn't a hard and fast rule. Bengen has since updated the number, believing that it's higher. However, it includes all costs, including taxes and advisor fees.

3) The inflation adjustment is supposed to be your own personal inflation rate. If your costs aren't increasing, you shouldn't adjust. It's based on maintaining your lifestyle. CPI may not reflect your inflation rate.

Also, there is a flaw in the 4% rule that is not immediately obvious. The study he ran is based on the best safe maximum withdrawal rate over a historical time period that included the Great Depression and other bad economic events. However, most people have a tendency to decide they want to retire when they meet a magic number (in your case, $1,000,000). And you're most likely to hit that number when the market is hitting all-time-highs.

In other words, if you follow the 4% Rule, and you retire as soon as you hit your number, you are much more likely to retire at a market peak. Few people retire at a market trough with 4% (even though that's the best survival case.)

The key takeaway should be that the 4% Rule is more of a Rule of Thumb, and should be used as guidance to know when you're in the ballpark of having enough assets to retire. Build in buffers and safeguards. This includes things like a paid-off house, maximizing social security, keeping a sufficient emergency fund, and having a retirement budget that allows slack so that you could reduce expenses (and your withdrawal rate) when the market conditions require it. And yes, it might mean going back to work to supplement your income or replenish your nest egg if necessary.

Some people insist on higher multiples, like 33X or 50X annual spending (3% rule or 2% rule). I think that's excessive. If your budget allows you to cut back so that you can live on 2% or 3% indefinitely, then you only need to do that when market conditions require. But you could still withdraw 4% at all other times. (In other words, just build slack in the form of discretionary spending into your plan. If something like a global pandemic comes up, cancel the trip to Europe that year. Or do 2 cruises instead of 3. Or switch from 1st class to economy. It doesn't actually take much.)

Another approach is to use a high multiple for your essential spending requirements, and a variable withdrawal method for your discretionary spending. Two examples of variable withdrawal methods are: 1) Using a flat 4% withdrawal rate off of each year's balance (also known as constant percentage method); or 2) Using a method such as Variable Percentage Withdrawal (VPW) (https://www.bogleheads.org/wiki/Variabl ... 0portfolio.)
Last edited by wolf359 on Mon May 17, 2021 10:20 am, edited 1 time in total.
Jack FFR1846
Posts: 13670
Joined: Tue Dec 31, 2013 7:05 am
Location: 26 miles, 385 yards west of Copley Square

Re: Clarification on 25x annual spending rule

Post by Jack FFR1846 »

4% comes from the Trinity study which found that there's a great chance of not running out of money in 30 years. Where I take issue with FIRE people is that they take those results and meld it into "I can live on 4% forever and never run out of money". Well, no. It's possible you retire at 29 and at 60 you're perilously close to running out of money. Then what?

I'm a huge proponent in setting up safety nets. What safety nets can you set up for retirement? Pad that 4%. I use a padding number of 2 times, thus my number is 50 times spending, which I do have. Another way to pad is to have at least some money in instruments that can't go down in value. I've got US Savings bonds in that category. Finally, don't have debt. As much as I can see the math in a bull market of keeping a mortgage, I can see during a bear market where one can see their portfolio lose value, lose your job and have a difficult time paying the bills. In that scenario, the bank comes to take your house. You gambled that the money you could have used to pay off the mortgage would earn more in the market. If that investment has now lost money and you're out of a job, you'd better have a good sized emergency fund or you'll be out of the house. I do also see people saying that because 2 spouses are working, they don't need an emergency fund. Well, alrighty then....since Covid only hit one indurstry, right?
Bogle: Smart Beta is stupid
AnEngineer
Posts: 554
Joined: Sat Jun 27, 2020 4:05 pm

Re: Clarification on 25x annual spending rule

Post by AnEngineer »

David Jay wrote: Mon May 17, 2021 10:09 am
AnEngineer wrote: Mon May 17, 2021 9:29 am
Jags4186 wrote: Mon May 17, 2021 9:24 am
AnEngineer wrote: Mon May 17, 2021 9:16 am
azanon wrote: Mon May 17, 2021 9:12 amBut actually using it as a system for retirement income, in the strictest sense, is pretty reckless.
Can you explain this?
While the 4% rule has a very high probability of success, it is not a 100% probability of success. Since you are 1 person living 1 life during 1 time period, it would be reckless to continue to increase your spending if your portfolio is getting destroyed. You may never recover and portfolio failure could lead to personal catastrophe.
If 100% is your standard then I suppose you consider driving to be reckless too? If your standard for "reckless" if out of step with everyone, fine, but it sounded like you had some complaint about how the 4% rule works in practice that doesn't align with the success rates reported in the studies.
Note that Jags is talking about continuing to spend the inflation-adjusted amount after a "black swan" type event in one's retirement years that has devastated one's portfolio. Because it is the only portfolio one has, it would be necessary to find a way to adjust expenses.

It feels like you read his comment differently, i.e. that the risk of retiring with 25x is too great a risk to take.

Or to use your example, he is not saying to give up driving due to risk. I believe he is saying to go ahead and take the road trip but if the "low tire pressure" light comes on while driving down the highway at 70MPH, it would be prudent to do something about it before losing control of the vehicle.
I think you're right on both counts. As Jags is putting it makes some sense: time to use the breaks if conditions don't look good. But the comment I was quoting from by azanon was saying that using the 4% rule is reckless. Maybe he meant something more specific, but I don't see it in what he wrote.
59Gibson
Posts: 377
Joined: Mon Dec 07, 2020 8:41 am

Re: Clarification on 25x annual spending rule

Post by 59Gibson »

50X for someone retiring in their 20s is probably prudent. Someone working til 50s- 60s to get to 50x to feel "safe" is extremely excessive. That's essentially planning for Armageddon, which is not a good use of time or energy.
Jags4186
Posts: 5785
Joined: Wed Jun 18, 2014 7:12 pm

Re: Clarification on 25x annual spending rule

Post by Jags4186 »

AnEngineer wrote: Mon May 17, 2021 10:25 am I think you're right on both counts. As Jags is putting it makes some sense: time to use the breaks if conditions don't look good. But the comment I was quoting from by azanon was saying that using the 4% rule is reckless. Maybe he meant something more specific, but I don't see it in what he wrote.
I know we’re all answering for everyone else which makes this confusing to follow and I apologize. The way I see it is that it is reckless to follow a rule blindly because it’s “supposed” to work. You may quickly find yourself past the point of no return. The penalty for messing up, even if it is a very small chance, is devastating. The 4% rule is a guide, not an individual plan.
AnEngineer
Posts: 554
Joined: Sat Jun 27, 2020 4:05 pm

Re: Clarification on 25x annual spending rule

Post by AnEngineer »

Jags4186 wrote: Mon May 17, 2021 10:40 am
AnEngineer wrote: Mon May 17, 2021 10:25 am I think you're right on both counts. As Jags is putting it makes some sense: time to use the breaks if conditions don't look good. But the comment I was quoting from by azanon was saying that using the 4% rule is reckless. Maybe he meant something more specific, but I don't see it in what he wrote.
I know we’re all answering for everyone else which makes this confusing to follow and I apologize. The way I see it is that it is reckless to follow a rule blindly because it’s “supposed” to work. You may quickly find yourself past the point of no return. The penalty for messing up, even if it is a very small chance, is devastating. The 4% rule is a guide, not an individual plan.
From my perspective, the key is that while overall the chance of failure may be low with 4%, once you've experienced a stock market crash you need to look at the conditional probability of success under those conditions. Combine that with the fact that, as wolf359 points out, early retirees are more likely to retire after the market has been abnormally good, that conditional probability of failure is likely to be unacceptably high. It'd be interested to see if someone has worked out the numbers, either based on historical data or Monte Carlo simulation.

BTW, thanks for the apology but you're fine. I didn't initially notice I wasn't responding to the same person (though I should have) and I appreciate David Jay interjecting the way he did.
dbr
Posts: 35886
Joined: Sun Mar 04, 2007 9:50 am

Re: Clarification on 25x annual spending rule

Post by dbr »

AnEngineer wrote: Mon May 17, 2021 10:48 am
Jags4186 wrote: Mon May 17, 2021 10:40 am
AnEngineer wrote: Mon May 17, 2021 10:25 am I think you're right on both counts. As Jags is putting it makes some sense: time to use the breaks if conditions don't look good. But the comment I was quoting from by azanon was saying that using the 4% rule is reckless. Maybe he meant something more specific, but I don't see it in what he wrote.
I know we’re all answering for everyone else which makes this confusing to follow and I apologize. The way I see it is that it is reckless to follow a rule blindly because it’s “supposed” to work. You may quickly find yourself past the point of no return. The penalty for messing up, even if it is a very small chance, is devastating. The 4% rule is a guide, not an individual plan.
From my perspective, the key is that while overall the chance of failure may be low with 4%, once you've experienced a stock market crash you need to look at the conditional probability of success under those conditions. Combine that with the fact that, as wolf359 points out, early retirees are more likely to retire after the market has been abnormally good, that conditional probability of failure is likely to be unacceptably high. It'd be interested to see if someone has worked out the numbers, either based on historical data or Monte Carlo simulation.

Yes, this is theoretically correct, including the recognition that early, or any retirement, might be taken because there has been a recent run-up in markets.

I think the big obstacle to evaluating conditional probabilities is that there is already huge uncertainty in these analyses and the data just doesn't exist to compute those things with useful margin of error.

There is a paper by Kitces that addresses the presumed contradiction that it can't be right to apply the same 4% estimate just before rather than just after a market crash. The argument is effectively a conditional probability one that if you know the market has just run up or just run down the immediate future returns should be modeled as less, respectively greater, than average. I don't have a link handy.

Also, there has been a post or posts here purporting to be a chart of safe withdrawal rate by year, and the number is highly variable. I don't know if it is correlated to returns just before retirement. In safe withdrawal rate studies such as FireCalc the bad years to retire are not necessarily correlated with market crashes but rather with being in secular bear markets or hit by the inflation peaks rising in the 1970's. That is why the mid sixties are such bad years to retire, and 1982 at the beginning of a secular bull market and decline of inflation is a good year to retire.
azanon
Posts: 3026
Joined: Mon Nov 07, 2011 10:34 am

Re: Clarification on 25x annual spending rule

Post by azanon »

AnEngineer wrote: Mon May 17, 2021 9:16 am
azanon wrote: Mon May 17, 2021 9:12 amBut actually using it as a system for retirement income, in the strictest sense, is pretty reckless.
Can you explain this?
Sure.

When someone say's they're going to use the "4% Rule" for retirement income, I assume they're implying a "Constant Dollar" method where they're going to increase subsequent years withdrawals by inflation from the initial year 1 amount. The problem with that is, "strictly speaking," you only make an initial year decision of how much to withdraw, then you are completely unopen/recalcitrant to adjusting your withdrawals based on how your actual portfolio balance is changing over time. So to me, that's inherently reckless to choose to use a system that doesn't adjust in any way to what is actually happening to your portfolio balance, whether it's done mechanically as part of a dynamic retirement system, or if you at least have some rules in your IPS that call for making adjustments annually, every 5 years, etc.

So yes, the very strict "constant dollar" and then hope and pray method, is a very bad one. Call that an opinion if you want, but I'm comfortable stating that as matter-of-fact though.
dbr
Posts: 35886
Joined: Sun Mar 04, 2007 9:50 am

Re: Clarification on 25x annual spending rule

Post by dbr »

azanon wrote: Mon May 17, 2021 11:18 am
AnEngineer wrote: Mon May 17, 2021 9:16 am
azanon wrote: Mon May 17, 2021 9:12 amBut actually using it as a system for retirement income, in the strictest sense, is pretty reckless.
Can you explain this?
Sure.

When someone say's they're going to use the "4% Rule" for retirement income, I assume they're implying a "Constant Dollar" method where they're going to increase subsequent years withdrawals by inflation from the initial year 1 amount. The problem with that is, "strictly speaking," you only make an initial year decision of how much to withdraw, then you are completely unopen/recalcitrant to adjusting your withdrawals based on how your actual portfolio balance is changing over time. So to me, that's inherently reckless to choose to use a system that doesn't adjust in any way to what is actually happening to your portfolio balance, whether it's done mechanically as part of a dynamic retirement system, or if you at least have some rules in your IPS that call for making adjustments annually, every 5 years, etc.

So yes, the very strict "constant dollar" and then hope and pray method, is a very bad one. Call that an opinion if you want, but I'm comfortable stating that as matter-of-fact though.
Indeed so. From the beginning these 4% results were intended and published as analyses of how portfolios behave and not as proposed plans for how people should manage investments in retirement. The results are probably useful for giving a person a ballpark from a distance regarding how they stand. At the time, historically, the significance was in showing how the actual statistics of variable returns increase the downside and the upside of what happens to portfolios compared to assuming fixed annual returns that are not how investments, stocks especially, actually behave.
Jags4186
Posts: 5785
Joined: Wed Jun 18, 2014 7:12 pm

Re: Clarification on 25x annual spending rule

Post by Jags4186 »

I think the reality is that while you may have a high likelihood of success with 4%, I don’t think most people including myself could handle miserly performance and continue spending at ever increasing rates.

Here’s the 4% rule in action over a 30 year period, starting with $1mm. It has a 96.7% rate of success. How many people would really be comfortable with performance underneath the black line I added in?

Image

Everything between the red line and the black line is a “success” but it seems to me to be a stressful experience.
Marseille07
Posts: 3994
Joined: Fri Nov 06, 2020 1:41 pm

Re: Clarification on 25x annual spending rule

Post by Marseille07 »

Jags4186 wrote: Mon May 17, 2021 11:39 am I think the reality is that while you may have a high likelihood of success with 4%, I don’t think most people including myself could handle miserly performance and continue spending at ever increasing rates.

Here’s the 4% rule in action over a 30 year period, starting with $1mm. It has a 96.7% rate of success. How many people would really be comfortable with performance underneath the black line I added in?

Image

Everything between the red line and the black line is a “success” but it seems to me to be a stressful experience.
A follow-up study has already been performed though. If you want to preserve the initial asset value, we pretty much have to go equities-heavy: https://i1.wp.com/earlyretirementnow.co ... table1.png
wolf359
Posts: 2399
Joined: Sun Mar 15, 2015 8:47 am

Re: Clarification on 25x annual spending rule

Post by wolf359 »

Jack FFR1846 wrote: Mon May 17, 2021 10:17 am 4% comes from the Trinity study which found that there's a great chance of not running out of money in 30 years. Where I take issue with FIRE people is that they take those results and meld it into "I can live on 4% forever and never run out of money". Well, no. It's possible you retire at 29 and at 60 you're perilously close to running out of money. Then what?
If the 4% Rule plan is going to fail, it is obvious in the first 5-10 years. If you can see disaster coming 20-25 years in advance, who would stay on that path until they're running out of money?

Put another way, would a money geek who obsesses over their safe withdrawal rate, and has the discipline, talent, and earning power to accumulate 25X their expenses by age 29, be unable to see this coming and not make adjustments? The failure rate is NOT the chance of failure. It's the chance that you might need to adjust your plan. If you build in some flexibility, you're fine. You can make the changes at age 35, not at 60.

There's an interesting read here: https://livingafi.com/2021/03/17/the-20 ... nt-update/

This was an early retirement success and failure. They didn't fail because they ran out of money. There are other things you need to do to prepare for retirement.
AnEngineer
Posts: 554
Joined: Sat Jun 27, 2020 4:05 pm

Re: Clarification on 25x annual spending rule

Post by AnEngineer »

azanon wrote: Mon May 17, 2021 11:18 am
AnEngineer wrote: Mon May 17, 2021 9:16 am
azanon wrote: Mon May 17, 2021 9:12 amBut actually using it as a system for retirement income, in the strictest sense, is pretty reckless.
Can you explain this?
Sure.

When someone say's they're going to use the "4% Rule" for retirement income, I assume they're implying a "Constant Dollar" method where they're going to increase subsequent years withdrawals by inflation from the initial year 1 amount. The problem with that is, "strictly speaking," you only make an initial year decision of how much to withdraw, then you are completely unopen/recalcitrant to adjusting your withdrawals based on how your actual portfolio balance is changing over time. So to me, that's inherently reckless to choose to use a system that doesn't adjust in any way to what is actually happening to your portfolio balance, whether it's done mechanically as part of a dynamic retirement system, or if you at least have some rules in your IPS that call for making adjustments annually, every 5 years, etc.

So yes, the very strict "constant dollar" and then hope and pray method, is a very bad one. Call that an opinion if you want, but I'm comfortable stating that as matter-of-fact though.
I think you're missing something about the 4% rule then. It shows that even if you don't adjust your withdrawals (except for inflation), then you still have a very large chance of success (the table by Marseille07 shows 95% success with 50% stocks over 30 years, 97% with 100% stocks). If you are willing to adjust spending down in the presence of poor returns, you'll do even better.

Your tone sounds like the 4% rule underpredicts the amount of money you need to retire, but your argument actually says it may overpredict it.
LiterallyIronic
Posts: 1503
Joined: Sat Dec 05, 2015 10:36 am

Re: Clarification on 25x annual spending rule

Post by LiterallyIronic »

bengal22 wrote: Sun May 16, 2021 6:15 pm Not sure why we forget the income part: ssa, pension, etc.
Because I expect those to be zero.
Independent George
Posts: 1279
Joined: Wed Feb 17, 2016 12:13 pm
Location: Chicago, IL, USA

Re: Clarification on 25x annual spending rule

Post by Independent George »

LiterallyIronic wrote: Mon May 17, 2021 11:57 am
bengal22 wrote: Sun May 16, 2021 6:15 pm Not sure why we forget the income part: ssa, pension, etc.
Because I expect those to be zero.
I don't think zero is a reasonable assumption. Pension numbers will differ based on the individual circumstances, but for Social Security, 50% - 75% of projected benefits is far more realistic - 50% as a worst case scenario, 75% as the amount after benefit cuts needed to make the system solvent under current projections.
sailaway
Posts: 3249
Joined: Fri May 12, 2017 1:11 pm

Re: Clarification on 25x annual spending rule

Post by sailaway »

Independent George wrote: Mon May 17, 2021 12:05 pm
LiterallyIronic wrote: Mon May 17, 2021 11:57 am
bengal22 wrote: Sun May 16, 2021 6:15 pm Not sure why we forget the income part: ssa, pension, etc.
Because I expect those to be zero.
I don't think zero is a reasonable assumption. Pension numbers will differ based on the individual circumstances, but for Social Security, 50% - 75% of projected benefits is far more realistic - 50% as a worst case scenario, 75% as the amount after benefit cuts needed to make the system solvent under current projections.
Personally, I figure it is zero for the first 10-20 years, might as well just use zero in the calculations. But then, my own SS will be rather puny between years teaching (but not long enough in one place for a pension) and years on a grad school stipend that was not classified as income and early retirement...
LiterallyIronic
Posts: 1503
Joined: Sat Dec 05, 2015 10:36 am

Re: Clarification on 25x annual spending rule

Post by LiterallyIronic »

Independent George wrote: Mon May 17, 2021 12:05 pm
LiterallyIronic wrote: Mon May 17, 2021 11:57 am
bengal22 wrote: Sun May 16, 2021 6:15 pm Not sure why we forget the income part: ssa, pension, etc.
Because I expect those to be zero.
I don't think zero is a reasonable assumption. Pension numbers will differ based on the individual circumstances, but for Social Security, 50% - 75% of projected benefits is far more realistic - 50% as a worst case scenario, 75% as the amount after benefit cuts needed to make the system solvent under current projections.
I don't have a pension, so that's automatically zero. For Social Security, even if stays 100% of benefits, I expect my benefits to be close enough to zero that it won't make a difference. I was in college until I was 32 and plan to retire at 50. When half of my 35 years for SS are zero (or extremely close to zero), I expect my SS benefit to be pretty much worthless. Especially if it gets slashed 25-50%.
wolf359
Posts: 2399
Joined: Sun Mar 15, 2015 8:47 am

Re: Clarification on 25x annual spending rule

Post by wolf359 »

Independent George wrote: Mon May 17, 2021 12:05 pm
LiterallyIronic wrote: Mon May 17, 2021 11:57 am
bengal22 wrote: Sun May 16, 2021 6:15 pm Not sure why we forget the income part: ssa, pension, etc.
Because I expect those to be zero.
I don't think zero is a reasonable assumption. Pension numbers will differ based on the individual circumstances, but for Social Security, 50% - 75% of projected benefits is far more realistic - 50% as a worst case scenario, 75% as the amount after benefit cuts needed to make the system solvent under current projections.
They're not forgotten. They're buffers.

I am claiming Social Security as late as possible, as cheap longevity insurance.

Realistically I expect Social Security to pay out 75% of expected benefits, in line with the Trustee's report. Therefore, I also am reserving some money in a traditional IRA to fund a QLAC annuity. This will be an additional longevity annuity that will help address the Social Security reduction.

Both the QLAC and Social Security are purely in the role of insurance. If there's an adverse investment event late in life, these income streams kicking in will reduce or eliminate the need for me to go back to work to put food on the table. Since I'm not counting on them as part of my primary retirement income plan, I do not include them in my estimates. I'm aware that their value isn't really zero, but they are for conservative planning purposes.

A paid off house is another buffer. I do not intend to use it, but if I had to, I could downsize and use the proceeds to replenish my investment nest egg. That smaller home could be used as the basis for a reverse mortgage.
tibbitts
Posts: 13846
Joined: Tue Feb 27, 2007 6:50 pm

Re: Clarification on 25x annual spending rule

Post by tibbitts »

LiterallyIronic wrote: Mon May 17, 2021 1:32 pm
Independent George wrote: Mon May 17, 2021 12:05 pm
LiterallyIronic wrote: Mon May 17, 2021 11:57 am
bengal22 wrote: Sun May 16, 2021 6:15 pm Not sure why we forget the income part: ssa, pension, etc.
Because I expect those to be zero.
I don't think zero is a reasonable assumption. Pension numbers will differ based on the individual circumstances, but for Social Security, 50% - 75% of projected benefits is far more realistic - 50% as a worst case scenario, 75% as the amount after benefit cuts needed to make the system solvent under current projections.
I don't have a pension, so that's automatically zero. For Social Security, even if stays 100% of benefits, I expect my benefits to be close enough to zero that it won't make a difference. I was in college until I was 32 and plan to retire at 50. When half of my 35 years for SS are zero (or extremely close to zero), I expect my SS benefit to be pretty much worthless. Especially if it gets slashed 25-50%.
SS benefits are heavily skewed toward lower accumulations, as the frequent discussions of break points illustrate. And while arguably the SSA's current plans would call for a across-the-board reduction, I believe almost everyone agrees that would be an unlikely scenario.
marcopolo
Posts: 4319
Joined: Sat Dec 03, 2016 10:22 am

Re: Clarification on 25x annual spending rule

Post by marcopolo »

tibbitts wrote: Mon May 17, 2021 1:57 pm
LiterallyIronic wrote: Mon May 17, 2021 1:32 pm
Independent George wrote: Mon May 17, 2021 12:05 pm
LiterallyIronic wrote: Mon May 17, 2021 11:57 am
bengal22 wrote: Sun May 16, 2021 6:15 pm Not sure why we forget the income part: ssa, pension, etc.
Because I expect those to be zero.
I don't think zero is a reasonable assumption. Pension numbers will differ based on the individual circumstances, but for Social Security, 50% - 75% of projected benefits is far more realistic - 50% as a worst case scenario, 75% as the amount after benefit cuts needed to make the system solvent under current projections.
I don't have a pension, so that's automatically zero. For Social Security, even if stays 100% of benefits, I expect my benefits to be close enough to zero that it won't make a difference. I was in college until I was 32 and plan to retire at 50. When half of my 35 years for SS are zero (or extremely close to zero), I expect my SS benefit to be pretty much worthless. Especially if it gets slashed 25-50%.
SS benefits are heavily skewed toward lower accumulations, as the frequent discussions of break points illustrate. And while arguably the SSA's current plans would call for a across-the-board reduction, I believe almost everyone agrees that would be an unlikely scenario.
For someone retiring at 50, it is not unreasonable to discount SS entirely. With 20 years to go before SS would start, the WR does not change that much (depending on anticipated expenses) when including vs not including it.
Once in a while you get shown the light, in the strangest of places if you look at it right.
User avatar
Godot
Posts: 604
Joined: Fri Jun 08, 2018 3:44 pm
Location: Little Beirut

Re: Clarification on 25x annual spending rule

Post by Godot »

Jags4186 wrote: Mon May 17, 2021 10:00 am
AnEngineer wrote: Mon May 17, 2021 9:29 am
Jags4186 wrote: Mon May 17, 2021 9:24 am
AnEngineer wrote: Mon May 17, 2021 9:16 am
azanon wrote: Mon May 17, 2021 9:12 amBut actually using it as a system for retirement income, in the strictest sense, is pretty reckless.
Can you explain this?
While the 4% rule has a very high probability of success, it is not a 100% probability of success. Since you are 1 person living 1 life during 1 time period, it would be reckless to continue to increase your spending if your portfolio is getting destroyed. You may never recover and portfolio failure could lead to personal catastrophe.
If 100% is your standard then I suppose you consider driving to be reckless too? If your standard for "reckless" if out of step with everyone, fine, but it sounded like you had some complaint about how the 4% rule works in practice that doesn't align with the success rates reported in the studies.
If the speed limit on a road is 65 MPH you might drive 65 MPH. If you’re an aggressive driver you might drive 80 MPH. If you’re a conservative driver you might drive 55 MPH. If it’s raining hard, regardless of how you normally drive, you slow down because conditions have changed. If it’s raining so hard you can’t see, you might pull over. You don’t just driving the same speed because there’s a sign on the side of the road that says 65 MPH. If you don’t slow down you have a high likelihood of ending up in a ditch.

Likewise, a retired person may be 60/40, 70/30, or 50/50 AA. They may all be withdrawing 4%, or maybe aggressive ones are withdrawing 5% or conservative ones are withdrawing 3%. But if the market crashes 50%, its reckless to continue going at the same speed until conditions improve.
That's a great analogy.
"If by Godot I had meant God I would have said God, and not Godot." | ― Samuel Beckett
azanon
Posts: 3026
Joined: Mon Nov 07, 2011 10:34 am

Re: Clarification on 25x annual spending rule

Post by azanon »

AnEngineer wrote: Mon May 17, 2021 11:57 am
azanon wrote: Mon May 17, 2021 11:18 am
AnEngineer wrote: Mon May 17, 2021 9:16 am
azanon wrote: Mon May 17, 2021 9:12 amBut actually using it as a system for retirement income, in the strictest sense, is pretty reckless.
Can you explain this?
Sure.

When someone say's they're going to use the "4% Rule" for retirement income, I assume they're implying a "Constant Dollar" method where they're going to increase subsequent years withdrawals by inflation from the initial year 1 amount. The problem with that is, "strictly speaking," you only make an initial year decision of how much to withdraw, then you are completely unopen/recalcitrant to adjusting your withdrawals based on how your actual portfolio balance is changing over time. So to me, that's inherently reckless to choose to use a system that doesn't adjust in any way to what is actually happening to your portfolio balance, whether it's done mechanically as part of a dynamic retirement system, or if you at least have some rules in your IPS that call for making adjustments annually, every 5 years, etc.

So yes, the very strict "constant dollar" and then hope and pray method, is a very bad one. Call that an opinion if you want, but I'm comfortable stating that as matter-of-fact though.
I think you're missing something about the 4% rule then. It shows that even if you don't adjust your withdrawals (except for inflation), then you still have a very large chance of success (the table by Marseille07 shows 95% success with 50% stocks over 30 years, 97% with 100% stocks). If you are willing to adjust spending down in the presence of poor returns, you'll do even better.

Your tone sounds like the 4% rule underpredicts the amount of money you need to retire, but your argument actually says it may overpredict it.
Why would I use a system with a 97% chance of success when I can use one with a 100% chance? I'll pass on a 3% chance of ruin.

And if you're going to introduce willingness of adjusting based on outcome, then my objection is completely removed. But then we're no longer talking about 4% rule/constant dollar method.
AnEngineer
Posts: 554
Joined: Sat Jun 27, 2020 4:05 pm

Re: Clarification on 25x annual spending rule

Post by AnEngineer »

azanon wrote: Tue May 18, 2021 9:39 am
AnEngineer wrote: Mon May 17, 2021 11:57 am
azanon wrote: Mon May 17, 2021 11:18 am
AnEngineer wrote: Mon May 17, 2021 9:16 am
azanon wrote: Mon May 17, 2021 9:12 amBut actually using it as a system for retirement income, in the strictest sense, is pretty reckless.
Can you explain this?
Sure.

When someone say's they're going to use the "4% Rule" for retirement income, I assume they're implying a "Constant Dollar" method where they're going to increase subsequent years withdrawals by inflation from the initial year 1 amount. The problem with that is, "strictly speaking," you only make an initial year decision of how much to withdraw, then you are completely unopen/recalcitrant to adjusting your withdrawals based on how your actual portfolio balance is changing over time. So to me, that's inherently reckless to choose to use a system that doesn't adjust in any way to what is actually happening to your portfolio balance, whether it's done mechanically as part of a dynamic retirement system, or if you at least have some rules in your IPS that call for making adjustments annually, every 5 years, etc.

So yes, the very strict "constant dollar" and then hope and pray method, is a very bad one. Call that an opinion if you want, but I'm comfortable stating that as matter-of-fact though.
I think you're missing something about the 4% rule then. It shows that even if you don't adjust your withdrawals (except for inflation), then you still have a very large chance of success (the table by Marseille07 shows 95% success with 50% stocks over 30 years, 97% with 100% stocks). If you are willing to adjust spending down in the presence of poor returns, you'll do even better.

Your tone sounds like the 4% rule underpredicts the amount of money you need to retire, but your argument actually says it may overpredict it.
Why would I use a system with a 97% chance of success when I can use one with a 100% chance? I'll pass on a 3% chance of ruin.

And if you're going to introduce willingness of adjusting based on outcome, then my objection is completely removed. But then we're no longer talking about 4% rule/constant dollar method.
I'm confused, now you're saying it's the fact that the success rate is only 97% (so 3.5% WR with 100% stocks 100% success rate is good enough?) that makes you dislike it, but earlier you said that it's "inherently reckless to choose to use a system that doesn't adjust in any way". Which is it?
delamer
Posts: 11486
Joined: Tue Feb 08, 2011 6:13 pm

Re: Clarification on 25x annual spending rule

Post by delamer »

azanon wrote: Tue May 18, 2021 9:39 am
AnEngineer wrote: Mon May 17, 2021 11:57 am
azanon wrote: Mon May 17, 2021 11:18 am
AnEngineer wrote: Mon May 17, 2021 9:16 am
azanon wrote: Mon May 17, 2021 9:12 amBut actually using it as a system for retirement income, in the strictest sense, is pretty reckless.
Can you explain this?
Sure.

When someone say's they're going to use the "4% Rule" for retirement income, I assume they're implying a "Constant Dollar" method where they're going to increase subsequent years withdrawals by inflation from the initial year 1 amount. The problem with that is, "strictly speaking," you only make an initial year decision of how much to withdraw, then you are completely unopen/recalcitrant to adjusting your withdrawals based on how your actual portfolio balance is changing over time. So to me, that's inherently reckless to choose to use a system that doesn't adjust in any way to what is actually happening to your portfolio balance, whether it's done mechanically as part of a dynamic retirement system, or if you at least have some rules in your IPS that call for making adjustments annually, every 5 years, etc.

So yes, the very strict "constant dollar" and then hope and pray method, is a very bad one. Call that an opinion if you want, but I'm comfortable stating that as matter-of-fact though.
I think you're missing something about the 4% rule then. It shows that even if you don't adjust your withdrawals (except for inflation), then you still have a very large chance of success (the table by Marseille07 shows 95% success with 50% stocks over 30 years, 97% with 100% stocks). If you are willing to adjust spending down in the presence of poor returns, you'll do even better.

Your tone sounds like the 4% rule underpredicts the amount of money you need to retire, but your argument actually says it may overpredict it.
Why would I use a system with a 97% chance of success when I can use one with a 100% chance? I'll pass on a 3% chance of ruin.

And if you're going to introduce willingness of adjusting based on outcome, then my objection is completely removed. But then we're no longer talking about 4% rule/constant dollar method.
To place any confidence in the difference between a 97% success rate and a 100% success rate given all the unknowns is misguided.

There are no guarantees.
One thing that humbles me deeply is to see that human genius has its limits while human stupidity does not. | | Alexandre Dumas, fils
dbr
Posts: 35886
Joined: Sun Mar 04, 2007 9:50 am

Re: Clarification on 25x annual spending rule

Post by dbr »

A better approach to issues like this 97% vs 100% is to test the outcomes by trying different amounts of withdrawal that score 100% and notice at what point the first failures start to appear. One can then set a contingency for withdrawal some margin less. It is also worthwhile to notice how fast the failure rate accelerates with increasing withdrawal, or with whatever one is attending to.

But in general one should imagine a margin applied to those numbers, maybe at least 5%, maybe 10%. If you are just at the edge of 100%, I would probably treat that as 90%-95% anyway. I don't know how far to move away from that edge. It is also not crazy to start at a 90%-95% success and see how it goes. Most of the outcomes in retirement withdrawal are not the worst case. Other people want a large margin against worst case issues. This could be drastically affected by how much other income one has. Also, in reality a person does not just set and hold a withdrawal rate. If for no other reason one does not because needs come and go with variation and some unpredictability. The best use of the results is to estimate from some time before retirement roughly what will be needed.
azanon
Posts: 3026
Joined: Mon Nov 07, 2011 10:34 am

Re: Clarification on 25x annual spending rule

Post by azanon »

delamer wrote: Tue May 18, 2021 10:29 am
azanon wrote: Tue May 18, 2021 9:39 am
AnEngineer wrote: Mon May 17, 2021 11:57 am
azanon wrote: Mon May 17, 2021 11:18 am
AnEngineer wrote: Mon May 17, 2021 9:16 am
Can you explain this?
Sure.

When someone say's they're going to use the "4% Rule" for retirement income, I assume they're implying a "Constant Dollar" method where they're going to increase subsequent years withdrawals by inflation from the initial year 1 amount. The problem with that is, "strictly speaking," you only make an initial year decision of how much to withdraw, then you are completely unopen/recalcitrant to adjusting your withdrawals based on how your actual portfolio balance is changing over time. So to me, that's inherently reckless to choose to use a system that doesn't adjust in any way to what is actually happening to your portfolio balance, whether it's done mechanically as part of a dynamic retirement system, or if you at least have some rules in your IPS that call for making adjustments annually, every 5 years, etc.

So yes, the very strict "constant dollar" and then hope and pray method, is a very bad one. Call that an opinion if you want, but I'm comfortable stating that as matter-of-fact though.
I think you're missing something about the 4% rule then. It shows that even if you don't adjust your withdrawals (except for inflation), then you still have a very large chance of success (the table by Marseille07 shows 95% success with 50% stocks over 30 years, 97% with 100% stocks). If you are willing to adjust spending down in the presence of poor returns, you'll do even better.

Your tone sounds like the 4% rule underpredicts the amount of money you need to retire, but your argument actually says it may overpredict it.
Why would I use a system with a 97% chance of success when I can use one with a 100% chance? I'll pass on a 3% chance of ruin.

And if you're going to introduce willingness of adjusting based on outcome, then my objection is completely removed. But then we're no longer talking about 4% rule/constant dollar method.
To place any confidence in the difference between a 97% success rate and a 100% success rate given all the unknowns is misguided.

There are no guarantees.
Really? There is no comparison at all to a system that cannot fail and one that can.

Strictly speaking (mathematically speaking), there are several methods that can't fail. While I'm not endorsing it, as just an example, the "constant percent" method cannot fail. If you only withdraw a percent of something indefinitely, it never hits zero.

I would never use a system that had a failure rate. To each his/her own.
TN_Boy
Posts: 2252
Joined: Sat Jan 17, 2009 12:51 pm

Re: Clarification on 25x annual spending rule

Post by TN_Boy »

azanon wrote: Wed May 19, 2021 11:30 am
delamer wrote: Tue May 18, 2021 10:29 am
azanon wrote: Tue May 18, 2021 9:39 am
AnEngineer wrote: Mon May 17, 2021 11:57 am
azanon wrote: Mon May 17, 2021 11:18 am

Sure.

When someone say's they're going to use the "4% Rule" for retirement income, I assume they're implying a "Constant Dollar" method where they're going to increase subsequent years withdrawals by inflation from the initial year 1 amount. The problem with that is, "strictly speaking," you only make an initial year decision of how much to withdraw, then you are completely unopen/recalcitrant to adjusting your withdrawals based on how your actual portfolio balance is changing over time. So to me, that's inherently reckless to choose to use a system that doesn't adjust in any way to what is actually happening to your portfolio balance, whether it's done mechanically as part of a dynamic retirement system, or if you at least have some rules in your IPS that call for making adjustments annually, every 5 years, etc.

So yes, the very strict "constant dollar" and then hope and pray method, is a very bad one. Call that an opinion if you want, but I'm comfortable stating that as matter-of-fact though.
I think you're missing something about the 4% rule then. It shows that even if you don't adjust your withdrawals (except for inflation), then you still have a very large chance of success (the table by Marseille07 shows 95% success with 50% stocks over 30 years, 97% with 100% stocks). If you are willing to adjust spending down in the presence of poor returns, you'll do even better.

Your tone sounds like the 4% rule underpredicts the amount of money you need to retire, but your argument actually says it may overpredict it.
Why would I use a system with a 97% chance of success when I can use one with a 100% chance? I'll pass on a 3% chance of ruin.

And if you're going to introduce willingness of adjusting based on outcome, then my objection is completely removed. But then we're no longer talking about 4% rule/constant dollar method.
To place any confidence in the difference between a 97% success rate and a 100% success rate given all the unknowns is misguided.

There are no guarantees.
Really? There is no comparison at all to a system that cannot fail and one that can.

Strictly speaking (mathematically speaking), there are several methods that can't fail. While I'm not endorsing it, as just an example, the "constant percent" method cannot fail. If you only withdraw a percent of something indefinitely, it never hits zero.

I would never use a system that had a failure rate. To each his/her own.
Every system has a non-zero failure rate.

Even the "constant percent" method has a failure mode. Ie. your withdrawal is down to $25 a year. Hard to live on that.

A lot can happen in a long retirement. I think looking for a "100%" method is not terribly useful. Maybe you have a wonderful withdrawal system but wind up with massive health care costs that you cannot pay for unless you have 10s of millions of dollars. Etc.

I think that is the larger point delamer is making. Looking for guarantees decades into the future ... there is no guarantee.
delamer
Posts: 11486
Joined: Tue Feb 08, 2011 6:13 pm

Re: Clarification on 25x annual spending rule

Post by delamer »

TN_Boy wrote: Wed May 19, 2021 11:48 am
azanon wrote: Wed May 19, 2021 11:30 am
delamer wrote: Tue May 18, 2021 10:29 am
azanon wrote: Tue May 18, 2021 9:39 am
AnEngineer wrote: Mon May 17, 2021 11:57 am

I think you're missing something about the 4% rule then. It shows that even if you don't adjust your withdrawals (except for inflation), then you still have a very large chance of success (the table by Marseille07 shows 95% success with 50% stocks over 30 years, 97% with 100% stocks). If you are willing to adjust spending down in the presence of poor returns, you'll do even better.

Your tone sounds like the 4% rule underpredicts the amount of money you need to retire, but your argument actually says it may overpredict it.
Why would I use a system with a 97% chance of success when I can use one with a 100% chance? I'll pass on a 3% chance of ruin.

And if you're going to introduce willingness of adjusting based on outcome, then my objection is completely removed. But then we're no longer talking about 4% rule/constant dollar method.
To place any confidence in the difference between a 97% success rate and a 100% success rate given all the unknowns is misguided.

There are no guarantees.
Really? There is no comparison at all to a system that cannot fail and one that can.

Strictly speaking (mathematically speaking), there are several methods that can't fail. While I'm not endorsing it, as just an example, the "constant percent" method cannot fail. If you only withdraw a percent of something indefinitely, it never hits zero.

I would never use a system that had a failure rate. To each his/her own.
Every system has a non-zero failure rate.

Even the "constant percent" method has a failure mode. Ie. your withdrawal is down to $25 a year. Hard to live on that.

A lot can happen in a long retirement. I think looking for a "100%" method is not terribly useful. Maybe you have a wonderful withdrawal system but wind up with massive health care costs that you cannot pay for unless you have 10s of millions of dollars. Etc.

I think that is the larger point delamer is making. Looking for guarantees decades into the future ... there is no guarantee.
That’s correct.

We need to define “success.” If success is not running your portfolio to zero before you die, that’s different than if success is being able to support a specific level of spending until you die.

Of course it is true that if you withdraw 5% or 10% or even 25% of your portfolio each year, you’ll never run out of money (assuming you have some diversity in your portfolio). However, you’ll probably have a highly variable level of withdrawals from year-to-year. So, therefore, you may not be able to support your preferred spending.
One thing that humbles me deeply is to see that human genius has its limits while human stupidity does not. | | Alexandre Dumas, fils
Marseille07
Posts: 3994
Joined: Fri Nov 06, 2020 1:41 pm

Re: Clarification on 25x annual spending rule

Post by Marseille07 »

azanon wrote: Wed May 19, 2021 11:30 am Really? There is no comparison at all to a system that cannot fail and one that can.

Strictly speaking (mathematically speaking), there are several methods that can't fail. While I'm not endorsing it, as just an example, the "constant percent" method cannot fail. If you only withdraw a percent of something indefinitely, it never hits zero.

I would never use a system that had a failure rate. To each his/her own.
There's a big difference between "constant-dollar" not failing (yet) vs "constant-percentage" not failing.

The former always carries a non-zero risk of failure. Some SWR just hasn't experienced it yet, given our historical data. The latter mathematically never fails.
Jack FFR1846
Posts: 13670
Joined: Tue Dec 31, 2013 7:05 am
Location: 26 miles, 385 yards west of Copley Square

Re: Clarification on 25x annual spending rule

Post by Jack FFR1846 »

59Gibson wrote: Mon May 17, 2021 10:27 am 50X for someone retiring in their 20s is probably prudent. Someone working til 50s- 60s to get to 50x to feel "safe" is extremely excessive. That's essentially planning for Armageddon, which is not a good use of time or energy.
Perhaps I should change my name to Armageddon Excessivicus mit Extremeicus.

Yes, I'm mixing pig Latin and gerbil German.
Bogle: Smart Beta is stupid
marcopolo
Posts: 4319
Joined: Sat Dec 03, 2016 10:22 am

Re: Clarification on 25x annual spending rule

Post by marcopolo »

Marseille07 wrote: Wed May 19, 2021 2:20 pm
azanon wrote: Wed May 19, 2021 11:30 am Really? There is no comparison at all to a system that cannot fail and one that can.

Strictly speaking (mathematically speaking), there are several methods that can't fail. While I'm not endorsing it, as just an example, the "constant percent" method cannot fail. If you only withdraw a percent of something indefinitely, it never hits zero.

I would never use a system that had a failure rate. To each his/her own.
There's a big difference between "constant-dollar" not failing (yet) vs "constant-percentage" not failing.

The former always carries a non-zero risk of failure. Some SWR just hasn't experienced it yet, given our historical data. The latter mathematically never fails.
Please define what you mean by success/failure.
Lets say you start with a $1M, and choose a 4% constant percentage approach, so $40k in the first year. If the constant percentage runs down the portfolio to say $394 such that this year's withdrawal is $15.76, do you still consider it a "success" since you have not technically run out of money?
Once in a while you get shown the light, in the strangest of places if you look at it right.
Marseille07
Posts: 3994
Joined: Fri Nov 06, 2020 1:41 pm

Re: Clarification on 25x annual spending rule

Post by Marseille07 »

marcopolo wrote: Wed May 19, 2021 2:43 pm Please define what you mean by success/failure.
Lets say you start with a $1M, and choose a 4% constant percentage approach, so $40k in the first year. If the constant percentage runs down the portfolio to say $394 such that this year's withdrawal is $15.76, do you still consider it a "success" since you have not technically run out of money?
Sure. I mean this discussion is academic in that the success criteria used by the Trinity Study is also simply having $0+ after 30 years.

I was just replying to the above poster saying SWR "cannot fail" is misguided, and even more misguided was to compare with "constant-percentage."
Post Reply