Acceptable "safe spending rate" for a long time horizon?

[inquisitiveinvestor]

Basically I'm trying to determine what level I can sustainably "live it up" to while retaining the ability to stop working at any point in time. I've saved up enough to where I could use the "traditional" SWR rate of 4% with some modest changes to my current spending habits (relocation within the same metro area). Although this is difficult to account for, this includes an assumed 10% of spending on health care (assuming health insurance subsidy of premiums above 9.5% of income through the PPACA exchange subsidies).

However, a lot of what I've read has me second guessing, as 3% is becoming the more commonly stated SWR for a typical retirement age, and many even reporting 1% or 2% as their SWR.

The likelihood that I would stop working completely and never go back (which would be a very early retirement) is relatively low, as I would probably want to do something part time for a while just to keep myself active and reduce the temptation to spend during idle time. I also want to avoid getting used to a higher spending level in the short term only to have to reduce it later, but would like to be able to enjoy as much as possible without endangering my financial independence.

[heeyy_joe]

3%.

[gauss]

I am planning around 2.5% for a 50 year time horizon, but this also assumes that both of our pensions will be solvent and that our accrued Social Security benefits will not be whacked by more than 33%.

The 2.5% falls out from our anticipated spending and our total assets and income streams.

That is to say that I did not target 2.5% ahead of time, but felt good about the result when thats what it turned out to be.


-gauss

[jerome len]

I can remember when everyone thought you could take out 5%, then 4%, now many say 3%........I'm taking 2%......to be safe.....and when I reach 75 and have enough, I'll change it to 3%.

[NanoSour]

Have a look at ESPlanner. This is exactly what it is designed to do.

esplanner.com

[Htown Harry]

Inquisitive, you may be interested in reading some of recent publications by academic researcher Wade Pfau (also search here for several threads discussing his articles.) He has a very readable writing style. With each article I have read, my understanding of the pluses and minuses of the 4% rule (and SWR calcualtions in general) has increased .

Articles and Columns

The 4% rule of the original Trinity study is based on the historical rates and sequences of market returns. In this article, he digs into the current low-interest, low real return bond market and extrapolates a higher number of "failures" of the 4% SWR when those diminished real bond returns are projected far out into the future:

Retirement Researcher Blog: New research article: The 4% Rule is Not Safe in a Low-Yield World

Will that be the case? Who knows, but "past performance does not guarantee future results" scenarios like that one lead some cautious posters to plan for lower SWR's.

My own approach, still in the formative stages, will be one where the withdrawal rate will ebb and flow a bit with the markets. Something like "Inflation-adjusted 4% when the market has a good year, less when it doesn't".

[audreyh1]

Quote:

Originally Posted by Htown Harry Inquisitive, you may be interested in reading some of recent publications by academic researcher Wade Pfau (also search here for several threads discussing his articles.) He has a very readable writing style. With each article I have read, my understanding of the pluses and minuses of the 4% rule (and SWR calcualtions in general) has increased . Articles and Columns The 4% rule of the original Trinity study is based on the historical rates and sequences of market returns. In this article, he digs into the current low-interest, low real return bond market and extrapolates a higher number of "failures" of the 4% SWR when those diminished real bond returns are projected far out into the future: Retirement Researcher Blog: New research article: The 4% Rule is Not Safe in a Low-Yield World Will that be the case? Who knows, but "past performance does not guarantee future results" scenarios like that one lead some cautious posters to plan for lower SWR's. My own approach, still in the formative stages, will be one where the withdrawal rate will ebb and flow a bit with the markets. Something like "Inflation-adjusted 4% when the market has a good year, less when it doesn't".

Don't forget that Pfau also assumed 1% in mutual fund fees, something that wasn't present in the original studies. He made several other tweaks based on non-US historical data and modeling.

Regardless, I only ever considered 4% safe for a 30 year time horizon, not for longer periods.

[Midpack]

Many consider 3% the unlimited duration WR based on past history. The (too) often cited "4% SWR" is based on a 30 year horizon, 95% prob of success with a 60% ± equity asset allocation.

Beyond years/duration, what you ultimately use depends on what probability of success you need to sleep at night, what AA you're comfortable with (how much in equity), how elastic your annual spending plan is and how robust your plan B options are among other things. Some people plan on 4-5%, others 2% or less, and everything in between. And inflation adjusted withdrawals (what SWR is based on) is only one approach, some prefer % remaining portfolio, SPIAs, spending dividends only (don't touch principal) to name a few.

Here are some previous threads on the subject if you're interested (and haven't seen them).

http://www.early-retirement.org/foru...swr-54608.html

http://www.early-retirement.org/foru...tes-61350.html

[ziggy29]

We've talked abut this a lot. As far as I can tell, 4% hasn't historically failed BUT it wasn't really intended for more than about a 30-year horizon in many studies, and a lot of folks seem to think the "game" has changed enough that the historical real returns over the last 80+ years which made that math work are not likely moving forward.

Hence, many people are looking at 3%, not 4% -- and some even less than that.

[shotgunner]

FIRECALC, is it a trusted tool? When I plug in some rather simple models, (example; $40K a year, $975,000 in assets, $16K SS starting in '19, 45% assets in market, 5 year treasury rates for fixed income) comes up with a WR just over 5% for 30 year horizon and 4.27% for 50 year horizon for 100% success probability using Investigate tab.

Worrying about running out of money is a real fear for many, dying without enjoying one's money while we are here is seldom spoken of, just saying.

[Katsmeow]

Quote:

Originally Posted by inquisitiveinvestor However, a lot of what I've read has me second guessing, as 3% is becoming the more commonly stated SWR for a typical retirement age, and many even reporting 1% or 2% as their SWR.

I am not entirely sure what time horizon you are considering a "long" one? You refer in your post to 3% for a typical retirement age. I'm not sure frankly that 3% is "more commonly stated SWR for a typical retirement age" which I would consider to be mid-60s. For that retirement age, usually the time horizon planned for is 30 years (which for the vast majority of people is overkill).

So, when you ask about the SWR for a long time horizon do you mean 30 years or 40 years or 50 years or what?

FWIW I think that life expectancy should be factored in. A man retiring at 65 has a 6% chance of living to 95, a woman 13%. The change that one of them will survive to 95 is 18%. I do think that one should plan for that possibility (especially women) however the degree of certainty that one might need for a plan might well be different for getting to say 90 than getting to 95 or 100.

[Katsmeow]

Quote:

Originally Posted by shotgunner FIRECALC, is it a trusted tool? When I plug in some rather simple models, (example; $40K a year, $975,000 in assets, $16K SS starting in '19, 45% assets in market, 5 year treasury rates for fixed income) comes up with a WR just over 5% for 30 year horizon and 4.27% for 50 year horizon for 100% success probability using Investigate tab.

Well, Firecalc is assuming that you first spend whatever you get from SS and then withdraw the balance you need from your portfolio.

Before taking SS, Firecalc will first draw from the portfolio and for that period of time withdrawals may exceed 4% by a substantial amount. However, once SS is taken then withdrawals may then go down to much lower than 4% as much of the spending is first covered by SS.

[Midpack]

Quote:

Originally Posted by shotgunner FIRECALC, is it a trusted tool? When I plug in some rather simple models, (example; $40K a year, $975,000 in assets, $16K SS starting in '19, 45% assets in market, 5 year treasury rates for fixed income) comes up with a WR just over 5% for 30 year horizon and 4.27% for 50 year horizon for 100% success probability using Investigate tab. Worrying about running out of money is a real fear for many, dying without enjoying one's money while we are here is seldom spoken of, just saying.

Depending on what you entered and which investigate choice you selected, the output can be a little confusing (you're not the first to ask) - example below (ignore the numbers themselves). The WR result includes any SS or pension, it's actually not the WR from your portfolio (or it may be until SS kicks in, and then less thereafter). So if the results shows 5% WR and you spend $40K/yr with $16K/yr from SS - you're actually withdrawing $24K/yr from your portfolio and your portfolio withdrawal rate is 3%. You may know all this, but just in case...again, you wouldn't be the first to ask. The first time I saw that result I thought to myself 'no way!'

Quote:

A spending level of $49,040 provided a success rate of 100.0% (111 total cycles, of which 0 failed). This spending level is 5.03% of your starting portfolio. (Your spending is assumed to come from any Social Security and pensions you entered, as well as from the portfolio.)

[shotgunner]

Quote:

Originally Posted by Midpack Depending on what you entered and which investigate choice you selected, the output can be a little confusing (you're not the first to ask) - example below (ignore the numbers themselves). The WR result includes any SS or pension, it's actually not the WR from your portfolio (or it may be until SS kicks in, and then less thereafter). So if the results shows 5% WR and you spend $40K/yr with $16K/yr from SS - you're actually withdrawing $24K/yr from your portfolio and your portfolio withdrawal rate is 3%. You may know all this, but just in case...again, you wouldn't be the first to ask. The first time I saw that result I thought to myself 'no way!'

Thanks, I guess I could have made my point more clear and that is an early retiree (many of us) could have a WR greater than 4% in the (early retirement years) before SS, any pensions or other source of income kicks in and still be 100% per FIRECALC. To hold the line to 4% and under WR before normal retirement age (if other sources of income are anticipated) could cause an individual to miss the opportunity to do things that money allows in years where they have a higher probability of being alive and in good health.

[Katsmeow]

I agree. It is very common that those retiring before SS or pensions come online may well have a withdrawal rate well above 4% early on. I am semi-retired and still do some work (DH is fully retired). We will have some higher expenses the next 3 or 4 years due to higher education costs for kids still at home. If I fully retire, I won't be eligible for SS for another few years (and may choose to defer beyond that) so our withdrawal rate for about 3 years will be really high. I expect that our portfolio, say, 5 years from now will be less than it is now as we deplete some of it to bridge that gap. However, both Firecalc and Fidelity RIP say we would be fine even so since our withdrawal rate after DH and I are both on SS will be less than 4%.

[ejman]

Quote:

Originally Posted by Katsmeow I am not entirely sure what time horizon you are considering a "long" one? You refer in your post to 3% for a typical retirement age. I'm not sure frankly that 3% is "more commonly stated SWR for a typical retirement age" which I would consider to be mid-60s. For that retirement age, usually the time horizon planned for is 30 years (which for the vast majority of people is overkill). So, when you ask about the SWR for a long time horizon do you mean 30 years or 40 years or 50 years or what? FWIW I think that life expectancy should be factored in. A man retiring at 65 has a 6% chance of living to 95, a woman 13%. The change that one of them will survive to 95 is 18%. I do think that one should plan for that possibility (especially women) however the degree of certainty that one might need for a plan might well be different for getting to say 90 than getting to 95 or 100.

I'm having a difficult time understanding the drive to have the capacity of very large drawdowns latter in life (say after age 95) by dramatically reducing drawdowns early on. I don't personally know any 95+ age people that spend much on anything beyond simple life maintenance. My parents are both in that category. My mother has had dementia for some time and all the money in the world wouldn't make one bit of difference. My father's world now consists of reading, family and an occasional glass of wine. He traveled and wined and dined most excellently but come the very early 90's the desire for such just vanished.

Am sure there are examples of 100+ year olds that still travel the world and have an interest in exotic vacations with all the trimmings but the combination of the chances of both making it to that age and having the desire for such activities seem very small. To deprive oneself over a long retirement for the very remote chance of such a situation seems an interesting choice.

[braumeister]

Quote:

Originally Posted by ejman I don't personally know any 95+ age people that spend much on anything beyond simple life maintenance.

Your comments make perfect sense, of course. But I think that what people are worried about is that 30-40 years from now the cost of simple life maintenance might be astronomical (in today's dollars).

[Htown Harry]

Quote:

Originally Posted by Midpack ...inflation adjusted withdrawals (what SWR is based on) is only one approach, some prefer % remaining portfolio, SPIAs, spending dividends only (don't touch principal) to name a few.

Quote:

Originally Posted by ejman I'm having a difficult time understanding the drive to have the capacity of very large drawdowns latter in life (say after age 95) by dramatically reducing drawdowns early on.

Ejman, you're not wrong in your assessment. Unfortunately, there is limited ability for general purpose retirement cash flow tools to adjust and model the "too old to get out and spend money" years.

One variation that can be used if you want to explore further would be to start retirement with a portfolio of $X, then immediately spend a portion of it on a deferred SPIA (longevity insurance). So if $Y is the cost of an annuity to provide payments beginning at age 85, you would run your model based on X - Y = Z, where Z is the money available for retirement that has a barely-tolerable probability of providing safe withdrawals from retirement to 85.

Firecalc does have an option to model some reductions.

Quote:

[ check here to cash flow using] Bernicke's Reality Retirement Plan: Start with the constant spending power model, but use Bernicke's "Reality Retirement Plan". Current age: Note: If you indicate you are 56 or older, then the spending will be reduced immediately. Ty Bernicke's Reality Retirement Planning: A New Paradigm for an Old Science describes extensive research showing that most people see significant reductions in spending with age (not related to reduced assets or income). If selected, this option will reduce your inflation-adjusted yearly spending by 2-3% per year starting at age 56, and then stabilizing at age 76 to keep up with inflation. You should read his article for details if you plan to use this option.

[photoguy]

Quote:

Originally Posted by shotgunner FIRECALC, is it a trusted tool? When I plug in some rather simple models, (example; $40K a year, $975,000 in assets, $16K SS starting in '19, 45% assets in market, 5 year treasury rates for fixed income) comes up with a WR just over 5% for 30 year horizon and 4.27% for 50 year horizon for 100% success probability using Investigate tab.

It's a *great* tool. But keep in mind that all models have limitations and assumptions. For example, it doesn't take into account market valuations (see all the threads on PE10 and expected returns). It also assumes that as long as you have any money left over, the run is a success. As a hopefull ER my success criteria for a 30 year run would probably be to have at least 50-75% (inflation adjusted) of my initial portfolio.

[ziggy29]

Quote:

Originally Posted by photoguy It also assumes that as long as you have any money left over, the run is a success. As a hopefull ER my success criteria for a 30 year run would probably be to have at least 50-75% (inflation adjusted) of my initial portfolio.

This is a good point. We may all differ in what we think are the acceptable drawdowns, but I think it's obvious that if even a "successful" run took us within a hair's width of failing before recovering, the stress we felt in fearing the money was about to run out before my lifespan ran out would feel like anything but a "successful" retirement. If we realized we had (say) only about 3 years left of needed expenses left to our name, even if it later recovered to (say) 10 years we'd live pretty freaked out.