4% SWR - revisited

[engr]

I was wondering how many individuals set their initial withdrawal rate greater than 4% (eg 5,6%) until social security kicks in and then reduce it back to 4% after that point. Does anyone know of any studies done on this topic?

 

[REWahoo]

I was wondering how many individuals set their initial withdrawal rate greater than 4% (eg 5,6%) until social security kicks in and then reduce it back to 4% after that point. Does anyone know of any studies done on this topic?

I did - although my WR is a little under 4% now that we are both drawing SS.

Can't site any studies, but I do know my numbers worked in FIRECalc...

 

[nun]

I was wondering how many individuals set their initial withdrawal rate greater than 4% (eg 5,6%) until social security kicks in and then reduce it back to 4% after that point. Does anyone know of any studies done on this topic?

In another thread I recently proposed a 4% WR from my accounts to ER. People were worried about it being unsustainable and pointed out that years with bad returns early on would be a severe blow to my plan. However, nobody made your point. With an initial 4% WR I can cover my expenses. I'll have to do that from 50 to 66 when I expect to get SS. If I inflate my budget by 3% a year to 66 I find that SS, a small pension and rent from an apartment I own will cover all my expenses.

 

[W2R]

I was wondering how many individuals set their initial withdrawal rate greater than 4% (eg 5,6%) until social security kicks in and then reduce it back to 4% after that point. Does anyone know of any studies done on this topic?

I do plan to withdraw less after I claim SS than I am now. I am already withdrawing less than 4%, though. I have been retired only about two years, and I am still trying to get used to the idea of withdrawing anything at all.

 

[Animorph]

I could be at 6% or so for a couple of years if DW decides to retire soon. That gets the last son through college. Then I run at about 4% until SS kicks in and maybe 2% by the time everything comes on line. You can run that kind of scenario through FIRECalc. You should also pretty much have the cash for those first high-expense years (assuming 2-3 years) in hand before retiring so you can avoid problems with the market tanking just when (or slightly before) you retire.

 

[pb4uski]

I was wondering how many individuals set their initial withdrawal rate greater than 4% (eg 5,6%) until social security kicks in and then reduce it back to 4% after that point. Does anyone know of any studies done on this topic?

Not aware of any studies, but keep in mind that 4% is from now to the end of time. In my case, my withdrawal rate will be close to 5% initially, decline to 4% when I start drawing an employer defined benefit pension at 60 and will be minimal once I turn 70 and begin collecting SS.

If the withdrawal rate was much higher than 5%, I'm not sure how comfortable I would feel.

 

[REWahoo]

...and will be minimal once I turn 70 and begin collecting SS.

Will this be true even after RMD's kick in?

 

[Animorph]

Will this be true even after RMD's kick in?

When RMD's kick in my taxable portfolio comes back to life. It's not a withdrawal, just a conversion from IRA to taxable.

 

[REWahoo]

It's not a withdrawal, just a conversion from IRA to taxable.

+1

 

[engr]

I keep reading about how everyone uses firecalc so I went to take another look. It asks what the initial portfolio value is but I don't see where you specify what amount is taxable, Roth or Traditional. Since each is taxed differently, I must be missing on how to specify each amount. Also, when additional monies are requested that will be added to the portfolio before retirement I enter the amount but this is how much money I will be adding to my company's 401K plan (traditional). Again, there must be a way to specify this along with how to enter the company match. I would appreciate any inputs from anyone.

 

[pb4uski]

When RMD's kick in my taxable portfolio comes back to life. It's not a withdrawal, just a conversion from IRA to taxable.

Agreed. However, from now until 70 I will be doing conversions within the 15% tax bracket to try to minimize RMDs.

 

[REWahoo]

I keep reading about how everyone uses firecalc so I went to take another look. It asks what the initial portfolio value is but I don't see where you specify what amount is taxable, Roth or Traditional. Since each is taxed differently, I must be missing on how to specify each amount. Also, when additional monies are requested that will be added to the portfolio before retirement I enter the amount but this is how much money I will be adding to my company's 401K plan (traditional). Again, there must be a way to specify this along with how to enter the company match. I would appreciate any inputs from anyone.

FIRECalc ignores taxes. See FIRECalc: Why another retirement calculator? for more info.

When entering future contributions, FIRECalc doesn't care if they come from you or from your employer.

 

[kyounge1956]

Not aware of any studies, but keep in mind that 4% is from now to the end of time. In my case, my withdrawal rate will be close to 5% initially, decline to 4% when I start drawing an employer defined benefit pension at 60 and will be minimal once I turn 70 and begin collecting SS.

If the withdrawal rate was much higher than 5%, I'm not sure how comfortable I would feel.

Not sure by what you mean by "from now to the end of time" but IIRC the 4% withdrawal rate is based on a study for a 30 year retirement and 95% success rate. A WR lower than 4% would be required to make a portfolio last forever. My apologies if I have misunderstood your meaning.

 

[smjsl]

Here is a study mentioned that suggests 4% for 30 years is not even close in today's environment, but should be more like 2% or even less. So, if you believe it, withdrawing more than 4% with goal of later withdrawing 4% would contradict this one; especially if your time horizon is longer than 30 years, which is the time interval usually studied for regular, not early, retirement:

http://www.early-retirement.org/for...ay-be-too-high-for-todays-retirees-58220.html

 

[donheff]

FIRECalc ignores taxes. See FIRECalc: Why another retirement calculator? for more info.

When entering future contributions, FIRECalc doesn't care if they come from you or from your employer.

And engr, that is a feature, not a flaw. It is much simpler and sensible to evaluate your likely tax situation yourself than rely on the sort of broad brush guess that FIRECALC would need to apply were it to evaluate taxable and non-taxable sources. Federal, State, local, employer payroll tax (for DW's small income stream in our case), taxable portions of pensions, capital gains... Trying to guesstimate all of this in FIRECALC would be impossible. Selecting a piece to guesstimate (e.g. FIT) would be more misleading than illuminating.

 

[Moneygrubber]

Withdrawal rates

Is the 4% withdrawal rate too conservative? I have run scenarios that show my portfolio increasing to the point where I croak with millions remaining, if the market performs on average. What if you have no plans to pass money to heirs? Does that change the calculation?
.

 

[timeasterday]

Is the 4% withdrawal rate too conservative? I have run scenarios that show my portfolio increasing to the point where I croak with millions remaining, if the market performs on average. What if you have no plans to pass money to heirs? Does that change the calculation?
.

It might be too conservative for some. We don't have kids so we are not planning to pass any money on to heirs. I am sure we will take out much more than 4% for a few years before I start taking SS, then reduce it. My wife is 7 years younger so another 7 years after my SS, her SS kicks in and at that point we are probably just taking out the RMD.

Personally I don't get too caught up in SWR talk. I have to take out what I need to live and I am pretty confident I will have more than enough.

 

[Alan]

I was wondering how many individuals set their initial withdrawal rate greater than 4% (eg 5,6%) until social security kicks in and then reduce it back to 4% after that point. Does anyone know of any studies done on this topic?

That seems to be a valid approach, and if you use a calculator like FIDO's Retirement Income Planner it will show you year by year where your income is coming from to match your budgeted expenses.

When we RE'ed a couple of years ago, 70% of our planned income came from non-COLA'ed pensions so our withdrawals each year have to make up for inflation for the whole sum, not just on the initial withdrawal. Consequently I set the target SWR at 3% and can see the projected withdrawals increase the WR percentage until we start collecting SS.

 

[Nords]

Here is a study mentioned that suggests 4% for 30 years is not even close in today's environment, but should be more like 2% or even less. So, if you believe it, withdrawing more than 4% with goal of later withdrawing 4% would contradict this one; especially if your time horizon is longer than 30 years, which is the time interval usually studied for regular, not early, retirement:

Maintaining flexibility with retirement spending is important.

I think there are two flaws with that study.

First, it's trying to determine the max withdrawal rate-- the maximum that a retiree could spend every year without paying any attention to any outside economic factors. It's similar to Raddr's hapless Y2K ER who's been blissfully ignorantly spending his 4% every year since Y2K, no matter how dire the projections seem.

Real human ERs don't do that. Some would say that humans can't do that, whether they ER or not. Everyone reduces their spending during bad economic times. "Unfortunately" for the research community, it's darn hard to model... except for Bob Clyatt's system.

Second flaw: Assuming that a 4% SWR means starting at 4%. Yeah, maybe that's what the procedure says, but the math is that some scenarios will start out at 5% or even 6% during down economies. The theory is that the economy will recover, so will the portfolio, and what started as a 6% withdrawal rate will ease off to less. The overall portfolio spending will run out at the same time as the rest because it's followed the overall 4% math.

However the math's not very reassuring. I don't know about you guys, but I'd be terrified to ER at a 6% initial withdrawal while expecting the bear market to go away in a few years. But my feelings are irrelevant because I'd never be able to convince my spouse to have that sort of faith in either the market or in me.

Yet Ty Bernicke has shown (anecdotally, anyway) that annual spending drops over ER. Another study recommends starting at 4%, monitoring the portfolio until it approaches "game over", and then shifting to annuities. Social Security might even be the only annuity necessary. Surely there's an intermediate "gray area" among these alternate scenarios where variable spending and some annuities would boost success.

I think the FIRECalc advantage is being able to run a portfolio and a withdrawal rate against history-- because we humans are too quick to assume that "this time it's different" (and its corollary, "... but it will never change again".) If FIRECalc gives encouraging results (success ratio over 80%) then I'd really scrub the numbers hard for variable spending. I'd annuitize a bare-bones income and invest the rest in value & dividend stocks. I'd definitely cut back to part-time work as soon as I could.

The "problem" with a 100% successful SWR is that it practically guarantees you will have at least a penny left over. In fact, the effort & assets required to boost the success rate from 95% to 100% means that at least 95% of the time you'll have a lot of money left over.

I much prefer using FIRECalc as an initial goal and then adding safety features like Otar's "red light" annuity system or variable spending.

Those insisting that 4% is over should read Bernstein's "Calculator from Hell" articles in the FAQ archives and then consider (at a minimum) Bob Clyatt's 4%/95% rule.

Speaking of "this time it's different": When my daughter was in high school she read a lot of Jane Austen. (I certainly never would have learned what follows, because I don't have the patience to read Austen.) Our daughter had heard us talking about 4%, so she was excited to see Jane referring to "four percenters". It turns out that 19th-century British investors would put their portfolio in long-term 4% bonds and live off their income. Due to the British colonial system and their gold standard, they were effectively maintaining inflation (in the British Isles, anyway) at zero percent-- for an entire century. Note that this modern research article recommends also investing a portion of an ER portfolio in inflation-adjusted bonds. The 19th-century four-percenters would've felt right at home with TIPS.

 

[fh2000]

I was wondering how many individuals set their initial withdrawal rate greater than 4% (eg 5,6%) until social security kicks in and then reduce it back to 4% after that point. Does anyone know of any studies done on this topic?

I am 55. Wife is 50. We are not retired yet but plan to be in 2-3 years, so this is a similar question that we are also looking for an answer. Wife said that when we get to our 70's, we will very likely spend less than our 50's/60's.

Inside FIRECalc, under Spending Model, it is listing Ty Bernicke's "Reality Retirement Planning: A New Paradigm for an Old Science" as an option. I find that if I switch from [Constant Spending Power] option to [Bernicke's Reality Retirement Plan] option, I could increase my initial WR from 4% to 5%, and FIRECalc still gives a much more favorable result.

I will likely do something similar to [Bernicke's Reality Retirement Plan] option, though smaller reduction than his 2-4% reduction each year.

 

[rescueme]

That seems to be a valid approach, and if you use a calculator like FIDO's Retirement Income Planner it will show you year by year where your income is coming from to match your budgeted expenses.

+1 (with comments).

You must be a FIDO (Fidelity) customer to have access to the program. In addition, FIDO offers three different forecast software programs, with Retirement Income Planner (RIP) being the most "robust".

As a side note, if you use it, you should also use the Fidelity "Full View" product, with all your various holdings (regardless of vendor). The output goes directly to RIP, but more importantly FV interfaces to M* to get a complete breakdown of your holdings. You can ceratinly manually add your info to RIP, but this ensues a better forecast plan, IMHO.

BTW, FV is actually Yodlee "under the covers" to acquire your account/holding information. If you are not comfortable with Yodlee accessing your accounts to gather info, you probably will have a problem with FV and will have to use manual "guesstimates" of your holding breakdown within RIP. Just an FYI...

 

[smjsl]

I think there are two flaws with that study.

First, it's trying to determine the max withdrawal rate-- the maximum that a retiree could spend every year without paying any attention to any outside economic factors.
[...]

Second flaw: Assuming that a 4% SWR means starting at 4%. Yeah, maybe that's what the procedure says, but the math is that some scenarios will start out at 5% or even 6% during down economies. The theory is that the economy will recover, so will the portfolio, and what started as a 6% withdrawal rate will ease off to less. The overall portfolio spending will run out at the same time as the rest because it's followed the overall 4% math.

Hi Nords, I am not following your second flaw. Where does the math suggest 5-6% withdrawal? I get a sense that it's related to first flaw you mentioned however (i.e. flexible spending is not accounted for).

Regarding the first flaw, I don't see it as a big flaw for the following reason. Imagine your expenses consisting of 2 parts - fixed part and flexible part. Fixed part is what you MUST spend to live on; i.e. bare-minimum expenses like shelter, food, health, necessary repairs, insurance, etc.. Second part is flexible-expenses - not must-haves, but nice-to-haves that can be cut during recessions, perhaps some hobbies, going out, and such.

What you are saying is that the study does not account for these two categories where second set of expenses could vary from $0 to $X. However, if you believe the math in the study, it does suggest that the fixed expenses can be at most 2% to be safe (1.8% in 2010). So, if a retiree in 2010 decided they can live on 4%, they really should be able to tighten their belts by a factor of ~2 in "bad" times according to the authors...

Actually, even more strongly, if 2% is the constant max SWR, then your tight years and spend-thrift years may need to average around 2% with the flexible plan as well, or at least it seems like a reasonable starting point (instead of 4%)... so, perhaps the implication then is that during good times you have to live on less than 2%, and during good times on more than 2% but you have to ensure it somehow averages out to something close to 2%? Well, probably a bit more than 2% since you get positive effect from tightening when your portfolio is doing bad, but I doubt this positive effect would get you to even 3%, let alone 4%... ?

If I follow the logic of the study, it's advantage over FIREcalc is that it explains why you get what you get in FIREcalc based on the outside economic factors at the time of retirement, and thus can better predict what you would get if you were to start your retirement today by taking both history and current conditions into account.

 

[panacea]

Unless I'm overlooking something, I don't think it matters what your initial draw rate is for a short period of time, assuming you can realistically get back to a 4% draw rate once Social Security, pension or other sources of income kick in. For example, I have $1 million dollars and I pull $60,000 (or 6%) at age 60. Then I do the same, or increase, the w/d for inflation at 61, 62, etc until age 66 when Social Security starts. Now let's say I only have $600,000 left at age 66 but I still need $60,000 but Soc Sec covers $36,000 so I only need $24,000 (4% of $600,000) from the portfolio I think that still works. I don't think it's any different than retiring at at 66 with $600,000 and applying for Social Security immediately. I've been thinking about this issue lately as well, so if someone can catch a flaw in my thinking please jump on it.

 

[jebmke]

Unless I'm overlooking something, I don't think it matters what your initial draw rate is for a short period of time, assuming you can realistically get back to a 4% draw rate once Social Security, pension or other sources of income kick in. For example, I have $1 million dollars and I pull $60,000 (or 6%) at age 60. Then I do the same, or increase, the w/d for inflation at 61, 62, etc until age 66 when Social Security starts. Now let's say I only have $600,000 left at age 66 but I still need $60,000 but Soc Sec covers $36,000 so I only need $24,000 (4% of $600,000) from the portfolio I think that still works. I don't think it's any different than retiring at at 66 with $600,000 and applying for Social Security immediately. I've been thinking about this issue lately as well, so if someone can catch a flaw in my thinking please jump on it.

It seems to me that the more you draw up front the more vulnerable you are to a bad sequence of returns. But the sequence problem exists for everyone -- it is just amplified if the draw is higher.

 

[ziggy29]

Unless I'm overlooking something, I don't think it matters what your initial draw rate is for a short period of time, assuming you can realistically get back to a 4% draw rate once Social Security, pension or other sources of income kick in. For example, I have $1 million dollars and I pull $60,000 (or 6%) at age 60. Then I do the same, or increase, the w/d for inflation at 61, 62, etc until age 66 when Social Security starts. Now let's say I only have $600,000 left at age 66 but I still need $60,000 but Soc Sec covers $36,000 so I only need $24,000 (4% of $600,000) from the portfolio I think that still works. I don't think it's any different than retiring at at 66 with $600,000 and applying for Social Security immediately. I've been thinking about this issue lately as well, so if someone can catch a flaw in my thinking please jump on it.

I'd say it partially depends on what Mister Market does while you are accelerating the depletion. If it picks that time to go south and stay south, it makes it less certain that the 4% will be as much as you'll need to be once other income sources kick in.