6.3% withdrawals, 100% survivable

[rigel]

Follow-up on my first post on variable withdraw strategy.

Using a flat 4% just isn't what I will be willing to do. If the market is down, I'm not comfortable taking 4% out at the bottom, after several down years (think 1929) I would take even less. In good times, I would like to take more. Mostly, I want to maximize the money I take out between 60-80 without going broke later and I also don't want to leave $Ms to someone else, I want to spend it, but without risking going broke.

So, I developed a stripped down, custom FIREcalc to test this - it doesn't have all the features, but it does test this theory.

Inputs: 100% stock portfolio, 40 year retirement (have down 30 yr also with similar results), CPI inflation

I look up the Market return minus inflation in the following table and get this year's withdrawal rate
:
Return - Inflation / Withdrawal
17% 10.0%
16% 9.0%
15% 8.0%
14% 7.5%
13% 7.0%
12% 6.5%
11% 6.0%
10% 5.5%
9% 5.0%
8% 4.5%
7% 4.5%
6% 4.0%
5% 4.0%
4% 3.5%
3% 3.0%
2% 2.5%

Then I look up age in a table like this (only part of the table)

Age Maximum Withdrawal
55 2.5%
56 2.5%
57 2.5%
58 3.0%
59 6.0%
60 9.0%
61 9.0%
62 9.0%
63 9.0%
64 10.0%
65 10.0%
66 10.0%

I then take the smaller number as the withdraw rate for that year.

If the real dollar value of the portfolio is less than 50% of the starting value, I over ride the withdraw and only take 1.5% This fixed going broke in 1929 and 1966 and makes a lot of sense to me!

If the real dollar value of the portfio is greater than 125%, than I add an additional 4% withdraw to the withdraw from above. If real dollar value is greater than 150%, then I add 10%. This helps spend money in the good times! That's about it. Here are the results:

Starting with 1871 data through 2006:

Number of Cycles 97
Number of Successful Cycles 97
Percent Surviability 100%

Min ending portfolio value in real dollars as percent of beginning portfolio 56 %
Median ending portfolio value in real dollars as % of beginning portfolio 105 %
Max ending portfolio value in real dollars as percent of beginning portfolio 247 %

Min avg withdrawal rate 3.84%
Median avg withdrawal rate 6.26%
Max avg withdrawal rate 13.63%

Worst cycle 3.84%
2nd worst cycle 3.89%
3rd worst cycle 4.03%
4th worst cycle 4.10%
5th worst cycle 4.11%

Cycles between 0 - 4 2
Cycles between 4 - 4.5 13
Cycles between 4.5 - 5 6
Cycles between 5 - 6 22
Cycles between 6 -7 13
Cycles between 7- 8 6
Cycles between 8 - 9 8
Cycles between 9+ 27

I think these rules are common sense, are not data mining, just trying to squeeze more money out of the portfolio by selling (spending) more in good times and selling less in bad (belt tightening in poor times).

The results I'm getting says one can safe squeeze, in most market conditions, a lot more than 4% from the portfolio.

Anyone else have some similar calculations?

 

[samclem]

If the real dollar value of the portfolio is less than 50% of the starting value, I over ride the withdraw and only take 1.5% This fixed going broke in 1929 and 1966 and makes a lot of sense to me!

If the real dollar value of the portfio is greater than 125%, than I add an additional 4% withdraw to the withdraw from above. If real dollar value is greater than 150%, then I add 10%. This helps spend money in the good times! That's about it. Here are the results:

Are the 50% amounts and 150% amounts relative to the original portfolio value corrected for inflation?

This seems like a reaonable approach, though I haven't done any calculations that would confirm your results. I would have a tough time withdrawing and immediately spending 10% of my porfolio under almost any circumstances, but could certainly see pting inot a "fun fund" to be spent over several years (that's just psychological, has nothing to do with whether your idea is sound). Some other thoughts/points for ya to chew on: It's not clear what you mean by the "flat 4%"strategy that you are rejecting--is it 4% of the starting value corrected for inflation every year, or is it 4% of the portfolio value every year? As you probably know, the variable withdrawal strategy (basing withdrawals on the ending value of the portfolio every year) is avaialbale in the advanced FIRECALC version, though there's nothing quite like the rules you've got here. We've talked a lot about varios versions of this idea, you might find some useful stuff using the search function. If you haven't been there yet, check out Gummy's site for a very good discussion of the variable withdrawal idea, with a simulator/calculator. I think you might be able to "trick" the Monte carlo simulator there to replicate something like your withdrawal plan and thereby get another check on its viability.
http://www.gummy-stuff.org/sensible_withdrawals.htm

 

[clifp]

Sounds interesting, but I don't understand what is magic about 59 when your max withdrawals jumps from 3% to 6%.

While I like the idea of being able to spend more when the market is doing well, it seems that your spending is going to have huge swings.

Also what is the max spending rate when the total value of your portfolio declines by say 10% in year, but the real value of the portfolio is greater than 50% and less than 125%?

 

[rigel]

Samclem - the 50% and 150% values are inflation adjusted percents of the original, starting portfolio value.

The 4% of the original portfolio, adjusted for inflation is a great starting point, however, for me, I just couldn't take that much out if the market is -10% for the year.

My model is based on this - the withdraw percents are against the starting portfolio value, inflation adjusted.

I'm not really a monte carlo guy, as I like the concept of history - market returns, inflation rates have some relationship.

I have followed FIRE since about 1997, and what the folks have done with it over the years is most excellent. Dory had a 1998 variable withdraw excel model, that I liked a lot.

I just wanted to explore a more variable, rule based withdraw strategy. Leaving millions to someone else is not my plan. I also don't want to be sitting on millions when I'm 90 and can't figure out how to enjoy it. So, I built this tool to explore the safety of maximizing withdraws between 60-80, but still keep out of trouble. Worse than having too much money at the end, is too little!

 

[rigel]

clifp - no magic - the withdraw strategies are configurable, not hard coded, so I can look at a wide variety of options.

For me personally - I plan to retire at 55, but my wife, being much younger than, will have the pleasure of continuing to work another 5 years - so, between 55-59, our money needs are less as we ease into retirement, while retaining some current payroll income.

If, in real dollars, the portfolio value is between 50% and 125%, I use an inflation adjusted market return, in your example, -10%, and see I can only take 2.5%, the lowest value in the table, based on age, I would like to take 10%, so I take the lower of the values - 2.5% of the original portfolio value, inflation adjusted.

 

[Rich_by_the_Bay]

Generally speaking, most articles on fixed percent withdrawals "allow" a higher overall income than an inflation adjusted initial 4%, at the risk of certain increased volatility -- I assume because the risk of running out of money is never a concern with the former. I'd probably be comfortable with 4.25% or 4.5% using fixed percent of total.

ESRBob's simple 95% rule can act as a safety net to avoid precipitous drops in income year-to-year, which the OP's plan virtually guarantees. I sure like the simplicty of the "95% rule" with a little common sense thrown in. It's been back-tested and passes my "sniff test" the way it is supported in his book.

Rigel's plan is interesting but I would not be comfortable with year to year variations in income of up to 50%. For that matter, I imagine there are many who could not quite make it on 2.5% to 3% in the early pre-age-60 years when active FIRE is the goal. Even if it is arithmetically correct, it doesn't match any real-world lifestyle I can foresee.

II reckon we've all looked for systems to allow a larger SWR then that measly 4-4.5% but -- especially if you plan to leave any kind of farewell present to your heirs -- I think that's as good as it gets.

But each to his or her own. If that plan meets your personal needs and goals, go for it. Too chaotic for me.

 

[Gone4Good]

I'm relatively certain that these crazy withdrawal scheme posts are jokes, but I'll bite anyway:

Have you tried to put in some actual numbers to see what your spending would look like under these rules?

If the real dollar value of the portfolio is less than 50% of the starting value, I over ride the withdraw and only take 1.5%

So you're planning to take 1.5% of a portfolio that is 50% smaller than what you started with? So if you started with a $2MM portfolio you think you can live on $15,000??

While its true one can theoretically achieve 100% success by drawing down a portfolio asymptotically to zero, spending is not infinitely flexible to the downside.

 

[rigel]

First, let me say that I don't appreciate being told that my post is a joke.

I started with the 1997 Dory36 Excel model of variable withdrawals. I have updated it with market returns, cpi, pii, from 1997 to 2006.

Second, your post contains a math error. 1.5% of 2MM is 30k.

Third, if your portfolio is down 50% in real dollars from the start of your retirement, aka, 1929, are you really blinding going to pull 4.0% adjusted for inflation out of your portfolio?

The in these models is trivial, I have not data mined, I have simply started with the premise: a 4% inflation adjusted withdraw. I felt that taking a variable withdraw of the original portfolio, inflation adjusted, based on market conditions, could safely get more from the portfolio. I have projected what I think I would really do in retirement and created some simple rules:

1) take a smaller than 4% withdraw if the market is down
2) take a larger than 4% withdraw if the market is up
3) take, on average, more money between 60-80 when I can use it, and less when I am 85
4) if the portfolio takes serious hits: 1929, 1966 - tighthen my belt
5) if the portfolio is basking in boom markets like post WWII or 1980-2000, take more money so you don't leave so much on the table.

I have done no gimmicks, no jokes - I posted results from my model. If you want to get 4.0% from your portfolio - great, I have no problem with that. But, I want to see if one can safely, methodically adjust withdraws to get more from the exact same portfolio by willing to be flexible - I think it can be done and my version of FIRE is showing it.

 

[Rich_by_the_Bay]

But, I want to see if one can safely, methodically adjust withdraws to get more from the exact same portfolio by willing to be flexible - I think it can be done and my version of FIRE is showing it.

Rigel, don't worry about the critical comments - you asked for feedback and you're getting it. I'm sure no one intends it to be personal.

But from your reaction above, all I can add is, be careful. Don't get too wedded to a plan which, while creative, has not been rigorously tested and which embeds an unusual amount of volatility of income. Not saying it's right or wrong for you, just that you don't want to get so wrapped in it that you ignore the tons and tons of previous conventional wisdom out there.

Good luck if you use it, and keep us posted.

 

[Gone4Good]

Second, your post contains a math error. 1.5% of 2MM is 30k.

No my math is fine.

$2,000,000 * 50% * 1.5% = $15,000


No offense meant in the prior post, it is sometimes hard to tell whether new posters are serious or not.

But my criticism was intended to be serious. It's important to look at the impact any withdrawal strategy will have on your life, especially in the downside scenarios. It seems to me that most people would have great difficulty living with the year-to-year spending volatility your plan would entail. And with a 100% equity allocation, having to cut spending down to 0.75% of your original portfolio value is a real possibility. Portfolio survivability is only one measure of success. Maintaining some minimum level of real spending power throughout your retirement is another equally important goal.

 

[rigel]

3 to go:

The numbers in the posting are options, variables, which are fully modifiable just like FIRE allows a wide range of situations to explore. It is the concept of withdrawing the money which is the point.

FIRE defines a floor, it looks for the worst case situations. That's great, but it doesn't provide guidelines for what to do if I don't retire in the worst years. How does one take withdraws and in what amounts - avoiding both going broke or having excess money when you are too old to spend it since you don't realy know which year you are in until it is history.

In my example, the 1.5% withdraw that I proposed, is only used when the real value of the portfolio has declined 50% - like from 1929-1933. In this model, with facing the real declines of 60% in portfolio value, I would not take 4% (adjusted for inflation) in 1933 just because the market was up. When faced with one's portfolio cut in half, I showed the withdraw not being 4% but 1.5% (not 0.75%).

If one retires (and manages to avoid) retiring in 1929, 1966, etc - the vast majority of the years where the market does not immediately crash, one should be able to safely take more than 4%.

From my model, the worst situation was the depression (and mid 1960's of course) - if retiring right then, one could average only a 3.8% withdraw - which my model shows, but you would get through it intact, with withdraw rates of 1.5% through only the worst of the worst part of the depression.

The rules in my model are to help guide when to take more as well as when to take less, based on real market conditions. In the typical year, when starting retirement, one should be able to get 5, 6 or even 7% on average from your portfolio - and safely, without risking going bust.

FIRE has done well defining the floor, the worst case withdraw strategy, get through that you get through anything. I want to find the ceiling as well - creating a withdrawal strategy that solves both.

 

[samclem]

Okay, here's another observation: You are obviously okay with considerable volatilty in your annual withdrawal amounts. Given that, why are you constructing this whole system (varying your annual withdrawals from) the beginning portfolio amount plus annual inflation? Our portfolios will not "know" what the inflation rate has been (unless you are invested only in inflation-adjusted securities). Most of us have constructed portfolios we hope will at least keep up with inflation, but equities, in particualr, can perform worse than inflation for long periods. So, if you can stand the volatility, why not simply peg your annual withdrawals to a percentage of the portfolio's year-end value? Again, ESRRob's "95%" plan does this, as does Gummy's "Sensible Withdrawal" scheme. There are ways of tweaking these plans (floors on annual withdrawals, "bonanzas" etc), to achieve your goals. To avoid leaving a lot of money behind, you could increase withdrawal percentages as you age.

 

[rigel]

Samclem - I'm ok with volatility, especially on the upside

I certainly agree that there are many approaches to deciding how to withdraw funds. The key, is to find one that your are comfortable with.

My spending, could be defined as those expenses that are relatively fixed (gasoline, utilities, property tax, etc) and variable expenses (travel, entertainment, toys, hobbies). If my fixed expenses are represented as 2.5% of the initial portfolio, then I would like a model that protects this as the floor (except in say a serious depression - not all expenses are fixed in that situation), but has upside volatility for extra fun things, extended international travel for example.

The point of tying back to a percent of the opening portfolio value adjusted for inflation - is it represents a specific standard of living amount regardless of the year. If 30k or 100k is what one needs today, unless something changes, that's what you will need in the future, after adjusting for inflation. It creates a constant reference. If one's fixed expenses require 50k, then you'll need that going forward as your floor, not to go below.

I've looked at the 95% and Gummy's and they certainly have their merits, but I haven't been able to get comfortable fitting them to how and when I want to spend money. So, I've been exploring variations on 4% of your initial portfolio adjusted for inflation. I view this, as simply a highly variable version of that, with a couple of extra twists to exploit upside potential if it happens, and one extra rule to protect on the downside in rare and serious downturns.

 

[Alan]

Rigel,

Have you read this article on some simple rules to follow for variable withdrawal rates?
http://www.fpanet.org/journal/articles/2006_Issues/jfp0306-art6.cfm?renderforprint=1


It is a complex read (at least it was for me), but I thought it made a lot of sense and was easy enough to follow, so I captured the 5 rules to follow for myself when I start my draw down.

Capital Preservation Rule.
The capital preservation rule applies when a current year's withdrawal rate—using the decision rules in effect—has risen more than 20 percent above the initial withdrawal rate.
The capital preservation rule expires 15 years before the maximum age to which the retiree wishes to plan; for example, a retiree assuming she would not live beyond age 100 would discontinue the capital preservation rule after age 85.
Under the capital preservation rule, is the current year's withdrawal is reduced by 10 percent. The other decision rules in effect are then applied to this decreased withdrawal amount.
This decreased withdrawal becomes the basis for determining the following year's withdrawal amount.

Portfolio Management Rule
Following years where an asset class has a positive return that produced a weighting exceeding its target allocation, the excess allocation is sold and the proceeds invested in cash to meet future withdrawal requirements.
Portfolio withdrawals are funded in the following order: (1) overweighting in equity asset classes from the prior year-end, (2) overweighting in fixed income from the prior year-end, (3) cash, (4) withdrawals from remaining fixed-income assets, (5) withdrawals from remaining equity assets in order of the prior year's performance.
No withdrawals are taken from any equity following a year with a negative return if cash or fixed-income assets are sufficient to fund the required withdrawal.

Inflation Rule
Yearly withdrawals increase by the annual rate of inflation as measured by the Consumer Price Index (CPI) except when the withdrawal rule freezes the withdrawals.
The maximum annual inflationary increase is 6 percent.
There is no "make-up" for a "capped" inflation adjustment

Withdrawal rate
Withdrawals increase from year to year in accordance with the inflation rule, except that there is no increase following a year where the portfolio's total return is negative and when that year's withdrawal rate would be greater than the initial withdrawal rate.
There is no "make-up" for a missed increase.

Prosperity Rule
The prosperity rule applies in years with a withdrawal rate more than 20 percent below the initial withdrawal rate.
Under the prosperity rule, the current year's withdrawal is increased by 10 percent. The other decision rules in effect are then applied to this increased withdrawal amount.
This increased withdrawal amount becomes the basis for determining the next year's withdrawal.

 

[jdw_fire]

My spending, could be defined as those expenses that are relatively fixed (gasoline, utilities, property tax, etc) and variable expenses (travel, entertainment, toys, hobbies). If my fixed expenses are represented as 2.5% of the initial portfolio, then I would like a model that protects this as the floor (except in say a serious depression - not all expenses are fixed in that situation), but has upside volatility for extra fun things, extended international travel for example.

The point of tying back to a percent of the opening portfolio value adjusted for inflation - is it represents a specific standard of living amount regardless of the year. If 30k or 100k is what one needs today, unless something changes, that's what you will need in the future, after adjusting for inflation. It creates a constant reference. If one's fixed expenses require 50k, then you'll need that going forward as your floor, not to go below.

If you used half of your portfolio to buy a CPI adjusted immediate annuity at 60yo the annuity would provide your required floor WD of 2.5% of initial portfolio with CPI adjustments. You could then spend the rest of your portfolio when and however you wanted knowing that you are insured that your floor income will be provided.

 

[youbet]

FIRE has done well defining the floor, the worst case withdraw strategy, get through that you get through anything.

But you don't believe it. You keep repeating that you'll dramatically cut spending following a deep market downturn.

My plan is different than yours in that I'm not really interested in providing an inheritance for anyone or dieing with millions. I want to spend more right now. I'm nine months into FIRE and am withdrawing on a fixed percentage plus inflation basis. Even though the market has dropped lately, and may continue to drop, we're continuing our discretionary spending and enjoying life a lot.

Why are you working so hard to have a plan that could cause you to cut deeply into spending during your prime, younger RE years?

 

[cute fuzzy bunny]

If you used half of your portfolio to buy a CPI adjusted immediate annuity at 60yo the annuity would provide your required floor WD of 2.5% of initial portfolio with CPI adjustments. You could then spend the rest of your portfolio when and however you wanted knowing that you are insured that your floor income will be provided.

Providing your personal rate of inflation and lifestyle inflation dont lose ground to the CPI, and presuming the insurance company doesnt go bankrupt, and presuming that your spouse doesnt mind half the income going away on your demise, and.........

 

[sgeeeee]

If you are willing to modify the spending model in FIRECalc, you can significantly change your initial withdrawal rate. Note, however, that initial withdrawal rate means something very different once you change the spending model. For example, if your spending model is to withdraw 9% of your portfolio for living expenses every year, you will never run out of money. But you may be very likely to be homeless and starving long before you die.

If you can separate how much of your spending is really needed vs how much is discretionary, then you can compute a higher SWR using firecalc:

http://early-retirement.org/forums/index.php?topic=2048.0

 

[youbet]

If you can separate how much of your spending is really needed vs how much is discretionary, then you can compute a higher SWR using firecalc:

That separation isn't always as easy as it sounds. Just because something, such as a beloved hobby, gets categorized as discretionary, doesn't mean it won't be badly missed when given up due to retirement budget problems. For folks who actually liked their jobs but retired in order to have more time to devote to "discretionary" budget items like hobbies and travel, they might wonder why they retired if they can't do those discretionary activities as planned!

My only point is that sometimes on this board we make it seem like separating what is discretionary and what is necessary is easy. And that giving up so-called discretionary budget items will be painless. Not necessarily so.

In the case we're discussing, computing a higher SWR by understanding "needed" vs "discretionary" budget items, it should be considered that the forfeiture of some of the "discretionary" items will quickly lead to an unhappy retirement. Bummer.

 

[Zipper]

Just take 5% of your 60/40 portfolio based on your year end balance. Simple, no fuss and it's worked for me since '97.

 

[Zathras]

For me, I don't see anything wrong with leaving others money as long as I don't deny myself quality of life.
That said, I don't see anything wrong with your plan AS LONG as you can truly be comfortable living with such a wild income ride.

 

[sgeeeee]

That separation isn't always as easy as it sounds. Just because something, such as a beloved hobby, gets categorized as discretionary, doesn't mean it won't be badly missed when given up due to retirement budget problems. For folks who actually liked their jobs but retired in order to have more time to devote to "discretionary" budget items like hobbies and travel, they might wonder why they retired if they can't do those discretionary activities as planned!

My only point is that sometimes on this board we make it seem like separating what is discretionary and what is necessary is easy. And that giving up so-called discretionary budget items will be painless. Not necessarily so.

In the case we're discussing, computing a higher SWR by understanding "needed" vs "discretionary" budget items, it should be considered that the forfeiture of some of the "discretionary" items will quickly lead to an unhappy retirement. Bummer.

I've always thought that one of the most difficult parts of retirement planning was estimating what I would need to spend in retirement to be satisfied. I knew what I spent while I was working, but my guess of what I would want to spend if I wasn't working was just that -- a guess. So I agree with you that separating required and discretionary spending could be difficult.

There are two important observations I would make, though: 1) Notice that when you use FIRECalc as proposed, discretionary budget is never eliminated. It is simply tied to a percentage of your overall portfolio. So it can shrink, but it won't go away. This makes the job a little bit easier. Say you have a travel budget and you think you could reduce it by 10% fairly easily (by choosing more frugal travel habits, by taking one less trip, . . . whatever). So you put 10% of your travel budget as discretionary. You will never have to reduce your travel budget by that much. The overall reduction will never be as high as 10%. 2) As hard as it may be to figure out how much of your spending you might consider discretionary, try to use one of the other proposed variable spending models and figure out how much your budgets might need to be cut.

But I don't do these simulations in order to see how much to spend each year. I don't even use the 4% rule. I saved enough money that I now spend what I want to spend. I do not anticipate having to reduce my spending during years of poor performance. But if some disastrous financial crises occured and I needed to contain spending, the required/discretionary simulations help to understand how much I might need to control.

 

[unclemick]

But I don't do these simulations in order to see how much to spend each year. I don't even use the 4% rule. I saved enough money that I now spend what I want to spend. I do not anticipate having to reduce my spending during years of poor performance. But if some disastrous financial crises occured and I needed to contain spending, the required/discretionary simulations help to understand how much I might need to control.

True, True - same here to some extent. Like when she suggested it might be time to remodel or Katrina - both treated as 'offers I couldn't refuse' - my el cheapo spending double or tripled irregardless of the market(2000 and 2005).

The simulation runs are a backdrop to see handgrenade wise how I'm doing. Also with the passage of time starting to hear footsteps( at age 63) so loosening up the purse strings as appropriate.

heh heh heh 8).

 

[youbet]

I've always thought that one of the most difficult parts of retirement planning was estimating what I would need to spend in retirement to be satisfied.

Absolutely. Even nine months into RE, it's still a moving target!

The simulation runs are a backdrop to see handgrenade wise how I'm doing. Also with the passage of time starting to hear footsteps( at age 63) so loosening up the purse strings as appropriate.

Wise thoughts from a man actually playing the game.....not just watching on TV!