Firecalc and SWR's

[Markedward]

I have discovered this web-site very recently and have been favorably impressed both with FireCalc and with some of the people who post here. As a quick intro - I am a 48 y/o Wall Street professional (23 years in the business) and I am seriously considering pulling the cord on the career, hence the interest that led to this site.

One initial point, I have read may posts regarding SWR's etc, and it seems that the 4% number has become somewhat of a Holy Grail, as in , you can't go wrong if you adhere to the 4 % WR. I would point out a few issues with this thinking:

1) This 4% number is based on historical analysis, and much of the data is not independent. In my case for example, I must plan for 45+ years in retirement, the data from 1871 dont really give me (2007-46-1871 = 90) 90 separate points from which to infer that I should be able to sustain a 4% withdrawal rate. This data obviously contains a very significant amount of overlapping information, for example, the 1871-1917 period has almost all of the same data as the 1872-1918 period. Really, there are only about 2 independent samples from which to draw data regarding historical performance, no statistician would want to draw firm conclusions from such a set of data, and I will not base my retirement decisions on any such analysis.

2) This historical data was compiled over a period of time when it is highly likely that stocks were inefficiently cheap (glad to discuss this in a separate topic if you would like). Cheap stocks lead to high returns. If I don't think that stocks are as cheap now as they were in the past, then I should not expect to earn the same kind of returns going forward.

3) Given that I am dubious as to future returns being equal to the past, I prefer to use a Monte Carlo approach, which FireCalc provides in a nice and easy to use fashion.

In my own case, using my rather lousy DB pension, SS at 80% (at 80% the SS system IS solvent for the foreseeable future), decent amount in 401K, I come up with something like a 3% SWR being one that I am more comfortable to live with. In my case, using my return assumptions and a 4% WR led to a 30% likelihood that my plan would fail. Given that I really, really don't want to have to go back tp work in 25 years if my plan should fail, I will only tolerate something like a 3% chance of failure via a Monte Carlo. In my case, I am using a 5.9% return with a 8.75% standard deviation in that return for the MC inputs.

So, be careful with that 4% assumption....In my MC testing , I would say that if you only need 30 years worth of coverage, 4% may be ok, depending on your other details, but if you are like me and want to plan for more like 45 or so, use that Monte Carlo tool with your own personal circumstances and see what you get.

 

[Cut-Throat]

One point is that I don't stocks were that cheap in September of 1929. Prices come and go and that is what FireCalc is run against.

I'd say whatever SWR you use, Rerun the Calculations every 5 years or so and adjust your spending. Spend more if the Chinese and Indians start fighting over our stocks. Spend less if we go into another depression.

 

[Markedward]

Good points Cut-Throat. Yes, it makes great sense to re-run the calcs, I would say re-run them every year. In my case for example, if the portfolio accrues "on plan" for 10 years, then I would be able to increase my 3%-proof withdrawals by almost 45% from the initial level. In fact, I would say that what we should be trying to do here is to keep the chance of failure constant, not necessarily the withdrawal amount.

 

[wabmester]

FIREcalc provides an interesting look into the past, and you're right that the future is guaranteed to be different.

4% is just a swag, but 3% is too low. For a 40-year period, a 100% TIPS portfolio (2.5% real) would give you an SWR of around 3.5%. I consider that a baseline.

The best-case 40-year withdrawl rate with a 90% stock allocation would have been about 10%.

So, those are pretty good bounds: 100% safe portfolio guarantees 3.5%, and an ultra-risky portfolio's best case might give you 10%.

Now, choose how much risk exposure you want, and place your bets.

 

[Markedward]

wab - I disagree, at least in part. Taxes must be taken into account, which is a big part of the reason for taking chances on stocks ,the 15% tax rate is the best thing going ex-Munis., In the case of TIPS, there are 2 issues:

1) Taxes: Tips currently yield around 2.5% over inflation, which would be a Breakeven inflation of 2.55 vs regular 10 yr Treasuries at around 5.05%. If we assume a 30% avg tax rate, that 5.05% turns into 3.53%, which, given a 2.75 inflation rate is .78% real after tax. Not horrible...BUT:

2) Unfortunately, our tax code taxes inflationary gains the same as real ones. So, lets say inflation accelerates to an ugly 8%, ,if TIPS still yielded 2.5 over imflation, then the yield would be 10.5%. After the 30% tax rate, that turns into 7.35% after tax, or negative .65% real. Not so good. And worse, if you had bought your 10 yr tip when it yielded 5.05 as in 1, the change to the high rate environment would cause severe holding period return problems, depending on the maturity of the TIP.

Not that Tips are bad, I just think they are best used in tax advantaged accounts such as a 401K.

 

[wabmester]

I don't think a 30% tax rate is realistic.   When you're retired, I assume you'll have no earned income.   You can expect a *much* lower marginal rate on that TIPS income.

 

[Markedward]

Not trying to brag, just illustrate my point, but at age 65, my 401K, plus SS, plus DB pension will put my income at around $250K. Granted, that is 17 years off, but, at that income level, your marginal tax rate is the full 35% at the Fed level.

To the extent that you have income to get you much over 100K, then 30% is an appropriate number to think about with respect to bond taxes.

 

[laurence]

Not trying to brag, just illustrate my point, but at age 65, my 401K, plus SS, plus DB pension will put my income at around $250K.

Then unless you pick up a coke habit you don't have to worry about a tenth of a % here or there!

Welcome to the board.

 

[wabmester]

Not to brag either, but my investments generated a 6-figure income last year in interest and dividends.   All of it was taxable.   I paid about $5000 in income tax.

Retirement gives you a lot of flexibility wrt taxes.    Now, I was a bit over the top in terms of expenses (2 houses, enough business income to allow me to write-off health insurance, etc), so your mileage may vary, but if you're smart enough to have $250K in investment income, I think you'll be smart enough to minimize your tax bite.

In any case, with a large income, it'll also be safe to assume that you have lots of wiggle room in terms of discretionary expenses.

Bottom line: don't wait until you reach the 3% SWR level before you pull the trigger.

 

[cute fuzzy bunny]

Yep, the tax man keeps passing a mirror under my nose too...and not for a coke habit...to see if i'm still breathing!

 

[retire@40]

Not trying to brag, just illustrate my point, but at age 65, my 401K, plus SS, plus DB pension will put my income at around $250K. Granted, that is 17 years off, but, at that income level, your marginal tax rate is  the full 35% at the Fed level.

To the extent that you have income to get you much over 100K, then 30% is an appropriate number to think about with respect to bond taxes.

First, if you can generate $250K of income per year, that's the equivalent of having over $6,000,000 in income-generating assets.  Most people here could get by with a 1% or less SWR at those levels, so there's plenty of slack even if you start at 3%.  We are not luxocrats here.

Second, your Federal marginal tax rate is not 35% at $250,000.

Third, at $100K of pure interest income, the marginal rate is 25%, with an effective tax rate of 17%, which is the more important number.  There may be additional state taxes, but you could always choose to move to a tax free state to avoid state taxes.

 

[Markedward]

you got me curious on the tax rates so I fired up Turbotax. At 200K Income, total Fed tax is 41782, and at 210K it is 45163, or, close to35%., From 100k to 110k it was 26%.

Anyway, my main point was that I find a 30% failure rate, as revealed via Monte Carlo to be unacceptable, in spite of the fact that Historical returns deem it close to 100% likely to succeed.

 

[preben]

With 45 years ahead of us I fully agree that relying on historical returns only is silly at best.
A more diversfied portfolio than SP500+US bonds only have historically done better. In addition the Firecalc models a person who blindly takes out the stash+inflation EVERY year, no matter WHAT the market does - nobody in their right mind would keep pulling $100k after their 2M stash have been cut in half (assuming 5% here just to annoy people focused on the 4% only for reasons mentioned above). Cheers!

 

[RASAP]

Ben,

You bring up a good point that I have been wondering about. If one is more prudent about how they withdraw funds from their port ("bucket method" for example) would not the potential for success increase? Has anyone factored in such withdrawal methods to see how the SWR is affected? It seems to me that we could see the "standard" 4% SWR number increase significantly.

 

[preben]

Yes it has been looked at multiple times actually. Gummy have done the "sensible withdrawals" (sorry, no link right here but a google Gummy pages should do it) and several studies Incl.: http://www.fpanet.org/journal/articles/2004_Issues/jfp1004-art6.cfm have shown up to 6% SWR based on historical nos (from a well diversified portfolio btw) - depending on how many "rules" one is willing to live with.

A fairly simply "rule" (partly in above FPA study) is to NOT increase for inflation in market down years. - Naturally that would be tough to do in a 70ties/80ties inflation scenario but at least one would then start cutting costs somewhere rather than blindly just adding inflation. Not adjusting for inflation would have been easy for me from 1998-2006 as I still spend almost exactly the same (but should be spending 30% more according to official nos.).

If one uses a split budget: 3% base costs/ok living + 3% discretionary (tropical holidays, fast cars, fancy restaurants Etc.) one should easily be able to freeze inflation increases, if not cut even more in bad times.

Cheers!

 

[Nords]

If one is more prudent about how they withdraw funds from their port ("bucket method" for example) would not the potential for success increase?  Has anyone factored in such withdrawal methods to see how the SWR is affected?

Well, yeah, Bob Clyatt wrote an entire book on the subject...

 

[dory36]

And you can use Bob's flexible withdrawals in FIRECalc as well -- see the checkbox for using portfolio value instead of starting portfolio in the advanced version.

His withdrawal strategy: According to the rule, each year's withdrawal is the greater of 95% of last year's withdrawal or 4% of the current portfolio as you started with.

FIRECalc uses whatever percentage withdrawal you start with, and allows you to set a different value than 95%. Note that "success" using the 95% rule means your portfolio retains the same value at the end of the term as you started with, rather than merely remaining "in the black".