« More on Early Retirement-Part 3 | Main | See Average Joe Retire »

March 26, 2012


Feed You can follow this conversation by subscribing to the comment feed for this post.

Kathy Sterndahl

That all seems pretty confusing, Syd. I spent years trying to figure out when I would be able to afford to retire. My crystal ball just wouldn't tell me what the future holds. And I had seen too many people who waited too long and then something happened to prevent them from enjoying their retirement. Finally, I just did it. I decided that I can always go back to work if I run out of money, or sell my mortgage-free house and rent. Of course, I live in Mexico where life can be very inexpensive if I chose to live that way.

Retired Syd

@Kathy: Excellent approach! This post is geared more for the Nervous Nellies out there. You are already ahead of the class.

New to this

First of all, Bravo!, for setting up a great framework to discuss this…(I think a version of this should find its way to a much broader audience as I know through my own circle of friends, there is tremendous inconsistencies in understanding)…

So having said all that nice stuff, the engineer in me requires some questions to test the “corner conditions” before I can sleep peacefully (aren’t u comforted to know the designer of the last airplane you flew in felt this way too)

Question 1) Let’s say Jane chose a random Tuesday to retire and the market was at whatever level it happened to be that day. Now, as luck would have it, the market was higher the next day, Wednesday. She really likes shoes, so declared that Tuesday was just a trial run and she really meant to retire on Wednesday.

Well, as luck would have it again, the market continued to go up for 10 straight days in a row and she repeated the above pattern each day.

Now you may say, “so what” that’s just rounding error. But if you extrapolate this same idea over longer periods of time, why wouldn’t Jane just continually “poll” the market for anytime in the future when her nest egg happened to be higher than on the day she originally set her lifetime budget?

She could always argue that by re-establishing her first day of retirement to the new “better” day that she was really behaving no differently than a brand new person who randomly chose this new day to retire.

Retired Syd

@New: So yes, your 10-day example is a rounding error. That's why I gave you your example with a 5-year run up. If Dick had decided to "reset" his retirement at the first day of Jane's he would NOT be operating with a portfolio that was as large as Jane's--he took out 5 years worth of living expenses. So his "new" 3% is going to be lower than Jane's (because his asset base is smaller than $1.2M.

You are only looking at the example over and over for the growth in the portfolio, and you are ignoring the fact that despite the market's rise, the portfolio is smaller by whatever the retiree took out over the x-years of whatever example you want to create.

So in your example above, when you extrapolate the market conditions over the long run, you ALSO have to extrapolate the withdrawals over that time period that the retiree took out of his portfolio. You can't just assume the nest egg gets bigger and bigger with the market, because it also goes down for withdrawals.

In the extrapolation (which is exactly what I was going for here in this example). Dick started with $1M and withdrew $161k over the years. He experienced the same growth as Jane, but his account is LOWER because of his annual withdrawals. So if he does what you are suggesting (resets his retirement at the 5 year mark when Jane retires), he's going to come up with an even SMALLER withdrawal amount.

See, you are only looking at one side of the equation (market growth) and ignoring the withdrawals the retiree would have taken over the years before the "reset" their retirement.


you lost me at 'hello'

Retired Syd

@alicia: You are not alone! Which is probably why according to a recent study 55% of Americans do not know how much money they need to save for retirement ant 45% haven't even tried to figure it out. It's complicated!

New at this

If your nest egg grows at a higher CAGR than your withdrawal (which is highly likely I hope) your portfolio will grow. My example holds, please re-read.

And as an example of the same issue in reverse, if Mary asks if she can retire in June with $1M and a goal of covering a $30K budget, Bengen (and Syd) would say she is safe. But if she chose to wait 6 months, and in the interim the market crashed 50%, Bengen (and Syd) would now explain that the 3% Rule would not support it. But what if she had already pulled the trigger back in June?

Retired Syd

@New: Ok, explain this to me mathematically, how can Dick's portfolio (which is identical to Jane's) POSSIBLY be as large as Jane's at year 5 when he has taken 5 years of withdrawals and Jane hasn't touched hers.

Retired Syd

@New: As to Mary, that's exactly what happened to Syd in real life! As long as the 50 years starting with my pre-crash balance is no worse than the worst 50 years, the rule says I'm ok.

If I had waited for 6 months when the market was much lower, all the rule says is STARTING AT THAT VALUE, if I experienced the worst 50 years FROM THAT DATE (not the date 6 months prior), I would be safe if I only withdrew 3% of that value.

You see we are all in the market together after that, so clearly Mary, retiring 6 months after I did will have a better 50 years than I had from my start date. So, she will not be in the worst 50 years if I am--if hers is worse than mine (the worst ever) hers is worse than in history, and the rule doesn't cover that.

The 3% rule doesn't cover a worse 50 years than what we've had in history. All bets are off then! That's a risk you take if you follow the rule (and also the risk that you live longer than 50 years!)

New at this

Syd - do u really not understand my example(s)? Please re-read any of the one's I've used, ignore the idea of comparing one investor against another, and try answering again. I know u can get there....

Retired Syd

@New: I did answer it. Now I give up.

Retired Syd

@New: Actually I'll try one more thing: In all the different dates you hypothetically choose in your questions, a retiree starting on each of those many choices can't ALL have the worst 50 years. Only the one with the worst 50 can have the worst 50 years, all the people that picked other dates have a better 50 years (or worst than history, which, as I said is not covered).

So at the end of 50 years, which ever date you started on that had the worst 50 years, that person dies with zero and all the people that followed the 3% rules on any other day you selected in your many questions had a better 50 years, so they die with money.

That's what the rule says.

If that didn't work then I really give up.

New to this

Thanks for the follow-up, Syd. As they say, men are from Mars and women are from Venus, and maybe that’s why we keep talking past each other…..

You seem to feel compelled to keep pointing out things in the Bengen article, which I promise I have read per your urging. (And I really do believe I understand every word he said)….

But what I am trying (un-successfully) to do is point out what I believe to be a shortcoming in his approach, and frankly it worries me enough to where I am hesitating to retire because of it....

So following your lead, I will try one more time with a hypothetical narrated discussion:

Here goes,

@New: So Syd, since retiring, have you ever experienced a moment in time, where if you had “recalculated” 3% of your current portfolio, you would have come up with a larger available spend than you had calculated on day 1?

Syd: Now that’s’ against the rules, you see your just supposed to do the calc on day one and inflation adjust it from there, blah, blah, etc, etc…..

@New: Ehem, Excuse me for interrupting, but I get all that. I’m trying to ask a different question, not clarify my understanding of what the Bengen article said. So, can I ask again…?

Syd: alright, the answer is No, market crashed as soon as I retired. This coupled with the fact that I was making withdrawals every year meant that I never had a day where my portfolio was higher since the day I retired.

@New: Ahhh, that explains why it’s been difficult for me to get you to see the scenario I am concerned about. Can I try again?

Syd: Sure. (jerk…)

@New: Great, Let's say, just for grins, that you Syd, retired into a raging bull market. (I get that you didn't but let's just pretend). Hypothetically, in that altered context, do you agree that it is mathematically possible, that you could now have a larger portfolio balance today than you did on the day you retired? And yes, this means even AFTER making withdrawals over the last several years.

Syd: Um, I don’t think so, you see I have been withdrawing annually from my portfolio…, and if you remember Dick, he said, she said, etc, etc, blah, blah……

@New: Ehem, excuse me for interrupting again….And I get that your just trying to teach me something, but please play along, just for a moment…….So what if the market had grown at, say, 20% per year, and you were only drawing out 3% per year…., do you see that in that instance your balance today would be higher than when you started?

Syd: Yeah, (you really are a jerk…)

@New: Now your turning red with frustration, but lets press on...So if under our hypothetical scenario where you recalculated 3% of the value of your hypothetical portfolio today, can you agree that it would result in a higher amount of money available for you to spend today than the amount that you calculated when you originally retired?

Syd: Your really are a jerk, but I see what you’re saying. But so what?

@New: So under this scenario, why would you not do a re-calc and “re-declare” your retirement date to today? How would that be any different from a brand new person with your identical portfolio, as of today, doing the math, today, and retiring, today…?

Syd: [please fill in the blank]

Retired Syd

@New: Yes, if the market increases 20% per year for a period of time, by all means, you can recalculate. Go for it. In that scenario, it would work.

Retired Syd

@New: Extra credit question: Do you understand why that does work?

Answer: Because the first 5 (or whatever years) of your retirement you experienced 20%/year growth. So you are NOT in the category of retirees that experienced the worst 50 years ever, quite the contrary, that was probably the best 50 years ever based on those 20%/year returns in those first 5 years. 3% was just the SAFE MAX--in the case you were the poor fellow that experienced the WORST 50 years. Clearly you did better than the poor fellow that retired in the worst 50 years.

You see how that works now?

New to this

So following that thought, could you just recalc and "reset" your retirement at any moment along the way - whenever 3% of your current portfolio would yield a bigger amount than your original calc...?



Your explanation was clear and appreciated. I accept the "rule" (and its assumptions and inherent limitations) for the helpful tool it is. I do not believe, nor do you suggest, that it is a guaranty.

Retired Syd

@New: I'll do you one better. Let's say you go 20 years and you have a great, great, great bull run on your measly 3% withdrawal rate adjusted for inflation each year. Now you have a really big nest egg. You can now "reset" to take out 4% of your really really big nest egg because you only have 30 years to live instead of 50 years.

The 3% rule was the minimum withdrawal ASSUMING YOU HAD THE WORST 50 YEARS. But it looks like you're not going to be that guy. You're the guy that had the best 50 years (so far). If you stuck to the 3% rate, you'd likely die with a great big pile of cash--one of the possible outcomes when using the 3% rule (if you are unfortunate enough to live over the best 50 years).

The 3% rule just said that you would not run out of money over the next 50 years if you only took 3% adjusted for inflation. Clearly, you were not going to run out of money in this scenario because the market did so well. So sure--reset, you are a very lucky guy.

Depending how much inflation was during those really really great years, though, you might already be at 4% (with your current withdrawal amount) by the time you inflate your current 3% dollar amount each year for inflation over the last 20 years. So it may not be as great as you are thinking it is. But yes, that works.

In the case that your portfolio did really really well during the first part of your retirement such that you want to re-set--that just means that you were one of the ones that were going to die with a pile of cash. If, on the contrary, you were one of the really really unlucky guys that retired in the worst 50 years, your original 3% would not have left you destitute. (And by the way, you would not have wanted to reset in that case, right? And you wouldn't have needed to, unless you thought your 50 years were looking even worse than any in recent history, of course.)


Wow, this exchange between Syd and @New is the best discussion I've ever seen on this topic. I feel myself painfully sharing in the frustration of trying to understand all this stuff.

Syd - Once you get this to a logical end, you need to bottle this for a publication of some sort.

Retired Syd

@Rick: That's exactly right. Another point that often gets lost in this discussion, is that just because you don't have the multiple needed for these safe withdrawal rates, that DOESN'T mean you don't get to retire. A) You might not experience the absolute worst time chunk in your own retirement, B) You may be able to go to Plan B, picking up some part-time work, cutting expenses, downsizing, moving to a cheaper location, selling your house and renting, renting out a room in your house, taking out a reverse mortgage, the list goes on and on. The rule is just a guidepost, not a law.

New at this

Great pivot, so say you are that unlucky guy in your last paragraph. And 6 months after you retire the market is +50% lower..., what would you do?

On the one hand, if you stick to the Bengen plan, you just stay the course, taking solace in the knowledge that history says you are ok.

But how does that reconcile with the fact that a brand new wannabe-retiree, with the identical assets of the above guy on that very same day, would be told he doesn't have enough to retire, per the 3% rule!

Retired Syd

@New: Good questions.

First, as you know from your interview with me above I AM THAT GUY, or gal as the case may be (although even though the market was 50% lower or whatever it was, my portfolio was not--I'm not 100% invested in the market as you, I'm sure would know.) I did decide to trust the original 3%, hoping that my 50 years are not worse than any in recent history. If I'm wrong and they are worse, I'll have to go to Plan B's (like I wrote about above), but if it is exactly the worst by historical standards, I should die and bounce the last check right on schedule.

I think by "the above guy" you are saying the guy that experienced the identical hit to his portfolio as I did but wants to retire then, 6 months later. The rule says that if he takes 3% of his current (lower asset base) he's ok if HE has the worst 50 years ever from his start date. As far as the rule is, only one of us will have the worst and the other will fare slightly better (remember, we're not assuming that the market does worse than ever during the 80 years studied. So only one of us can experience the worst by historic standards.)

In this case that you are proposing, if we each were drawing the 3% at our start date, and I had the worst (because I retired when the market was higher, so my results would be the one that lost, he would do better as he started when the market was lower), I would die with zero and he would die with some cash.

Which means, that yes, he would have been ok even drawing a higher rate. He fell above the 3% rule, so he would have (in retrospect) been able to take out more--maybe even what I was taking. The 3% is just the safest rate. In the end, he would have been ok with a little higher, because he didn't experience the worst 50 years like I did.

The thing is you don't know which bucket you'll be in until it's all over, so if you want to be absolutely safe, you only start with 3%.

The guy that retired 6 months after me WOULD have been ok with a higher withdrawal rate, because his 50 years are better than my 50 years.

Some people will have a stellar 50 years (and in hindsight could have withdrawn 6%!). You just don't know which one you are going to be, so if you want to be absolutely safe, you go with the 3%. In many cases, as you illustrate, it would be overkill.

Overkill is what the rule is intended to almost guarantee. Undekill is what the rule is intended to prevent.


I don't like those SWR estimates for a couple of reasons:
- you can pull the plug far quicker if your necessary expenses are pretty low, you don't mind cutting back on discretionary expenses and / or are ok with working a bit
- they don't take into account that every older (say 75-80+) person I've known has drastically lower expenses than most people in their middle years (excluding long term care or increased medical costs - where YMMV - especially if you live in a country with UHC.)

I also find the whole "when / how much to retire" discussion so incredibly fear based (you're an exception). Not saying you shouldn't be realistic, but the way I figure it, you just have to have saved more than 90% of the population - which isn't that hard to do for most middle class people.

Then there's the whole depression fear out there too - it makes no sense. My grandfather *made* most of his money in the great depression years. Makes me think Buffett is right and now is a good time to invest in single family homes in some depressed areas.

New at this

Thanks but Let me clarify what I mean by "above guy"

Guy1 retires in June with $1M. Calc at 3% = $30K.

6 months after horrific crash, in December, his portfolio is worth $600K.

He notes that $30k/$600K = 5%, but since history says he'll be ok he stays the course and stays committed to taking out $30K, inflation adjusted every yr going forward.

Now, along comes Guy2 in December. He has $600K and wants to take out $30K each year to spend. He calcs $30k/$600k = 5%. But since this is more than 3%, he concludes he doesn't yet have enough so continues to work.

How do these 2 stories, both following Bengens 3% rule, reconcile?

Retired Syd

@Jacq: YOU are a breath of fresh air!

The comments to this entry are closed.


Enter your email address:

Delivered by FeedBurner

Twitter Updates

    follow me on Twitter