changing with the times

expat

Confused about dryer sheets
Joined
Jun 6, 2003
Messages
2
I have a question about Safe Withdrawal rates and adaptability (or conditional probability if you prefer). Please excuse me if it's well-covered or old hat to you guys.

It falls naturally into 2 parts (the second came to me as I thought of the first):

1. let's say you have ERed with a 40-year horizon and a SWR to go with that. After, say, 5 years, should you worry or even panic if your withdrawals now exceed the SWR for the portfolio you now have over 35 years? Many 40-year cases will dip into this state after 5 years but still survive in the long run: do you just hope you've got one of those and not a "FAIL" run, or do you modify your withdrawal regularly?

2. same starting situation, but this time, after 5 years you are still on a SWR for the portfolio you have, over 35 years. Is that good enough? I think not, because your situation has now changed: you now know that you're not one of those who died during that 5 years. Whatever reasoning led to to a 40-year horizon 5 years ago, would not lead you to a 35-year horizon now, but to something longer.

This being my first post I'd like to say hello and especially thanks to Dory36 and all the others who have put so much into this (and in such a good humour compared to Certain Other Places). I have certainly taken much out of it while lurking.
 
Let try to answer this:

1. Simple answer - you are safe. The situation is the same as every period in histrory with a bear market. The next year or two after starting withdrawls will determine whether you port value is up or down. Historically your portfolio has always recovered so the assumption is it will again.

2. This is a hard issue from a pyschological point. 5 years ago you said you were only going to live 40 more years. Now 5 years later you still think you are going to live 40 more years. My question would be what has changed to make you think you are going to live longer now then 5 years ago?

It could be you underestimated your life span 5 years ago, in which case your assumptions used to determine your SWR were WRONG. If they were wrong then maybe your SWR its not really safe.

Or possibly, your life span estimates were correct and its just very difficult to realize that - hey, five years are gone already and I've only got a short 35 more years here at most.

I have an estiamate for my age of death and its very sobering (you have to determine this in order to ER). Mine is based on my family history (parents, grandparents age of death) plus 5 to 10 years. This number is one of many reasons I desire ER. Once you face your own mortality - and realize how little time you have it opens up a lot of issues. I don't want to spend the time I have left working for the man! I have things I want to do. That is why FIRE is important to me.
 
These are really good, relevant questions that I've puzzled over myself.

As a retired person ages and their portfolio value changes, there are a couple of factors at work that affect the "true" safe withdrawal rate (SWR) and the SWR indicated by FIRECalc or similar calculators. Sometimes these factors work in the same "direction" and sometimes they tend to cancel.

One factor that always tends to increase the "indicated" SWR is that the withdrawal period keeps shrinking (assuming that the user plans on dying at the same age, which I like to assume to be 100). This tends to increase the "indicated" SWR -- especially as the withdrawal period gets down to 10 years or less.

Another factor (as Expat recognizes) is that the action of markets will cause the user's assets (the principal) to deviate from the value assumed in the "worst case" FIRECalc scenario. And here's a sort of paradox: If markets are good, then the principal amount will be greater than anticipated in the "worst case scenario," and re-running the analysis will tend to indicate that the SWR could be increased (even if the same withdrawal period is used). But if markets have performed exceptionally well, they may be experiencing a "bubble" that increases their chances of sub-par performance in the future! Thus, re-running FIRECalc with the new principal amount would tend to produce an overly optimistic SWR.

Conversely, if markets perform very poorly and the principal drops below the amount in the "worst case scenario," then the chances for better (or at least "normal") market performance in the future are presumably better. And yet a re-running of FIRECalc would tend to indicate a lower SWR.

I think that the best approach to this dilemma is to apply "reasonably conservative" judgement to fit the specific circumstances. If the principal amount increases more than planned, then try to maintain the same SWR, at least for a few years, before re-analyzing. And if the principal amount delines more than planned as the result of market action (as it has done since March 2000 for people with a heavy allocation to stocks) then cut back on withdrawals -- perhaps to the percentage of the principal represented by the original FIRECalc analysis.

Dory 36 please note that this is a reason why it would be convenient for FIRECalc to give the user the option of selecting a withdrawal amount that is a particular percentage of the principal amount. It is also a reason why it is a real service to have FIRECalc available, to allow a user to experiment with changing financial circumstances. Hopefully, they will apply judgment in doing so and not take the output of FIRECalc as an absolute guarantee of anything! 8)
 
Hi Expat,

I would like to add two slightly different points of view to help answer your questions about life expectancy, and about possible modifications to SWR as one actualizes what was formerly a hypothetical set of possibilities.

Like expectancy as usually expressed-eg 18 years at age 60- is derived from a decay curve. Knowing nothing about ones individual health, if you have an 18 year expectancy at 60, at 65 you have not 13 years, but something more than that-maybe 14.5 for an off the top example. This is because the original 18 year estimate at age 60 included some number of men who died between 60 and 65. Since you weren't one of them(hooray!!), you have those years as a freebie, so to speak.

Also in thinking about life expectancy-don't forget that if you have a female partner retiring with you, her life expectancy is apt to be quite a bit longer.

Your other question of should you modify take-out down the road, to reflect your actual experience, or just continue full speed ahead? You may know that this question has been heavily debated on other early retirement boards. Personally, if my experience were somewhat better than expected five years into a 40 year plan (defined as my nominal withdrawal amount has fallen as a % of my nominal fund), I think I would not alter anything. This can make up to some extent for the fact that there is a lot of slop in making these projections. Also, fluctuations will occur!

In the opposite situation where market conditions or inflation adjustments have pushed my withdrawal rate higher, I at least would try to cut back. If I had planned a Carribean vacation, I might go camping somewhere instead. Because I know I would sit on that beach and worry my head off; and or get p*ssed at every cent my wife spent, or every tip I felt compelled to distribute.

After all, it is the underlying free cash generation of the portfolio that determines what can be taken from it. SWR's are abstracted from that. As Polish semanticist Korzybski wrote, "the map is not the territory."I believe that there is a fair amount of art in deciding these questions, especially when the fund is not clearly more than enough.

I at least would feel a lot more comfortable taking out 4% of a portfilio with an average PE of 10, than from one with an average PE of 30! On the other hand, one reason why there are so many early retirements is that PE's have expanded, and lifted "returns" above the growth of free cash generation in the underlying businesses.

Only The Shadow Knows--and he isn't talking!

Mikey
 
This is a deep theoretical question. It is fascinating, and without a clear answer.

You can sustain any withdrawal rate below 100% if you just readjust each year.

On the other hand, if in the first five years of retirement your withdrawal rate jumps from 4% to 10%, I think you would be foolish not to make some adjustments. Similarly, if your rate goes to 1% you definitely have some room to maneuver.

The calculators are there to show what happens when you sustain a single dollar withdrawal through thick and thin (you can put in some adjustments, I know, but that's the basic idea). When you live your life you need to be more flexible.
 
Let's say you have ERed with a 40-year horizon and a SWR to go with that. After, say, 5 years, should you worry or even panic if your withdrawals now exceed the SWR for the portfolio you now have over 35 years?
Hi Expat -- welcome from a former expat!

Others have provided some good thoughts, so let me just add one clarification regarding the workings of the program.

Assuming you pick a 100% safe rate for the 40 year span, then the withdrawals will be safe for any subset of that 40 year period -- even though your withdrawals may increase as a percentage of some future portfolio balance to what would seem to be an unsafe rate.

This seems implausible, but if there were any outcome during those 40 years that would cause the portfolio to fail, then you wouldn't have gotten a 100% survival figure.

This means that you could rerun the program whenever you want -- even during a "bubble", and restart your withdrawals at the new rate -- while remaining at the 100% survivable rate. All you are doing is incrementally decreasing your potential estate.

All this assumes, of course, that the future holds nothing worse for us than we've experienced in the past, and that your investments match the market.

Dory36
 
This is true for everyone using 100% SWR. Anyone using less than 100% may have their odds change with time as some of the possible scenarios drop out of consideration. If someone's portfolio has fallen dramatically right after they started withdrawals at the 95% SWR, the odds that they are in one of those rare 5% periods have dramatically increased. IMO, it may be better to stay with the 100% SWR figures in high P/E markets, since there is a statistical correlation between P/E levels and future return. 100% SWR will never vary from 100%, as long as nothing worse happens in the future than has happened in the past. So far, nothing is worse in the current bear market than has happened in the past, and there are signs of incipient economic recovery. The 100% SWR is still intact.
 
A couple of addiditional comments:

1. What Mikey says about life expectancy is correct -- that as a person ages it declines less than one year per year. But I think that it is prudent, for purposes of financial planning, to assume that you will live longer than your "life expectancy." In the first place, it is just an average (or perhaps a median?) based on historic data. A person's odds of living longer are on the order of 50-50, and somewhat more if you consider the liklihood of future medical advances that will extend life expectancy further.

2. As has been noted in previous posts, even a "100% successful" safe withdrawal rate is based on past financial conditions that may not be reproduced in the future. Barring any major catastrophe like a nuclear war (in which all bets are off) I would not expect the U.S. stock market to collapse the way that it did in the 1930's, because we know more than we did then about how to counteract such events. (For example, by having federal insurance on bank deposits and brokerage balances, and by maintaining the growth of the money sup0ply.)

On the other hand, there is a definite possibility that the the underlying growth of the real economy that drives long-term market returns (especially on stocks) will decline. To account for these factors, I tend to favor an allocation to stocks that is on the "low side" of what FIRECalc indicates is optimum, and a significant allocation to TIPs.
 
Like expectancy as usually expressed-eg 18 years at age 60- is derived from a decay curve. Knowing nothing about ones individual health, if you have an 18 year expectancy at 60, at 65 you have not 13 years, but something more than that-maybe 14.5 for an off the top example. This is because the original 18 year estimate at age 60 included some number of men who died between 60 and 65.

Be careful here. The life expectancy used when calculating a SWR is not a probablity. The number you decide on for life expectancy determines only one thing - how long you want your money to last. Just because median life span is, lets say 85 years, this really means nothing concerning your life expectancy. Remember as you age your life expectancy continues to go up. The life expectancy of a 100 year old has to be above the median or 85!

Lets say you use the median life expectancy age in your SWR calc. All things being equal you have a 50/50 chance of living beyond 85 and therfore a chance your port won't be 100% safe.

For me I've used family history and then added 10 years to account for improvements in health care. This puts my SWR life expectancy around 100, while actually it is probably much closer to 90. Of course I could be dead tomorrow :'(. If I live to be 120 :D my port may not make it, thats a chance I'm willing to take because I don't want to have to work even longer so that my portfolio can handle the extra 20 years that are very unlikely. It all comes down to what you are comfortable with.
 
Dory, I must not be understanding what you mean when you say you can re-run the program during a "bubble" (or trough), and remain at a 100% survival. If I start FireCalc with all the default options, and just change the starting withdrawal to $25k, then I get 100% success. If I had done that on Jan 2000, and suffered the resulting 45% drop in the S&P, then, ignoring withdrawals and rebalancing, I'd have about $480k ($650k * (75% * 65% + 25%) ). Of course, I'd actually have quite a bit less (maybe $425k) because my 25% money market wouldn't have covered my $25k withdrawal every year, and rebalancing would have increased my stock market losses. Now, if I run FireCalc again with $480k and $25k withdrawal with 27 years remaining, then I get an 89% success rate. if I re-run with $425k, then I get a 75%.

In general doesn't FireCalc assume that the market will follow the same path it has historically? Historically, the crash of '29-32 wasn't followed by another 50% drop, but if you start from a 50% drop, then FireCalc will use the great crash as one of it's possibilities, wiping out your 100% successful portfolio.
 
What I meant about the 100% safe rate being OK to calculate at a "bubble" was this:

If for your investment mix etc a $40,000 withdrawal showed as 100% safe for 40 years with a starting portfolio of $1 million, then the effects of the dramatic declines in the Great Depression have already been factored in, and your plan worked anyway. You'd still take an inflation-adjusted $40,000 a year out, and should be OK for 40 years. That's pretty much the whole concept of the safe withdrawal rates.

Now, suppose that $1 million portfolio grew to $2 million after you retired. The safe withdrawal rate you had previously chosen is not relevant -- you can start a new safe withdrawal rate using the larger portfolio. The balance of your portfolio at some past time is not relevant.

That's what I meant about rerunning the program.

This may seem like something for nothing, but what you are really doing is decreasing your likely estate.

Does that help?

Dory36
 
This is just an "attitude thing". My wife and I always
assume an early demise. Thus, looking way out
into the future has little interest for us. To put it another way, looking at our situation if I live to (say
80) seems to have little relevance. On the other hand,
only a fool would not allow for the possibility.
 
johngalt,

I have to agree with your outlook on life beyond [say age 85]. For the majority, these people aren't "living"; most are just "waiting to die".

I would say that at this point, I really won't care all that much what my portfolio looks like. Of course, maybe in 35 years I won't feel the same way. I'll cross that bridge when I get to it!

Red
 
Fascinating and throught-provoking replies.

I was indeed thinking roughly in terms of a target period inspired by one's life expectancy (though I'd call it Russian Roulette to use that figure itself!), so exactly as Ted says, whatever made you pick 40 years, back 5 years ago, would not make you pick 35 years now, though I wouldn't pick 40 now either, more like 36 say - same behaviour as life expectancy, it stretches but not at 1 year expected/1 year survived.

Dory36, thanks for reminding me that 100% safe rate is indeed safe! I suppose that is a sound argument for not messing with substantially lower percentages just to retire earlier - that's Russian Roulette too.

Thanks people.
 
If the stock market does really well, you can use the Pay Out Period Reset Method to increase your withdrawals while staying within the 100% SWR. Whenever your portfolio increases in value, you can take the 100% SWR percentage from the new portfolio high. For example, Bill G can take 3 1/2 % of his 36 billion this year. If his portfolio increases to 100 billion in 5 years, he can take 3 1/2 % of 100 billion, inflation adjusted, for the rest of his life and still be in the 100% SWR.

I can certainly agree with enjoying life while you are still healthy enough to do so. :)

Check out the math with this downloadable Excel calculator:

http://rehphome.tripod.com/popr.html
 

Latest posts

Back
Top Bottom