Bogleheads

[johndcraig]

This was also posted on the other board, but since it's out of commission on Diehards.org I'm also posting it here. Interested to see what you all think.

Here are three scenarios that tie together a couple of recent threads. In each scenario we assume that the cost of living has increased over the first 10 years so that twice as much cash is needed to cover expenses in Year 10 as was needed in Year 1. The withdrawal rate being used is based upon Trinity and other studies that show 4% adjusted for inflation to be a safe withdrawal rate over a 40 year period. The individuals are all 55 in Year 1, and each want a plan that will survive until age 95.

Part I.

Scenario 1 -- poor returns. Alex begins retirement in Year 1 with 100k. He needs 4% adjusted for inflation for annual living expenses. By Year 10 his annual needs and withdrawal is 8k and his portfolio balance is still 100k (Similar to the 70s). The Year 10 balance of 100k is 50k in Year 1 dollars.

Scenario 2 - base period. Ben begins Year 1 retirement with100k, with the same needs and withdrawals as Alex. In Year 10 his balance is 200k, which in Year 1 dollars is 100k.

Scenario 3 - good returns. Chuck also begins Year 1 retirement with 100k and withdraws the same as A and B. At Year 10 his portfolio has grown to 400k (similar to the 90's). The 400k balance in Year 1 dollars is 200k.

Question: Should Alex, Ben and Chuck each continue to withdraw 8k adjusted by inflation in future (30) years? If not, what should they withdraw and why?

Part II.

In Year 10 Art, Bert, and Carl all begin retirement at age 65. In Year 1, 10 Years before retirement, Art had a much larger portfolio than Carl, but now A, B and C each has a portfolio of 200k and each need 8K for living expenses. Should Art in Scen1, Bert in Scen2, and Carl in Scen3 each withdraw 4% (adjusted by future inflation) of their 200k for 30 years to get the required 8K? If not, what different amount should they withdraw and why?


There are usually those who will take issue with the example chosen. Of course the example is extremely simplified to demonstrate specific points. If anyone does spot errors in this methodology that would materially change the conclusions, please let me know what they are so the example can be changed.

John

[ccbwc]

I maintain a 3% withdrawal rate right now. If my portfolio goes up, then I get more spending money. As I age, I will increase my withdrawal percentage.

Regards
_________________
Chip Plumb

[bob90245]

[johndcraig]

Bob

I got it from another study you posted elsewhere (I don't remember where). I doubt that a change from 4% to 3.4% would materially change what this thread is intended to illustrate. Do you think it would? Otherwise it is necessary to keep it as simple as possible.

If you like, go ahead and make the change to 3.4%. Then how do you answer the above questions?

John

[bob90245]

[johndcraig]

Thanks Bob

You have answered Part II, but you may have oversimplified Part I.. If Art is withdrawing the correct amount of 4% of 200k, doesn’t that mean the correct amount for Alex should be 4% of 100k (so much for SWR)? If, say, 6% is ok for Alex, then why not 6% for Art? You also didn’t mention Chuck. If Carl is withdrawing the correct amount of 4% of 200k, doesn’t that give Chuck the green light to double his withdrawals to 16k or 4% of 400k?

John

PS Thanks Mel, I forgot about the edit function

[Mel Lindauer]

[bob90245]

OK, let me provide more detail for part I.

In scenario 1, Alex should not have waited 10 years to allow his portfolio to dwindle by half. So my previous comments apply primarily to his case.

In scenario 3 (part I), Chuck need not have waited until year 10 to boost his withdrawal rate. Again, following one of the Variable Withdrawal Strategies would have permitted Chuck to withdraw a bit more as the portfolio size swelled (and the corresponding CURRENT withdrawal rate declined).

So it cuts both ways. Tighten one's belt, as necessary, if returns are poor. Or if the markets give returns more generous than anticipated, spend a bit more.

[johndcraig]

Bob

Obviously, it was not just the 4% adjusted withdrawals that reduced Alex’s portfolio. It was primarily the market and inflation that hurt him, not his 4% withdrawals. He didn’t know what would happen in the market from one year to the next or what inflation would be. If Art’s 4% is the SWR in Year 10, why would a greater withdrawal by Alex be ok? If it’s ok for Alex to withdraw more than 4% in Year 10, why isn’t it also ok for Art?

You say it is OK for Chuck to withdraw a “bit more”. If it is ok for Carl to withdraw 4% in Year 10, why shouldn’t it be ok for Chuck to withdraw the full 16k?

John

[bob90245]

[Phil]

In addition to Art, Bert and Carl, we also have Dave, who is married to Sue. Dave also happens to have 200K to fund retirement, and the estimated combined living expenses of Dave and Sue are 8K.

Carl is discussing his plans with Dave, and offering him advice. Carl notes that he had originally planned to take out 8K for expenses, adjusted for inflation, but just learned that the 4% rule only applies to horizons as long as 30 years. With a distinct possibility of a 40 year remaining lifetime – or more, Carl is telling Dave he will have to tighten the belt a bit, and live on 6.8K per year. Carl hopes that his portfolio will increase in the future, so that 3.4% of the larger amount will be closer to his original expense projection, but ruefully notes that his portfolio could take a hit, and prudence suggests that he take 3.4% of the lower amount, which could be even less than the new lower target.

"Plus", Carl says to Dave, "you have it even worse. With two people, you have to worry about the money lasting until the second of you passes away, so your time frame is longer than ours."

"No", Dave responds, "I'm not worried, because I got advice from Phil. I'm planning to spend 8K per year, adjusted for inflation. If the stock or bond markets drop, I'll continue to spend the same. And not only that, the 3.4% SWR you are using reduces the probability of exhausting your funds to a very low level, but not to zero. My likelihood if running out is closer to zero."

"Sorry, Dave, that cannot be", Carl says. "Haven't you heard that there's no free lunch? The only way you can take a higher withdrawal than us is to take more risk, and that will increase the chance that you will run out of money."

Dave responds – "I agree, Carl, that there is no free lunch. You can reduce your chance of bankruptcy to an acceptable level by lowering your withdrawal rate. The price you pay is lower income during retirement, the flipside is that there will probably be some money for heirs. I can understand why they would want you to do that, but we've already given what we choose to give to our heirs. Our retirement fund is for our retirement. I don't plan to tighten my belt, and face the chance of tightening it even more later just so there can be money left for heirs."

"Sounds impossible, Dave. How do you think you can manage this?" asks Carl.

"Very simple", responds Dave.

"I purchased an annuity from Vanguard. For 200K, they will pay me 8K per year, increase the payment each year for inflation, and make the payments as long as I or Sue are still alive."

Can someone explain to me why Dave, whose 200K can fund 8K per year, adjusted for inflation, whether the markets go up or down, should follow the plan of Carl, who also starts with 200K, but can safely take out only 6.8K per year?

[johndcraig]

Sorry if the facts are confusing

All 6 individuals have 30 years left in year 10 under the facts. They were all 55 in Year 1 but three didn't retire until age 65.

In both scenario 1 and 3 it was the market that primarily accounted for the low and high balance at Year 10. Small withdrawal adjustments in the first 10 years would not have had a great impact.

You seem to have little or no concern with Carl's 4% withdrawal despite the large run up in the prior 10 years. You do, however, have concern with the flip side where Alex just had 10 bad years. Some would say that SWR theory would indicate the opposite and things would more likely improve for Alex after 10 bad years and worsen for Carl after 10 good years.

John

[BrianTH]

[johndcraig]

[bob90245]

[BrianTH]

[runchman]

I think what you are pointing out is the paradox that if x% swr is 'safe' looking forward when you retire, why can't you then treat each subsequent year as a 'new' retirement and adjust your w/d upward after good market years.

My answer is that since x% was established by simulating *not* increasing your w/d after good years, those good years give a boost to your portfolio, without which the established 'safe' x% would have been lower.

So somehow you are cheating the statistics by raising your w/d on good years, although intuitively it doesn't make sense - hence the paradox.

Take a limit case, how about on a continual basis you examine your portfolio value, and ratchet up your withdrawal rate so you're 4% is based on the daily peak value. Should be safe going forward based on the study, right? Yet somehow it fails the common sense test and would not be a good thing to do.

Perhaps we need a trinity study with a sliding scale of 'safe w/d amount determination', with an associated safe rate.

establish once at start of retirement = 4% swr
establish based on 5-year peak port value = 3.4% swr
...
establish based on daily portfolio value = 2.8% swr

Somehow I think the 'take x% at the start and adjust for inflation' makes more sense.

- John

[BrianTH]

[runchman]

[BrianTH]

[runchman]

Oh great, just what I need, another source of diversions from work, another site to read!

- John

[johndcraig]

What I was trying to get at with this thread

Backtesting withdrawal rates shows that 4% is safe for 30/40 years based on various studies. Of course, safe means only that the backtesting worked in the past and not that it will work in the future. Accordingly, the talk about it actually being safe is wrong IMO. Alex’s situation demonstrates this and shows that most people aren’t buying into the “safe” withdrawal rate. After 10 years, market declines resulted in Alex now withdrawing at an 8% rate. If we look at the 70’s, the 80’s and 90’s market would have bailed one out in these circumstances, but good sense tells us that that is too much of a gamble and that Alex should probably take a more conservative approach.

The point that most didn’t pick up on was Carl’s situation. People were conservative when dealing with Alex, and even Chuck, but not with Carl. Some were hesitant to suggest that Chuck should now withdraw 16k/year (4%), but had no problem saying the same percentage withdrawal was right for Carl. I understand a typical conservative reaction for Chuck, but I have a tougher time understanding why it doesn’t carry over to Carl. My guess is that this goes back to people believing in theSWR “rule”, so no one questioned the straightforward 4% withdrawals for the 3 Year10 retirees. IMO, common sense worked for Alex and Chuck, but SWR “rule” took over when it came to Carl.

Well, that’s my take-away on this. Clear as mud?

John

[HockeyMike35]

[johndcraig]

A few points

Mike your poll is qualified to mean “do you think it is a good conservative starting point for most retires?” No problem, I would answer yes to that. 4% is much better than suggesting some precise number as others have attempted to do. It seems somewhat conservative and it might prove a helpful rule-of-thumb.

What is also true IMO is that the Trinity and similar studies are simply historical studies with some relevance and importance for future planning. Certainly they do not come up with a “safe” withdrawal rate. For example, they are based upon historical returns that average 10.6% and nearly all experts believe that number way too high. They also don’t work as a practical matter, as nearly everyone has pointed out with respect to Alex’s Scenario 1. So sure, 4% is a good conservative starting point, but that is all.

I think what this thread illustrates, yet most are not fully willing to accept, is that the 4% “rule” must be totally flexible and continually reconsidered in light of all of the facts. It is not a “rule” and it should never be considered as anything precise. When push comes to shove, I think most people agree that flexibility is needed; the problem I have is that they still seem to cling to the “rule”. Facts proved some adjustments were necessary for Alex, and everyone agreed. No one yet is willing to even consider that Carl’s situation may be the riskiest of the six scenarios, so just maybe Carl should be considering this as he goes forward. When it comes to Carl, flexibility goes out the window and everyone runs back to the comfort of the 4% Safe withdrawal rate rule.

I would not suggest changing the 4% rule-of-thumb for Carl. It would be wrong IMO to suggest, as others have in the past, that some other precise lower number should be used instead. What Carl should do, however, is: consider the fact that he is likely in the riskiest position of all; stay flexible and conservative with regard to overall allocations; and realize that there is nothing “safe” about the 4% rule-of-thumb that he is using.

IMO the biggest problem is that many take comfort in a “Safe” withdrawal rate when it is nothing more than a convenient rule-of-thumb.

John

[johndcraig]

To all titled “Published author in finance”

How about it? What are your views on this matter? What do you believe to be the appropriate context in which one should consider and use the 4% SWR?

John

[Norbert Schlenker]

[DRiP Guy]

I can't believe I read this thread. I want my 5 minutes back... red herrings, straw men and false assumptions.

4% is a rule of thumb, for a well-diversified portfolio. It has worked over the scenarios back-tested for, so if things in future are like past, it is a pretty good place to start planning. That's it.

I see NO GROUP OF PEOPLE CLAIMING ANYTHING MORE!!!!!

THERE NEVER WAS!!!!!

JDC, precisely what aspect of the process of establishing the information is in error? I don't mean what you think someone might someday misinterpret it to mean thereby misapplying it; I mean what 'error' in calculation, or in methodology is inherent to the current information or how it is presented or contextualized?

Every time I have seen respected authors or posters on forums use the information from Trinity or other historical SWR explorations, it is with all due caution to it's nature being of historical information, not of some magical and hugely precise predictive powers.

If you have a better model, an specific analytical criticism tracible to a SPECIFIC claim or method or author or article, perhaps a math error to point, out, etc, I know that I and many others would be completely delighted to hear of it, and to discuss it at length.

edit - three paragraphs have been removed, let's please focus on the substance of an argument and leave out the speculations over motive - Alex

[johndcraig]

Moderators, what about it?

First a couple of quick comments on Dripguy’s post

[bob90245]

[Norbert Schlenker]

Readers of this thread might find a post I made at FWF last summer worth thinking about. It discusses the sensitivity of withdrawal rates to various inputs.

[DRiP Guy]

[TnGuy]

You folks may have already seen this story, but I just ran across it in my travels and thought it would fit into the discussion on this thread:

Keeping the gold in the golden years - How to spend more money in retirement without going broke


David
_________________
"Money will not make you happy. And happy will not make you money." - Groucho Marx

[DRiP Guy]

[DRiP Guy]

[HockeyMike35]

[johndcraig]

Mike

[HockeyMike35]

[johndcraig]

Historical average returns

[Taylor Larimore]

It crosses my mind that you have not read the Trinity Study. Here is a link:

Retirement Savings: Choosing a Withdrawal Rate That Is Sustainable

Best wishes.
Taylor

[HockeyMike35]

[bob90245]

[johndcraig]

Taylor and Mike

Yes I have read the study Taylor, but I’m not sure what point you are making.

Mike, If the past average was 10.6% and the expected future average returns are 3% less (if you like, assume real returns so you can avoid inflation), then given any past 30 year period, high or low, is the return for that past period more likely inflated by some percent or not? Are you suggesting that in the future the 3% average reductions will all come from the high return periods? Is there some reason to think that the future return reduction percentages will not come from both the high and low periods? If lower returns do come from both periods, how much value are the historical studies?

The worst 30 year periods generally come from a poor period followed by a good period. So even if the reduction does come only from good periods, lower recovery in the good periods will put the 4% in jeopardy. Consider the 70’s where the recovery occurred in the 80’s and 90’s. If those later good two decade’s returns were substantially reduced, how certain would one be that the 4% over the entire period would be safe?

I am not saying Trinity type studies are valueless, but if future equity returns drop across the board, the value of such studies is certainly diminished. In the same way, MC simulations based on average returns that are too high lose their value.

John

PS Mike, yes it is semantics, that is why examples are helpful.

[johndcraig]

Bob90

I posted my last post before reading yours. Remember the main point of this thread is to question the reliability of historical data, and not to determine a SWR. It is not about 3.27% vs 4%. As far as I’m concerned, 4% is as good a rule of thumb as any when one accepts that it has low reliability. It is the degree of reliability of historical that is important IMO, and not some attempt to derive a perfect SWR.

John

[HockeyMike35]

So if you want to use a 50/50 portfolio. You believe that the next 30 years equity returns will be 3.5% lower then the worst in history and inflation, volitility and all other factors will be the same. Then lower your withdrawl rate to 3.27%. I do not share such a pessimistic view of the future. This example is an excellent illistration of how conservative the 4% number is.

I think it also important to note Bob's findings with a 50/50 portfolio.

[Orion]

On the other hand, if you believe that life on Earth will be wiped out by an asteroid impact in a year, then 100% should be very safe. "for discussion purposes", as said above, we can generate any answer we wish.

However, I don't know any way to prove that such contrivances are better than history, so I agree with Taylor that: "In my opinion, the Trinity Study will give us as good an answer as we're going to get."

At least until something better comes along. (But I can't predict when that will happen either.)

[bob90245]

[johndcraig]

Bob

Of course, I believe that lower future returns must mean a lower SWR. My concern is that everyone is looking for mathematical formulae that will provide them with the answers to AA, SWR, etc. My view is that the important thing is not to get sucked into believing that such formula answers exist. The emphasis therefore must be on the lack of reliability of the data and the dangers of accepting formula answers. Rules of thumb that are agreed to be merely helpful but not reliable are what make sense to me. That’s why I think 4% is as good as any.

When a person shifts thinking from, I will adopt 4% as a “safe withdrawal rate” to, I will generally plan at 4%, but continually monitor the situation and risk, then that person has the right mind set IMO.

John

[bob90245]

John, I guess I don't understand the nuance you are trying to convey.

If you believe that "lower future returns must mean a lower SWR" (than 4%), then why call for people to "generally plan at 4%"? Wouldn't that be setting them up for failure? Especially if you believe that 4% SWR "has low reliability" for the future.

[Norbert Schlenker]

[HockeyMike35]