Bogle's 10-Year Forecast

[daylatedollarshort]

Here is a related article form the WSJ -

Say Goodbye to the 4% Rule for Retirement - WSJ

"If you had retired Jan. 1, 2000, with an initial 4% withdrawal rate and a portfolio of 55% stocks and 45% bonds rebalanced each month, with the first year's withdrawal amount increased by 3% a year for inflation, your portfolio would have fallen by a third through 2010, according to investment firm T. Rowe Price Group. And you would be left with only a 29% chance of making it through three decades, the firm estimates."

 

[Fermion]

Here is a related article form the WSJ -

Say Goodbye to the 4% Rule for Retirement - WSJ

"If you had retired Jan. 1, 2000, with an initial 4% withdrawal rate and a portfolio of 55% stocks and 45% bonds rebalanced each month, with the first year's withdrawal amount increased by 3% a year for inflation, your portfolio would have fallen by a third through 2010, according to investment firm T. Rowe Price Group. And you would be left with only a 29% chance of making it through three decades, the firm estimates."

Through 2010. But the article is from 2013 and look what the market did from 2010 to 2015...it went way up. I am curious what formula they used to estimate the 29% chance. Did they have other decades of dot com and mortgage crashes to compare against?

 

[razztazz]

Here is a related article form the WSJ -

Say Goodbye to the 4% Rule for Retirement - WSJ

"If you had retired Jan. 1, 2000, with an initial 4% withdrawal rate and a portfolio of 55% stocks and 45% bonds rebalanced each month, with the first year's withdrawal amount increased by 3% a year for inflation, your portfolio would have fallen by a third through 2010, according to investment firm T. Rowe Price Group. And you would be left with only a 29% chance of making it through three decades, the firm estimates."

Why a steady 3% inflation? Some of those years (maybe many) had much lower inflation and at least one year was slightly negative.

But along the same lines, someone else (rddr or something like that) has done a similar study and gotten the same results: Year 2000 retiree ain't gonna make it.

I don't understand the hand wringing from these people (except that scare articles sell ...?) In the past 1906, 1937, 1966, 1967 and maybe another have failed the 4% regimen. 2000 while not good and quite possibly a failure, is just another data point among the 4 or 5 that have already flopped.

OK, so we have ONE MORE fail sequence. The 4% rule is no more dead because of 2000 than it was because of 1966. I think it's the principle of recency at play.

It might be dead going forward due to low dividend payouts, chronically high PEs or something else nobody has seen yet, but you can't tell just from 2000.

 

[daylatedollarshort]

Through 2010. But the article is from 2013 and look what the market did from 2010 to 2015...it went way up. I am curious what formula they used to estimate the 29% chance. Did they have other decades of dot com and mortgage crashes to compare against?

I don't know what formula they used. The Betterment CEO (not an unbiased source as he is shilling his own SWR methodology) is saying something similar in a more recent article because rates are currently so low compared to when the 4% rule was developed (5 year Treasuries in 2002 were 4.5%, today 2%):

The 4% retirement rule is broken ... and here's why

They could all be wrong about the future. Economic prediction often are, but the change in Treasury rates is a fact, as well as Shiller's current PE ratios for stocks.

 

[ERD50]

Originally Posted by daylatedollarshort View Post
Here is a related article form the WSJ -

Say Goodbye to the 4% Rule for Retirement - WSJ

"If you had retired Jan. 1, 2000, with an initial 4% withdrawal rate and a portfolio of 55% stocks and 45% bonds rebalanced each month, with the first year's withdrawal amount increased by 3% a year for inflation, your portfolio would have fallen by a third through 2010, according to investment firm T. Rowe Price Group. And you would be left with only a 29% chance of making it through three decades, the firm estimates."

....

I don't understand the hand wringing from these people (except that scare articles sell ...?) In the past 1906, 1937, 1966, 1967 and maybe another have failed the 4% regimen. 2000 while not good and quite possibly a failure, is just another data point among the 4 or 5 that have already flopped.

OK, so we have ONE MORE fail sequence. The 4% rule is no more dead because of 2000 than it was because of 1966. I think it's the principle of recency at play.

It might be dead going forward due to low dividend payouts, chronically high PEs or something else nobody has seen yet, but you can't tell just from 2000.

Agreed. Their assumptions seem pretty fuzzy.

Look at this - I ran a 3.5% WR (35,000 from 1,000,000 portfolio for easy math) for 30 years in FIRECalc, and it has 100% success.

Now trim the time frame down to 10 years. The drop in these same 100% historically successful retirements was:

The lowest and highest portfolio balance at the end of your retirement was $413,411 ​

That's ~ a 59% drop, and it still provided a 100% historically successful 30 year retirement - so what's the big deal about a 33.3% drop?

-ERD50

 

[razztazz]

Agreed. Their assumptions seem pretty fuzzy.

Look at this - I ran a 3.5% WR (35,000 from 1,000,000 portfolio for easy math) for 30 years in FIRECalc, and it has 100% success.

Now trim the time frame down to 10 years. The drop in these same 100% historically successful retirements was:

The lowest and highest portfolio balance at the end of your retirement was $413,411 ​

That's ~ a 59% drop, and it still provided a 100% historically successful 30 year retirement - so what's the big deal about a 33.3% drop?

-ERD50

Good point. These people like to point out that 4% rule is dead but they don't indicate what is safe. I think one of these writers has said something like 3% is the new 4% but his assumption was a 1% management fee!

You got a 100% safe with 3.5%. A long time ago I got 3.3%. If I am not mistaken, Bengen found that 3.5% was a "forever withdrawl rate."

So, now all that needs to be discerned is where 2000 is on the spectrum of failures. Is it worse than 1906, 1966...? Maybe 3.5% would be just swell?

 

[statsman]

Good point. These people like to point out that 4% rule is dead but they don't indicate what is safe. I think one of these writers has said something like 3% is the new 4% but his assumption was a 1% management fee!

You got a 100% safe with 3.5%. A long time ago I got 3.3%. If I am not mistaken, Bengen found that 3.5% was a "forever withdrawl rate."

So, now all that needs to be discerned is where 2000 is on the spectrum of failures. Is it worse than 1906, 1966...? Maybe 3.5% would be just swell?

Under what range of AAs would a 3.5% withdrawal rate be considered 100% safe? Also, and I apologize ahead of time, is the 3.5% withdrawal rate based on the initial portfolio value, or is the 3.5% rate based on the yearly balance?

 

[Markola]

Why a steady 3% inflation? Some of those years (maybe many) had much lower inflation and at least one year was slightly negative.

That's what I've been wondering. Similarly, must one's personal inflation rate be necessarily 3%? The "Your Money or Your Life" book was pretty persuasive that retirees can control most of their inflation rate through substitution effects, e.g. if banana prices are up, buy apples. That's the point about living in the nice RV in New Mexico and even the healthcare wild card could be managed somewhat through international options. So, if one's personal inflation can be managed to be closer to 1% but the models generally assume steady 3%, then market returns of, say, 5% versus 7% would seem to be a steady state. I guess I, too, am trying to quibble with the WSJ article because, from what I read on this board, folks have a great many lifestyle levers to pull well short of portfolio failure and going back to w*rk.


Sent from my iPad using Early Retirement Forum

 

[Fermion]

TIPS get you 30 years at a 3% SWR if your spending is low enough that taxes are not a factor. Imagine never ever worrying about stocks OR bonds. The only thing you would need to worry about is a collapse of government.

 

[razztazz]

Under what range of AAs would a 3.5% withdrawal rate be considered 100% safe? Also, and I apologize ahead of time, is the 3.5% withdrawal rate based on the initial portfolio value, or is the 3.5% rate based on the yearly balance?

Sorry but I no longer have that Bengen study I mentioned. It was done probably early 2000's though.

The 3.5% was as per normal. Initial portfolio value upped for inflation. And "forever" meant 70 years for purposes of the study. He assumed you'd have to work part of your life to get the money in the first place, and you're only going to live so long. I don't recall anything special or noteworthy about allocations so unless I dreamed this whole thing he must have used pretty off-the-rack AAs. 60/40 etc

I forget the details and techniques he used to project 70 years out but I believe he did try to compensate for the fact there are not a lot rolling 70 periods of market history.

 

[razztazz]

Under what range of AAs would a 3.5% withdrawal rate be considered 100% safe? Also, and I apologize ahead of time, is the 3.5% withdrawal rate based on the initial portfolio value, or is the 3.5% rate based on the yearly balance?

http://www.fa-mag.com/news/article-290.html

Here's the Bengen article I was remembering. Nov 2000 is the date on it. The charts and graphs unfortunately have been stripped

Salient bits:

In my earlier writing on withdrawals from retirement investment portfolios (which appeared in the Journal of Financial Planning), I assumed that clients would retire at age 65, live "only" to age 95, and therefore require only 30 years of withdrawals from their investment portfolios. I determined "safe" withdrawal rates, which would allow retirees to stretch their resources, based on historical investment returns and rates of inflation, until at least age 95.

After much thought, I decided the most useful analysis I could perform for the Methuselah client would be to determine safe withdrawal rates for much longer periods of time, up to at least 75 years.

As you might expect, as desired portfolio longevity increases, the safe maximum withdrawal rate decreases. It is also clear from the chart that beyond a certain portfolio longevity-about 65 years-there is very little decline in the safe maximum withdrawal rate, no matter how much the longevity is increased. In fact, the curve in the chart seems to approach an "asymptote," or ultimate value, of about 3.5%.

 

[NW-Bound]

http://www.fa-mag.com/news/article-290.html

Here's the Bengen article I was remembering. Nov 2000 is the date on it. The charts and graphs unfortunately have been stripped

Salient bits:

I have to commend you on your excellent memory. It is as good as mine.

First of all, FIRECalc shows that a 3.5% WR will nearly deplete a 50/50 portfolio in 30 years in the worst case. It did not survive 40 years. So, how do we reconcile it with Bengen's result?

Here's the flaw in his simulation. He duplicated the period of 1964-1999 over and over into infinity. We all know that the interval of 1983-2000 was the most exuberant period for investors. It ended in an extremely high note, and by replaying this over and over, surely our eternal-life retiree will live happily over and over. Had he chosen another interval that ended in a low note to repeat over and over, the result would have been that our "Methalusah" retiree would be broke shortly after 30 years. If he were to repeat this 2000 article today, people would immediately ask about using a period that ended in 2009. Remember that from 2000 to 2009, the S&P lost 1/2 of its value after inflation, and that is with a WR of 0!

Except:

The Ibbotson Associates data, on which I base my analysis, covers only 74 years of security returns and inflation, for the years 1926 through 1999. How should I model future investment returns (and inflation) to provide for time horizons as long as 75 years?

I decided to extend my model by using a past period of historical returns and repeat it over and over, as long as necessary. I settled on the 36-year period of 1964-1999 as being representative of my hypothetical future. Thus, for the period 2000 through 2035, I used the actual Ibbotson returns for large-cap stocks (LCS), intermediate-term government bonds (ITG), as well as the measure of inflation (CPI), for the years 1964-1999.​

For mortal retirees, a 3.5% WR will most likely survive a 30-year retirement. Nothing that Bogle said challenges that. It's just that we cannot expect to spend and still see our stash grows, and to be able to leave behind a big pile for our heirs.

And about what Shiller said, he was referring to young investors who are still accumulating, not geezers like us who were fortunate to work, save, and invest during the exuberant past which is not likely to repeat.

 

[petershk]

The thing I keep coming back to (and my personal poor record enforces) is that the biggest drag is my own decision making. While a 60/40 or 50/50 portfolio may not survive today at 4%... Taking action to try and beat what is given is on average much less safe. But that's really really hard to actually act on when you see a 20% drop or a 30% gain. Those swings cause action... Action causes fees, taxes and, frequently, below market returns.

I think that part of bogle's message is often lost. And btw I am not a believer that people like Buffett are lucky... I think they work at it 12-14 hours a day and have the right temperament. I don't work on it that much and don't know if I have the temperament so in moving to "don't look at it" portfolio albeit it's taking a bit of time.

Sent from my HTC One_M8 using Early Retirement Forum mobile app

 

[razztazz]

I have to commend you on your excellent memory. It is as good as mine.

First of all, FIRECalc shows that a 3.5% WR will nearly deplete a 50/50 portfolio in 30 years in the worst case. It did not survive 40 years. So, how do we reconcile it with Bengen's result?

Here's the flaw in his simulation. He duplicated the period of 1964-1999 over and over into infinity. We all know that the interval of 1983-2000 was the most exuberant period for investors. It ended in an extremely high note, and by replaying this over and over, surely our eternal-life retiree will live happily over and over. Had he chosen another interval that ended in a low note to repeat over and over, the result would have been that our "Methalusah" retiree would be broke shortly after 30 years. If he were to repeat this 2000 article today, people would immediately ask about using a period that ended in 2009. Remember that from 2000 to 2009, the S&P lost 1/2 of its value after inflation, and that is with a WR of 0!

Except:

The Ibbotson Associates data, on which I base my analysis, covers only 74 years of security returns and inflation, for the years 1926 through 1999. How should I model future investment returns (and inflation) to provide for time horizons as long as 75 years?

I decided to extend my model by using a past period of historical returns and repeat it over and over, as long as necessary. I settled on the 36-year period of 1964-1999 as being representative of my hypothetical future. Thus, for the period 2000 through 2035, I used the actual Ibbotson returns for large-cap stocks (LCS), intermediate-term government bonds (ITG), as well as the measure of inflation (CPI), for the years 1964-1999.​

For mortal retirees, a 3.5% WR will most likely survive a 30-year retirement. Nothing that Bogle said challenges that. It's just that we cannot expect to spend and still see our stash grows, and to be able to leave behind a big pile for our heirs.

And about what Shiller said, he was referring to young investors who are still accumulating, not geezers like us who were fortunate to work, save, and invest during the exuberant past which is not likely to repeat.

You are correct. I had been trying, without success, to relocate this article for years, to reconcile the parts I remembered with what my lying eyes tell me when I run FireCalc.

I remember in the early 2000's during the prolonged crash, rerunning scenarios in FireCalc but cutting things off at 1996. That's just when the bubble that led to the 2000-2003 crash started to inflate. I figured --nobody at any time-- should depend on abnormalities or one-off situations like that to secure their futures.

The cavalry doesn't always arrive in time. Where there is a will there is not always a way. My father was that way. "Don't worry. Nothin's gonna happen!" "Shhhht! You worry too much! His final years were very bad. Very sad.

 

[DEC-1982]

and on the other side ...

Ignore the retirement alarmists: The 4% rule is imminently safe

" In fact, Kitces went back to look at how those who retired in 2000 — the very same ones thatTheWall Street Journal used to question the rule — are faring today. The results: If they retired in 2000 with $1 million, they would still have about $900,000 left.However, we need to adjust that number for inflation: $900,000 today is worth about $650,000 in constant 2000 dollars. In other words, about one-third of the nest egg for these retirees has been depleted.

But, as Kitces points out, this assumes these couples truly did adjust their withdrawals for inflation every year. "A growing base of research suggests that retiree spending in real dollars tends to decline in later years," Kitces writes, "which means in practice, a 2000 retiree today is probably even better off and spending even less as a current withdrawal rate than these calculations would suggest." "

 

[daylatedollarshort]

I think one thing to keep in mind is that Firecalc and other similar tools assume the future U.S. investment environment will be like the past regardless of globalization, flash boys, the changes in China's economy and impact on the U.S., global wealth inequality, the advent of derivative type investments, historically low interest rates, different types and sizes of wars, the rise of mutual funds, technological changes, the rise of the supercorporations and trade pact transcending country boundaries, and many other factors.

If the U.S. investing future is like the past then Firecalc's numbers, assuming the programming is accurate, may be correct. If you believe there is less than 100% chance that the U.S. investing environment will behave like the past, then it might be good to take the Firecalc results and multiply them by a factor of less than 100%.

 

[explanade]

and on the other side ...

Ignore the retirement alarmists: The 4% rule is imminently safe

" In fact, Kitces went back to look at how those who retired in 2000 — the very same ones thatTheWall Street Journal used to question the rule — are faring today. The results: If they retired in 2000 with $1 million, they would still have about $900,000 left.However, we need to adjust that number for inflation: $900,000 today is worth about $650,000 in constant 2000 dollars. In other words, about one-third of the nest egg for these retirees has been depleted.

But, as Kitces points out, this assumes these couples truly did adjust their withdrawals for inflation every year. "A growing base of research suggests that retiree spending in real dollars tends to decline in later years," Kitces writes, "which means in practice, a 2000 retiree today is probably even better off and spending even less as a current withdrawal rate than these calculations would suggest." "

So maybe the 2000 retirees withdrew a lower percentage after late 2008 for a couple of years and/or withdrew based on their year ending 2008 and 2009 balance rather than their 2000 starting balance?

 

[petershk]

I think one thing to keep in mind is that Firecalc and other similar tools assume the future U.S. investment environment will be like the past regardless of globalization, flash boys, the changes in China's economy and impact on the U.S., global wealth inequality, the advent of derivative type investments, historically low interest rates, different types and sizes of wars, the rise of mutual funds, technological changes, the rise of the supercorporations and trade pact transcending country boundaries, and many other factors.

If the U.S. investing future is like the past then Firecalc's numbers, assuming the programming is accurate, may be correct. If you believe there is less than 100% chance that the U.S. investing environment will behave like the past, then it might be good to take the Firecalc results and multiply them by a factor of less than 100%.

You mean like Vietnam, the Cuban missile crisis, leaving the gold standard, 2 world wars, a great depression, the internet, television, S&L scams, the nifty 50... Stuff like that?

I expect the future will look a lot like the past in those terms. How the market will look...no idea. But then... Not really much alternative.

It's the creation of wealth through business, innovation and infrastructure that drives investment returns. Now if those things go away in not sure what investment is "safe." People say gold maybe... But if you're in a society that is collapsing gold won't buy much either. Cash is useless when that happens. Maybe something like farmland and animals so you can at least make food...

I'm optimistic that the system will keep working and churning out amazing things that improve our lives.

Sent from my HTC One_M8 using Early Retirement Forum mobile app

 

[DEC-1982]

So maybe the 2000 retirees withdrew a lower percentage after late 2008 for a couple of years and/or withdrew based on their year ending 2008 and 2009 balance rather than their 2000 starting balance?

There are two events, the 2000 crash and the 2008 crash. It's hard to say what specific steps people took. I am guessing anybody who has assets in the market and sees the market drop by 50% is going to change his or her spending behavior.

I re-read the article and understood it as follows:

if they increased their 2000 spending by inflation every year, then starting from a $1million portfolio in 2000, they would still have $900,000 today.

But the reality is that they did NOT increase their spending every year, so they would have more than $900,000 today.

On the lack of spending increases with inflation, this pretty much matches what the Bureau of Labor Statistics reports, although one cannot assume people had money to spend and didn't; maybe they had no money to spend and therefore spending went down.

This thread is about Mr Bogle and his 10-year estimate. I don't know if I got off-topic; I just wanted to end the day on a happy note.

 

[daylatedollarshort]

You mean like Vietnam, the Cuban missile crisis, leaving the gold standard, 2 world wars, a great depression, the internet, television, S&L scams, the nifty 50... Stuff like that?

I expect the future will look a lot like the past in those terms. How the market will look...no idea. But then... Not really much alternative.

There is an alternative, especially for essential retirement expenses. There is no one size fits all AA for every retirement investor, but personally I am risk adverse and more interested in capital preservation than growth. John Bogle says "invest we must" but his company is not going to push investments they do not make profits from, even if those profit margins are less than his competitors. He is still in business to make money.

 

[samclem]

There is an alternative, especially for essential retirement expenses. There is no one size fits all AA for every retirement investor, but personally I am risk adverse and more interested in capital preservation than growth. John Bogle says "invest we must" but his company is not going to push investments they do not make profits from, even if those profit margins are less than his competitors. He is still in business to make money.

Bogle is retired from Vanguard, it's "his company" only in the sense that he started it. He doesn't make one more cent if people invest with Vanguard. He does make money if people buy his books, same as the folks "pushing" (your term) that liability matching strategy.

I think one thing to keep in mind is that Firecalc and other similar tools assume the future U.S. investment environment will be like the past . . . .

No, FIRECalc makes no such assumptions at all. The program shows how a given set of investments, withdrawal strategies, etc would have performed in the past. It is left to the user to make any assumptions about the future. The instructions in FIRECalc are clear on this limitation/virtue.

 

[samclem]

. . . " In fact, Kitces went back to look at how those who retired in 2000 — the very same ones thatTheWall Street Journal used to question the rule — are faring today. The results: If they retired in 2000 with $1 million, they would still have about $900,000 left.However, we need to adjust that number for inflation: $900,000 today is worth about $650,000 in constant 2000 dollars. In other words, about one-third of the nest egg for these retirees has been depleted."

Hmmm, they've logged 15 years of retirement in one of the biggest drops on US equity value ever, and only burned through 35% of their money. At this rate, their dough will only last another 29 years, and if they were 65 when they retired they'll be flat broke when they are 109 years old unless they pare back their spending. Sounds like they are doing pretty darn well, and should hope for a continuation of this "bad luck".

 

[NW-Bound]

Hmmm, they've logged 15 years of retirement in one of the biggest drops on US equity value ever, and only burned through 35% of their money. At this rate, their dough will only last another 29 years, and if they were 65 when they retired they'll be flat broke when they are 109 years old unless they pare back their spending. Sounds like they are doing pretty darn well, and should hope for a continuation of this "bad luck".

Strictly speaking, if a retiree spent 35% of his stash on the first 15 years of retirement and if this 15-year period repeats, he will spend more than 35% of the initial value this 2nd time. It's because he now has only 65% of the stock dividends and bond yields that he had the 1st time, and his 4% WR is now 6.2% WR.

Still, it is likely that he will make it another 15 years. Thus, the 4% WR succeeds once more, and the 2000-2015 period did not fall out of the range of past market performance.

But, but, but how many of us are prepared to see our "incredible shrinking portfolio" right before we croak? When people think of 30-year retirement survival, they are still thinking of dying and leaving behind the same stash as they start their retirement with. They are still seeing in their mind the runs of FIRECalc that soar up and to the right of the chart.

People have to be reminded every so often that 4% WR means they may end their life being thousandaire and no longer a millionaire. And that should be considered a success, not a failure.

 

[daylatedollarshort]

Apologies in advance if the article in the link below has been posted already here, but it was a new article to me on the same subject and addresses the question of how accurate have Bogle's forecasts been in the past:

Bogle: Stocks 6%, Bonds 3%

"Formally, the relationship between Bogle's 10-year stock forecasts and subsequent market returns is 0.81. The even-simpler bond model has an even-higher correlation of 0.95, illustrating once again that with investment projections, less can be more. (Greater complexity does not necessarily mean greater sophistication; as Mark Twain wrote, sometimes people compose long letters because they lack the time--and perhaps the skill--to draft a short one.)

Thus, the evidence suggests that while Bogle's forecasts aren't reliable guides for a single, specified time period, they will likely be close to the mark over a sufficiently long term."

 

[CaliforniaMan]

...Formally, the relationship between Bogle's 10-year stock forecasts and subsequent market returns is 0.81. The even-simpler bond model has an even-higher correlation of 0.95...

This quote from the article does not make any sense to me. Correlation of what to what? Such high correlations seems to indicate he is correlating a time series, which of course you cannot do. Maybe I am missing something?