cancel
Showing results for 
Search instead for 
Did you mean: 
     
Highlighted
Contributor ○○○

Re: Bengen's 4% Distribution Method

I thought I’d do a comparison of the various withdrawal methods I've tested.

The initial investment was $500,000 in Vanguard Wellington for all methods.  1968 Wellington is the worst case scenario.  With a 75/25 asset allocation Bengen had two of the 55 rolling 30-year retirement periods fall short by one year.  1968 Wellington fell short by over 5 years.

 Method 1:  Initial withdrawal is investment times 4%.  Subsequent withdrawals are increased by inflation.

Method 2:  Initial withdrawal is investment times 4%.  Subsequent withdrawals are increased by inflation minus 1 percentage point (i.e., if inflation is 3.4% the 2.4% is used). 

Method 3:  Initial withdrawal is investment times 4%.  Subsequent withdrawals is either the beginning of year balance times 4% or the previous withdrawal whichever is larger.

Method 4:  Initial withdrawal is investment times 4%.  Subsequent withdrawals are one-half beginning of year balance times 4% plus one-half of the previous withdrawal adjusted for inflation.

Method 5:  All withdrawals are beginning of year balance times the modified required minimum distribution rate.  The required minimum distribution rate is the MRD rate plus 1.24.

Results of the test: None of the methods were ideal.  All resulted in a loss of purchasing power.  The only difference was how much and when.

Method 1 maintained purchasing power until 1992 when the partial payment resulted is a 70% drop.  In 1993 the loss was 100%.  The total of all withdrawals was $1,114,000.  Ending balance was $0.

Method 2 resulted in a slow decrease in purchasing power.  Half way thru in 1982 the loss was 12%.  The retiree needed $55,431 to maintain his standard of living and he got $48,633.   At the end in 1997 the loss of purchasing power was 24%.   Total withdrawals were $1,359,000. Ending balance was $294,000.

Method 3 rapidly lost purchasing power during the first half of retirement.  By 1982 purchasing power was down 61%.  The retiree needed $55,431 and he got $21,737.  Things then started getting better and at the end the retiree had 5% more purchasing power than he had when he started.  Total withdrawals were $1,157,000.  Ending balance was $2,899,000.

Method 4 was similar to method 3.  By 1982 purchasing power decreased 57%.  The retiree needed $55,431 and he got $23,641. At the end of retirement purchasing power had increased 4%. Total withdrawals were $1,145,000.  Ending balance was $3,132,000.

Method 5 was similar to methods 3 and 4 just not as drastic.  By mid-retirement in 1982 the retiree’s purchasing power decreased 49%.  He needed $55,431 to have maintained his standard of living and got $28,148.  Things got better and his purchasing power at the end of retirement was a whopping 47% higher than when he started.  Total withdrawals were $1,345,000 and the ending balance was $1,079,000.

Conclusion:  Spend it now or save it for later.  Withdrawals were somewhat volatile with methods 4 and 5.  Methods 1,2 and 3 were specifically designed to let withdrawals increase but not decrease.  As we talked about earlier if the money is discretionary spending then the retiree would select a different option than if it was for fixed expenses.  Thinking they were very different methods I was surprised at how similar the withdrawal patterns were for methods 3, 4 and 5.

Table1: Percent of Required Withdrawal

year

Method 1 Withdraw

Method 2 Withdraw

Method 3 Withdraw

Method 4 Withdraw

Method 5 Withdraw

1968

100%

100%

100%

100%

100%

1972

100%

96%

85%

83%

88%

1977

100%

92%

60%

55%

72%

1982

100%

88%

59%

43%

51%

1987

100%

84%

70%

66%

65%

1992

30%

80%

79%

79%

90%

1997

0%

76%

105%

104%

53%

 

Table 2: Withdrawal

year

Method 1 Withdraw

Method 2 Withdraw

Method 3 Withdraw

Method 4 Withdraw

Method 5 Withdraw

1968

$20,000

$20,000

$20,000

$20,000

$20,026

1969

$20,940

$20,740

$20,719

$20,829

$21,192

1970

$22,238

$21,818

$20,719

$20,225

$19,188

1971

$23,484

$22,822

$20,719

$19,999

$20,023

1972

$24,259

$23,347

$20,719

$20,042

$21,408

1973

$25,083

$23,907

$20,719

$20,687

$23,355

1974

$27,266

$25,748

$20,719

$19,982

$20,307

1975

$30,619

$28,658

$20,719

$18,080

$16,517

1976

$32,732

$30,349

$20,719

$17,799

$20,377

1977

$34,336

$31,532

$20,719

$18,932

$24,759

1978

$36,636

$33,330

$20,719

$18,914

$23,350

1979

$39,934

$35,996

$20,719

$19,193

$24,218

1980

$45,245

$40,423

$20,719

$20,525

$25,204

1981

$50,900

$45,072

$21,737

$22,873

$26,598

1982

$55,431

$48,633

$21,737

$23,641

$28,148

1983

$57,537

$49,994

$25,664

$25,614

$30,049

1984

$59,723

$51,394

$30,445

$29,150

$32,340

1985

$62,053

$52,885

$32,354

$32,051

$35,063

1986

$64,411

$54,366

$39,922

$37,542

$38,275

1987

$65,119

$54,420

$45,377

$42,843

$42,168

1988

$67,984

$56,270

$45,377

$45,898

$46,482

1989

$70,976

$58,183

$49,625

$50,217

$51,508

1990

$74,241

$60,278

$57,930

$56,973

$57,498

1991

$78,769

$63,352

$57,930

$58,963

$64,581

1992

$24,392

$64,683

$63,968

$64,473

$72,986

1993

$0

$65,912

$66,279

$68,560

$82,216

1994

$0

$67,032

$72,230

$73,822

$93,091

1995

$0

$68,172

$72,230

$74,865

$105,800

1996

$0

$69,194

$87,876

$85,503

$120,880

1997

$0

$70,786

$98,019

$96,940

$136,910

Total Withdrawals

$1,114,307

$1,359,295

$1,157,327

$1,145,138

$1,344,518

End Bal

$0

$294,032

$2,898,941

$3,132,475

$1,078,774

 

0 Kudos
Highlighted
Contributor ○○○

Re: Bengen's 4% Distribution Method

It’s raining outside.  I decided I wanted to see the comparison over a period with very good returns. 

The initial investment was $500,000 in Vanguard Wellington for all methods.  Starting year is 1990.  There is one additional Method.  It’s the 4% Rule with Kitces ratchet-up modification.

 Method 1:  Initial withdrawal is investment times 4%.  Subsequent withdrawals are increased by inflation.

Method 2:  Initial withdrawal is investment times 4%.  Subsequent withdrawals are increased by inflation minus 1 percentage point (i.e., if inflation is 3.4% the 2.4% is used). 

Method 3:  Initial withdrawal is investment times 4%.  Subsequent withdrawals is either the beginning of year balance times 4% or the previous withdrawal whichever is larger.

Method 4:  Initial withdrawal is investment times 4%.  Subsequent withdrawals are one-half beginning of year balance times 4% plus one-half of the previous withdrawal adjusted for inflation.

Method 5:  All withdrawals are beginning of year balance times the modified required minimum distribution rate.  The required minimum distribution rate is the MRD rate plus 1.24.

Method 6.  Initial withdrawal is investment times 4%.  Subsequent withdrawals are increased by inflation.  Every third year if the beginning of year balance is 50% greater than the original investment then the withdrawal is increased by 10%.

Results of the test:  Much better than 1968.  None of the withdrawal methods failed. In general it was bigger withdrawals and bigger ending balances.  Low inflation and high returns make a big difference.

Method 1:  Total withdrawals were $904,000 compared to $1,114,000  but purchasing power stayed at 100% for all 30 years.  Withdrawals completely depleted the 1968 investment but the same investment in 1990 had an ending balance of $3,752,000.

Method 2:  If you like ending balances then this method wins during good years but withdrawals were disappointing.  This is the only method that failed to keep up with inflation.  In most of the other methods withdrawals exceeded inflation.  Mid-retirement purchasing power was down 13%.  The retiree needed $29,245 but only got $25,503.  By the end of the retirement it was down 25%.   Because withdrawals didn’t keep up with inflation total withdrawals were only $774,000 but the ending balance was the highest of all methods at $4,105,000.

Method 3:  After 9 years the retiree is starting to get spoiled.  In 1999 he needed $29,245 to maintain his standard of living and he got $43,067.  More than 60% higher than he needed thru the halfway mark.  And, the last half of retirement was even better.  His final year was 103% higher than needed.  At the end of his retirement his withdrawals were more than twice what was required to have maintained his standard of living.  Total withdrawals were $1,482,000.  Ending balance was $2,195,000.

Method 4: One of two methods where a withdrawal was below $20,000 but it quickly made up for it.  At the half way point it caught up with method three by allowing a withdrawal that was 61% more than required.  During the last year its withdrawal was 105% more than required.  Total withdrawals were $1,452 and its ending balance was $2,352,000.

Method  5: The other method where a withdrawal was below $20,000 but this method had the biggest withdrawals.  In 2004 at the halfway point withdrawals were 128% higher than needed.  The retiree needed $29,245 to maintain his beginning standard of living.  He got $66,607.  At the end of retirement his withdrawal was 157% more than required.  Total withdrawals were $2,007,000.  Ending balance was the lowest at $796,000.

Method 6:  There wasn’t a single year in the 1968 retirement where the 50% rule would have been used.  Very different for the 1990 retirement.  The rule kicked in on the second review and retiree got a 10% bonus every time after that.  In 2004 his withdrawal was 46% above requires.  $41,818 compared to $29,245.  At the end of retirement they were 136% bigger.  Total withdrawals were $1,430,000.  Its ending balance was $2,562,000.

Conclusion:  The methods that worked best for the 1968 retirement were the ones that worked worst for the 1990 retirement and vise-versa.  We cannot see the future so we need to pick a method that works for both.

Table 1: Percent of Required Withdrawal

year

Method 1 Withdraw

Method 2 Withdraw

Method 3 Withdraw

Method 4 Withdraw

Method 5 Withdraw

Method 6 Withdraw

1990

100%

100%

100%

100%

100%

100%

1994

100%

96%

108%

104%

118%

100%

1999

100%

92%

166%

155%

208%

121%

2004

100%

87%

161%

161%

228%

146%

2009

100%

83%

182%

162%

206%

161%

2014

100%

79%

184%

181%

287%

195%

2019

100%

75%

203%

205%

257%

236%

 

Table 2: Withdrawals.

year

Method 1 Withdraw

Method 2 Withdraw

Method 3 Withdraw

Method 4 Withdraw

Method 5 Withdraw

Method 6 Withdraw

1990

$20,000

$20,000

$20,000

$20,000

$20,026

$20,000

1991

$21,220

$21,020

$20,000

$19,940

$19,089

$21,220

1992

$21,878

$21,461

$22,084

$21,323

$23,186

$21,878

1993

$22,512

$21,869

$22,882

$22,430

$24,543

$22,512

1994

$23,120

$22,241

$24,937

$24,017

$27,360

$23,120

1995

$23,744

$22,619

$24,937

$24,293

$26,760

$26,119

1996

$24,338

$22,958

$30,339

$27,701

$35,077

$26,772

1997

$25,141

$23,486

$33,840

$31,385

$40,293

$27,655

1998

$25,569

$23,651

$40,034

$36,230

$48,937

$30,938

1999

$25,978

$23,793

$43,067

$40,308

$54,014

$31,433

2000

$26,679

$24,197

$43,168

$42,725

$55,623

$32,282

2001

$27,586

$24,778

$45,751

$45,464

$60,473

$36,717

2002

$28,027

$24,927

$45,761

$46,502

$61,673

$37,305

2003

$28,700

$25,276

$45,761

$44,735

$56,462

$38,200

2004

$29,245

$25,503

$47,176

$46,980

$66,607

$42,818

2005

$30,211

$26,090

$50,348

$50,110

$72,310

$44,231

2006

$31,238

$26,716

$51,630

$52,444

$75,352

$45,735

2007

$32,019

$27,116

$58,472

$56,946

$86,579

$51,566

2008

$33,331

$27,957

$60,831

$60,991

$91,185

$53,681

2009

$33,365

$27,705

$60,831

$53,937

$68,883

$53,734

2010

$34,266

$28,176

$60,831

$54,988

$81,164

$60,704

2011

$34,780

$28,317

$60,831

$56,963

$86,523

$61,614

2012

$35,823

$28,884

$60,831

$58,328

$86,232

$63,463

2013

$36,432

$29,086

$60,831

$60,983

$92,933

$70,996

2014

$36,978

$29,231

$68,067

$66,970

$106,217

$72,060

2015

$37,274

$29,173

$71,761

$71,841

$110,124

$72,637

2016

$37,535

$29,085

$71,761

$72,845

$103,694

$80,460

2017

$38,323

$29,405

$73,335

$76,281

$107,835

$82,150

2018

$39,128

$29,729

$80,765

$82,039

$115,551

$83,875

2019

$39,872

$29,996

$80,765

$81,838

$102,495

$94,015

Total Withdrawals

$904,312

$774,447

$1,481,627

$1,451,535

$2,007,199

$1,429,888

End Bal

$3,751,804

$4,105,274

$2,194,522

$2,352,350

$795,850

$2,562,074

 

Highlighted
Explorer ○

Re: Bengen's 4% Distribution Method

Thanks for providing all this. I'm going to study it to see what insights can be mined for my own situation. For those who do have some discretionary income, one reason not to follow any rule as more than a guideline is that spending needs and desires vary a lot from year to year. (In my case, it varies a lot with whether I do or don't have a girlfriend.) And we usually prefer not to take more income than we want to spend, so as to defer taxes or avoid a higher tax bracket.

Rule 6 has a  lot to recommend it, given that we never know what lies ahead. In bad times, it assures that we won't be skimping during the years when we're more likely to be alive. But if things go well, it takes good advantage of that pretty soon, letting us withdraw as much as we're likely to want to spend.    

0 Kudos
Highlighted
Participant ○

Re: Bengen's 4% Distribution Method


@Mustang wrote:

I thought I’d do a comparison of the various withdrawal methods I've tested.

The initial investment was $500,000 in Vanguard Wellington for all methods.  1968 Wellington is the worst case scenario.  With a 75/25 asset allocation Bengen had two of the 55 rolling 30-year retirement periods fall short by one year.  1968 Wellington fell short by over 5 years. 

Taking an initial withdrawal of 4%, with later withdrawals adjusted for inflation, lasted longer when starting for 1968 than it would have for some other starting years.  The number of years depend on details of how the calculation is done.  My calculation adjust for inflation using the average CPI-U over 12 month periods, rather than the change for a month, such as December, from a year ago.  I spent during each month of the year and not only during one month.

What has Bengen written, and it what article did he write it, that a 75/25 asset allocation sometimes lasted less than 30 years?  You made a similar statement in another post. In a reply to that post I provided a link to an article and a quote from an article where Bengen stated that 75/25 lasted one or two years shorter than 50/50 but the shortest that 75/25 lasted was 32 years.

0 Kudos
Highlighted
Contributor ○○○

Re: Bengen's 4% Distribution Method

@PatMorgan 

You are correct.  Thanks for reminding me.  I did remember it incorrectly.  I remembered the one year short part and the part that it impacted two retirement periods but not the rest. 

I have a PDF copy of a section of the October 1994 issue of The Journal of Financial Planning.  I was busy at the time and didn’t go back and read it.  On page 175, Bengen writes,” The only penalties occur in portfolio year 1966, which is shortened by one year, from 33 to 32 years, and in 1969, which is shortened from 36 years to 34. All the other scenario years have equal or greater longevity.”

I also remembered that one of Kitces’ tables showed 1960 going to zero.  I went back and looked that up too.  He was using 4.5% not 4%.

Now that I’ve looked it up maybe I’ll remember it.  Hopefully, but my memory just isn’t as good as it used to be.

0 Kudos
Highlighted
Contributor ○○○

Re: Bengen's 4% Distribution Method

After comparing 6 methods here is what I dislike the most about each.

#1 Initial 4%, subsequent withdrawals increased by inflation.    1968 retirement: Ran out of money with 6 years to go.  1990 retirement:  Left a lot on the table, $3.75 million, that could have been used.

#2 Initial 4%, subsequent withdrawals increased by inflation – 1% point.  1990 retirement: Unnecessarily cut withdrawals and left even more on the table, $4.1 million, that could have been used.  Total withdrawals were the least of any method and were $130,000 less than #1 method.

#3, 4, and 5.  1968 retirement: Really bad mid-retirement performance.  Started catching up during last half.

#3  The greater of 4% times EOY balance or last withdrawal. 1968 retirement. No increase for 12 years (1969-1980). In 1982 withdrawals were only 39% of what was needed. They lagged behind until the very last year   

#4 One-half EOY times 4% plus one-half of #1 method.   1968 retirement:  Worse than #3 method.  No increase for 12 years with volatility.  Withdrawals were lower than the initial withdrawal 7 times with the lowest of $17,800 in 1976.  They didn’t catch up until the last year.  1990 retirement:  After 1992 withdrawals took off but they were up and down.

#5 Modified RMD (1.24): 1968 retirement:  Withdrawals fell behind what was needed to maintain a standard of living in 1970 and didn’t catch up until 1993.  Withdrawals were less than half what was needed until 1985.  Ups and downs with 2 years below the initial amount with the lowest of $16,517 in 1975.  1990 retirement:  Very good withdrawals but ending balance was the smallest of all the methods.

#6 Initial 4%, subsequent withdrawals increased by inflation, ever third year a 10% bonus if ending balance is 50% bigger than initial investment. 1968 retirement: Never used.  The ending balance never reached 50%.  1990 retirement:  Received a bonus every three years after the sixth year.  Withdrawals steadily increased but not as fast as #3 method.

Total withdrawals. 1968 retirement: #1, #3 and #4 approximately the same. #2 and #5 approximately the same but roughly $200K higher.  1990 retirement:  #3, #4 and #6 approximately the same.  #5 roughly $550K higher.  #1 roughly $550K lower.  #2 was roughly $700K lower.

P.S.  I hope it doesn't rain tomorrow.  I need to mow the lawn.

0 Kudos
Highlighted
Explorer ○

Re: Bengen's 4% Distribution Method

Mustang, that's a nice summary of the performances for a great period and a hideous period. To add to the picture, I'll report Pfau's results for a 50th percentile period, taken from the 2018 table that I've cited. (It includes only four of the methods you've tested, and is based on a 50/50 allocation.)

BASIC BENGEN
Real spending in 10/20/30 years, 50th percentile: $4,000/$4,000/$4000
Remaining wealth after 30 years, 50th percentile $124,740

ENDOWMENT FORMULA – WEIGHTED AVERAGE
Real spending in 10/20/30 years, 50th percentile: $4,030/$4,160/$4,280
Remaining wealth after 30 years, 50th percentile: $122,410

BASIC KITCES
Real spending in 10/20/30 years, 50th percentile: $4,000/$5,320/$7,090
Remaining wealth after 30 years, 50th percentile: $79,500

MODIFIED RMD RULE
Real spending in 10/20/30 years, 50th percentile: $5,080/$6,650/$6,220
Remaining wealth after 30 years, 50th percentile: $41,570

0 Kudos
Highlighted
Contributor ○○○

Re: Bengen's 4% Distribution Method


@ElLobo wrote:

Now, my final question for this thread.  You and Bengen have been withdrawing at a real, inflation adjusted 4%, in 1968 dollars.  Ten years into your retirement, you're still withdrawing at that real 4% rate, in 1968 dollars, but your current rate of withdrawal, in 1978/79 dollars, is up to 10%.  So, what does Bengen and all the retirement withdrawal studies say about retirement portfolio survivability whenever a real, inflation adjusted 10% is taken out, starting in 1979?  Ifn it's less than 10 years, there's a good chance you'll spend down your portfolio, and Bengen did, before those 20 years come about.


It has been years since I have read Bengen and the Trinity studies.  My memory isn't as good as it used to be.  So I am going back and re-reading them.  I'll let Bengen answer you himself.  This is from his 1994 article published in the Journal of Financial Planning (pgs 177-178).

"The "black hole" group is in a very uncomfortable situation. As an example, the client who retired in 1929 with $500,000 in a retirement fund saw that fund dwindle to less than $200,000 by the end of 1932. Although his withdrawals have also declined from $20,000 in 1929 to $15,300 in 1932, owing to deflation, those withdrawals now equal about 7.6 percent of his portfolio, whereas he began by withdrawing only 4 percent. In this situation, with stocks having performed so dismally so early in retirement, it may be tempting to switch all investments to bonds in order to salvage what is left of the original capital.’

"If we eliminate stocks completely, investing only in intermediate-term bonds, his money will be exhausted in 1946, after only 17 more years. If we invest in 25-percent stocks, the money will last till 1950; 50 percent in stocks, 1957. But if we had left the allocation at 75-percent stocks, the client would still have $1.7 million in 1992...'

“This same analysis can be repeated for all the other "black hole" clients who were unfortunate enough to begin their retirements in 1937, 1946, 1969, 1973, 1974---the years of major and minor events. This is a testament to the enormous recovery power of the stock market---and the need to avoid emotion when investing.’

“However, the client has another option to improve the situation for the long term, and that is to reduce---even if temporarily---his level of withdrawals. If the client can manage it without too much pain, this may be the best solution, as it does not depend on the fickle performance of markets, but on factors the client controls completely: his spending.’

“As an example, let us return to the 1929 retiree. At the end of 1930, as he is about to make his second annual withdrawal, the market has already declined about 30 percent from the end of 1928, and there looks like more trouble ahead. If he reduces his 1930 withdrawal by only 5 percent, and continues to withdraw at this reduced level during retirement, by 1949 he will have 20 percent more wealth than otherwise, which can be passed on to his heirs. After 30 years, the wealth is 25 percent greater…’

“Thus the "black hole" client has at least two alternatives to improve his portfolio longevity, with an infinite number of permutations of the two possible. The one alternative he cannot afford, and which we as advisors must work hard to dissuade him from doing, is to pull back from the stock market and retreat to bonds.’

I added the bold print.  Pulling back from stock would have caused his 1929 retiree to exhaust his portfolio in 17 years.  Staying in stocks made it last over 63 years.  We also know that cutting back just a little on spending (which is one of Bengen's two alternatives) saved the 1968 retirement period and we know that the market recovered.  The last 8 years of the 1968 retirement period was the first 8 years of the 1990 retirement period.

0 Kudos
Highlighted
Contributor ○○○

Re: Bengen's 4% Distribution Method


@PatMorgan wrote:

Taking an initial withdrawal of 4%, with later withdrawals adjusted for inflation, lasted longer when starting for 1968 than it would have for some other starting years.  The number of years depend on details of how the calculation is done.  My calculation adjust for inflation using the average CPI-U over 12 month periods, rather than the change for a month, such as December, from a year ago.  I spent during each month of the year and not only during one month.


I used the inflation Year to Year inflation numbers from the link below.  How is that wrong?  Why would average be better than year to year?

https://www.thebalance.com/u-s-inflation-rate-history-by-year-and-forecast-3306093

0 Kudos
Highlighted
Frequent Contributor

Re: Bengen's 4% Distribution Method

I wasn't looking for Bengen's answer to my question.  I knew what the answer was going to be.  I was looking for YOUR thoughts!  After all, you've said all along that you weren't willing to take a cut in your withdrawal from one year to the next in either real, inflation adjusted OR nominal terms! 8-))

“However, the client has another option to improve the situation for the long term, and that is to reduce---even if temporarily---his level of withdrawals. If the client can manage it without too much pain, this may be the best solution, as it does not depend on the fickle performance of markets, but on factors the client controls completely: his spending."

Ifn I'm not mistaken, I think I said this to you on Friday:

"What you don't want to recognize is the YOU, and you alone, determine your 'rate of withdrawal', your 'method'. That's not in dispute. Nevertheless, whether or not that method survives retirement ISN'T up to you, it's completely determined by everyone ELSE, aka the 'markets', going forward, where it counts!"

 

 

ElLobo, de la casa de la toro caca grande
0 Kudos
Highlighted
Contributor ○○○

Re: Bengen's 4% Distribution Method

@ElLobo   I don't think a loss of purchasing power to someone who needs it is ideal. But sometimes that has to happen.

You already knew that I was willing to accept cuts if they appear necessary.  I preferred a modified Bengen 4% rule where inflation -1% point is used.  That method eventually ended up cutting purchasing power around 20% but it was a gradual cut over a 25 year period.  Since the future is unknown a lot can happen in 25 years.

That is similar to the 5% cut per year that Bengen discussed in his paper.  He picked a flat rate but there are many financial experts that advise not taking inflation increases when the market is down so this isn't anything new.

What isn't acceptable is a method where the cuts are 40-60% during the first part of retirement if the money is required for core expenses.  If they were to cover discretionary expenses then the cuts would be fine but not for core living expenses. I have said that many times.  So how much of a cut I find acceptable depends completely on what the money is used for.

0 Kudos
Highlighted
Frequent Contributor

Re: Bengen's 4% Distribution Method


@Mustang wrote:

@ElLobo   I don't think a loss of purchasing power to someone who needs it is ideal. But sometimes that has to happen.

You already knew that I was willing to accept cuts if they appear necessary.  I preferred a modified Bengen 4% rule where inflation -1% point is used.  That method eventually ended up cutting purchasing power around 20% but it was a gradual cut over a 25 year period.  Since the future is unknown a lot can happen in 25 years.

That is similar to the 5% cut per year that Bengen discussed in his paper.  There are many financial experts that advise not taking inflation increases when the market is down so this isn't anything new.

What isn't acceptable is a method where the cuts are 40-60% during the first several years if the money is required for core expenses.  If they were to cover discretionary expenses then the cuts would be fine but not for core living expenses. I have said that many times.  So how much of a cut I find acceptable depends completely on what the money is used for.


As Bengen stated, there are two alternatives to spending down a portfolio, cut back on the withdrawals or move (retreat!) to bonds.  And as he also stated, there is an infinite number of permutations of those two possible.  I apologize to you for thinking you were interested in exploring these alternatives in a logical manner, rather than simply explaining to us what you have chosen for your heirs.

ElLobo, de la casa de la toro caca grande
0 Kudos
Highlighted
Follower ○○

Re: Bengen's 4% Distribution Method

Each individual or person has to adapt the withdrawal rate to their own individual circumstances.  Four  percent is only a starting point not a hard rule.  Thanks for your thoughts on this topic.


Sent from Yahoo Mail for iPhone
0 Kudos
Highlighted
Contributor ○○○

Re: Bengen's 4% Distribution Method


@ElLobo wrote:


As Bengen stated, there are two alternatives to spending down a portfolio, cut back on the withdrawals or move (retreat!) to bonds.  And as he also stated, there is an infinite number of permutations of those two possible.  I apologize to you for thinking you were interested in exploring these alternatives in a logical manner, rather than simply explaining to us what you have chosen for your heirs.


Are you OK?  Bengen didn't say anything like that.  He actually said the opposite about bonds.  This is exactly what I posted a little earlier tonight.  It is a direct quote from his 1994 paper.

“Thus the "black hole" client has at least two alternatives to improve his portfolio longevity, with an infinite number of permutations of the two possible. The one alternative he cannot afford, and which we as advisors must work hard to dissuade him from doing, is to pull back from the stock market and retreat to bonds."

@RickW2  Yes. It has to be adapted for the length of the retirement period and mid-course corrections are necessary.  The vast majority of this discussion has been on alternatives and modifications to Bengen's withdrawal method.  One I really like is Kitce's ratchet-up method.

0 Kudos
Highlighted
Participant ○

Re: Bengen's 4% Distribution Method

Mustang,

Earlier, you suggested and then discarded a modified Bengen where you start following Bengen rule but then review it every five years using a new SWR for a shorter time horizon.   What did you find that you did not like about that?   Just curious.

 

0 Kudos
Highlighted
Frequent Contributor

Re: Bengen's 4% Distribution Method


@Mustang wrote:

@ElLobo wrote:


As Bengen stated, there are two alternatives to spending down a portfolio, cut back on the withdrawals or move (retreat!) to bonds.  And as he also stated, there is an infinite number of permutations of those two possible.  I apologize to you for thinking you were interested in exploring these alternatives in a logical manner, rather than simply explaining to us what you have chosen for your heirs.


Are you OK?  Bengen didn't say anything like that.  He actually said the opposite about bonds.  This is exactly what I posted a little earlier tonight.  It is a direct quote from his 1994 paper.

“Thus the "black hole" client has at least two alternatives to improve his portfolio longevity, with an infinite number of permutations of the two possible. The one alternative he cannot afford, and which we as advisors must work hard to dissuade him from doing, is to pull back from the stock market and retreat to bonds."

@RickW2  Yes. It has to be adapted for the length of the retirement period and mid-course corrections are necessary.  The vast majority of this discussion has been on alternatives and modifications to Bengen's withdrawal method.  One I really like is Kitce's ratchet-up method.


I'm OK, but on vacation now.  Is English a second language for you?  How is what I wrote 'opposite' from what Bengen wrote about bonds?  I simply paraphrased your Bengen quote, which I was very familiar with from my old days of looking at this stuff!  Of course you must realize that the 'black hole' is his early reference to 'Sequence Of Return' risks while portfolio 'longevity' is, in our terminology, portfolio 'survivability!  And 'spending down a portfolio' is colloquial for a lack of longevity!  8-))

"As Bengen stated, there are two alternatives to spending down a portfolio, cut back on the withdrawals or move (retreat!) to bonds.  And as he also stated, there is an infinite number of permutations of those two possible."

“Thus the "black hole" client has at least two alternatives to improve his portfolio longevity, with an infinite number of permutations of the two possible. The one alternative he cannot afford, and which we as advisors must work hard to dissuade him from doing, is to pull back from the stock market and retreat to bonds."

 

ElLobo, de la casa de la toro caca grande
0 Kudos
Highlighted
Contributor ○○○

Re: Bengen's 4% Distribution Method


@Saratoga wrote:

Mustang,

Earlier, you suggested and then discarded a modified Bengen where you start following Bengen rule but then review it every five years using a new SWR for a shorter time horizon.   What did you find that you did not like about that?   Just curious.

 


That was a theory I hadn’t tested.  I’m just an xx-GI trying to figure this stuff out.  I mess around with testing theories because I want to see what is going on in the background.  I get a hint of that by reading an expert’s explanation or doing a little testing of my own.  Just looking at summary numbers in a table really doesn’t tell me much.  Besides, it gives me something to do other than watch the news.

I dropped the theory because someone posted a link to Kitces’ ratchet-up method.  He thoroughly tested it using historical data. Much more than I could ever do.  I got interested in it and didn’t go back to the restart theory.

I view withdrawals as something needed to pay bills. That is a completely different view from someone who thinks of them as discretionary spending.  Since I have five spreadsheets, two funds with three start dates, already set up so I decided to test the restart theory.

Here are the final outcomes using Bengen’s initial 4% withdrawal method and never reviewing or changing anything.

           Fund                      Starting Year                  Final Withdrawal              Final Balance 

          Wellington                1968                                    $93,324                     -$886,134      

          Wellington                1971                                    $84,324                    $1,443,886  

          Wellesley                  1971                                    $84,324                    $2,993,190   

          Wellington                1990                                    $39,872                     $3,754,304

          Wellesley                  1990                                   $39,874                     $2,606,074

A couple of quick observations:  Withdrawals are unaffected by returns.  They are exactly the same regardless of fund.  They are effected by inflation.  They were far less during the low inflation years than the high inflation years but all withdrawals have the purchasing power-- $20,000 as of the start date.

The starting investment was $500,000 in all tests.  Wellesley outperformed Wellington starting in 1971. Wellington outperformed Wellesley starting in 1990. And there are a couple of glaring problem. Wellington 1968 failed.  A fund cannot go negative.  And looking at the other final balances, money that could be used isn’t.

What does restarting every five years do?  Here are the rates.  1968 - 4.0%, 1972 - 4.5%, 1977 - 5.0%, 1982 – 6.0%, 1987 -7.0%, and 1992 – 7.5%.  Here are the final results.

              Fund                      Starting Year                   Final Withdrawal              Final Balance 

              Wellington                1968                             $93,093                            $1,565,093

              Wellington                1971                            $95,772                             $4.222.988  

              Wellesley                  1971                           $107,370                            $4,161,240  

              Wellington                1990                             $80,307                             $2,436,218

              Wellesley                  1990                               $64,484                             $1,173,129

Final Withdrawals are higher and final balances are higher.  How can that be?  Returns didn’t change.  Like some of the other methods, this method robs from the middle years.  I’ll use the 1968 retirement as an example.  In this discussion the term “needed money” means inflation adjusted.

The first 9 years weren’t that bad.  Withdrawals in the 9th year were $29,107 to the $32,732 that was needed.  Then the sequence of return failure hit.  In 1973-1974 losses were 11.8% and 17.7% while inflation hit 8.7% and 12.5%.  The 2nd restart hit hard.  The retiree needed $34,396 and got $21.735.  Everything was even worse leading up to the third restart.  The good returns of 1979-1980 (13.5% and 22.6%) were completely eaten up by inflation of 9%, 13.3%, 12.5% and 8.9% (1978-1981).  By 1982 $55,431 was required to maintain the standard of living.  The withdrawal was $27,321. 

After that inflation went down and returns went up.  For example, returns were 28.5% in 1985 and inflation was only 3.8%.  The next restart equalized things.  Needed income was $65,119 and the withdrawal was $64,192.  Same with the last restart in 1992, needed income was $81,211 and the withdrawal was $81,009. 

If the retiree has a lot of discretionary expenses that can be eliminated then the system would work.  I he doesn’t it won’t.  Because the income is so volatile.  I prefer the much better tested Kitces’ ratchet-up method.  It and the standard Bengen 4% method didn’t work in 1968 but that is an extremely rare event that didn’t happen using Bengen’s, the Trinity authors’, Pfau’s or Kitces’ test data.  For it I prefer using Bengen’s method using an annual adjustment of one percentage point less than inflation.

0 Kudos
Highlighted
Frequent Contributor

Re: Bengen's 4% Distribution Method

"A fund cannot go negative."

True 'dat.

But, since the value of your portfolio is the number of shares you own times the current NAV, the number of shares you own CAN go to zero, in fact, DOES go to zero whenever you sell your last share.  That's called 'spending down your portfolio'.  Specifically, that's what happened to Bengen, in 1992, whenever $48,784 was taken out.  That's what was left in that portfolio at that point in time.  Your 'withdrawal' should have been $157,538 (inflation adjusted $40,000 in 1968 dollars), plus a bit for 1992 inflation!  8-))

ElLobo, de la casa de la toro caca grande
0 Kudos
Highlighted
Contributor ○○○

Re: Bengen's 4% Distribution Method


@ElLobo wrote:

"As Bengen stated, there are two alternatives to spending down a portfolio, cut back on the withdrawals or move (retreat!) to bonds.  And as he also stated, there is an infinite number of permutations of those two possible."

“Thus the "black hole" client has at least two alternatives to improve his portfolio longevity, with an infinite number of permutations of the two possible. The one alternative he cannot afford, and which we as advisors must work hard to dissuade him from doing, is to pull back from the stock market and retreat to bonds."

 


Communication is hard and misunderstandings do arise but those two paragraphs do not say the same thing.  Bengen was talking about improving the portfolio's longevity.  The first paragraph says he needs "to cut back on the withdrawals or move into (retreat!) bonds."  It is missing the key phrase, "dissuade him from" and without it the entire meaning is changed.

With your insistence on using yield instead of return I assumed you meant what you typed.

Do you always resort to insults and name calling when you are wrong?

Highlighted
Participant ○

Re: Bengen's 4% Distribution Method

Mustang,

Thanks.   Just to be sure, would you state once more your method of 5 year review you used in the test?

0 Kudos
Announcements