UPDATE - Assumed Portfolio Return in Retirement

A place to talk about speculative investing ideas for the optional Variable Portfolio

Moderator: Global Moderator

EdwardjK
Executive Member
Executive Member
Posts: 151
Joined: Mon Oct 11, 2010 8:16 pm

UPDATE - Assumed Portfolio Return in Retirement

Post by EdwardjK » Mon May 28, 2018 10:02 pm

Hi. I've built a spreadsheet that takes into account my starting portfolio value, estimated spending in retirement, inflation, and Social Security income. Using Solver, the spreadsheet calculates the average annual return needed so that I exhaust my portfolio at age 100. Not that my wife and I will last that long, but it's good to plan that long.

So I am asking what others are using as their assumed portfolio return in retirement to compare to what my spreadsheet says I need. 4% ? 5% ? Something else?

Your feedback is appreciated.

*** AUTHOR UPDATE: May 31, 2018 ***

Based upon everyone's feedback, I created my own "Monte Carlo" model to test my retirement portfolio.

I started by compiling index returns using the Callan Periodic Table of Investment Returns from 1980 to 2017. From these returns, I calculated annual returns using a mix resembling my portfolio (40% S&P 500, 10% small cap, 10% international, 10% REIT and 30% cash).

Then I randomly sorted the 37 years of returns into 255 possible outcomes and added them to my spreadsheet. Again, the spreadsheet starts with my current portfolio value, deducts annual expenses, and adds portfolio returns; accounts Social Security and includes the requisite inflation impact on expenses. I also "stressed" the model by adding an additional 20% to my annual expenses.

The results show that my portfolio will be exhausted by age 95 in 14 out of 255 cases, or in 5.5% of the cases. The worst case shows exhaustion at age 88. IT IS IMPORTANT TO NOTE THAT, IN THIS CASE, THERE WERE FOUR (4) LOSS YEARS IN THE FIRST TEN (10) YEARS AND THE WORST YEAR WAS YEAR 1 (The worst year returns were actually from 2008).

So what does this tell me...

1. Although there are no guarantees, I feel confident that I will maintain my lifestyle and not run out of money before death
2. If, in the first 10 years of retirement, I experience multiple years of losses, then I may need to cut back my lifestyle and expenses
3. To the extent I am healthy, then I will not hesitate to go to dinner, travel and otherwise enjoy my life in retirement
4. It was a smart move to purchase a 10 annual payment long term care policy with a 5% annual inflation rider. Only 4 payments left.

So that's my story. More feedback welcome and appreciated.
Last edited by EdwardjK on Thu May 31, 2018 5:59 pm, edited 4 times in total.
User avatar
mathjak107
Executive Member
Executive Member
Posts: 4456
Joined: Fri Jun 19, 2015 2:54 am
Location: bayside queens ny
Contact:

Re: Assumed Portfolio Return in Retirement

Post by mathjak107 » Tue May 29, 2018 2:21 am

EdwardjK wrote:
Mon May 28, 2018 10:02 pm
Hi. I've built a spreadsheet that takes into account my starting portfolio value, estimated spending in retirement, inflation, and Social Security income. Using Solver, the spreadsheet calculates the average annual return needed so that I exhaust my portfolio at age 100. Not that my wife and I will last that long, but it's good to plan that long.

So I am asking what others are using as their assumed portfolio return in retirement to compare to what my spreadsheet says I need. 4% ? 5% ? Something else?

Your feedback is appreciated.
average returns mean little when spending down . it is all about the sequence of whatever markets ,rates and inflation do . the same excact average returns can see a 15 year difference in how long the money will last just based on the order those 3 come in at . that 15 years difference spans the best outcome to the worst outcome all with the same average return .

retirement planning is all about basing your draw on the worst outcomes we have had to date . the likes of those who retired in 1907,1929,1937,165/1966 .

about 4% inflation adjusted made it through most of the time frames to date with at least 40% equities . so we are really not interested in any average's when spending down , it does nothing for us . averages only means something when we are accumulating money and not decumulating it .
Last edited by mathjak107 on Tue May 29, 2018 3:20 am, edited 1 time in total.
User avatar
mathjak107
Executive Member
Executive Member
Posts: 4456
Joined: Fri Jun 19, 2015 2:54 am
Location: bayside queens ny
Contact:

Re: Assumed Portfolio Return in Retirement

Post by mathjak107 » Tue May 29, 2018 2:27 am

how far off can these simple how long will my money last calculators be ? these are the ones that are actually reverse amortization calculators which never have you spending down in a down year. every year is subject to the same positive average return .


the 17-year period (1987-2003), the S&P 500's average return was 13.47%. It doesn't make any difference if we look at the returns from 1987 to 2003 or from 2003 to 1987.

But when taking withdrawals, the sequence of returns makes all the difference. The same initial capital, the same withdrawal amount, the same returns - but a different sequence produces dramatically different results.

if we leave the rate of inflation growth with the principal and spend everything else .

For a $100,000 portfolio adjusted for inflation over those 17 years, the difference is a remaining balance of $76,629 to a deficit of $187,606. depending on the order those gains and losses came in .

never ever use a calculator that simply asks you to fill in a projected average return .
User avatar
mathjak107
Executive Member
Executive Member
Posts: 4456
Joined: Fri Jun 19, 2015 2:54 am
Location: bayside queens ny
Contact:

Re: Assumed Portfolio Return in Retirement

Post by mathjak107 » Tue May 29, 2018 3:25 am

DR MOSHE MILEVSKY wrote a very interesting paper which shook the financial planning world , RETIREMENT RUIN AND THE SEQUENCE OF RETURNS.

he came up with a draw rate that would drain all scenarios so he could see how many years difference there was with the same average return only in a different order .

he took an example of a constant 7% return year after year based on a 7% average return over 30 years and drew 7% or 950 a month out starting at age 65 . the money was exhausted by age 86. this is typical of what uninformed folks do in an excell spread sheet or reverse amortization calculator when they have to enter a growth rate so they use an average.

next he took the same 7% average return and made it happen in different orders.

he made year 1 up 7% ,year 2 minus -13% and year 3 up 27% and repeated that pattern . the same 7% average return went broke at 83.


again , same 7% average return ,making the first year up 7% , next year up 27% and 3rd year minus -13%. you went broke at age 90. your money lasted 7 years longer than the example above with the same 7% average return.


next he did first year minus -13% , 2nd year up 7% and 3rd year up 27% . you were broke by 81.

that is also the same 7% average return

lastly he made 1st year up 27% ,2nd up 7% and 3rd year down 13% , same 7% average return and you lasted until 95.

that is almost 10 years longer than just figuring a constant 7% year after year ,.

the variation on the same 7% average return in how long your money will last is largely controlled by the order of those gains and losses.

in this case the same 7% average return most folks just throw in an excell spreadsheet and spend down didn't last until 86.5 like the spread sheet said. they went broke based on the order of that 7% average anywhere from 81 to 95.

as you see being down early on can drastically reduce your failure age by a lot even though the average return over time is identical. it isn't just the sequence of returns that has this effect ,it is the sequence of inflation too. now you have the two working against each other as well altering outcomes even more.
User avatar
ochotona
Executive Member
Executive Member
Posts: 3353
Joined: Fri Jan 30, 2015 5:54 am

Re: Assumed Portfolio Return in Retirement

Post by ochotona » Tue May 29, 2018 4:49 am

Edward, mathjak is right, you cannot plan a retirement using a simple spreadsheet. You must use Monte Carlo simulation. Fortunately this is available for free in different places. A good planning program such as can be found at schwab.com will give you Monte Carlo simulation, also portfolio visualizer. Com offers a free Monte Carlo simulator.
User avatar
mathjak107
Executive Member
Executive Member
Posts: 4456
Joined: Fri Jun 19, 2015 2:54 am
Location: bayside queens ny
Contact:

Re: Assumed Portfolio Return in Retirement

Post by mathjak107 » Tue May 29, 2018 4:59 am

actually monte carlo is one way , actual historical is another . a very good one is firecalc and that is historical data .

firecalc's default is actual historical . but they do offer a monte carlo simulator as a click on option. the fidelity planner uses monte carlo . monte carlo attempts to find even worse case scenarios but pretty much the two methods come out very close in results since the 1965/1966 group was about as bad as you can get .

i like fidelity's a bit better , it is stricter in it's defaults . such as unless you manually over ride settings firecalc assumes the same inflation on all items

fidelity assumes 5.50% on healthcare and long term care costs . they did use 7.50% on healthcare but it did come down a bit .

retirement planning is like building a house in a hurricane area . you can build your home to standards that will withstand the worst storm your area ever had . it does not mean a stronger storm can't destroy it , but it does mean you at least are at the gate able to withstand the worst of the worst .

i would not build a home by averaging out the wind speed from the least to the greatest . that would not stand up to anything worse than average .

so we plan around the worst , then hope for the best and take raises along the way as things actually play out .

any good retirement planner is always based on the worst outcomes and what it would have taken retirees to survive most of them . generally a 90% passing grade is good enough .

when we add in statistical life expectancy to the success rate the fact is statistically most of us won't last 30 years in retirement based on retiring in our 60's . so that ups the success rate from 90% to pretty close to 97% or so since in practice odds are our money will not need to be based on 30 years but good planning says we plan for the worst cases anyway .
User avatar
ochotona
Executive Member
Executive Member
Posts: 3353
Joined: Fri Jan 30, 2015 5:54 am

Re: Assumed Portfolio Return in Retirement

Post by ochotona » Tue May 29, 2018 6:53 am

Portfolio charts.com by Tyler will give you historical retirement feasibility studies. Very informative graphics.
User avatar
sophie
Executive Member
Executive Member
Posts: 1959
Joined: Mon Apr 23, 2012 7:15 pm

Re: Assumed Portfolio Return in Retirement

Post by sophie » Tue May 29, 2018 7:16 am

I've been doing my forecasting with both a spreadsheet and cfiresim (online tool). The results are very close, which is reassuring.

Don't forget to include taxes and a cushion for unexpected expenses in your retirement budget.
User avatar
mathjak107
Executive Member
Executive Member
Posts: 4456
Joined: Fri Jun 19, 2015 2:54 am
Location: bayside queens ny
Contact:

Re: Assumed Portfolio Return in Retirement

Post by mathjak107 » Tue May 29, 2018 7:53 am

ochotona wrote:
Tue May 29, 2018 6:53 am
Portfolio charts.com by Tyler will give you historical retirement feasibility studies. Very informative graphics.
if i remember , tylers charts miss all the worst case scenario data by starting in the 1970's . it is the first 15 years of every worst case scenario case we had that made the entire 30 year retirement a failure .
User avatar
mathjak107
Executive Member
Executive Member
Posts: 4456
Joined: Fri Jun 19, 2015 2:54 am
Location: bayside queens ny
Contact:

Re: Assumed Portfolio Return in Retirement

Post by mathjak107 » Tue May 29, 2018 7:56 am

here are the worst cases ever for retirees . every one of them was killed in the first 15 years . by the end of a 30 year time frame the returns looked quite average on all of them . but with the excessive spending required the first 15 years even the greatest bull market in history could not save them


the 30 year results are pretty average :

1907 stocks returned 7.77% -- bonds 4.250-- rebalanced portfolio 7.02- - inflation 1.64--

1929 stocks 8.19% - - bonds 1.74%-- rebalanced portfolio 6.28-- inflation 1.69--

1937 stocks 10.12 - - bonds 2.13 - rebalanced portfolio -- 7.24 inflation-- 2.82

1966 stocks 10.23 - -bonds 7.85 -- rebalanced portfolio 9.56- - inflation 5.38

for comparison the 140 year average's were:

stocks 8.39--bonds 2.85%--rebalanced portfolio 6.17% inflation 2.23%



lets look at the first 15 years in those time frames determined to be the worst we ever had.

1907--- stocks minus 1.47%---- bonds minus .39%-- rebalanced minus .70% ---inflation 1.64%

1929---stocks 1.07%---bonds 1.79%---rebalanced 2.29%--inflation 1.69%

1937---stocks -- 3.45%---bonds minus 3.07%-- rebalanced 1.23%--inflation 2.82%

1966-stocks minus .13%--bonds 1.08%--rebalanced .64%-- inflation 5.38%
User avatar
Cortopassi
Executive Member
Executive Member
Posts: 3338
Joined: Mon Feb 24, 2014 2:28 pm
Location: https://www.jwst.nasa.gov/content/webbL ... sWebb.html

Re: Assumed Portfolio Return in Retirement

Post by Cortopassi » Tue May 29, 2018 8:28 am

Edward,

Not the simple answer you were hoping for, huh? ;)

If you want a reasonably simple assessment, and a way to track your accounts, I have been using personalcapital.com for about 9 months. If has a very flexible retirement simulator, which uses Monte Carlo analysis and shows you a median expectation and 10% expectation with your portfolio allocation (pretty much worst case, hopefully). You can put in SS numbers, expected life, expected income/expense events, etc.

Part of their disclaimer:

"This Retirement Planner tool is intended to provide a basic sense of how your portfolio may grow and support your stated spending goals. It also gives you an easy way to see how adjustments to the most important variables (saving rate, spending rate and retirement age) can impact your outcome. All results are presented in today’s dollars, inflation adjusted at 4.0% unless changed by the user. Projections utilize Monte Carlo simulation, with the median scenario and 10% worst-case scenarios highlighted. The portfolio survival percentage represents the percent of the simulations that ended age 92, or the selected life expectancy age, with a balance above $0. Growth and volatility assumptions are based on historical asset class returns and correlations using the users aggregated asset allocation or a moderate, well-diversified asset allocation with 8.4% annual return and 11.4% standard deviation of return, if not available."
User avatar
Cortopassi
Executive Member
Executive Member
Posts: 3338
Joined: Mon Feb 24, 2014 2:28 pm
Location: https://www.jwst.nasa.gov/content/webbL ... sWebb.html

Re: Assumed Portfolio Return in Retirement

Post by Cortopassi » Tue May 29, 2018 8:30 am

And not that I care, but if anyone wants to actually sign up for personal capital, which is free, you get $20 and I get $20 if you use this link.

http://share.personalcapital.com/x/gUwtX9

Seriously, I am not doing this to make money. I really like their tools. They do try, for a while, to spam you to use their advisor services, but otherwise I have switched to using them vs. Quicken.
User avatar
Tyler
Executive Member
Executive Member
Posts: 2066
Joined: Sat Nov 12, 2011 3:23 pm
Contact:

Re: Assumed Portfolio Return in Retirement

Post by Tyler » Tue May 29, 2018 9:00 am

mathjak107 wrote:
Tue May 29, 2018 7:53 am
ochotona wrote:
Tue May 29, 2018 6:53 am
Portfolio charts.com by Tyler will give you historical retirement feasibility studies. Very informative graphics.
if i remember , tylers charts miss all the worst case scenario data by starting in the 1970's . it is the first 15 years of every worst case scenario case we had that made the entire 30 year retirement a failure .
The effect of the sampling timeframe on withdrawal rates is covered in detail here: https://portfoliocharts.com/withdrawal-rates-faq/
User avatar
mathjak107
Executive Member
Executive Member
Posts: 4456
Joined: Fri Jun 19, 2015 2:54 am
Location: bayside queens ny
Contact:

Re: Assumed Portfolio Return in Retirement

Post by mathjak107 » Tue May 29, 2018 9:37 am

here is the good news :

as michael kitces extracted from all the worst cases , it all boils down to the first 15 years averaging at at least a 2% real return to have 4% inflation adjusted hold .

it really does not matter what investments you use as long as the math holds . it can be invested in beaver cheese and it will not matter .

of course if you want to see if how you would have done against the worst case scenarios you would need the exact data on those investments or else you can not compare to the past , with gold we can't really do that so just like when we use other investments in our portfolio that did not exist you in all honesty cannot stress test the portfolio .

but thanks to michael kitce's work we don't have to . we can monitor our own actual results going forward with our actual portfolio's . all we really need from the past is the math and sequencing of the failures .

if five years in you are not seeing at least a 2% real return a red flag should go up that says monitor this and a pay cut may be coming . remember a 2% real return my leave you with a buck in the 31st year so you really need to do better . if you are not seeing at least 2% real return as an average that red flag has to go up .

so far i have been retired 3 years and even after pulling 6 figures a year to live on we are a few hundred thousand higher than the day we retired so there is a nice cushion built up .

so it is not just about the swr it is about the balance too that goes with that swr rate . kitces found a 60/40 mix not only provided a 4% safe withdrawal rate over 30 years but it left you with more than you started with over 90% of the 117 rolling 30 year time frames to date and 67% of those time frame left you with double what you started with .

so just looking at an swr rate holding in a portfolio is only part of the story , the balance left for heirs and a cushion is also something of concern .

but thanks to the numbers crunching of the old data it is now translated in to a actual average real return we need to see as a min over the first 15 years . .
Last edited by mathjak107 on Tue May 29, 2018 10:23 am, edited 14 times in total.
User avatar
ochotona
Executive Member
Executive Member
Posts: 3353
Joined: Fri Jan 30, 2015 5:54 am

Re: Assumed Portfolio Return in Retirement

Post by ochotona » Tue May 29, 2018 9:37 am

The 1966 begin date for retirement is seen as one of the very worst in recent decades. Maybe 1929 would be worse.
User avatar
mathjak107
Executive Member
Executive Member
Posts: 4456
Joined: Fri Jun 19, 2015 2:54 am
Location: bayside queens ny
Contact:

Re: Assumed Portfolio Return in Retirement

Post by mathjak107 » Tue May 29, 2018 9:38 am

ochotona wrote:
Tue May 29, 2018 9:37 am
The 1966 begin date for retirement is seen as one of the very worst in recent decades. Maybe 1929 would be worse.
actually 1965/1966 was the worst ever because of inflation .
User avatar
ochotona
Executive Member
Executive Member
Posts: 3353
Joined: Fri Jan 30, 2015 5:54 am

Re: Assumed Portfolio Return in Retirement

Post by ochotona » Tue May 29, 2018 10:00 am

Yeah I guess a dime went a long way in the 1930s due to deflation!
User avatar
mathjak107
Executive Member
Executive Member
Posts: 4456
Joined: Fri Jun 19, 2015 2:54 am
Location: bayside queens ny
Contact:

Re: Assumed Portfolio Return in Retirement

Post by mathjak107 » Tue May 29, 2018 10:07 am

deflation played a big roll in 1929 . markets only had to come back within 18% of the old high and it was the same in dollars .

when it comes to 1965/1966 inflation was 2.50% -3.50% . who would have guessed in 3 years time it would have doubled and by 1974 it would be 11%. it was crushing to a retiree. but with inflation so low who ever expected a 4x increase coming .
User avatar
Tyler
Executive Member
Executive Member
Posts: 2066
Joined: Sat Nov 12, 2011 3:23 pm
Contact:

Re: Assumed Portfolio Return in Retirement

Post by Tyler » Tue May 29, 2018 12:36 pm

EdwardjK wrote:
Mon May 28, 2018 10:02 pm
Hi. I've built a spreadsheet that takes into account my starting portfolio value, estimated spending in retirement, inflation, and Social Security income. Using Solver, the spreadsheet calculates the average annual return needed so that I exhaust my portfolio at age 100. Not that my wife and I will last that long, but it's good to plan that long.

So I am asking what others are using as their assumed portfolio return in retirement to compare to what my spreadsheet says I need. 4% ? 5% ? Something else?

Your feedback is appreciated.
I see two things here worth more discussion.

First is the idea that retirement portfolio performance can be accurately modeled with simple average returns. That’s definitely not true, and doing it that way without accounting for sequence of returns will drastically over-state your true safe withdrawal rate. For more on that, read this: https://portfoliocharts.com/2015/11/17/ ... ates-work/

The second is the idea of using an alternative withdrawal strategy designed to deplete your portfolio at age 100. That sounds a lot like Variable Percentage Withdrawal (VPW) that is popular among some Bogleheads. Basically, it constantly adjusts withdrawals with an amortization-based goal to deplete your portfolio on a certain date. You might try searching Bogleheads.org for “VPW” or playing with the VPW setting on this retirement spending calculator: https://portfoliocharts.com/portfolio/r ... -spending/
User avatar
mathjak107
Executive Member
Executive Member
Posts: 4456
Joined: Fri Jun 19, 2015 2:54 am
Location: bayside queens ny
Contact:

Re: Assumed Portfolio Return in Retirement

Post by mathjak107 » Tue May 29, 2018 1:27 pm

i use bob clyatt's withdrawal method myself . it works great .

it rewards you when markets are up and does not require much adjusting if it is down .

you just look at your year end balance and you get either 4% of it or if markets are down :

you take the higher of 4% of the balance or what you took last year less 5% ..

that is it , there is nothing else to it . it stress tested out to 40 years just fine .

personally i would never want any draw method that tried to guess the future and deplete my portfolio by a certain date. as you saw in milvesky's work the same return can see a 15 year difference in how long the money lasted just based on sequences .
User avatar
ochotona
Executive Member
Executive Member
Posts: 3353
Joined: Fri Jan 30, 2015 5:54 am

Re: Assumed Portfolio Return in Retirement

Post by ochotona » Tue May 29, 2018 1:39 pm

Why not just run a new retirement plan every year and just accept the specified new payout? Why does it have to be complicated?
User avatar
mathjak107
Executive Member
Executive Member
Posts: 4456
Joined: Fri Jun 19, 2015 2:54 am
Location: bayside queens ny
Contact:

Re: Assumed Portfolio Return in Retirement

Post by mathjak107 » Tue May 29, 2018 1:50 pm

ochotona wrote:
Tue May 29, 2018 1:39 pm
Why not just run a new retirement plan every year and just accept the specified new payout? Why does it have to be complicated?
because the income can vary by a lot year to year just doing it off the cuff . that is just the opposite of what a retirement income stream should be . it needs to be safe ,secure and consistent . if markets are down 20% you should not have a plan that cuts you in draw so much you can't pay bills .

so it needs to have a structure and form to it that tests well . that is the issue trying to use something like an rmd schedule . it does the opposite of what it should .
stuper1
Executive Member
Executive Member
Posts: 1365
Joined: Sun Mar 03, 2013 7:18 pm

Re: Assumed Portfolio Return in Retirement

Post by stuper1 » Tue May 29, 2018 2:53 pm

mathjak107 wrote:
Tue May 29, 2018 1:27 pm
i use bob clyatt's withdrawal method myself . it works great .

it rewards you when markets are up and does not require much adjusting if it is down .

you just look at your year end balance and you get either 4% of it or if markets are down :

you take the higher of 4% of the balance or what you took last year less 5% ..

that is it , there is nothing else to it . it stress tested out to 40 years just fine .

personally i would never want any draw method that tried to guess the future and deplete my portfolio by a certain date. as you saw in milvesky's work the same return can see a 15 year difference in how long the money lasted just based on sequences .
This simple withdrawal method sounds good. Just a question to help me understand. Say the market was way up in the preceding year, so that my 4% withdrawal was a big number compared to previous years. Then, this year the market is way down. Would I really only take 5% off my big number from last year? Then, what if the market was way down for a second year, would I only take another 5% off the next year?
User avatar
mathjak107
Executive Member
Executive Member
Posts: 4456
Joined: Fri Jun 19, 2015 2:54 am
Location: bayside queens ny
Contact:

Re: Assumed Portfolio Return in Retirement

Post by mathjak107 » Tue May 29, 2018 3:01 pm

yes , that is exactly how it works . that 5% drop off that big number is all you take . if markets are down the following year then once again you get the higher of 4% of the balance or 5% off the previous years draw .

all inflation adjusting is built in .

it really works well .

with the standard 4% constant dollar draw we typically hear of (4% rule ) you can die with way to much money unspent as 90% of the time with a 60/40 mix you die with more than you started .

so you need to take raises . a rule of thumb is anytime your balance is 50% higher than you started take a 10% raise plus the inflation raise. repeat every 3 years if you are still up more than 50% .

that is a lot of work compared to the method i use.

we just use that amount to set goal posts for spending each year . we can easily spend 2 or 3x our budget , we love travel and spending on our hobbies so over spending is easy . we both love photography so we are always going some where so we need goal posts .
stuper1
Executive Member
Executive Member
Posts: 1365
Joined: Sun Mar 03, 2013 7:18 pm

Re: Assumed Portfolio Return in Retirement

Post by stuper1 » Tue May 29, 2018 3:16 pm

mathjak107 wrote:
Tue May 29, 2018 1:27 pm

you just look at your year end balance and you get either 4% of it or if markets are down :

you take the higher of 4% of the balance or what you took last year less 5% ..
I want to make sure I understand this right, so please don't think I'm quibbling over words. Is the first line of your description above somewhat redundant, or am I missing something? If the method was simply stated this way would it be accurate: at the end of the year take the higher of 4% of the balance or what you took last year less 5% ?
Post Reply