C
Cut-Throat
Guest
nm
haha
Give me a museum and I'll fill it. (Picasso) Give me a forum ...
Re: SWR should not be constant over a Retirement L
I think your idea is interesting. But many people do like to keep traveling; my wife's father is 90 and they still go to Europe, Hawaii, etc. Other than being a bit deaf from 10 years in the artillery, you can't really tell that he isn't a much younger man. He ran and finished marathons until he was 70.
A detail question- in your graph, what do the colors on the bars mean?
Mikey
I think your idea is interesting. But many people do like to keep traveling; my wife's father is 90 and they still go to Europe, Hawaii, etc. Other than being a bit deaf from 10 years in the artillery, you can't really tell that he isn't a much younger man. He ran and finished marathons until he was 70.
A detail question- in your graph, what do the colors on the bars mean?
Mikey
cute fuzzy bunny
Give me a museum and I'll fill it. (Picasso) Give me a forum ...
Re: SWR should not be constant over a Retirement L
Ugh. I'm with John. I must have flown over two million miles in my time, i've been all over europe and on almost every island in the carribbean. I havent seen the inside of an airplane in 3 years, and I doubt I ever will again.
Of course, with lake tahoe, napa valley, san francisco, the marin headlands/muir woods, point reyes and big sur within a few hours drive, maybe I wont need to.
Ugh. I'm with John. I must have flown over two million miles in my time, i've been all over europe and on almost every island in the carribbean. I havent seen the inside of an airplane in 3 years, and I doubt I ever will again.
Of course, with lake tahoe, napa valley, san francisco, the marin headlands/muir woods, point reyes and big sur within a few hours drive, maybe I wont need to.
OP
OP
Cut-Throat
Guest
Re: SWR should not be constant over a Retirement L
Mikey,
Clearly your father-in-law is the exception. I've got friends of mine that are close to 70 that walking to the trout stream is a marathon in itself. I travel a lot myself and love it - However, I see very few old folks doing it. I'm playing the averages. Many people do like to travel in their old age, but then again I'm guessing they are though buying furniture and Brand new cars, the latest computers. Demographic spending patterns support my argument.
A trip to Woodstock in 69 seemed great at the time, but I'm not sure I'd relish the accomodations today.
The Orange- Red on the Graph is the Taxable portion, The Yellow is my Tax Deferred Funds and the Blue is my Wife's Tax Deferred Funds.
Mikey,
Clearly your father-in-law is the exception. I've got friends of mine that are close to 70 that walking to the trout stream is a marathon in itself. I travel a lot myself and love it - However, I see very few old folks doing it. I'm playing the averages. Many people do like to travel in their old age, but then again I'm guessing they are though buying furniture and Brand new cars, the latest computers. Demographic spending patterns support my argument.
A trip to Woodstock in 69 seemed great at the time, but I'm not sure I'd relish the accomodations today.
The Orange- Red on the Graph is the Taxable portion, The Yellow is my Tax Deferred Funds and the Blue is my Wife's Tax Deferred Funds.
Re: SWR should not be constant over a Retirement L
Here are some more links along these lines.
The first talks about Utility Theory.
therealchips
http://nofeeboards.com/boards/viewtopic.php?p=13392&13392
This one talks about Sensible Withdrawal Rates.
gummy
http://home.golden.net/~pjponzo/sensible_withdrawals.htm
Let me mention a trick that you can use with FIRECalc:
Normally, FIRECalc makes withdrawals based upon a percentage of a portfolio's initial balance. It will adjust withdrawals to match inflation unless you choose otherwise. The most frequently mentioned withdrawal rate using this strategy is 4% (plus inflation).
The most commonly mentioned alternative strategy is to withdraw a constant percentage of a portfolio's current balance. Your portfolio is guaranteed never to run out of money although the balances and withdrawal amounts can become ridiculously low. The most frequently mentioned withdrawal rate using this alternative strategy is 5%.
FIRECalc can calculate the results of the alternative strategy. In fact, you can mix the two approaches in any way that you wish. Simply add the percentage of the current balance to the expenses and put the percentage of the initial balance in its usual place. Most of the time, there will be a balance left over. (It is in nominal dollars). You can see what would have happened in the past by reading the FIRECalc balances for each year.
To convert from nominal dollars to real dollars, look at the stock market data at Professor Shiller's database. It is the source of input data for FIRECalc and all other simulators that use historical sequences going back to 1871. It includes CPI information. You can use the CPI information to convert from nominal dollars to real dollars.
http://www.econ.yale.edu/~shiller/
http://www.econ.yale.edu/~shiller/data.htm
Have fun.
John R.
Here are some more links along these lines.
The first talks about Utility Theory.
therealchips
http://nofeeboards.com/boards/viewtopic.php?p=13392&13392
This one talks about Sensible Withdrawal Rates.
gummy
http://home.golden.net/~pjponzo/sensible_withdrawals.htm
Let me mention a trick that you can use with FIRECalc:
Normally, FIRECalc makes withdrawals based upon a percentage of a portfolio's initial balance. It will adjust withdrawals to match inflation unless you choose otherwise. The most frequently mentioned withdrawal rate using this strategy is 4% (plus inflation).
The most commonly mentioned alternative strategy is to withdraw a constant percentage of a portfolio's current balance. Your portfolio is guaranteed never to run out of money although the balances and withdrawal amounts can become ridiculously low. The most frequently mentioned withdrawal rate using this alternative strategy is 5%.
FIRECalc can calculate the results of the alternative strategy. In fact, you can mix the two approaches in any way that you wish. Simply add the percentage of the current balance to the expenses and put the percentage of the initial balance in its usual place. Most of the time, there will be a balance left over. (It is in nominal dollars). You can see what would have happened in the past by reading the FIRECalc balances for each year.
To convert from nominal dollars to real dollars, look at the stock market data at Professor Shiller's database. It is the source of input data for FIRECalc and all other simulators that use historical sequences going back to 1871. It includes CPI information. You can use the CPI information to convert from nominal dollars to real dollars.
http://www.econ.yale.edu/~shiller/
http://www.econ.yale.edu/~shiller/data.htm
Have fun.
John R.
dory36
Early-Retirement.org Founder, Developer of FIRECal
- Joined
- Jun 23, 2002
- Messages
- 1,841
Re: SWR should not be constant over a Retirement L
Dory36
John, I think I follow you on this, but can you give an example?
Dory36
dory36
Early-Retirement.org Founder, Developer of FIRECal
- Joined
- Jun 23, 2002
- Messages
- 1,841
Re: SWR should not be constant over a Retirement L
Another modification of the constant percentage of current portfolio approach was proposed, and I think used, by Galeno on TMF. His approach basically gave him a yearly spending amount that was the running average of about 3-4 years of withdrawals.
Another modification of the constant percentage of current portfolio approach was proposed, and I think used, by Galeno on TMF. His approach basically gave him a yearly spending amount that was the running average of about 3-4 years of withdrawals.
unclemick
Give me a museum and I'll fill it. (Picasso) Give me a forum ...
Re: SWR should not be constant over a Retirement L
There are plans and then there is real life - still have my 1982 plan(on graph PAPER) to retire early at age 63 in 2007 with ballpark 1.3 mil in my 401k.
After layoff in 93 at age 49, became an ER when moving/new job began to look icky. 93-2003 in ER would look weird on a speadsheet as to income/s, expenses, portfolio, etc. Our SWR was - whatever was needed.
Plans are plans - SWR and retirement calculators are a worthwhile craft to master and use as a source of insight. BTY- our 'utility' function morphed over the years from travel to basically no travel(arthritis and age? ala Cut-Throat).
I go with variable SWR nowadays(partly because in our case we can). The floor is the portfolio yield (div.+int) and the ceiling is my take on current SWR. I like JWR1945's old posted rule of thumb portfolio yield plus maybe 1% as a kind of cross check which puts us nowadays ballpark 3.5 -4%.
So I have a plan - and we're in real life ER. Interesting.
There are plans and then there is real life - still have my 1982 plan(on graph PAPER) to retire early at age 63 in 2007 with ballpark 1.3 mil in my 401k.
After layoff in 93 at age 49, became an ER when moving/new job began to look icky. 93-2003 in ER would look weird on a speadsheet as to income/s, expenses, portfolio, etc. Our SWR was - whatever was needed.
Plans are plans - SWR and retirement calculators are a worthwhile craft to master and use as a source of insight. BTY- our 'utility' function morphed over the years from travel to basically no travel(arthritis and age? ala Cut-Throat).
I go with variable SWR nowadays(partly because in our case we can). The floor is the portfolio yield (div.+int) and the ceiling is my take on current SWR. I like JWR1945's old posted rule of thumb portfolio yield plus maybe 1% as a kind of cross check which puts us nowadays ballpark 3.5 -4%.
So I have a plan - and we're in real life ER. Interesting.
OP
OP
Cut-Throat
Guest
Re: SWR should not be constant over a Retirement L
JWR
Yes, I am confused on this as well. An example would be nice
JWR
Yes, I am confused on this as well. An example would be nice
dory36
Early-Retirement.org Founder, Developer of FIRECal
- Joined
- Jun 23, 2002
- Messages
- 1,841
Re: SWR should not be constant over a Retirement L
As I see it, the whole idea of any SWR approach is to test feasibility of a plan, not to map out your spending for the next 45 years.
Dory36
Yep.
As I see it, the whole idea of any SWR approach is to test feasibility of a plan, not to map out your spending for the next 45 years.
Dory36
Re: SWR should not be constant over a Retirement L
Here is an example. I used FIRECalc's default settings except for the Withdrawals on the first line and the expenses, which are close to the bottom. The default setting for Withdrawals is $30000. The default setting for the expenses is 0.18%.
The year 1965 was an especially bad year for retirees. Using FIRECalc's default settings, the balance after 10 years was $501223. After 20 years it was $346372. It failed at 26 years. That is, it had a negative balance at 26 years. These are the results of withdrawals based upon the initial balance of one's portfolio.
The $30000 withdrawal amount is 4.62% (more precisely, 4.61538%) of the $650000 initial balance. To calculate withdrawals that are 4.62% of the current balance, put $0 for Withdrawals (instead of $30000) and 4.80% for the expenses (since 4.62 + 0.18 = 4.80). That is, you add 4.62% to the normal expenses of 0.18% for a total of 4.80%.
For a retirement portfolio starting in 1965, the balance after 10 years would have been $556980. After 20 years it would have been $1253064. After 30 years it would have been $2546642.
Looking at Professor Shiller's database, the CPI in January 1965 was 31.2. (January 1965 is read from the listing as 1965.01. February is 1965.02 and so forth.) The CPI in January 1975 was 52.1. In January 1985 the CPI was 105.5. In January 1995 the CPI was 150.3. To adjust for inflation divide the index value for 1965 by the index value of the later years. Then multiply the (nominal) balances by these fractions. This converts the balances to real dollars. The fraction for 1975 was 31.2 / 52.1 = 0.5988. For 1985 it was 31.2 / 105.5 = 0.2957. For 1995 it was 31.2 / 150.3 = 0.2076.
After adjusting for inflation, the default portfolio ($30000 annual withdrawals plus inflation with an initial balance of $650000) would have had a (real) balance of $300132 in 1975 and $102422. (It failed at year 26.)
After adjusting for inflation, a portfolio that withdrew 4.62% of the portfolio's current balance would have had a (real) balance of $333520 in 1975 and $370531 in 1985 and $528682 in 1995. Notice that the withdrawal amount would have decreased to $15409 (or 4.62% of $333520) in 1975 in terms of buying power (i.e., real dollars). On the other hand, the balance never falls to zero and it lasts indefinitely. By 1995, it would have recovered most of its buying power in spite of having provided an income stream for three decades. That is, 4.62% of $528682 is $24425.
Now for something really fancy. Put $15000 into Withdrawals. That is 2% of the initial balance of $650000. Put 2.80% into expenses. That is 2.62% above the normal expenses of 0.18%. This corresponds to withdrawing $15000 plus inflation and 2.62% of the portfolio's current balance each year. The first year is the same as before (in both examples). From then on, this is a mixture of the two approaches. For a retirement portfolio that started in 1965, the nominal balance after 10 years would have been $533890. After 20 years, it would have been $913051. After 30 years, it would have been $1603048.
The inflation-adjusted balances would have been $319693 in 1975 and $269989 in 1985 and $332792 in 1995. The withdrawal amount in 1975 would have been $15000 + 2.62% of $269989 = $15000 + $7074 = $22074. The withdrawal amount in 1995 would have been $15000 + 2.62% of $332792 = $15000 + $8719 = $23719 in real dollars (that is, after adjusting for inflation). By using this mixed approach, we get a little less in 1995 but a lot more in 1975 than if we had stuck to using only the current balance. Of course, the portfolio would have failed if we had kept our withdrawals based only upon the initial balance.
I hope that this gives you an idea of what FIRECalc can do in addition to its originally planned for capabilities. It is a powerful analysis tool.
Have fun.
John R.
Here is an example. I used FIRECalc's default settings except for the Withdrawals on the first line and the expenses, which are close to the bottom. The default setting for Withdrawals is $30000. The default setting for the expenses is 0.18%.
The year 1965 was an especially bad year for retirees. Using FIRECalc's default settings, the balance after 10 years was $501223. After 20 years it was $346372. It failed at 26 years. That is, it had a negative balance at 26 years. These are the results of withdrawals based upon the initial balance of one's portfolio.
The $30000 withdrawal amount is 4.62% (more precisely, 4.61538%) of the $650000 initial balance. To calculate withdrawals that are 4.62% of the current balance, put $0 for Withdrawals (instead of $30000) and 4.80% for the expenses (since 4.62 + 0.18 = 4.80). That is, you add 4.62% to the normal expenses of 0.18% for a total of 4.80%.
For a retirement portfolio starting in 1965, the balance after 10 years would have been $556980. After 20 years it would have been $1253064. After 30 years it would have been $2546642.
Looking at Professor Shiller's database, the CPI in January 1965 was 31.2. (January 1965 is read from the listing as 1965.01. February is 1965.02 and so forth.) The CPI in January 1975 was 52.1. In January 1985 the CPI was 105.5. In January 1995 the CPI was 150.3. To adjust for inflation divide the index value for 1965 by the index value of the later years. Then multiply the (nominal) balances by these fractions. This converts the balances to real dollars. The fraction for 1975 was 31.2 / 52.1 = 0.5988. For 1985 it was 31.2 / 105.5 = 0.2957. For 1995 it was 31.2 / 150.3 = 0.2076.
After adjusting for inflation, the default portfolio ($30000 annual withdrawals plus inflation with an initial balance of $650000) would have had a (real) balance of $300132 in 1975 and $102422. (It failed at year 26.)
After adjusting for inflation, a portfolio that withdrew 4.62% of the portfolio's current balance would have had a (real) balance of $333520 in 1975 and $370531 in 1985 and $528682 in 1995. Notice that the withdrawal amount would have decreased to $15409 (or 4.62% of $333520) in 1975 in terms of buying power (i.e., real dollars). On the other hand, the balance never falls to zero and it lasts indefinitely. By 1995, it would have recovered most of its buying power in spite of having provided an income stream for three decades. That is, 4.62% of $528682 is $24425.
Now for something really fancy. Put $15000 into Withdrawals. That is 2% of the initial balance of $650000. Put 2.80% into expenses. That is 2.62% above the normal expenses of 0.18%. This corresponds to withdrawing $15000 plus inflation and 2.62% of the portfolio's current balance each year. The first year is the same as before (in both examples). From then on, this is a mixture of the two approaches. For a retirement portfolio that started in 1965, the nominal balance after 10 years would have been $533890. After 20 years, it would have been $913051. After 30 years, it would have been $1603048.
The inflation-adjusted balances would have been $319693 in 1975 and $269989 in 1985 and $332792 in 1995. The withdrawal amount in 1975 would have been $15000 + 2.62% of $269989 = $15000 + $7074 = $22074. The withdrawal amount in 1995 would have been $15000 + 2.62% of $332792 = $15000 + $8719 = $23719 in real dollars (that is, after adjusting for inflation). By using this mixed approach, we get a little less in 1995 but a lot more in 1975 than if we had stuck to using only the current balance. Of course, the portfolio would have failed if we had kept our withdrawals based only upon the initial balance.
I hope that this gives you an idea of what FIRECalc can do in addition to its originally planned for capabilities. It is a powerful analysis tool.
Have fun.
John R.
OP
OP
Cut-Throat
Guest
Re: SWR should not be constant over a Retirement L
JWR,
Great Stuff - Now I follow you !!
Thanks!!
JWR,
Great Stuff - Now I follow you !!
Thanks!!
dory36
Early-Retirement.org Founder, Developer of FIRECal
- Joined
- Jun 23, 2002
- Messages
- 1,841
Re: SWR should not be constant over a Retirement L
Thanks John -- that's really useful.
Dory36
Thanks John -- that's really useful.
Dory36
Re: SWR should not be constant over a Retirement L
You are certainly welcome.
Let me point out that you can also do this with the Retire Early Safe Withdrawal Calculator.
FIRECalc is much more user friendly, of course.
Neither calculator has a built-in option of presenting balances in terms of real (inflation adjusted) dollars. Dory36 has put such that capability on his to-do list. But he has a ton of other things on his to-do list as well.
Have fun.
John R.
You are certainly welcome.
Let me point out that you can also do this with the Retire Early Safe Withdrawal Calculator.
FIRECalc is much more user friendly, of course.
Neither calculator has a built-in option of presenting balances in terms of real (inflation adjusted) dollars. Dory36 has put such that capability on his to-do list. But he has a ton of other things on his to-do list as well.
Have fun.
John R.
Re: SWR should not be constant over a Retirement L
For unclemick:
If you look at the data at Professor Shiller's site, you will see that real dividend amounts have generally increased with inflation plus a little bit more. Again, I am talking about amounts, not yields. Let me add that dividend behavior has been quite erratic. It has not increased smoothly with time, unless you are looking in the very long-term of 50 years or so.
To the extent that dividend amounts continue to increase with inflation, dividends place a floor underneath withdrawal rates. You can collect dividends indefinitely because you never have to sell any shares. (Interest acts the same, when you use real dollars instead of nominal dollars.) If you allow your final balance to decrease to zero, then the withdrawal rate is something higher than the floor provided by the initial dividend yield. Empirically, this has worked out to be something close to 1% under stressful conditions.
We can depend upon this logic and we know what to look for when it fails. That is, we know to look at real dividend amounts and how well they behave.
Have fun.
John R.
For unclemick:
If you look at the data at Professor Shiller's site, you will see that real dividend amounts have generally increased with inflation plus a little bit more. Again, I am talking about amounts, not yields. Let me add that dividend behavior has been quite erratic. It has not increased smoothly with time, unless you are looking in the very long-term of 50 years or so.
To the extent that dividend amounts continue to increase with inflation, dividends place a floor underneath withdrawal rates. You can collect dividends indefinitely because you never have to sell any shares. (Interest acts the same, when you use real dollars instead of nominal dollars.) If you allow your final balance to decrease to zero, then the withdrawal rate is something higher than the floor provided by the initial dividend yield. Empirically, this has worked out to be something close to 1% under stressful conditions.
We can depend upon this logic and we know what to look for when it fails. That is, we know to look at real dividend amounts and how well they behave.
Have fun.
John R.
Re: SWR should not be constant over a Retirement L
Looking at mixed withdrawal strategies is helpful when you think about how real retirees are likely to behave. When stock prices fall, real retirees are likely to cut back on their spending whether or not it makes sense intellectually.
For example, I have a Federal Civil Service pension that is more than sufficient to meet my needs and which includes increases (but never any decreases) to match inflation. Even then, I found myself cutting back on my spending during the bear market. My reaction was instinctive, not logical. I do not expect to use the money myself. It will just go to my kids later on.
Have fun.
John R.
Looking at mixed withdrawal strategies is helpful when you think about how real retirees are likely to behave. When stock prices fall, real retirees are likely to cut back on their spending whether or not it makes sense intellectually.
For example, I have a Federal Civil Service pension that is more than sufficient to meet my needs and which includes increases (but never any decreases) to match inflation. Even then, I found myself cutting back on my spending during the bear market. My reaction was instinctive, not logical. I do not expect to use the money myself. It will just go to my kids later on.
Have fun.
John R.
Re: SWR should not be constant over a Retirement L
Unclemick, what does your portfolio consist of, and how do you draw down when you're in your floor scenario? Do you spend dividends plus sell off 1% of the total portfolio from the best performing asset?
Re: SWR should not be constant over a Retirement L
1. Any idea how much one's income is likely to vary while withdrawing dividends plus 1%?
2. How safe would this have been in the past?
3. What would have been the minimum stock market exposure required to make this work in the past century?
Thanks. This has great appeal.
Three follow-up questions:
1. Any idea how much one's income is likely to vary while withdrawing dividends plus 1%?
2. How safe would this have been in the past?
3. What would have been the minimum stock market exposure required to make this work in the past century?
Thanks. This has great appeal.
cute fuzzy bunny
Give me a museum and I'll fill it. (Picasso) Give me a forum ...
Re: SWR should not be constant over a Retirement L
Now this makes me want firecalc to have a detailed display option that shows your actual withdrawal/costs amount per year so I can see what it is (was).
Maybe instead of showing
12312 123123 12312312 as the port size per year it could show
12312 (30000) 123123 (32000)...with the withdrawal amount in parens. Or do I just need another cup of coffee?
What I'm looking for in specific is to see the actual amount I would have had to live on in these scenarios as a percent instead of an inflation adjusted amount. When I did the firecalc run it showed I was 100% no problem, but reported the maximum withdrawal as a fairly lowish number.
Now this makes me want firecalc to have a detailed display option that shows your actual withdrawal/costs amount per year so I can see what it is (was).
Maybe instead of showing
12312 123123 12312312 as the port size per year it could show
12312 (30000) 123123 (32000)...with the withdrawal amount in parens. Or do I just need another cup of coffee?
What I'm looking for in specific is to see the actual amount I would have had to live on in these scenarios as a percent instead of an inflation adjusted amount. When I did the firecalc run it showed I was 100% no problem, but reported the maximum withdrawal as a fairly lowish number.
cute fuzzy bunny
Give me a museum and I'll fill it. (Picasso) Give me a forum ...
Re: SWR should not be constant over a Retirement L
Nevermind, I'm feeling industrious today and did it in excel.
I did my base calcs with my existing portfolio %'s, zero for withdrawal, and 4.5% expenses. Report was 100% successful (which it always will be at zero withdrawal).
Drawing down the actual withdrawal amounts in excel showed that except for the naughty years 1929 and 1965 and a couple of others, the withdrawal was in excess of the hard inflation adjusted amount I had previously run with. In the bad years I'd have had to live off of about 2/3 my current expenditure rate. Not horrible.
Making a 10k per year base adjustment and keeping the 4.5% on top of that gave me 100% (not a given in this case) and a usable withdrawal in 100% of the years projected.
What I dont have and am not sure how to calc without drawing in historic inflation tables is whether the last 5-10 years worth of withdrawals would have been high enough to overcome inflation. In other words, would I have descended into eating cat food and washing off plastic wrap. My dad, a depression era child, does rinse off plastic wrap and dries out paper towels and reuses them. Boy I hope we never see those times again.
Nevermind, I'm feeling industrious today and did it in excel.
I did my base calcs with my existing portfolio %'s, zero for withdrawal, and 4.5% expenses. Report was 100% successful (which it always will be at zero withdrawal).
Drawing down the actual withdrawal amounts in excel showed that except for the naughty years 1929 and 1965 and a couple of others, the withdrawal was in excess of the hard inflation adjusted amount I had previously run with. In the bad years I'd have had to live off of about 2/3 my current expenditure rate. Not horrible.
Making a 10k per year base adjustment and keeping the 4.5% on top of that gave me 100% (not a given in this case) and a usable withdrawal in 100% of the years projected.
What I dont have and am not sure how to calc without drawing in historic inflation tables is whether the last 5-10 years worth of withdrawals would have been high enough to overcome inflation. In other words, would I have descended into eating cat food and washing off plastic wrap. My dad, a depression era child, does rinse off plastic wrap and dries out paper towels and reuses them. Boy I hope we never see those times again.
Re: SWR should not be constant over a Retirement L
For BobSmith:
I cannot answer your questions using FIRECalc. I have to use one of my modified versions of the Retire Early Safe Withdrawal Calculator. The unmodified calculator is available at www.retireearlyhomepage.com . I have posted detailed instructions as to how to make my modifications are among several posts at the SWR Research Group at www.nofeeboards.com . These calculators are similar to FIRECalc, but they also have some important differences, at least using their default settings. Among them is how withdrawals are applied. (Withdrawals are split into two parts, one at the very beginning of the year and the other at the very end of the year. Gains or losses in the market can affect results.) The default expense ratio is 0.20% (instead of 0.18%).
The one thing that might stand out is that I routinely use an initial balance of $100000 with those calculators in contrast to the $650000 that FIRECalc has for its default value.
In addition, I have set things up so as to show balances. I have not set them up to show the amount withdrawn each year. You have to hand-calculate that. On a positive note, I have data summaries similar to FIRECalc both in real dollars and in nominal dollars.
I set conditions up using an initial balance of $100000 and 0.20% expenses for a portfolio of 80% stocks and 20% commercial paper (re-balanced annually at zero cost). I withdrew all dividends (i.e., in terms of the model inputs, I set the dividend reinvestment percentage equal to zero).
The first thing that I did was to identify the maximum withdrawal rate that would have had 100% survival for 30-year retirements. It was 2.3% (plus inflation). At 2.4%, there was a single failure in 1929. I looked at balances after 10 years and selected 1911, 1929, 1937 and 1966 for further examination.
In what follows, I am using real dollars. They are already adjusted for inflation.
1) With a starting balance of $100000 in 1911, the balance after 10 years (or 1921) would have been $28932. After 20 years (in 1931), the balance would have been $38739. After 30 years (in 1941), the balance would have been $11830.
2) With a starting balance of $100000 in 1929, the balance after 10 years (or 1939) would have been $54051. After 20 years (in 1949), the balance would have been $17160. After 30 years (in 1959), the balance would have been $4218. The portfolio would have run out of money by 32 years.
3) With a starting balance of $100000 in 1937, the balance after 10 years (or 1947) would have been $39187. After 20 years (in 1957), the balance would have been $38080. After 30 years (in 1967), the balance would have been $28847.
4)With a starting balance of $100000 in 1966, the balance after 10 years (or 1976) would have been $43634. After 20 years (in 1986), the balance would have been $27647. After 30 years (in 1996), the balance would have been $20336.
[more follows]
For BobSmith:
I cannot answer your questions using FIRECalc. I have to use one of my modified versions of the Retire Early Safe Withdrawal Calculator. The unmodified calculator is available at www.retireearlyhomepage.com . I have posted detailed instructions as to how to make my modifications are among several posts at the SWR Research Group at www.nofeeboards.com . These calculators are similar to FIRECalc, but they also have some important differences, at least using their default settings. Among them is how withdrawals are applied. (Withdrawals are split into two parts, one at the very beginning of the year and the other at the very end of the year. Gains or losses in the market can affect results.) The default expense ratio is 0.20% (instead of 0.18%).
The one thing that might stand out is that I routinely use an initial balance of $100000 with those calculators in contrast to the $650000 that FIRECalc has for its default value.
In addition, I have set things up so as to show balances. I have not set them up to show the amount withdrawn each year. You have to hand-calculate that. On a positive note, I have data summaries similar to FIRECalc both in real dollars and in nominal dollars.
I set conditions up using an initial balance of $100000 and 0.20% expenses for a portfolio of 80% stocks and 20% commercial paper (re-balanced annually at zero cost). I withdrew all dividends (i.e., in terms of the model inputs, I set the dividend reinvestment percentage equal to zero).
The first thing that I did was to identify the maximum withdrawal rate that would have had 100% survival for 30-year retirements. It was 2.3% (plus inflation). At 2.4%, there was a single failure in 1929. I looked at balances after 10 years and selected 1911, 1929, 1937 and 1966 for further examination.
In what follows, I am using real dollars. They are already adjusted for inflation.
1) With a starting balance of $100000 in 1911, the balance after 10 years (or 1921) would have been $28932. After 20 years (in 1931), the balance would have been $38739. After 30 years (in 1941), the balance would have been $11830.
2) With a starting balance of $100000 in 1929, the balance after 10 years (or 1939) would have been $54051. After 20 years (in 1949), the balance would have been $17160. After 30 years (in 1959), the balance would have been $4218. The portfolio would have run out of money by 32 years.
3) With a starting balance of $100000 in 1937, the balance after 10 years (or 1947) would have been $39187. After 20 years (in 1957), the balance would have been $38080. After 30 years (in 1967), the balance would have been $28847.
4)With a starting balance of $100000 in 1966, the balance after 10 years (or 1976) would have been $43634. After 20 years (in 1986), the balance would have been $27647. After 30 years (in 1996), the balance would have been $20336.
[more follows]
Re: SWR should not be constant over a Retirement L
For Bob_Smith (continued):
In all of these cases the withdrawal amount would have been $2300 plus inflation every year plus all of the dividends on the stocks in the portfolio. Withdrawals from dividends are restricted to the 80% part of the portfolio that consists of stocks and they do not include interest from the commercial paper.
1) The dividend yield in 1911 was 5.07%. In 1921 it was 7.11%. In 1931 it was 6.05%. In 1941 it was 6.38%.
2) The dividend yield in 1929 was 3.46%. In 1939 it was 4.11%. In 1949 it was 6.16%. In 1959 it was 3.16%.
3) The dividend yield in 1937 was 4.15%. In 1947 it was 4.69%. In 1957 it was 3.82%. In 1967 it was 3.41%.
4) The dividend yield in 1966 was 2.94%. In 1976 it was 3.80%. In 1986 it was 3.81%. In 1996 it was 2.26%.
To help you interpret these numbers, remember that my starting balance is $100000, not the $650000 that FIRECalc has as its default value. This means that 4% of the initial balance is $4000. Since we are always withdrawing $2300 (plus inflation) from the portfolio, dividends must provide $1700 to reach 4% of the initial balance. If the current balance is $40000, stocks amount to $32000 (which is 80% of $40000). If you get $1700 from $32000 of stocks, the dividend yield must be 5.3125% (or 1700 / 32000) or more.
Finally, I went back to the problem that you are interested in.
If you withdraw 1% of the initial balance (plus inflation) and all of the dividends, your portfolio would never have run out of money (or, at least, for 60 years). Your withdrawals would have been $1000 each year (which is 1% of $100000) plus all dividends. If your current balance ever fell to $40000, it would have required a dividend yield of 9.375% (or 3000 / 32000) to bring total withdrawal amount up to $4000. (That is, $1000 from the balance and $3000 from stock dividends.) If your balance were $80000 or more, stocks would total $64000 (or more) and a dividend yield of 4.6875% (or 3000 / 64000) would be sufficient to bring the total withdrawal amount to $4000 (which is 4% of the $100000 initial balance).
These are in real dollars. They are already adjusted for inflation.
1) With a starting balance of $100000 in 1911, the balance after 10 years (or 1921) would have been $37269. After 20 years (in 1931), the balance would have been $78687. After 30 years (in 1941), the balance would have been $60173.
2) With a starting balance of $100000 in 1929, the balance after 10 years (or 1939) would have been $68221. After 20 years (in 1949), the balance would have been $38672. After 30 years (in 1959), the balance would have been $77998.
3)With a starting balance of $100000 in 1937, the balance after 10 years (or 1947) would have been $51034. After 20 years (in 1957), the balance would have been $82632. After 30 years (in 1967), the balance would have been $110175.
4)With a starting balance of $100000 in 1966, the balance after 10 years (or 1976) would have been $60153. After 20 years (in 1986), the balance would have been $55585. After 30 years (in 1996), the balance would have been $91219.
[more follows]
For Bob_Smith (continued):
In all of these cases the withdrawal amount would have been $2300 plus inflation every year plus all of the dividends on the stocks in the portfolio. Withdrawals from dividends are restricted to the 80% part of the portfolio that consists of stocks and they do not include interest from the commercial paper.
1) The dividend yield in 1911 was 5.07%. In 1921 it was 7.11%. In 1931 it was 6.05%. In 1941 it was 6.38%.
2) The dividend yield in 1929 was 3.46%. In 1939 it was 4.11%. In 1949 it was 6.16%. In 1959 it was 3.16%.
3) The dividend yield in 1937 was 4.15%. In 1947 it was 4.69%. In 1957 it was 3.82%. In 1967 it was 3.41%.
4) The dividend yield in 1966 was 2.94%. In 1976 it was 3.80%. In 1986 it was 3.81%. In 1996 it was 2.26%.
To help you interpret these numbers, remember that my starting balance is $100000, not the $650000 that FIRECalc has as its default value. This means that 4% of the initial balance is $4000. Since we are always withdrawing $2300 (plus inflation) from the portfolio, dividends must provide $1700 to reach 4% of the initial balance. If the current balance is $40000, stocks amount to $32000 (which is 80% of $40000). If you get $1700 from $32000 of stocks, the dividend yield must be 5.3125% (or 1700 / 32000) or more.
Finally, I went back to the problem that you are interested in.
If you withdraw 1% of the initial balance (plus inflation) and all of the dividends, your portfolio would never have run out of money (or, at least, for 60 years). Your withdrawals would have been $1000 each year (which is 1% of $100000) plus all dividends. If your current balance ever fell to $40000, it would have required a dividend yield of 9.375% (or 3000 / 32000) to bring total withdrawal amount up to $4000. (That is, $1000 from the balance and $3000 from stock dividends.) If your balance were $80000 or more, stocks would total $64000 (or more) and a dividend yield of 4.6875% (or 3000 / 64000) would be sufficient to bring the total withdrawal amount to $4000 (which is 4% of the $100000 initial balance).
These are in real dollars. They are already adjusted for inflation.
1) With a starting balance of $100000 in 1911, the balance after 10 years (or 1921) would have been $37269. After 20 years (in 1931), the balance would have been $78687. After 30 years (in 1941), the balance would have been $60173.
2) With a starting balance of $100000 in 1929, the balance after 10 years (or 1939) would have been $68221. After 20 years (in 1949), the balance would have been $38672. After 30 years (in 1959), the balance would have been $77998.
3)With a starting balance of $100000 in 1937, the balance after 10 years (or 1947) would have been $51034. After 20 years (in 1957), the balance would have been $82632. After 30 years (in 1967), the balance would have been $110175.
4)With a starting balance of $100000 in 1966, the balance after 10 years (or 1976) would have been $60153. After 20 years (in 1986), the balance would have been $55585. After 30 years (in 1996), the balance would have been $91219.
[more follows]
Re: SWR should not be constant over a Retirement L
In terms of my initial remarks, for those years that turned out to have the lowest withdrawal rates historically, you would have been able to withdraw 1% (roughly) above the initial dividend yield and then increased that amount to match inflation for the next 30 years.
Finally, I am not sure what you are asking for in your third question:
Along similar lines, you might prefer to withdraw a fraction of the dividends (such as 50% or 75%) instead of all of them.
Have fun.
John R.
In terms of my initial remarks, for those years that turned out to have the lowest withdrawal rates historically, you would have been able to withdraw 1% (roughly) above the initial dividend yield and then increased that amount to match inflation for the next 30 years.
Finally, I am not sure what you are asking for in your third question:
You would have to specify a withdrawal amount of some kind.
Along similar lines, you might prefer to withdraw a fraction of the dividends (such as 50% or 75%) instead of all of them.
Have fun.
John R.
Re: SWR should not be constant over a Retirement L
Thanks John! It appears would it would have been a rough ride with significant fluctuations in income with an 80% stock portfolio.
I'll probably plan, at least initially, to withdraw around 3.5% in the first year, and the same plus inflation in subsequent years, but I'll also implement Gummy's strategy to some extent ("decide upon some minimum annual withdrawal rate ... just enough to pay the bills ...then only withdraw beyond that if the market is good.."). I haven't determined my bread and water budget yet, but I'm working on it. I would gladly experience significant deprivation if the alternative was a return to work.
At this point I'm hoping to buy a boatload of LT individual TIPS if they get back to 2.5% (about 35%) plus 40% TSM, 5% REITS fund, and the rest in a ST Corp Bond fund, CDs, and cash.
Your "3% SWR for 56 years" has great appeal and is a very helpful article. The switching spooks me, but I may go that route at some point, especially if TIPS get back to around 2.5%. It makes a lot of sense and appeals to my relatively low risk tolerance. I have read only about half of the articles at the SWR Research Group, and some of it exceeds the limits of my math skills, but I continue to learn there and I plan to continue reading. Thanks for all of your research - and your willingness to share it.
Thanks John! It appears would it would have been a rough ride with significant fluctuations in income with an 80% stock portfolio.
I'll probably plan, at least initially, to withdraw around 3.5% in the first year, and the same plus inflation in subsequent years, but I'll also implement Gummy's strategy to some extent ("decide upon some minimum annual withdrawal rate ... just enough to pay the bills ...then only withdraw beyond that if the market is good.."). I haven't determined my bread and water budget yet, but I'm working on it. I would gladly experience significant deprivation if the alternative was a return to work.
At this point I'm hoping to buy a boatload of LT individual TIPS if they get back to 2.5% (about 35%) plus 40% TSM, 5% REITS fund, and the rest in a ST Corp Bond fund, CDs, and cash.
Your "3% SWR for 56 years" has great appeal and is a very helpful article. The switching spooks me, but I may go that route at some point, especially if TIPS get back to around 2.5%. It makes a lot of sense and appeals to my relatively low risk tolerance. I have read only about half of the articles at the SWR Research Group, and some of it exceeds the limits of my math skills, but I continue to learn there and I plan to continue reading. Thanks for all of your research - and your willingness to share it.
