Yep, RMD works as well. In the limit, RMD is the same as VPW if you assume a 0% expected return. Because of that, the withdrawal profile looks a bit different in that it has you withdrawing much less in the earlier years than VPW does and quite a bit more in later years.
Yet another withdrawal strategy thread
- Thread starter NameRedacted
- Start date
Yep, RMD works as well. In the limit, RMD is the same as VPW if you assume a 0% expected return. Because of that, the withdrawal profile looks a bit different in that it has you withdrawing much less in the earlier years than VPW does and quite a bit more in later years.
One other thing you can do that I haven't yet played with. Generally with VPW, you set it for a fixed number of years of retirement. Some people add an extra 5 years to it as margin. An alternative would be to start with a fixed number of years, then each year update the remaining number of years based on mortality tables (with margin added)... It's really easy to do with the PMT function in excel.
Pilot2013
Full time employment: Posting here.
Yeah, that makes sense. I guess I wasn't thinking about that as I was looking at it. Thankfully, folks here can point out blind spots.
I would probably word it a little differently. You have to withdraw the full RMD amount required and you will pay taxes on this full amount. After that, any excess can then be deposited into a taxable account for future use.
Every Jan. we sell whatever we expect to need for the upcoming year and rebalance as needed. So, at the beginning of the year, we'll have almost 2 year's of expenses in cash, mainly in a 1% savings account. Ever month or two, we'll transfer to a local checking account to pay bills.
I can't see selling investments every month. It would be very expensive for us and would be a ton of w*rk and stress. I do a monthly rollup of expenses and investible assets, but this is mainly out of curiosity and to verify transactions in major accounts. I thought of changing to only a quarterly rollup, but I like watching my investments, at least during a bull market.
I can't see selling investments every month. It would be very expensive for us and would be a ton of w*rk and stress. I do a monthly rollup of expenses and investible assets, but this is mainly out of curiosity and to verify transactions in major accounts. I thought of changing to only a quarterly rollup, but I like watching my investments, at least during a bull market.
LOL!
Give me a museum and I'll fill it. (Picasso) Give me a forum ...
- Joined
- Jun 25, 2005
- Messages
- 10,252
Since the OP says they like math, I suggest they read this series on sustained withdrawal rates:
https://earlyretirementnow.com/2016/12/07/the-ultimate-guide-to-safe-withdrawal-rates-part-1-intro/
I think the OP's method may even have been covered somewhere in those articles.
https://earlyretirementnow.com/2016/12/07/the-ultimate-guide-to-safe-withdrawal-rates-part-1-intro/
I think the OP's method may even have been covered somewhere in those articles.
NameRedacted
Recycles dryer sheets
- Joined
- Oct 23, 2016
- Messages
- 236
Thanks for all the responses guys! Lots of good ideas as usual. I really appreciate the wealth of experience and fresh ideas.
I'll do one big post to respond to all the different points.
1. Several people suggested it is all too complicated or over-detailed.
Well, this would only take 10-15 minutes to set up in a spreadsheet and after that there would be very little additional work. Also, a couple of people implied it sounds too much like work - but in my opinion this is the interesting kind of work. If all I had to do all day was stuff like this I wouldn't be retiring. This is way better than going to meetings and having to repeatedly explain how and why everything works the way it does to the nuance-deficient guys.
Additionally, I'm a firm believer in the need to keep my mind active in retirement. So, I'm not going to shy away from complexity.
2. There were a few comments about withdrawing monthly instead of annually and how that would increase costs.
What I'm trying to do is keep money in the market as long as possible on the grounds that on average it grows x% a year where x is usually positive. I think this outweighs the extra 'trading' costs. Most of the money on the equity side of the portfolio is invested in individual stocks with an average of about $10,000 in each one. Trading costs are $10 per buy/sell. So 12 trades a year would be $120. That's less than 0.01% of the portfolio. Granted, it's more than if I sold a year's worth of stock all at once but I'm expecting to gain more than that by leaving the money in the market longer. Also, and this is a little dig at those people that keep their money in low cost index funds, 0.01% is way less than the expense ratios of index funds.
3. I see various people suggesting lower withdrawal rates.
One person even mentioned that they are withdrawing 2%. Unless you're planning to leave a lot to your kids or want to create some kind of charitable fund that is way to low. 4% is supposed to survive the worst possible scenarios over the last 100 years or so with a very high probability of success. I'm not sure what's to be gained by living on half that for the rest of my life just in case some extremely low probability event occurs and even if that were to happen - I'd still be able to fall back on social security AND before anyone tells me that I'll never see a dime of social security, people have been telling me that forever. I just don't believe them.
Now I see other people are withdrawing amounts between 3-4% and those seem more reasonable. It would be interesting to know from anyone that has been doing this a while how that is working out for you. Are you finding that your portfolio is growing really large or that you are scrimping to get by on 3% or that 3% is plenty and you don't really have anything you want to spend the additional money on that 4% would provide?
I did see OverThinkMuch's post and it seems to be working for him but 2.5% is too low for me. If I ever get down to about 60-65% of my initial draw I'd be looking for a part time job.
It's a bit paradoxical really in that we'd like to travel earlier in retirement while we can but doing that requires more money. So I need a higher withdrawal rate now than later, but logic tells me that it's safer to withdraw a higher rate later when there are fewer years left for it to all go pear-shaped.
This brings me to the one thing that does concern me about my plan. I think I'm probably adjusting the withdrawal rate upwards too aggressively early on. A 4.68% rate with 30 years still to go and a 5.4% rate with 20 years left both seem high to me. I think I may dial those down a bit.
And on that topic Rodent's suggestion provides food for thought.
4. Vanguard floor to ceiling dynamic spending and VPW
I read over the 'links' provided by Rodent and they were really useful. Thankyou very much for that. I also looked at VPW suggested by big-papa which is also extremely useful.
It seems like I'm using the floor and ceiling method with a very high range and a monthly withdrawal strategy rather than a yearly strategy. The Vanguard article has a range of 2.5% as the floor and a ceiling of 5% but later on discusses a range of 10% either way which is close enough to mine to garner useful data. That results in no failures but 25% of the time the withdrawal amount is 2/3 or less of the initial withdrawal in real terms. That makes me think that I might have better chances limiting the ceiling to a value lower than 12%/year, perhaps 6% but keeping the floor at 12%.
Furthermore as an added bonus it discusses 'safe' withdrawal rates using the floor/ceiling methods for varying retirement lengths and AAs. I put "safe" in quotes because it sees an 85% failure rate with these numbers. It doesn't quite cover my allocation which is 70/30 but is close enough to get a feel as to what to expect.
However, this Bogleheads link - thanks big-papa has completely different and more comprehensive numbers:
Now when I look at the VPW table here:
https://www.bogleheads.org/wiki/Var...able-percentage_withdrawals_during_retirement
These are the two sets of numbers compared to mine. I am 55 by the way.
40 years: Vanguard: 4.1% Bogleheads: 4.8% mine: 3.96%
30 years: Vanguard: 4.8% Bogleheads: 5.3% mine: 4.68%
20 years: Vanguard: 6.3% Bogleheads: 6.3% mine: 6.12%
10 years: Vanguard: 10.5% Bogleheads: 8.8% mine: 6.84%
Additionally, I'm adjusting my rate every month and I don't think they are doing that at all, which should work against my method. On the other hand Vanguard are using a more constrained ceiling and floor and Bogleheads are using none at all.
That's quite a large difference between the two at least initially. I think that I'd prefer a higher safety margin than 85% so based solely on the Vanguard information I'm thinking of lowering the ceiling to 6%/year (0.5%/month) but leave the floor at 12%, plus I think I need to reduce the change in the rate of withdrawal during the early retirement years and increase it during the later retirement years. So instead of increasing the rate by 0.0005% a month I think a better bet would be:
0.00025%/month for years 1-10
0.00050%/month for years 11-20
0.00075%/month for years 21-30
0.00100%/month for years 31-40
This would mean the following withdrawal rates
Year 0 - 3.96% /year
Year 10 - 4.32% /year
Year 20 - 5.04% /year
Year 30 - 6.12% /year
Year 40 - 7.56% /year
For anyone that's still following me - I calculated that 0.00025% of my current portfolio is $3.60 so this would give me roughly an additional $36/year to spend after 1 year.
However, the Bogleheads information implies this is unnecessary and that I should be fine. Although that methodology does imply that you should have a base income from SS or pensions which I don't have yet. So. I think I may just make the adjustments derived from the Vanguard data.
5. Variable withdrawal vs fixed spending
While this may seem to be a problem I don't think it is. many people have variable income in real life. Some people have variable overtime or variable commission. Others run a small business with seasonal income. Others have a spouse that takes a part-time job for a while. Still others lose their job or get promoted. It's just life. People cope. In many respects we are better off than the average guy. A huge war chest is a sobering factor to any worries on that score. I agree with audreyh1 on this I think that keeping a year's worth of expenses readily accessible should help with any short sharp shocks.
I'll do one big post to respond to all the different points.
1. Several people suggested it is all too complicated or over-detailed.
Well, this would only take 10-15 minutes to set up in a spreadsheet and after that there would be very little additional work. Also, a couple of people implied it sounds too much like work - but in my opinion this is the interesting kind of work. If all I had to do all day was stuff like this I wouldn't be retiring. This is way better than going to meetings and having to repeatedly explain how and why everything works the way it does to the nuance-deficient guys.
Additionally, I'm a firm believer in the need to keep my mind active in retirement. So, I'm not going to shy away from complexity.
2. There were a few comments about withdrawing monthly instead of annually and how that would increase costs.
What I'm trying to do is keep money in the market as long as possible on the grounds that on average it grows x% a year where x is usually positive. I think this outweighs the extra 'trading' costs. Most of the money on the equity side of the portfolio is invested in individual stocks with an average of about $10,000 in each one. Trading costs are $10 per buy/sell. So 12 trades a year would be $120. That's less than 0.01% of the portfolio. Granted, it's more than if I sold a year's worth of stock all at once but I'm expecting to gain more than that by leaving the money in the market longer. Also, and this is a little dig at those people that keep their money in low cost index funds, 0.01% is way less than the expense ratios of index funds.
3. I see various people suggesting lower withdrawal rates.
One person even mentioned that they are withdrawing 2%. Unless you're planning to leave a lot to your kids or want to create some kind of charitable fund that is way to low. 4% is supposed to survive the worst possible scenarios over the last 100 years or so with a very high probability of success. I'm not sure what's to be gained by living on half that for the rest of my life just in case some extremely low probability event occurs and even if that were to happen - I'd still be able to fall back on social security AND before anyone tells me that I'll never see a dime of social security, people have been telling me that forever. I just don't believe them.
Now I see other people are withdrawing amounts between 3-4% and those seem more reasonable. It would be interesting to know from anyone that has been doing this a while how that is working out for you. Are you finding that your portfolio is growing really large or that you are scrimping to get by on 3% or that 3% is plenty and you don't really have anything you want to spend the additional money on that 4% would provide?
I did see OverThinkMuch's post and it seems to be working for him but 2.5% is too low for me. If I ever get down to about 60-65% of my initial draw I'd be looking for a part time job.
It's a bit paradoxical really in that we'd like to travel earlier in retirement while we can but doing that requires more money. So I need a higher withdrawal rate now than later, but logic tells me that it's safer to withdraw a higher rate later when there are fewer years left for it to all go pear-shaped.
This brings me to the one thing that does concern me about my plan. I think I'm probably adjusting the withdrawal rate upwards too aggressively early on. A 4.68% rate with 30 years still to go and a 5.4% rate with 20 years left both seem high to me. I think I may dial those down a bit.
And on that topic Rodent's suggestion provides food for thought.
4. Vanguard floor to ceiling dynamic spending and VPW
I read over the 'links' provided by Rodent and they were really useful. Thankyou very much for that. I also looked at VPW suggested by big-papa which is also extremely useful.
It seems like I'm using the floor and ceiling method with a very high range and a monthly withdrawal strategy rather than a yearly strategy. The Vanguard article has a range of 2.5% as the floor and a ceiling of 5% but later on discusses a range of 10% either way which is close enough to mine to garner useful data. That results in no failures but 25% of the time the withdrawal amount is 2/3 or less of the initial withdrawal in real terms. That makes me think that I might have better chances limiting the ceiling to a value lower than 12%/year, perhaps 6% but keeping the floor at 12%.
Furthermore as an added bonus it discusses 'safe' withdrawal rates using the floor/ceiling methods for varying retirement lengths and AAs. I put "safe" in quotes because it sees an 85% failure rate with these numbers. It doesn't quite cover my allocation which is 70/30 but is close enough to get a feel as to what to expect.
However, this Bogleheads link - thanks big-papa has completely different and more comprehensive numbers:
Now when I look at the VPW table here:
https://www.bogleheads.org/wiki/Var...able-percentage_withdrawals_during_retirement
These are the two sets of numbers compared to mine. I am 55 by the way.
40 years: Vanguard: 4.1% Bogleheads: 4.8% mine: 3.96%
30 years: Vanguard: 4.8% Bogleheads: 5.3% mine: 4.68%
20 years: Vanguard: 6.3% Bogleheads: 6.3% mine: 6.12%
10 years: Vanguard: 10.5% Bogleheads: 8.8% mine: 6.84%
Additionally, I'm adjusting my rate every month and I don't think they are doing that at all, which should work against my method. On the other hand Vanguard are using a more constrained ceiling and floor and Bogleheads are using none at all.
That's quite a large difference between the two at least initially. I think that I'd prefer a higher safety margin than 85% so based solely on the Vanguard information I'm thinking of lowering the ceiling to 6%/year (0.5%/month) but leave the floor at 12%, plus I think I need to reduce the change in the rate of withdrawal during the early retirement years and increase it during the later retirement years. So instead of increasing the rate by 0.0005% a month I think a better bet would be:
0.00025%/month for years 1-10
0.00050%/month for years 11-20
0.00075%/month for years 21-30
0.00100%/month for years 31-40
This would mean the following withdrawal rates
Year 0 - 3.96% /year
Year 10 - 4.32% /year
Year 20 - 5.04% /year
Year 30 - 6.12% /year
Year 40 - 7.56% /year
For anyone that's still following me - I calculated that 0.00025% of my current portfolio is $3.60 so this would give me roughly an additional $36/year to spend after 1 year.
However, the Bogleheads information implies this is unnecessary and that I should be fine. Although that methodology does imply that you should have a base income from SS or pensions which I don't have yet. So. I think I may just make the adjustments derived from the Vanguard data.
5. Variable withdrawal vs fixed spending
While this may seem to be a problem I don't think it is. many people have variable income in real life. Some people have variable overtime or variable commission. Others run a small business with seasonal income. Others have a spouse that takes a part-time job for a while. Still others lose their job or get promoted. It's just life. People cope. In many respects we are better off than the average guy. A huge war chest is a sobering factor to any worries on that score. I agree with audreyh1 on this I think that keeping a year's worth of expenses readily accessible should help with any short sharp shocks.
pb4uski
Give me a museum and I'll fill it. (Picasso) Give me a forum ...
ERed at 56 five years ago. Our withdrawals have been between 4.5%-5.5% of our beginning of year balance over the last few years.
I assess the risk of financial ruin by focusing on our ultimate withdrawal rate and periodic re-analyses of whether we still "have enough". Using Quicken Lifetime Planner, I can look at the projected withdrawals once our SS and my pension are online in relation to the projected amount of our nestegg... once those flows start our WR (current year withdrawal divided by prior year end balance) moderates from a projected 4.8% in 2018 to about 2.8%... very comfortable for a 66 yo IMO.
As far as withdrawals, I take our taxable account distributions in cash and a monthly transfer from our online savings account (the cash part of our 60/35/5 AA) and then occasionally additional transfers only if our local bank account is getting low (under $5k).
I also do an analysis annually whether our retirement assets will support our life plans and lifestyle using Firecalc, Quicken Lifetime Planner, Financial Engines and other similar tools.
Our real life experience from the last 5 years has been unbelievable. Our portfolio has supported us, plus buying a winter condo in cash, plus buying two new cars, building a new two-car garage with bonus attic loft, etc. and our net retirement assets (retirement accounts less mortgage, car loans) is 8% higher than when I retired (17% if I include the value of the winter condo in our retirement assets).
I'll likely ratchet up our spending (more splurges in everyday spending, more travel, more gifts and charitable giving) some between now and 70 as investment performance allows to avoid leaving the kids a boatload of money.
Last edited:
Regarding VPW, there are a number of things to understand about it.
1. The table given over at bogleheads assumes only certain, limited assets for a portfolio. If you have anything other than what they have, you'll need to create your own backtest.
The formula for the % withdrawal in any given year in Excel is given by
PMT(expected return, # years remaining, -1, 0, 1)
2. What to use for expected return?
There was a large discussion over there about this. The author himself never really believed that this should be "expected return" but rather a "reasonable return assumption". Others, including myself, believe that expected return is what goes here. What can you use for this?
- average return as far back as you have data for
- 1/CAPE for the stock portion of your portfolio and best guess for bond portion, though there are many models out there for it as well.
- You can also update the expected return every year if that's what you want to do. For example, by using something like 0.8/CAPE each year, you'll find that withdrawals are somewhat smoothed vs. the pure form of VPW
- You can add, say, 0.5% for the first 10 years expected return to have a little more (on average) to spend on travel, etc.
- What I did once was a backtest for a retirement starting in the late 1960's for 40 years. Adjusted the expected returns such that the first year's withdrawal and final withdrawal were about the same. This is the same sequence of return that's known to be "worst case" for the SWR method. In the case of VPW, this results in largest ratio of maximum withdrawal to minimum withdrawal, so you have a feel for what the lowest withdrawal might ever be based on history. Could it be worse? Sure, no guarantees and worse means that other withdrawal methods are also affected strongly.
- Because whatever you put in will never be exactly what happens to you in real life, if you undershoot what happens is that you have lower withdrawals in early retirement, but higher withdrawals later in retirement vs. the nominal case. Vice versa if you chose too high an expected return. This system always guarantees that you do not run out of money before the last year. What it doesn't guarantee is the smoothest ride possible to get there nor will it guarantee that the withdrawal in the worst case year will be enough to live on. No variable method can guarantee that.
3. There is nothing preventing you from using this method to calculate the monthly withdrawal. In the PMT formula, change the number of periods and interest rate to reflect a monthly withdrawal.
4. You can adjust with expected return to perfectly match the first withdrawal of the Vanguard floor/ceiling method and then compare that method vs. any form of VPW over any historical period you like.
Last edited:
Lsbcal
Give me a museum and I'll fill it. (Picasso) Give me a forum ...
NameRedacted, I posted a modified VPW model just now. Have you considered taking your model and using the VPW spreadsheet to simulate for a bad period such as starting in 1966?
P.S. I haven't read this thread entirely as I just got back from a vacation.
P.S. I haven't read this thread entirely as I just got back from a vacation.
Last edited:
+1 See my posting just above yours....
oscar1
Recycles dryer sheets
- Joined
- Jul 25, 2013
- Messages
- 139
Like the Boggle-heads simplistic approach. Every Friday morning $xxx.xx shows up as an auto deposit from my Schwab account into my bank debit account from which bills, eating out, most all living expenses are paid through out the week. Excess is moved to a bank savings account which serves as the emergency account. $xxx.xx is estimated annually in December based on current run rate. At times when the market is exploding, I stock up on cash because the inevitable is coming. I could improve my margins with some risk but I like to sleep well at night knowing I have enough cash to ride out a 2 year recession should it arrive. All non-cash investments are in diversified index funds which should last several decades as long as weekly $xxx.xx does not get out of control. I also give myself a nice raise and bonus each January just to stick to the company's I labored over for so many years (no anomosity there).
urn2bfree
Full time employment: Posting here.
- Joined
- Feb 14, 2011
- Messages
- 852
I recall somebody advocating a simpler system of taking the arithmetic mean of the classic Trinity 4% rule and the RMD amount each year.
pb4uski
Give me a museum and I'll fill it. (Picasso) Give me a forum ...
That sounds good on average. 
Chuckanut
Give me a museum and I'll fill it. (Picasso) Give me a forum ...
Good choice since you have included one of those relentless, powerful Bear markets that takes no prisoners and doesn't give a d#!&$m about your plans and the way the experts think it should be 'this time'.
Chuckanut
Give me a museum and I'll fill it. (Picasso) Give me a forum ...
I do something similar, and it's about as close as I will get to trying to 'time the market'.
This year, with the market indexes bumping up against new highs, I took out more than I will need. It goes into my pile of spendable cash, but it won't be spent (I hope) this year. So, if the market decides to drop by 20% before I do my next annual withdrawal, I will have it there to buffer my spending decrease. I lived through the 73-74 crash (a whopping 45% drop for those of you who have forgotten). It took 20 years for the market to recover on an inflation adjusted basis. A few years later, we saw a 20% drop in one day. Are we having fun yet?
On important factor that many models don't consider is that I am not getting any younger or healthier. If the market decides to take a huge haircut for the next 2-6 years, I don't want to stop doing the things that I may not be able to do when the market recovers. So, I keep a big buffer. That may be a drag on returns, but it also insures that I can do many things I want to do while I have the health and vigor to do them.
Last edited:
