The Variable Savings Rate (VSR) -- an accumulation-time prelude to VPW

Non-investing personal finance issues including insurance, credit, real estate, taxes, employment and legal issues such as trusts and wills
Post Reply
Topic Author
longinvest
Posts: 4710
Joined: Sat Aug 11, 2012 8:44 am

The Variable Savings Rate (VSR) -- an accumulation-time prelude to VPW

Post by longinvest »

NOTE

The VSR spreadsheet, called "Accumulation Worksheet", is now part of the VPW Accumulation And Retirement Worksheet.

INSTRUCTIONS

See this post.

LINKS

The spreadsheet can be used online or downloaded.

The links can be found on the Boglehead wiki VPW page: VPW Accumulation And Retirement Worksheet.

SCREENSHOT

Image


A frequent question, in our forums, is "How much should I save?" A good answer to this is "As much as you can, but not so much as to live like a pauper." But, this is imprecise.

In a recent thread, I proposed an interesting calculation that leads to a variable savings rate which adapts to the saver's situation, retirement horizon, and current portfolio balance. Here's a copy the main parts of my post.

Let's pick a family with two earners making $31,000 each, for a household income of $62,000. That's the approximate US median household income.

In their 20s, they pay down debt, accumulate a down payment, and buy a modest house at the end of their 20s with a 30-year mortgage. They don't start saving and investing before the age of 30. Then, they work, live below their means, save, and invest from age 30 to 64. At age 65, they retire. Each would qualify for a annual $15,500 Social Security pension at age 66, but they decide that one of them will claim at age 65 and get $14,000 and the other will delay until 70 to get $20,000.

We'll assume, for simplicity, that their expenses are relatively level all along. Their mortgage is modest, and once it's paid, an equivalent HELOC is used to spread the cost of maintenance (kitchen remodel, roof, etc.).

How much should they save to retire with dignity? I'll use a similar calculation to what I've shown in the post The Mathematics of Retirement Investing.

One of the spouses likes aggressive investing, but the other is risk averse. They compromise to holding a 50/50 stocks/bonds portfolio (with stocks subdivided between domestic and international, and bonds subdivided between nominal and inflation-indexed) all life long, during both work years and retirement. They know that market returns fluctuate, but, for planning purpose, they use a real 3.5% growth trend for such a balanced portfolio.

To retire with dignity, they'll need to:
  • Accumulate ($20,000 X 5) = $100,000 to fill the gap in Social Security payments from age 65 to 69.
  • Save and invest enough in a tax-deferred accounts to equalize ($62,000 - savings) with ($14,000 + $20,000 + portfolio withdrawals).
At a real 3.5% growth rate, investing $1000 per year for 35 years, from age 30 to 64, would grow to a total of $66,674 at age 65. Accumulating $100,000 requires $1,500 in annual savings and investing. At age 65 with a 50/50 portfolio, the VPW table allows for a 4.8% withdrawal percentage. Multiplying $66,674 by 4.8% gives $3,200. In other words, each additional $1,000 in annual savings would project into portfolio withdrawals fluctuating from $3,200 in retirement; that's 3.2 times the annual savings.

So we want to find the additional savings such that ($62,000 - $1,500 - additional savings) = ($14,000 + $20,000 + 3.2 X additional savings).

Code: Select all

S = additional savings
$62,000 - $1,500 - S = $14,000 + $20,000 + 3.2 S
$60,500 - S = $34,000 + 3.2 S
$60,500 - $34,000 = 3.2 S + S
$26,500 = 4.2 S
S = $26,500 / 4.2 = $6,310
So, the household must save ($1,500 + $6310) = $7,810 annually into a tax-deferred account, from 30 to 64, living on the equivalent of a ($62,000 - $7,810) = $54,190 pre-tax income to preserve a relatively similar (but fluctuating) standard of living in retirement.

What the calculations above are suggesting is a dynamic savings rate. It projects variable percentage withdrawals back into accumulation years. Similar to VPW, it requires annual calculations without taking the past into account. Let me illustrate this.

At age 30, starting with a $0 portfolio, according to the above calculations, they need to save $7,810, or $300/bi-weekly pay, into a tax-deferred account. That's a 12.6% savings rate.

Every year, they redo the above calculations based on their current portfolio balance and current income, using an updated investment horizon. At age 35, for example, maybe they've been unlucky and their portfolio has only grown to $37,500 (inflation-adjusted) . The household income is still $62,000. Let's do the calculations, keeping the constant real 3.5% portfolio growth trend but for a 30 year investment horizon from age 35 to 64.

In 30 years, $37,500 would grow to $105,255. This would cover the Social Security gap and allow for an additional ($5,255 X 4.8%) = $252 VPW portfolio withdrawal. In 30 years, investing $1000 per year would grow to a total of $51,623 and allow for a ($51,623 X 4.8%) = $2,478 VPW withdrawal.

We want to find the additional savings such that ($62,000 - additional savings) = ($14,000 + $20,000 + $252 + 2.478 X additional savings).

Code: Select all

S = additional savings
$62,000 - S = $14,000 + $20,000 + $252 + 2.478 S
$62,000 - S = $34,252 + 2.478 S
$62,000 - $34,252 = 2.478 S + S
$27,748 = 3.478 S
S = $27,748 / 3.478 = $7,978
So, the household must save $7,978, or $307/bi-weekly pay, at age 35 into a tax-deferred account, living on the equivalent of a ($62,000 - $7,978) = $54,022 pre-tax income. That's a 12.9% savings rate.

See how the savings rate fluctuates according to market returns. Lower market returns lead to a slightly higher savings rate, naturally letting them buy slightly more investment assets when they're down. Higher market returns would lead to a slightly lower savings rate, naturally letting them buy slightly less investment assets when they're up.

Note that there's no need to predict future returns. Instead, I suggest using VPW's growth trend. It's actually important to use a constant growth trend. Saving more when markets are high, and less when they're low (as would happen if one used "good" future return predictions*) wouldn't be a smart thing to do!

* I personally don't believe that anybody is able to make good future return predictions.

Of course, many will consider it generally imprudent to target a retirement age of 65, because one or both spouses could lose their job before that, or they could wish to retire earlier. That's fine. One could easily redo the above calculations with an earlier target retirement age. Note, too, that the above calculations determine a minimal savings rate. One could save more, too, but this would be a choice of targeting a higher standard of living in retirement than when working.
Last edited by longinvest on Sat Feb 08, 2020 9:28 am, edited 12 times in total.
Bogleheads investment philosophy | One-ETF global balanced index portfolio | VPW
Topic Author
longinvest
Posts: 4710
Joined: Sat Aug 11, 2012 8:44 am

Re: The Variable Savings Rate (VSR) -- an accumulation-time prelude to VPW

Post by longinvest »

Of course, the above calculations apply to a single person too. Also, the above couple could have chosen to skip on home ownership and simply rented all lifelong. This would have allowed them to start saving and investing in their twenties, instead of putting all the money on the house down payment.
Bogleheads investment philosophy | One-ETF global balanced index portfolio | VPW
User avatar
Svensk Anga
Posts: 887
Joined: Sun Dec 23, 2012 5:16 pm

Re: The Variable Savings Rate (VSR) -- an accumulation-time prelude to VPW

Post by Svensk Anga »

Your example supports the fairly common recommendation to save about 12%. Nice.

I see a couple possible tweaks:

Your example couple gets a raise in after-tax income upon retirement for two reasons. 1.) FICA taxes stop along with earned income. 2.) Their SS income should be at least partially tax-free. So a somewhat lower savings rate is possible if after-tax income is to be held constant.

I would have accelerated the savings rate when the mortgage was paid off. (In fact, I did.) But it is understandable if they have deferred spending desires that they can now satisfy in their remaining working years and into retirement. The example has some lifestyle inflation when the house down payment saving and mortgage payments stop. This lifestyle inflation is likely significant for the median income couple paying off a house, but maybe not so much for higher earners who were satisfied with modest house.

The example given works for the median income couple (or single). For those of us well beyond the second SS bend point, SS will replace less of the former salary and so the saving rate will have to be higher.
Topic Author
longinvest
Posts: 4710
Joined: Sat Aug 11, 2012 8:44 am

Re: The Variable Savings Rate (VSR) -- an accumulation-time prelude to VPW

Post by longinvest »

Svensk Anga wrote: Fri Nov 16, 2018 2:01 pm The example given works for the median income couple (or single). For those of us well beyond the second SS bend point, SS will replace less of the former salary and so the saving rate will have to be higher.
The method is meant to adapt to the couple's situation over time. Every year, they redo the calculations using their current salaries and portfolio balance, and get a new savings rate. So, if their salaries increase (faster than inflation), their savings rate will increase accordingly.

It's true that the method is conservative. Equating gross salary minus tax-deferred savings to retirement income will possibly lead to a higher net income in retirement, but the volatility of market returns makes fun of any attempt at false precision. I think that the calculations are good enough.

The important thing is that it's simple and adaptive. If one spouse unexpectedly loses his job, VSR will determine a new savings rate and result into a smaller spending budget. When a new job is found, VSR will provide a new appropriate savings rate, taking into account the updated salaries and portfolio balance.
Bogleheads investment philosophy | One-ETF global balanced index portfolio | VPW
Topic Author
longinvest
Posts: 4710
Joined: Sat Aug 11, 2012 8:44 am

Re: The Variable Savings Rate (VSR) -- an accumulation-time prelude to VPW

Post by longinvest »

I have uploaded version 1.0 of the Variable Savings Rate spreadsheet.

It is available online and to download from the links in the first post.

Main changes:
  • Initial version.
As usual, comments are welcome.

Enjoy!
Bogleheads investment philosophy | One-ETF global balanced index portfolio | VPW
Topic Author
longinvest
Posts: 4710
Joined: Sat Aug 11, 2012 8:44 am

Re: The Variable Savings Rate (VSR) -- an accumulation-time prelude to VPW

Post by longinvest »

The Variable savings rate (VSR) spreadsheet is easy to use:

Each year of accumulation:
  1. The following data must be provided:
    • Age, salary, as well as portfolio balance and stock allocation.
    • Desired retirement age.
    • Defined Benefit Pension #1: Social Security annual contribution (OASI tax), start age of payments, and monthly payment amount, all of which can be estimated using Neurosphere's Social Security Estimator.
    • Optional: Defined Benefit Pensions #2 and #3 information.
    • Savings frequency: Monthly, semimonthly, or biweekly.
  2. The VSR spreadsheet calculates:
    1. The amount that the investor must systematically add to the portfolio on the selected frequency (monthly, semimonthly, or biweekly).
    2. The required flexibility that the investor must exhibit given the current asset allocation. In particular, the spreadsheet calculates:
      • The impact, on the portfolio, of a 50% loss in stocks.
      • The additional savings that the investor must be able to do (e.g. implying a reduction in spending budget) after a market downturn.
    3. The projected annual retirement income (before and after loss).
Every year, a new savings amount is calculated, based on the evolving situation of the investor such as age, salary, and portfolio balance.

It is important for the investor to consider the required flexibility information provided in the spreadsheet, and keep enough room in the spending budget to easily reduce expenses by the suggested amount in the future, after a market downturn.

Calculations must imperatively be repeated every year, as well as everytime a significant change happens to the investor's situation (like an increase in salary).

Here's a screenshot:

Image

We see that the spreadsheet suggests, in the illustrated case, to save and invest in the portfolio $556 per month. This represents a (($556 X 12) / $40,000) = 17% savings rate. The spreadsheet also suggests to keep enough flexibility in the budget to easily reduce spending and increase savings by $25 per month.

The spreadsheet also provides a projected annual retirement income: $31,213. This is the equilibrium income. It's equal to the projected retirement income composed of Social Security and variable portfolio withdrawals. It's also the effective income available this year after removing savings and Social Security tax: ($40,000 - ($556 X 12) - $2,120).

Note that $31,213 is a pre-tax income. Part of this income will be used to pay taxes. What remains can be used for other expenses. During accumulation, it is assumed that savings are invested within a tax-deferred account, such as a 401K or an IRA account.

Additional assumptions:
  • It is assumed that asset allocation remains unchanged during accumulation and retirement.
  • During retirement, it is assumed that the investor will use the Variable-percentage withdrawal (VPW) method to take withdrawals from the stocks/bonds portfolio.
Lastly, it's the opinion of the author of these lines that a US investor could consider adopting a very simple, yet robust strategy: (1) invest all savings into a globally diversified all-in-one balanced index fund or ETF (60/40 stocks/bonds) like Vanguard's LifeStrategy Moderate Growth Fund, (2) use the VSR spreadsheet once a year to determine the new amount of regular savings, and (3) set automatic contributions to retirement accounts on payday accordingly. As our mentor Jack Bogle used to say: "When there are multiple solutions to a problem, choose the simplest one."

Links to the VSR spreadsheet are found in this post.

I invite forum members to try it and provide their feedback on this thread.
Last edited by longinvest on Sun May 26, 2019 11:56 pm, edited 2 times in total.
Bogleheads investment philosophy | One-ETF global balanced index portfolio | VPW
AlohaJoe
Posts: 6038
Joined: Mon Nov 26, 2007 2:00 pm
Location: Saigon, Vietnam

Re: The Variable Savings Rate (VSR) -- an accumulation-time prelude to VPW

Post by AlohaJoe »

I think it is a good approach. I am always somewhat disappointed how little good guidance there is for savers. As you say, it is usually vague advice like "as much as you can but not too much".

I have a similar-ish spreadsheet I've been creating.

https://docs.google.com/spreadsheets/d/ ... sp=sharing

My main motivation for it was to include a few things that I felt simpler suggestions often lacked: Couples can have different ages, different retirement dates, different incomes, different Social Security amounts, and different Social Security claiming dates.

I said mine "similar-ish" because there are a few points where I took a different approach:
  • I use the current TIPS 30-year real yield for the discount rate (it is pulled dynamically from the Treasury website in cell A35).
  • I rely on annuity pricing (cell A20) rather than VPW
There's a "cake chart" to help readers visualise the sources of income each year (i.e. the layers of the cake). There's also a simulation tab showing how thing would have historically played out for someone following the advice of the spreadsheet.

This is an example of what the "cake chart" looks like:

Image

And a prototypical case is:
  • Person A retires, while Person B decides to keep working a few more years. Person A does not claim Social Security yet. Bridge #1 is required to make up the income shortfall from Person A no longer having a salary.
  • Person B also retires. Neither one is claiming Social Security yet, so Bridge #2 is required to make up the full income shortfall.
  • Person B claims Social Security; Person A continues to defer. Bridge #3 is required to make up the shortfall between one Social Security payment and the full income desired.
  • Finally, Person A also claims Social Security. This is "steady state" and I assume they buy an annuity to make up the income shortfall between their two Social Security payments and their desired income.
While I assume the entire portfolio is in TIPS & annuities, it is important that readers keep in mind you don't have to invest in those things. But the spreadsheet is telling you the risk-free cost of retirement. If you want to take on risk -- by saving less or investing in riskier things -- that's perfectly fine. One secondary goal of the spreadsheet was to help make more explicit how much risk someone is taking when the embark on a savings plan. If the spreadsheet says "you need to save $14,000 this year to avoid risk in your retirement plan" and you only save $12,000 then you know you are saying "I hope the market bails me out for that extra $2,000!"

Conversely, it also shows how expensive it is to remove risk. Being risk-free is very expensive!

Readers may notice that my spreadsheet gives fairly high savings rates: $16,000 a year at age 23 (albeit with a goal of early retirement at age 55). I added a Simulation tab that shows those kinds of numbers are actually pretty plausible, even when invested in a 60/40 portfolio (i.e. they do not result in dramatic oversaving).

I was mostly playing around with the spreadsheet for my own edification. But I discovered/realised how much complexity there is even in my simplified scenario. What happens if Person A doesn't retire entirely but continues to work part-time? What happens if there are pensions involved? What if one person is adjusted by inflation and the other isn't? What happens if expenses are much higher for the first few years of retirement and then drop off? (Maybe that is showing the shift from paying expensive private health insurance to qualifying for Medicare. Maybe that shows the last few years of paying off a mortgage or college education for children.)

It also, for me, drove home how financially challenging it is to retire early or to ignore Social Security in one's planning. It is easy to say those kind of things when a family is making $150,000 or $250,000 a year. It is much harder even on $80,000 a year.
Topic Author
longinvest
Posts: 4710
Joined: Sat Aug 11, 2012 8:44 am

Re: The Variable Savings Rate (VSR) -- an accumulation-time prelude to VPW

Post by longinvest »

Aloha,
AlohaJoe wrote: Sat May 25, 2019 9:53 am I said mine "similar-ish" because there are a few points where I took a different approach:
  • I use the current TIPS 30-year real yield for the discount rate (it is pulled dynamically from the Treasury website in cell A35).
  • I rely on annuity pricing (cell A20) rather than VPW
If you invest your money into fluctuating assets and want to use "future return predictions" as discount rate, you have to predict their future returns over your various horizons which can easily span over more than 50 or 60 years. Nobody can do that. Even the longest-maturity TIPS is limited to 30 years, and its total real return isn't known (even within a tax-sheltered account) because of the uncertainty of coupon reinvestment rates. As for annuity prices, they fluctuate over decades.

The VSR and VPW approach is to (1) acknowledge that future returns are uncertain and make no attempt at predicting them, (2) use a "not too bad" but constant growth trend, which is a wild-ass guess (WAG) of a middle-of-the-road portfolio return (higher than low returns, lower than high returns) to calibrate savings or withdrawals, and (3) simply adjust savings or withdrawals yearly as events unfold.

In the new VSR spreadsheet, I've included an estimate of the short-term budget flexibility required due to the investor's asset allocation choice. It's impossible to accurately predict long-term flexibility requirements. Anyway, humans have the ability to adapt over time.

If you wish to explain your method and spreadsheet in details, I invite you to start a new thread about it (you can link to it from this thread, of course). I'd like to keep this thread targeted at helping users of the VSR spreadsheet and improving it.

Cheers,

longinvest
Bogleheads investment philosophy | One-ETF global balanced index portfolio | VPW
Topic Author
longinvest
Posts: 4710
Joined: Sat Aug 11, 2012 8:44 am

Re: The Variable Savings Rate (VSR) -- an accumulation-time prelude to VPW

Post by longinvest »

AlohaJoe wrote: Sat May 25, 2019 9:53 amI was mostly playing around with the spreadsheet for my own edification. But I discovered/realised how much complexity there is even in my simplified scenario. What happens if Person A doesn't retire entirely but continues to work part-time? What happens if there are pensions involved? What if one person is adjusted by inflation and the other isn't? What happens if expenses are much higher for the first few years of retirement and then drop off? (Maybe that is showing the shift from paying expensive private health insurance to qualifying for Medicare. Maybe that shows the last few years of paying off a mortgage or college education for children.)
I haven't addressed, so far, the problem of retirement planning for couples in the VSR and VPW spreadsheet, other than suggest, in our wiki, to use the younger's spouse age to determine VPW withdrawal rates.

Planning for couples is challenging. For one thing, there's always the possibility of separation of spouses due to health issues (one spouse ending up in a long-term care facility), or divorce. Such an adverse event can't be efficiently planned for; insurance companies don't sell "divorce insurance" that replaces the ex-spouse's financial contribution. One must simply acknowledge the risk and plan to adapt by reducing spending if it ever happens. It would be inefficient for a couple to live on the equivalent of a post-divorce single budget just so that spouses don't experience a financial loss if it happens.

The other difficulty is that Social Security is not perfectly similar to a joint-life annuity; on the death of one spouse, one of the two Social Security pensions is lost. Luckily, the survivor gets to keep the highest of the two pensions. So, it might make sense for the spouse with the higher Social Security payment to delay it, and for the other to claim it as early as possible. Yet, financial planning should account for the drop in income on the death of one spouse. Just to make matters worse, all income will now be taxed into the hand of a single person, which might further reduce after-tax income available for spending.

The current solution we use in my couple is to make two separate calculations, one for each spouse, and not plan for the "single-life" part of future pensions (e.g. not bridge the gap in income between retirement and start of a single-life pension, consider as if it didn't exist). When we get the unplanned money, we'll use it as "extra, unplanned for" spending for the couple. We won't use it for regular spending (unless, maybe, its an expense we know the surviving spouse will stop). My wife and I are aware that, because of taxes, net income will be lower. But, there will be one less mouth to feed, and the surviving spouse could always downsize the home.

In other words, we only need to maintain two copies of the VSR spreadsheet, one per spouse, and fill them appropriately to exclude single-life pensions (or part of pensions that are partly joint-life). It's simple enough and it works for us.

For a US couple, each spouse could put 50% of the higher of the two Social Security pensions into her personal VSR spreadsheet.
Bogleheads investment philosophy | One-ETF global balanced index portfolio | VPW
User avatar
Horton
Posts: 941
Joined: Mon Jan 21, 2008 3:53 pm

Re: The Variable Savings Rate (VSR) -- an accumulation-time prelude to VPW

Post by Horton »

The discount rate is an interesting topic. I prefer to use risk free current market assumptions, rather than historical averages. Sounds like this is the approach AlohaJoe uses. It’s consistent with the lifecycle model and is the approach ESPlanner / MaxiFi use. The latter tools are not free but do permit a lot of flexibility when modeling: Social Security, pensions, annuities, ages of spouse / children, college savings, longevity of couples, asset allocation, etc.

That said, I can attest that there is a steep learning curve when using ESPlanner / MaxiFi, and the VSR may be a more straightforward tool for many to gauge progress.
Topic Author
longinvest
Posts: 4710
Joined: Sat Aug 11, 2012 8:44 am

Re: The Variable Savings Rate (VSR) -- an accumulation-time prelude to VPW

Post by longinvest »

Horton wrote: Sat May 25, 2019 2:28 pm The discount rate is an interesting topic. I prefer to use risk free current market assumptions, rather than historical averages.
If your goal is to delay spending from younger and healthier years to older years, using a so-called "risk-free" rate is a good method. It's also a good way to hoard money and an awesome approach for financial advisers to keep Assets Under Management (AUM) as big as possible for the longest time possible. Keeping as much AUM as possible is also why we continue to hear about taking constant inflation-adjusted withdrawals from a portfolio of fluctuating assets regardless of market returns. It's also why we always hear that future returns in all markets will be low and that one should pick a very low SWR to be "safe".

I know of nobody who can predict stocks, bonds, TIPS, savings accounts, and other financial assets total returns over my hopefully multi-decades investment horizons (yes, it isn't a single horizon, but hundreds of horizons).

Good luck!
Bogleheads investment philosophy | One-ETF global balanced index portfolio | VPW
Topic Author
longinvest
Posts: 4710
Joined: Sat Aug 11, 2012 8:44 am

Re: The Variable Savings Rate (VSR) -- an accumulation-time prelude to VPW

Post by longinvest »

longinvest wrote: Sat May 25, 2019 2:39 pm
Horton wrote: Sat May 25, 2019 2:28 pm The discount rate is an interesting topic. I prefer to use risk free current market assumptions, rather than historical averages.
If your goal is to delay spending from younger and healthier years to older years, using a so-called "risk-free" rate is a good method. It's also a good way to hoard money and an awesome approach for financial advisers to keep Assets Under Management (AUM) as big as possible for the longest time possible. Keeping as much AUM as possible is also why we continue to hear about taking constant inflation-adjusted withdrawals from a portfolio of fluctuating assets regardless of market returns. It's also why we always hear that future returns in all markets will be low and that one should pick a very low SWR to be "safe".

I know of nobody who can predict stocks, bonds, TIPS, savings accounts, and other financial assets total returns over my hopefully multi-decades investment horizons (yes, it isn't a single horizon, but hundreds of horizons).

Good luck!
Let me add that using a changing rate, like the so-called "risk-free" rate, leads to a very bad investing approach. When the rate is low, one saves more and "buys more" highly-priced assets. When the rate increases (and asset prices deflate), one saves less and "buys less" low-priced assets. "buy more high, buy less low" is a form of "buy high, sell low", the worst investing method ever.

It pays to think ahead of time about the consequences of one's choices. The choice of using a constant internal growth trend, in VPW and VSR, is fundamental to the approach; it isn't an arbitrary choice.
Bogleheads investment philosophy | One-ETF global balanced index portfolio | VPW
User avatar
Horton
Posts: 941
Joined: Mon Jan 21, 2008 3:53 pm

Re: The Variable Savings Rate (VSR) -- an accumulation-time prelude to VPW

Post by Horton »

longinvest wrote: Sat May 25, 2019 2:39 pm
Horton wrote: Sat May 25, 2019 2:28 pm The discount rate is an interesting topic. I prefer to use risk free current market assumptions, rather than historical averages.
If your goal is to delay spending from younger and healthier years to older years, using a so-called "risk-free" rate is a good method. It's also a good way to hoard money and an awesome approach for financial advisers to keep Assets Under Management (AUM) as big as possible for the longest time possible. Keeping as much AUM as possible is also why we continue to hear about taking constant inflation-adjusted withdrawals from a portfolio of fluctuating assets regardless of market returns. It's also why we always hear that future returns in all markets will be low and that one should pick a very low SWR to be "safe".

I know of nobody who can predict stocks, bonds, TIPS, savings accounts, and other financial assets total returns over my hopefully multi-decades investment horizons (yes, it isn't a single horizon, but hundreds of horizons).

Good luck!
Actually, by using MaxiFi and similar methods supported by the lifecycle theory, I’ve come to discover in the last year that I can (a) spend more now and (b) invest less aggressively while still being able to retire when I want.

I anticipate that VSR/VPW would tell me something similar as well. Perhaps it would paint an even more optimistic picture, but I know that I’m risk averse. In any event, this is a big improvement over the “4% rule.”

Again, I think your tool is useful and apologize if you took my comments as criticism. VPW at least has been labeled as “developed by the Bogleheads”, so I just wanted to throw out another perspective for those who use the tool.

Thanks again for all your hard work!
User avatar
Horton
Posts: 941
Joined: Mon Jan 21, 2008 3:53 pm

Re: The Variable Savings Rate (VSR) -- an accumulation-time prelude to VPW

Post by Horton »

longinvest wrote: Sat May 25, 2019 2:50 pm
longinvest wrote: Sat May 25, 2019 2:39 pm
Horton wrote: Sat May 25, 2019 2:28 pm The discount rate is an interesting topic. I prefer to use risk free current market assumptions, rather than historical averages.
If your goal is to delay spending from younger and healthier years to older years, using a so-called "risk-free" rate is a good method. It's also a good way to hoard money and an awesome approach for financial advisers to keep Assets Under Management (AUM) as big as possible for the longest time possible. Keeping as much AUM as possible is also why we continue to hear about taking constant inflation-adjusted withdrawals from a portfolio of fluctuating assets regardless of market returns. It's also why we always hear that future returns in all markets will be low and that one should pick a very low SWR to be "safe".

I know of nobody who can predict stocks, bonds, TIPS, savings accounts, and other financial assets total returns over my hopefully multi-decades investment horizons (yes, it isn't a single horizon, but hundreds of horizons).

Good luck!
Let me add that using a changing rate, like the so-called "risk-free" rate, leads to a very bad investing approach. When the rate is low, one saves more and "buys more" highly-priced assets. When the rate increases (and asset prices deflate), one saves less and "buys less" low-priced assets. "buy more high, buy less low" is a form of "buy high, sell low", the worst investing method ever.

It pays to think ahead of time about the consequences of one's choices. The choice of using a constant internal growth trend, in VPW and VSR, is fundamental to the approach; it isn't an arbitrary choice.
When rates are low, your liability (retirement spending) is high. That’s why you need to save more.

What happens if rates do not rise to historical levels? You may wind up having to save more later in your career or delay retirement.
Topic Author
longinvest
Posts: 4710
Joined: Sat Aug 11, 2012 8:44 am

Re: The Variable Savings Rate (VSR) -- an accumulation-time prelude to VPW

Post by longinvest »

Horton wrote: Sat May 25, 2019 3:03 pm I anticipate that VSR/VPW would tell me something similar as well. Perhaps it would paint an even more optimistic picture, but I know that I’m risk averse. In any event, this is a big improvement over the “4% rule.”
Horton, if you're very risk averse, I suggest that you investigate VSR and VPW using a conservative portfolio. I would be careful, though, not to try too much to avoid one type of risk, because one can often unawarely be loading up on other harmful risks by avoiding a specific risk. It's often best to mitigate risks, instead of trying to completely avoid them. Anyway, that's the approach I take.
Bogleheads investment philosophy | One-ETF global balanced index portfolio | VPW
Topic Author
longinvest
Posts: 4710
Joined: Sat Aug 11, 2012 8:44 am

Re: The Variable Savings Rate (VSR) -- an accumulation-time prelude to VPW

Post by longinvest »

Horton wrote: Sat May 25, 2019 3:11 pm
longinvest wrote: Sat May 25, 2019 2:50 pm Let me add that using a changing rate, like the so-called "risk-free" rate, leads to a very bad investing approach. When the rate is low, one saves more and "buys more" highly-priced assets. When the rate increases (and asset prices deflate), one saves less and "buys less" low-priced assets. "buy more high, buy less low" is a form of "buy high, sell low", the worst investing method ever.

It pays to think ahead of time about the consequences of one's choices. The choice of using a constant internal growth trend, in VPW and VSR, is fundamental to the approach; it isn't an arbitrary choice.
When rates are low, your liability (retirement spending) is high. That’s why you need to save more.

What happens if rates do not rise to historical levels? You may wind up having to save more later in your career or delay retirement.
Horton, the math is what it is. Using a changing risk-free rate to drive a savings method or a withdrawal method, like VSR and VPW, leads to counterproductive investing behavior. It just cannot be otherwise.

Future market total-returns are uncertain. It's probably best to learn to deal with it.
Bogleheads investment philosophy | One-ETF global balanced index portfolio | VPW
abracadabra11
Posts: 230
Joined: Sat May 01, 2010 2:09 pm

Re: The Variable Savings Rate (VSR) -- an accumulation-time prelude to VPW

Post by abracadabra11 »

Are you taking requests for the spreadsheet?

If so, can you expand the available retirement age selections to include 40?
Topic Author
longinvest
Posts: 4710
Joined: Sat Aug 11, 2012 8:44 am

Re: The Variable Savings Rate (VSR) -- an accumulation-time prelude to VPW

Post by longinvest »

abracadabra11 wrote: Sat May 25, 2019 4:23 pm Are you taking requests for the spreadsheet?

If so, can you expand the available retirement age selections to include 40?
Abra Cadabra,

Can you explain the magic that allowed you to become so highly efficient at what you do to generate so much wealth that you're able to contemplate retirement at age 40, while at the same time, you're hoping to quit at 40?

I can't imagine my preferred idols (movie stars, artists, business leaders, and so many others) quitting at 40 and depriving humanity of their talents!

This spreadsheet is meant as a Bogleheads tool. It shouldn't be a tool that discourages average young savers by letting them think that targeting retirement at 40 is normal. I wouldn't want to encourage, either, young over-enthusiasts to deprive their spouse and children from the pleasures money is meant to buy. Life's about the journey, not about the grim final destination.

If you let me in on your magic, I'll let you in on my trick to modify the spreadsheet yourself.
:sharebeer
Bogleheads investment philosophy | One-ETF global balanced index portfolio | VPW
abracadabra11
Posts: 230
Joined: Sat May 01, 2010 2:09 pm

Re: The Variable Savings Rate (VSR) -- an accumulation-time prelude to VPW

Post by abracadabra11 »

longinvest wrote: Sat May 25, 2019 5:20 pm
abracadabra11 wrote: Sat May 25, 2019 4:23 pm Are you taking requests for the spreadsheet?

If so, can you expand the available retirement age selections to include 40?
Abra Cadabra,

Can you explain the magic that allowed you to become so highly efficient at what you do to generate so much wealth that you're able to contemplate retirement at age 40, while at the same time, you're hoping to quit at 40?

I can't imagine my preferred idols (movie stars, artists, business leaders, and so many others) quitting at 40 and depriving humanity of their talents!

This spreadsheet is meant as a Bogleheads tool. It shouldn't be a tool that discourages average young savers by letting them think that targeting retirement at 40 is normal. I wouldn't want to encourage, either, young over-enthusiasts to deprive their spouse and children from the pleasures money is meant to buy. Life's about the journey, not about the grim final destination.

If you let me in on your magic, I'll let you in on my trick to modify the spreadsheet yourself.
:sharebeer
There's no magic. 42 is my planned retirement. High savings rate + military pension make for a pretty easy formula. For some members, retiring at 38 is even a possibility.

Will I do other activities after 42? Absolutely, but I have no intention on relying on them to support my family's spending.
Topic Author
longinvest
Posts: 4710
Joined: Sat Aug 11, 2012 8:44 am

Re: The Variable Savings Rate (VSR) -- an accumulation-time prelude to VPW

Post by longinvest »

abracadabra11 wrote: Sun May 26, 2019 4:27 pm
longinvest wrote: Sat May 25, 2019 5:20 pm If you let me in on your magic, I'll let you in on my trick to modify the spreadsheet yourself.
There's no magic. 42 is my planned retirement. High savings rate + military pension make for a pretty easy formula.
Thanks. I'm sending you a private message.
:sharebeer
Bogleheads investment philosophy | One-ETF global balanced index portfolio | VPW
Topic Author
longinvest
Posts: 4710
Joined: Sat Aug 11, 2012 8:44 am

Re: The Variable Savings Rate (VSR) -- an accumulation-time prelude to VPW

Post by longinvest »

I have uploaded version 1.2 of the Variable Savings Rate spreadsheet.

It is available online and to download from the links in the first post.

Main changes:
  • Add support for defined benefit pensions with and without cost of living adjustments.
As usual, comments are welcome.

Enjoy!
Bogleheads investment philosophy | One-ETF global balanced index portfolio | VPW
Topic Author
longinvest
Posts: 4710
Joined: Sat Aug 11, 2012 8:44 am

Re: The Variable Savings Rate (VSR) -- an accumulation-time prelude to VPW

Post by longinvest »

I've announced on the VPW thread the release of a VPW Accumulation And Retirement Worksheet that includes both an Accumulation sheet and a Retirement sheet. The VSR spreadsheet is now part of this new worksheet.

Enjoy!
Bogleheads investment philosophy | One-ETF global balanced index portfolio | VPW
azanon
Posts: 3033
Joined: Mon Nov 07, 2011 10:34 am

Re: The Variable Savings Rate (VSR) -- an accumulation-time prelude to VPW

Post by azanon »

So just skimming over this, is this technique a way of implementing "Consumption Smoothing", which i've read about from time to time? one link to a definition: https://www.investopedia.com/terms/c/co ... othing.asp
Topic Author
longinvest
Posts: 4710
Joined: Sat Aug 11, 2012 8:44 am

Re: The Variable Savings Rate (VSR) -- an accumulation-time prelude to VPW

Post by longinvest »

azanon wrote: Tue Jun 11, 2019 8:05 am So just skimming over this, is this technique a way of implementing "Consumption Smoothing", which i've read about from time to time? one link to a definition: https://www.investopedia.com/terms/c/co ... othing.asp
Your question is somewhat similar to asking if VPW is a way of implementing SWR (for which the answer is no, because VPW delivers fluctuating withdrawals and never prematurely depletes the portfolio).

Consumption smoothing has a predictive aspect to it and has sequence of return problems. VSR doesn't.

VSR is about trying to estimate a reasonable level of savings, while staying flexible and reassessing the situation every year. After low returns, savings are somewhat adjusted up. After high returns, savings are somewhat adjusted down. The idea is not to over-save or under-save too much. It's very similar to VPW, but it applies to portfolio contributions, instead of portfolio withdrawals.

The nice thing is that it eliminates the use of a "Target Number" for retirement planning (like seeking a $1,000,000 portfolio or 25 times expenses). Instead, it lets its user choose a retirement age and adjusts savings (and, therefore spending) accordingly*, subject to annual adjustments. Among its features is informing its user of the required flexibility to face a 50% stocks crash with the chosen asset allocation and retirement plan.

* Taking into account the user's current age, salary, portfolio balance and allocation, planned retirement age, and future pensions like Social Security.
Bogleheads investment philosophy | One-ETF global balanced index portfolio | VPW
mouth
Posts: 307
Joined: Sun Apr 19, 2015 6:40 am

Re: The Variable Savings Rate (VSR) -- an accumulation-time prelude to VPW

Post by mouth »

Hello and thanks for all the work longinvest. I'm sorry if this is answered elsewhere but what is the intent of the "Annual Contribution" cell in the Accumulation sheet?

If it matters ...
I ask in the context that I am 48yo, currently employed, and receiving a COLA adjusted military pension right now. So I've got that entered starting at 49yo (since you throw an error starting it at 48) but I can't quite untangle the spaghetti to understand the purpose of the Annual Contribution cell.

Related, since I am technically receiving my pension right now, do I add that to the Annual Salary cell?
Topic Author
longinvest
Posts: 4710
Joined: Sat Aug 11, 2012 8:44 am

Re: The Variable Savings Rate (VSR) -- an accumulation-time prelude to VPW

Post by longinvest »

mouth wrote: Sat Sep 26, 2020 7:04 am Hello and thanks for all the work longinvest. I'm sorry if this is answered elsewhere but what is the intent of the "Annual Contribution" cell in the Accumulation sheet?

If it matters ...
I ask in the context that I am 48yo, currently employed, and receiving a COLA adjusted military pension right now. So I've got that entered starting at 49yo (since you throw an error starting it at 48) but I can't quite untangle the spaghetti to understand the purpose of the Annual Contribution cell.

Related, since I am technically receiving my pension right now, do I add that to the Annual Salary cell?
Mouth,

Before retirement, your military pension is part of your (total) Annual Salary. You can't start a pension before retirement in the worksheet. Your military pension is also a Defined Benefit Pension that starts at the same age as retirement. You should set its Annual Contribution to $0, as you don't have to contribute anything into it.

Another pension that you should include is Social Security (SS). You can estimate your future benefits using Neurosphere's latest estimator (you must first download it or copy it; you obviously can't put your own numbers into a publicly shared spreadsheet). You enter your historical earnings and use your current work salary (do not include your military pension) for the current year and all future work years. Do not change the default inflation assumption (which is "zero" to calculate using current dollars). I suggest to delay SS to age 70. You'll find an estimate of your annual old-age and survivor taxes (OASI) on the "Taxes Paid" sheet. That's the number to use as Annual Contribution for Social Security. You could also estimate this amount by looking at your own tax records for 2019, or maybe look at your recent paycheck slips.

Let's say, for example, that you intended to retire from your current work at age 55. You'd set Retirement Age to 55. You'd set the Annual Salary to the sum of your current work salary and military pension. You'd also use your current work salary from age 48 to 54 in Neuroshpere's estimator.

The objective of the worksheet is to equalize the annual amount available for taxes and expenses after subtracting pension and portfolio contributions during accumulation with the annual sum of pension payments and portfolio withdrawals during retirement.
Bogleheads investment philosophy | One-ETF global balanced index portfolio | VPW
mouth
Posts: 307
Joined: Sun Apr 19, 2015 6:40 am

Re: The Variable Savings Rate (VSR) -- an accumulation-time prelude to VPW

Post by mouth »

Thanks so much for your reply. That all makes sense :sharebeer
tj
Posts: 4817
Joined: Thu Dec 24, 2009 12:10 am

Re: The Variable Savings Rate (VSR) -- an accumulation-time prelude to VPW

Post by tj »

So when I use the spreadsheet and it says my monthly contribution should be $0 - how can that be right? It says that for any asset allocation, too. I set the retirement date for 57 - I don't need to put another penny into retirement savings? That seems odd.
Topic Author
longinvest
Posts: 4710
Joined: Sat Aug 11, 2012 8:44 am

Re: The Variable Savings Rate (VSR) -- an accumulation-time prelude to VPW

Post by longinvest »

tj wrote: Sat Sep 26, 2020 10:11 pm So when I use the spreadsheet and it says my monthly contribution should be $0 - how can that be right? It says that for any asset allocation, too. I set the retirement date for 57 - I don't need to put another penny into retirement savings? That seems odd.
Tj, you didn't tell us about the numbers you've put into the Accumulation worksheet. This could happen, for example, when the Portfolio Balance gets disproportionately big relatively to the Annual Salary and Defined Benefit Pensions after receiving a huge inheritance. It could also happen as the result of a significant drop in annual salary. It could also be the result of entering incorrect numbers into the spreadsheet. It's difficult to say without specific numbers.
Bogleheads investment philosophy | One-ETF global balanced index portfolio | VPW
tj
Posts: 4817
Joined: Thu Dec 24, 2009 12:10 am

Re: The Variable Savings Rate (VSR) -- an accumulation-time prelude to VPW

Post by tj »

longinvest wrote: Sun Sep 27, 2020 10:04 am
tj wrote: Sat Sep 26, 2020 10:11 pm So when I use the spreadsheet and it says my monthly contribution should be $0 - how can that be right? It says that for any asset allocation, too. I set the retirement date for 57 - I don't need to put another penny into retirement savings? That seems odd.
Tj, you didn't tell us about the numbers you've put into the Accumulation worksheet. This could happen, for example, when the Portfolio Balance gets disproportionately big relatively to the Annual Salary and Defined Benefit Pensions after receiving a huge inheritance. It could also happen as the result of a significant drop in annual salary. It could also be the result of entering incorrect numbers into the spreadsheet. It's difficult to say without specific numbers.
So, the inputted portfolio balance was roughly 9x the inputted annual salary for age 35. The monthly defined benefit pensions are about 4% of the inputted annual salary - when using the minimum ages with early start reductions.

Increasing the salary does cause a monthly contribution to populate.

As a newish federal employee, I will be seeing some substantial pay increases for the first few years in my position, as my income goes up, the portfolio contributions do start ot become necessary, but my projected SS and FERS (and associated annual contributions) would also increase.

For SS, it's pretty easy to go to ssa.tools and use benefits that have been earned so far. For FERS, i mean, I don't have enough time in service to receive any sort of pension yet. Perhaps I should just leave that blank until I do qualify for a pension - even though federal employment and the associated pension are probably on the safer side.
Topic Author
longinvest
Posts: 4710
Joined: Sat Aug 11, 2012 8:44 am

Re: The Variable Savings Rate (VSR) -- an accumulation-time prelude to VPW

Post by longinvest »

tj wrote: Sun Sep 27, 2020 11:24 am
longinvest wrote: Sun Sep 27, 2020 10:04 am
tj wrote: Sat Sep 26, 2020 10:11 pm So when I use the spreadsheet and it says my monthly contribution should be $0 - how can that be right? It says that for any asset allocation, too. I set the retirement date for 57 - I don't need to put another penny into retirement savings? That seems odd.
Tj, you didn't tell us about the numbers you've put into the Accumulation worksheet. This could happen, for example, when the Portfolio Balance gets disproportionately big relatively to the Annual Salary and Defined Benefit Pensions after receiving a huge inheritance. It could also happen as the result of a significant drop in annual salary. It could also be the result of entering incorrect numbers into the spreadsheet. It's difficult to say without specific numbers.
So, the inputted portfolio balance was roughly 9x the inputted annual salary for age 35. The monthly defined benefit pensions are about 4% of the inputted annual salary - when using the minimum ages with early start reductions.

Increasing the salary does cause a monthly contribution to populate.

As a newish federal employee, I will be seeing some substantial pay increases for the first few years in my position, as my income goes up, the portfolio contributions do start ot become necessary, but my projected SS and FERS (and associated annual contributions) would also increase.

For SS, it's pretty easy to go to ssa.tools and use benefits that have been earned so far. For FERS, i mean, I don't have enough time in service to receive any sort of pension yet. Perhaps I should just leave that blank until I do qualify for a pension - even though federal employment and the associated pension are probably on the safer side.
Tj, having a portfolio equal to 9 times your annual salary at age 35 is extraordinarily exceptional.

The worksheet is right; until your salary increases (or markets drop, depending your your portfolio's allocation), your workplace savings are sufficient if you stay at the same job and retire at age 57. But, 22 years is a long time...

Here's a suggestion.

I'm assuming that you spend less than what remains of your paycheck after workplace savings and taxes. You could take advantage of this to aim for an earlier financial independence age. You could target age 50 or 45, for example, setting it as Retirement Age in the worksheet. You'd choose the age such that it results into saving a little less than the amount you don't currently need (you want to keep some flexibility for market fluctuations). Note that you'd also have to estimate the impact of (hypothetically) stopping to work this early on your work pension and Social Security to enter the reduced amounts in the worksheet. This can be tricky, sometimes. Eventually, when you'd reach financial independence age, you could switch to using the Retirement worksheet to determine how much you can spend (after estimating and subtracting taxes from the amount suggested by the worksheet). You'd spend that amount off your paycheck and continue investing whatever remains. Continuing to work, once financially independent, would allow you to slowly and safely increase your standard of living without fearing early work termination. Financial independence can bring peace of mind and provide flexibility.

You're in a very nice situation. Congratulations!
Bogleheads investment philosophy | One-ETF global balanced index portfolio | VPW
tj
Posts: 4817
Joined: Thu Dec 24, 2009 12:10 am

Re: The Variable Savings Rate (VSR) -- an accumulation-time prelude to VPW

Post by tj »

Thanks for the tip about lowering the retirement age. That is a useful tool and does indeed cause the monthly contribution to populate.
Topic Author
longinvest
Posts: 4710
Joined: Sat Aug 11, 2012 8:44 am

Re: The Variable Savings Rate (VSR) -- an accumulation-time prelude to VPW

Post by longinvest »

In some situations (higher income), there might be limits on how much can be contributed to tax-sheltered retirement accounts. In another thread, I suggested an adjustment to the suggested portfolio contribution amount when a significant part of it would be invested after paying taxes (and when mortgage payments represent a significant part of the spending budget).

I think that most users of the spreadsheet shouldn't make such adjustments; they should maximize the use of (traditional) tax-sheltered accounts and aim to keep relatively low mortgage payments by amortizing the mortgage over 25 to 30 years and not overextend their finances buying a too expensive home. There's a risk of ending up "house poor" when applying an adjustment to retirement savings for mortgage payments.

My suggestion is to discount the part of portfolio contributions exposed to taxes by the current average tax rate of the investor (that's not the marginal tax rate!). This average tax rate can be estimated by dividing last year's total taxes by last year's salary, in most cases. As an example, if suggested annual portfolio contributions are $22,000, but only $18,000 can be contributed to a traditional 401k, the excess $4,000 would be discounted by the average tax rate (let's say 25%) to $3,000 (for total portfolio contributions of $21,000 instead of $22,000). This is an imperfect calculation, as taxable investments are exposed to taxes. Let me repeat that it's best not to make such an adjustment, especially when the difference is small like in this example; it's only $1,000/year, less than 5% of the suggested $22,000 annual portfolio contribution. The accumulating investor should be able to afford making the entire $4,000 taxable account contribution.

See the linked thread for the (risky to end up house poor) suggestion to adjust portfolio contributions for the principal part of mortgage payments. I don't recommend it, but it could be useful for those who pay their mortgage quickly and will have many mortgage-free years left of portfolio contributions before retirement (or financial independence).
Bogleheads investment philosophy | One-ETF global balanced index portfolio | VPW
Topic Author
longinvest
Posts: 4710
Joined: Sat Aug 11, 2012 8:44 am

Re: The Variable Savings Rate (VSR) -- an accumulation-time prelude to VPW

Post by longinvest »

In the main VPW thread, forum member Bentonkb asked a question about the Accumulation sheet:
Bentonkb wrote: Mon Mar 08, 2021 7:00 pm I’m working my way through the formulas in the Accumulation worksheet and trying to figure things out. It has come down to the Additional Savings calculation in cell B76 – I think I understand the purpose of the calculation, but the formula doesn’t make any sense to me and neither does the numerical result.

B75/(1+PMT(Lists!B134,100-B12,-FV(Lists!B134,B12-B8,-1,0,0),0,1))

Cell B75 is the Unallocated Income, which seems to be current living expenses (earnings minus savings) minus the sum of all the pensions and the Initial Portfolio Withdrawal. I’ve been thinking of Unallocated Income as the expected funding shortfall during the first year of retirement if the worker doesn’t make any additional contributions into the retirement savings account.

The denominator looks to be 1 plus the additional after-retirement income generated by increasing the annual pre-retirement savings by $1. Please explain how this calculation gives you the suggested amount of extra money that the worker needs to save this year to keep his retirement on track. I’m missing some important piece of information that will help me understand.

In the scenario I was playing with, the return rate is 3.8%, the worker’s age is 49 with a target retirement at 55. The Unallocated Income is $45,176, which gives an Additional Savings of $34,886.

FV(Lists!B134,B12-B8,-1,0,0) is the future value of an extra dollar going into the savings account from now until retirement day.

PMT(Lists!B134,100-B12,-FV(Lists!B134,B12-B8,-1,0,0),0,1) is the extra annual retirement income from the extra dollar of savings.

One dollar of extra savings between now and retirement becomes $6.59 in the bank, which becomes $0.29 of extra annual retirement income.

How does $45,176/(1 + 0.29) give you the required Additional Savings?

I’m struggling with both the reasoning behind the formula and the result. Saving $34,886/year from age 49 to 55 doesn’t seem like it should provide an extra $45,176/year in income from 55 to 100.

$34,886 per year from 49 to 55 should put $230,007 in the bank, which would only produce an extra $10,289 in retirement income, not the whole $45,176 of Unallocated Income.

Is it possible that the formula should be (Unallocated Income)/PMT(Lists!B134,100-B12,-FV(Lists!B134,B12-B8,-1,0,0),0,1)?

That would make more sense because if the Unallocated Income is big and the extra income from extra savings is small, then the worker will have to save a lot more each year.

Hopefully I’ve provided enough detail so that you can figure out what I’m failing to understand. Thanks for your patience.
Bentonkb,

The Accumulation sheet's goal is to balance the amount available for taxes and expenses during accumulation and retirement.

Starting with $0 at age 49 and contributing $34,829/year for 6 years (at end of year) into a portfolio with a 3.8% growth rate results into a $229,864 portfolio at age 55. As the investor's available income (before savings) is $45,176, this leaves ($45,176 - $34,829) = $10,347 for taxes and expenses during accumulation.

Starting with a $229,864 portfolio at age 55 with an unrealistic constant 3.8% growth rate and taking $10,347/year withdrawals at the beginning of the year depletes the portfolio after a last withdrawal at age 99. (See the wiki page for instructions about how to use VPW during retirement and dampen the financial risk related to living beyond age 100).

Saving less than $34,829 would have resulted into a higher amount than $10,347 left for taxes and expenses during accumulation, and a lower amount than $10,347 left for taxes and expenses during retirement. Saving more than $34,829 would have resulted into a lower amount than $10,347 left for taxes and expenses during accumulation, and a higher amount than $10,347 left for taxes and expenses during retirement. The $10,347 amount is the equilibrium amount for taxes and expenses during accumulation and retirement.

In real life returns fluctuate. That's why the Accumulation sheet provides an estimate of the immediate impact of a -50% stock loss and qualifies it as required flexibility. It's important to regularly update the worksheet's inputs and adapt savings according to the resulting suggestion.

The first principle of the Bogleheads investment philosopy is to develop a workable plan. This involves living below one's means so that enough money is left available for investing. The Accumulation Worksheet provides a dynamic estimate of how much to reasonably save during the current year given the investor's situation (age, salary, portfolio size and allocation, future pensions, and target retirement/financial-independence age).
Bogleheads investment philosophy | One-ETF global balanced index portfolio | VPW
Bentonkb
Posts: 17
Joined: Sun Mar 07, 2021 6:47 am

Re: The Variable Savings Rate (VSR) -- an accumulation-time prelude to VPW

Post by Bentonkb »

longinvest wrote: Tue Mar 09, 2021 12:01 am In the main VPW thread, forum member Bentonkb asked a question about the Accumulation sheet:
Bentonkb wrote: Mon Mar 08, 2021 7:00 pm I’m working my way through the formulas in the Accumulation worksheet and trying to figure things out. It has come down to the Additional Savings calculation in cell B76 – I think I understand the purpose of the calculation, but the formula doesn’t make any sense to me and neither does the numerical result.

B75/(1+PMT(Lists!B134,100-B12,-FV(Lists!B134,B12-B8,-1,0,0),0,1))

Cell B75 is the Unallocated Income, which seems to be current living expenses (earnings minus savings) minus the sum of all the pensions and the Initial Portfolio Withdrawal. I’ve been thinking of Unallocated Income as the expected funding shortfall during the first year of retirement if the worker doesn’t make any additional contributions into the retirement savings account.
Bentonkb,

The Accumulation sheet's goal is to balance the amount available for taxes and expenses during accumulation and retirement.

Starting with $0 at age 49 and contributing $34,829/year for 6 years (at end of year) into a portfolio with a 3.8% growth rate results into a $229,864 portfolio at age 55. As the investor's available income (before savings) is $45,176, this leaves ($45,176 - $34,829) = $10,347 for taxes and expenses during accumulation.
Oh, now I can see where my thinking was off. The Additional Savings calculation is not supposed to provide extra retirement income to make up the entire Unallocated Income. It is only supposed to make up the difference between the Additional Savings and the Unallocated Income! Of course, I should have seen that. If the worker can do without the $34,829 in their current budget, they don't need the whole $45,179 to be spendable income after retirement. The extra income requirement is just the difference.

Thanks for clarifying.
Topic Author
longinvest
Posts: 4710
Joined: Sat Aug 11, 2012 8:44 am

Re: The Variable Savings Rate (VSR) -- an accumulation-time prelude to VPW

Post by longinvest »

Bentonkb wrote: Tue Mar 09, 2021 6:18 am If the worker can do without the $34,829 in their current budget, they don't need the whole $45,179 to be spendable income after retirement. The extra income requirement is just the difference.
:thumbsup
Bogleheads investment philosophy | One-ETF global balanced index portfolio | VPW
tj
Posts: 4817
Joined: Thu Dec 24, 2009 12:10 am

Re: The Variable Savings Rate (VSR) -- an accumulation-time prelude to VPW

Post by tj »

longinvest wrote: Sun Sep 27, 2020 12:04 pm
tj wrote: Sun Sep 27, 2020 11:24 am
longinvest wrote: Sun Sep 27, 2020 10:04 am
tj wrote: Sat Sep 26, 2020 10:11 pm So when I use the spreadsheet and it says my monthly contribution should be $0 - how can that be right? It says that for any asset allocation, too. I set the retirement date for 57 - I don't need to put another penny into retirement savings? That seems odd.
Tj, you didn't tell us about the numbers you've put into the Accumulation worksheet. This could happen, for example, when the Portfolio Balance gets disproportionately big relatively to the Annual Salary and Defined Benefit Pensions after receiving a huge inheritance. It could also happen as the result of a significant drop in annual salary. It could also be the result of entering incorrect numbers into the spreadsheet. It's difficult to say without specific numbers.
So, the inputted portfolio balance was roughly 9x the inputted annual salary for age 35. The monthly defined benefit pensions are about 4% of the inputted annual salary - when using the minimum ages with early start reductions.

Increasing the salary does cause a monthly contribution to populate.

As a newish federal employee, I will be seeing some substantial pay increases for the first few years in my position, as my income goes up, the portfolio contributions do start ot become necessary, but my projected SS and FERS (and associated annual contributions) would also increase.

For SS, it's pretty easy to go to ssa.tools and use benefits that have been earned so far. For FERS, i mean, I don't have enough time in service to receive any sort of pension yet. Perhaps I should just leave that blank until I do qualify for a pension - even though federal employment and the associated pension are probably on the safer side.
Tj, having a portfolio equal to 9 times your annual salary at age 35 is extraordinarily exceptional.

The worksheet is right; until your salary increases (or markets drop, depending your your portfolio's allocation), your workplace savings are sufficient if you stay at the same job and retire at age 57. But, 22 years is a long time...

Here's a suggestion.

I'm assuming that you spend less than what remains of your paycheck after workplace savings and taxes. You could take advantage of this to aim for an earlier financial independence age. You could target age 50 or 45, for example, setting it as Retirement Age in the worksheet. You'd choose the age such that it results into saving a little less than the amount you don't currently need (you want to keep some flexibility for market fluctuations). Note that you'd also have to estimate the impact of (hypothetically) stopping to work this early on your work pension and Social Security to enter the reduced amounts in the worksheet. This can be tricky, sometimes. Eventually, when you'd reach financial independence age, you could switch to using the Retirement worksheet to determine how much you can spend (after estimating and subtracting taxes from the amount suggested by the worksheet). You'd spend that amount off your paycheck and continue investing whatever remains. Continuing to work, once financially independent, would allow you to slowly and safely increase your standard of living without fearing early work termination. Financial independence can bring peace of mind and provide flexibility.

You're in a very nice situation. Congratulations!
Another question...how would I account for needing to save for increased expenses in retirement in the spreadsheet? For example, most employees have subsidized health insurance from their employers while they are working. If I was to target an age 45 or age 50 early retirement date, I would need to fund 15 or 20 years of health insurance before I can start Medicare.

Also, how do I determine if a segment of my assets are considered separate from the portfolio? For example, maybe I want to buy real estate someday, but I don't know when that will be. I don't maintain a separate "down payment fund", but I could certainly sell off assets someday to make a down payment. I guess I could decrease the porfolio balance in some future year whenever that happens?
Topic Author
longinvest
Posts: 4710
Joined: Sat Aug 11, 2012 8:44 am

Re: The Variable Savings Rate (VSR) -- an accumulation-time prelude to VPW

Post by longinvest »

tj wrote: Sun Mar 21, 2021 12:51 pm
longinvest wrote: Sun Sep 27, 2020 12:04 pm Tj, having a portfolio equal to 9 times your annual salary at age 35 is extraordinarily exceptional.
...
You're in a very nice situation. Congratulations!
Another question...how would I account for needing to save for increased expenses in retirement in the spreadsheet? For example, most employees have subsidized health insurance from their employers while they are working. If I was to target an age 45 or age 50 early retirement date, I would need to fund 15 or 20 years of health insurance before I can start Medicare.
I don't know how much a new Toyota Corolla will cost in 15 to 20 years. Will it still be available to buy? Good question. It's similar with insurance premiums. I don't know how much they'll cost. Life is uncertain. It's best to accept it and deal with it by having a lot of flexibility in one's budget, when retiring.

Given your portfolio size, I suggest to keep an eye on the Retirement Worksheet and switch to using it when it becomes reasonable to do so (e.g. when the numbers start making sense) while still accumulating, as I explained in a post of the main VPW thread:
longinvest wrote: Tue Jan 14, 2020 8:31 pm You would start simulating retirement using the Retirement sheet, calculating a hypothetical "retirement" gross (pre-tax) income based solely on your investments and future pensions (ignoring current work income). You would then estimate the amount of taxes (federal, state, anticipated medical premiums, etc.) you would have to pay on that hypothetical retirement income, and use the remaining hypothetical "net retirement income" amount as a guide for your after-tax budget for the current year. If the resulting net amount is smaller than your current work net income, you would invest the work income surplus into your portfolio. If your portfolio has grown so much that the resulting amount is bigger than your current work net income, you could safely extract the missing net income difference from your portfolio to increase your current after-tax spending budget.
As for your second question:
tj wrote: Sun Mar 21, 2021 12:51 pm Also, how do I determine if a segment of my assets are considered separate from the portfolio? For example, maybe I want to buy real estate someday, but I don't know when that will be. I don't maintain a separate "down payment fund", but I could certainly sell off assets someday to make a down payment. I guess I could decrease the porfolio balance in some future year whenever that happens?
There's an infinite number ways to make things complicated. I tend to seek simplicity, instead. I would hold a single undivided (balanced) portfolio.

I don't want to get into a buy vs rent calculation. I'll simply say that the renter has easily known current housing expenses. The owner, on the other hand, has transferred part of the investment portfolio into a real estate investment. Each mortgage payment is made of two parts: interest, which could be considered as a current housing cost, and capital which represents an investment into real estate. The real estate investment attracts its own taxes (municipal taxes). It also exposes the owner to difficult to determine in advance maintenance costs. The renter can easily pay for as little housing as needed at any point in time. The owner often has to buy bigger in anticipation of future family needs, or for attractiveness later when selling. It's difficult to determine who currently pays more for housing, the renter or the owner. There are just too many factors at play, many of which will only be known after the fact, at an unknown future date, once the net proceeds of selling the home are known.

Generally speaking, going from renting a small apartment to owning a big house represents an increase in housing expenses. The size of the down payment plays a relatively minor role. A bigger down payment reduces the portfolio size on one side, but it also reduces monthly interest expenses on the other side. The VPW Worksheet automatically adjusts its projected amount, available for taxes and expenses, based on the portfolio size.

One could theoretically consider the capital part of mortgage payments as part of retirement savings; it's an investment into real estate. But this can be risky, as it can hide having paid too much for housing, which could lead to becoming house poor in retirement.
Bogleheads investment philosophy | One-ETF global balanced index portfolio | VPW
SnowBog
Posts: 1558
Joined: Fri Dec 21, 2018 11:21 pm

Re: The Variable Savings Rate (VSR) -- an accumulation-time prelude to VPW

Post by SnowBog »

Replying here as requested (vs. the VPW thread).

To recap our situation with highly variable income:
  • We do not know in advance how much we'll earn in a year
  • We have a high level of confidence we'll make at least $200k gross
  • Our expenses fit comfortably within that range (expenses [w/o taxes] area actually closer to $100k, but are likely closer to $150k in early retirement inclusive of taxes and healthcare)
  • But in a good year, we could make 2x (or more)
For us, I don't understand why we'd use the $100k as you mentioned below... While we might have a year that's all we make, that would understate our expenses and undercut our savings. In my view, this would be the same as if someone was laid off during the year - it could happen (and that's why we have an EF) - but presumably the "model" is based on "expected income".

But if I'm following the logic, then we can safely "ignore" our variable income (as we save - not spend anything extra) and instead use the accumulation sheet with a "base salary" that covers expenses (including taxes).

That makes sense to me, and aligns with how I've been trying to use the accumulation sheet. We normally save a lot more than is "recommended", but that's mainly to maximize our savings in our top earning years.

Thanks for all your assistance and contributions on this tool!
longinvest wrote: Sun Mar 21, 2021 4:49 pm
SnowBog wrote: Sat Mar 20, 2021 10:40 pm In our personal case - our joint base income is approximately $200k, but our variable income can range anywhere from -$100k to an extra $300k+ (last several years its been in the + $50k - $200k range). Our "expenses" [excluding taxes] remain relatively flat. Similar to above, in a "low income" year - VPW seems to imply we don't need to save any more... In a "high income year" - VPW suggests we need to save a lot more... Bottom line - its hard to "read" VPW to determine if we've saved enough, as the "equilibrium" is based off [in our case] a flawed assumption that our current "income" has any relevance on our retirement needs (we can live on less than our base salary - so even that is higher than needed).

As such - other than a sanity check ensuring we save more than VPW suggests (using our "expenses" as "annual salary" in VPW), we largely ignore the Accumulation tab, and use other "expense based" planning tools instead. But I'll concede its probably more useful for those with less variability in their income...

Are we looking at this wrong? Should we be using something like "expenses" (as I'm using now), or "average last 3|5|10" salaries?
SnowBog, I suggest asking question about the Accumulation Worksheet on its own thread after taking the time to carefully read the explanations I've provided in various posts.
[...]
As for the household income which fluctuates between $100,000 and $500,000, it's probable that the couple doesn't know one year in advance that there will be a $400,000 bonus accumulated by the end of year. So, the couple shouldn't plan for more than a $100,000 income. Extra income will go to pay related taxes and the rest could simply be dumped into the portfolio, helping to reach financial independence earlier. Eventually, the couple will be able to switch to using the Retirement Worksheet while still in the accumulation phase. I've discussed this earlier:
longinvest wrote: Tue Jan 14, 2020 8:31 pm You would start simulating retirement using the Retirement sheet, calculating a hypothetical "retirement" gross (pre-tax) income based solely on your investments and future pensions (ignoring current work income). You would then estimate the amount of taxes (federal, state, anticipated medical premiums, etc.) you would have to pay on that hypothetical retirement income, and use the remaining hypothetical "net retirement income" amount as a guide for your after-tax budget for the current year. If the resulting net amount is smaller than your current work net income, you would invest the work income surplus into your portfolio. If your portfolio has grown so much that the resulting amount is bigger than your current work net income, you could safely extract the missing net income difference from your portfolio to increase your current after-tax spending budget.
Topic Author
longinvest
Posts: 4710
Joined: Sat Aug 11, 2012 8:44 am

Re: The Variable Savings Rate (VSR) -- an accumulation-time prelude to VPW

Post by longinvest »

SnowBog,
SnowBog wrote: Sun Mar 21, 2021 8:09 pm We have a high level of confidence we'll make at least $200k gross
That's effectively the amount I would use in the worksheet, in that case.
SnowBog wrote: Sun Mar 21, 2021 8:09 pm But if I'm following the logic, then we can safely "ignore" our variable income (as we save - not spend anything extra) and instead use the accumulation sheet with a "base salary" that covers expenses (including taxes).
Mostly, but I would probably enjoy part of the extra money with occasional non-recurring fun expenses, too.
SnowBog wrote: Sun Mar 21, 2021 8:09 pm Thanks for all your assistance and contributions on this tool!
Thanks for the nice comment.
Bogleheads investment philosophy | One-ETF global balanced index portfolio | VPW
SnowBog
Posts: 1558
Joined: Fri Dec 21, 2018 11:21 pm

Re: The Variable Savings Rate (VSR) -- an accumulation-time prelude to VPW

Post by SnowBog »

longinvest wrote: Sun Mar 21, 2021 9:52 pm
SnowBog wrote: Sun Mar 21, 2021 8:09 pm But if I'm following the logic, then we can safely "ignore" our variable income (as we save - not spend anything extra) and instead use the accumulation sheet with a "base salary" that covers expenses (including taxes).
Mostly, but I would probably enjoy part of the extra money with occasional non-recurring fun expenses, too.
Thanks for the followup!

We planned to "enjoy" a little more last year with two vacations scheduled (hadn't done more than a weekend road trip in a few years)... :oops: Maybe after COVID...
j7se
Posts: 36
Joined: Fri Jan 01, 2016 11:16 pm

Re: The Variable Savings Rate (VSR) -- an accumulation-time prelude to VPW

Post by j7se »

What about known inheritances or the sale of property, businesses, or other assets at specific time periods? Would you just use a discount rate to find the PV of those assets now? Or do you have a better technique to account for these?

Thanks
Topic Author
longinvest
Posts: 4710
Joined: Sat Aug 11, 2012 8:44 am

Re: The Variable Savings Rate (VSR) -- an accumulation-time prelude to VPW

Post by longinvest »

j7se wrote: Sun Apr 25, 2021 5:14 pm What about known inheritances or the sale of property, businesses, or other assets at specific time periods? Would you just use a discount rate to find the PV of those assets now? Or do you have a better technique to account for these?
J7se, I'd never count on an inheritance before the money has made into my account. Once it does, the spreadsheet will automatically adapt its suggestions.

For a private business, its fair market value should probably be counted as part of the investor's wealth, but it's worse than investing into an illiquid publicly-traded single stock; the real market price which another investor would be willing to actually pay today for it is probably quite uncertain. This simple spreadsheet doesn't have a magic wand to solve such difficult problems.
Bogleheads investment philosophy | One-ETF global balanced index portfolio | VPW
j7se
Posts: 36
Joined: Fri Jan 01, 2016 11:16 pm

Re: The Variable Savings Rate (VSR) -- an accumulation-time prelude to VPW

Post by j7se »

I don’t think it has to be complicated. My suggestion would be to add one line on the defined benefit portion to either ask if the payment is one time or monthly or to just add a line for a one time payment. This also helps those who are contemplating a pension cash out.
Topic Author
longinvest
Posts: 4710
Joined: Sat Aug 11, 2012 8:44 am

Re: The Variable Savings Rate (VSR) -- an accumulation-time prelude to VPW

Post by longinvest »

j7se wrote: Mon Apr 26, 2021 1:37 am I don’t think it has to be complicated. My suggestion would be to add one line on the defined benefit portion to either ask if the payment is one time or monthly or to just add a line for a one time payment. This also helps those who are contemplating a pension cash out.
J7se, if one knows the current (after transaction fees) market value of the business, one can simply include it as part of the portfolio balance and consider it as stocks for asset allocation purpose. There's no need for any special entry. There's a big risk of being overoptimistic about the real market value of a business, though, which should be the amount someone else would actually have the means and be willing to pay for it today.

For a house, on the other hand, it's already being consumed as imputed rent, so I wouldn't add its current market value as part of the portfolio. Once it's sold, the proceeds (after transaction fees) can be invested into the portfolio and the spreadsheet will automatically adjust its suggestions accordingly.
Bogleheads investment philosophy | One-ETF global balanced index portfolio | VPW
slickwillie
Posts: 14
Joined: Wed Aug 03, 2011 2:22 pm

Re: The Variable Savings Rate (VSR) -- an accumulation-time prelude to VPW

Post by slickwillie »

I searched the thread using 65.7 but didn't see this has been addressed. The accumulation worksheet has a fixed pension reduction to 65.7%. I assume this is to account for inflation? What metric is being used to account for inflation rate? Thanks.
Topic Author
longinvest
Posts: 4710
Joined: Sat Aug 11, 2012 8:44 am

Re: The Variable Savings Rate (VSR) -- an accumulation-time prelude to VPW

Post by longinvest »

slickwillie wrote: Sat Jul 03, 2021 4:20 pm I searched the thread using 65.7 but didn't see this has been addressed. The accumulation worksheet has a fixed pension reduction to 65.7%. I assume this is to account for inflation? What metric is being used to account for inflation rate? Thanks.
Slickwillie, this percentage varies based on asset allocation and accounts for the inflation target of the central bank (which is 2%). Calculations to derive the pension spending ratio are provided in this post and this post. The final formula is surprisingly simple.
Bogleheads investment philosophy | One-ETF global balanced index portfolio | VPW
Post Reply