A nearly riskless and stable spending rule for retirement? Lets discuss the ARVA strategy.

JohnnyPHX

Thinks s/he gets paid by the post
Joined
Mar 1, 2015
Messages
1,662
Location
Phoenix AZ
Sorry this is a long one. Going down the rabbit hole.
Lately I have been looking at various retirement withdrawal strategies and looking for options to potentially lower risk during the decumulation phase. So many opinions and options out there. The 4% rule is widely discussed and most here use it as more of a guideline. The goal is to know how much we can safely spend during retirement and not run out of money. Nobody can really predict what the future holds for any of us. We hope for the best but try to plan for the worst.

My Google news feed had this article show up in it recently. "This Might Be the Key to Sustainable Spending in Retirement". Stefan Sharkansky PhD. recommends the Annually Recalculated Virtual Annuity (ARVA) method for spending in retirement. With this approach, your retirement portfolio should contain only two types of assets — a ladder of Treasury Inflation-Protected Securities (TIPS) and a low-cost stock market index fund.

I searched for ARVA discussions on this forum and only found some brief discussions on the subject from many years ago.

https://larrysiegeldotorg.wordpress...nly-spending-rule-article-youll-ever-need.pdf
This is the original document titled THE ONLY SPENDING RULE ARTICLE YOU’LL EVER NEED. It is 26 pages long.
To sum it up, retirement spending is best approached like an annuitization problem, using either real annuities or a virtual annuity spending rule like ARVA. If your portfolio uses risk-free assets such as laddered TIPS, your withdrawals will be stable but relatively modest due to their safety. If you use risky assets, spending should be recalculated regularly based on your portfolio’s value, allowing you to adjust for market changes and avoid running out of money. Seeking higher returns means accepting more uncertainty in how much you can spend, so risk tolerance and asset allocation should be chosen thoughtfully.

Stefan Sharkansky PhD wrote The Only Other Spending Rule Article You Will Ever Need. It notes "We expand on ideas in existing literature to provide a spending formula that is flexible enough to adapt to a retiree’s evolving circumstances, yet which can be implemented by a practitioner or sophisticated retiree using only a spreadsheet."

Then I discovered a website by Stefan, with a new retirement tool.
The Best Third

The website states "Its backed by cutting-edge financial research", the tool puts advanced statistical methodology at your fingertips, guiding you through a proven strategy that maximizes your retirement security while keeping it simple enough for anyone to follow."

The about us section states "The Best Third got started when Irene asked her husband Stefan how much they could safely withdraw from their savings each year when they retire. Stefan is a PhD statistician who specializes in finance. He discovered that most of the conventional advice on retirement spending, like the “4% Rule”, will neither protect you from running out of money, nor provide enough for you to enjoy the best quality of life that you can really afford. He found some authoritative research papers with sound approaches, but more work was needed to turn the theory into a practical solution. So he wrote a paper which describes a practical and validated implementation of the best retirement spending strategy. That paper has been published in the Financial Analysts Journal, a leading peer-reviewed journal for finance professionals.

Their FAQ states:

How does The Best Third get paid?​

Right now, we don’t. The Best Third is completely free to use, and we don’t receive compensation of any kind—from users, advisors, or anyone else.
We’re still in the early stages, testing and refining the service based on real feedback from retirees and advisors. Our priority is to make The Best Third as helpful, accurate, and easy to use as possible before we think about charging for it.
In the future, we may explore different ways to support the service sustainably. But for now, our focus is simple: learn from users and keep improving.

Interesting subject I think. I am still investigating it all at this point. Would love to hear others feed back on the whole subject.
 
Ok I thought this might have piqued some interest here today but zero response. I know it was a bit long but I think it might have caught a few peoples attention and look at the new calculator model The Best Third. I went and played around with it. It's really well laid out and flows nicely with well placed drop down menus. It's simplistic in feel but it allows for a lot of custom inputs. Then it takes all your input and gives you a few options for calculation methods. The results are a chart and a list of annual tbill amounts you should purchase so that you annual spending amount is covered by you pensions, SS and the TBills. It suggest total amount of TBills you need to purchase and the remainder is left in the stock market fund for additional growth. It's an interesting concept that I am still researching.
 
I would have tried it, but it asks for registration before doing anything.

I tried reading one of the links, but it's late.

I do think, if a person has all their spending covered by: pension, SS and TBills. They are golden and can invest the rest. Of course maybe their spending is too low ?

What of the case where 4% (rule this person is trying to improve upon) is just enough to cover spending. Does that mean it's all Pension, SS and TBills and zero stocks ?
 
Registration was pretty easy. Just need to provide email and some basic info. Call me Mr Smith now... 😜
I don't know much about TBills but it is advising me I need to pay more now for them than what they will be worth later? I am not sure I follow that logic?
What they are stating they are improving over the 4 percent rule is a steady, lower risk never run out of money system. From what I am gathering is you buy 30 years worth of TBills, an amount for each year to cover the difference between your spending needs and your pension and SS. Like an annuity would do. Once you calculate how much you need in TBills the rest can stay in the market for additional growth for blow that dough expenses I guess.
 
Last edited:
In terms of publications, ARVA is one of the oldest amortization based withdrawal methods out there.

In terms of what the most recent paper is doing, there are many of us who have been taking this approach, long before this paper was published - it's not a new idea. A very simple and accessible way to do this would be:

1. Cash floor consisting of some combination of SS, Pension, TIPS ladder, SPIA
2. Using whatever your favorite AA is (up to 100% stock like the paper suggests), use VPW to perform the amortization to calculate your withdrawals.

VPW: https://www.bogleheads.org/wiki/Variable_percentage_withdrawal

What I have been doing in retirement is very similar.
1. Cash floor is TIPS ladders, SS (once they start, an SS Bridge till then), dividends from stock in my taxable account
2. 100% stock for the remainder

In practice #1 covers all of our nondiscretionary spending, including baseline taxes. #2 is for any lumpy, discretionary, or contingencies. Today, #2 has only ever been used to pay for the incremental taxes associated with my annual Roth conversions. I use my own amortization spreadsheet for #2 (because I don't agree with some of the assumptions that VPW uses) but at this point all it does is set a budget for the max we could sell in a given year - we're likely to never come close to that.

Even if we had to regularly sell stock for nondiscretionary spending, we still would have used such a method.

Cheers
 
Last edited:
There are a lot of withdrawals strategies and each have pros and cons. There is one size fits all withdrawal strategy. IMHO most withdrawal strategies are trying to either "die with zero" or cutting it too close to goldilocks with a lot of assumptions/calculations. Or in cases like VPW, sacrifice quality of life. We are not OK with either case.

We are using "camel" strategy! Don't try to google it, it's a made up name.

Camel strategy: Have a larger than required risky portfolio with a low (<3%) withdrawal rate. Our aim is to maintain the current lifestyle if things go south or inflate our lifestyle if ride one of the better SORR line.
 
I do think, if a person has all their spending covered by: pension, SS and TBills. They are golden and can invest the rest. Of course maybe their spending is too low ?
This is our 1st world issue... Both my pension and SSDI are enough to cover living expenses, as will DW upcoming pension, So that's covered x3. We need to figure what to do with the excess over the several years while doing Roth conversions before she starts her SS.
 

This is from above article about ARVA. I think it kinda sums up what I might be after in my desire for a less risky spend down phase.
From the article:
We expand on ideas in existing literature to provide a spending formula that is flexible enough to adapt to a retiree’s evolving circumstances, yet which can be implemented by a practitioner or sophisticated retiree using only a spreadsheet. We show both theoretically and empirically how this framework would offer a more efficient use of retirement assets to provide a more economically comfortable retirement for many retirees than do the methods most commonly used by practitioners today.

An ARVA portfolio consists of a riskless asset—in practice a ladder of TIPS—and/or risky assets, such as stocks and conventional bonds. The TIPS ladder guarantees a pre-defined real dollar income stream. Withdrawals are made from the risky portion of the ARVA portfolio as if it is amortized like a fixed-term annuity. Unlike standard annuities, however, the time to maturity may be adjusted periodically and the periodic withdrawals vary with the market values of the risky assets. depleted during the retiree’s lifetime. At the same time, ARVA positions the retiree to spend more of their available wealth with more control over their legacy. While an ARVA enables more total withdrawals than today’s methods that are typically described as “safe,” that advantage comes with the trade-off that withdrawals vary from year to year with fluctuations in the risky asset values.
 
How about the Firecalc strategy? Run Firecalc every year, go to the investigate tab and find your maximum spending rate and then try and spend that amount.

That was gonna be my approach!
That was kinda my thinking too. But I am just looking for something that might help us be even more comfortable over the long haul with our spending and not leaving too much on the table. Not a die with zero mentality but just knowing what we can comfortably splurge on and not worry over. It tough going from frugal mindset to OK we can splurge some now.
Recent example I wanted to get a new bicycle to get some variety in my exercising. It was really hard for me dust off the cobwebs in my wallet and spend 1k. For many a drop in the bucket but it was a weird feeling for me. I think have a better long term plan knowing we have taken some risk off the table might help me. We have a spread sheet of projected long term spending and if we delay SS a few years it would cover our essential spending.
 
Last edited:

This is from above article about ARVA. I think it kinda sums up what I might be after in my desire for a less risky spend down phase.
From the article:
We expand on ideas in existing literature to provide a spending formula that is flexible enough to adapt to a retiree’s evolving circumstances, yet which can be implemented by a practitioner or sophisticated retiree using only a spreadsheet. We show both theoretically and empirically how this framework would offer a more efficient use of retirement assets to provide a more economically comfortable retirement for many retirees than do the methods most commonly used by practitioners today.

An ARVA portfolio consists of a riskless asset—in practice a ladder of TIPS—and/or risky assets, such as stocks and conventional bonds. The TIPS ladder guarantees a pre-defined real dollar income stream. Withdrawals are made from the risky portion of the ARVA portfolio as if it is amortized like a fixed-term annuity. Unlike standard annuities, however, the time to maturity may be adjusted periodically and the periodic withdrawals vary with the market values of the risky assets. depleted during the retiree’s lifetime. At the same time, ARVA positions the retiree to spend more of their available wealth with more control over their legacy. While an ARVA enables more total withdrawals than today’s methods that are typically described as “safe,” that advantage comes with the trade-off that withdrawals vary from year to year with fluctuations in the risky asset values.
Yep, that's essentially the gist of it.

One thing the author only mentioned, but deliberately didn't use in his example - is that this doesn't have to be a "die with zero" strategy. At least for the stock portion, if you're amortizing using the Excel (or Googlesheets) PMT function, one of the parameters is the terminal value of the portfolio. A die with zero plan populates that with $0, whereas somebody who wants to leave a fixed legacy, or just keep some amount in reserve, can target any terminal value they wish. The withdrawal calculation then takes that into account.

(VPW, as it is currently written, doesn't have this feature)

Cheers.
 
How about the Firecalc strategy? Run Firecalc every year, go to the investigate tab and find your maximum spending rate and then try and spend that amount.

That was gonna be my approach!
That one actually has a name: Retire again (and again)
 
That was kinda my thinking too. But I am just looking for something that might help us be even more comfortable over the long haul with our spending and not leaving too much on the table. Not a die with zero mentality but just knowing what we can comfortably splurge on and not worry over. It tough going from frugal mindset to OK we can splurge some now.
Recent example I wanted to get a new bicycle to get some variety in my exercising. It was really hard for me dust off the cobwebs in my wallet and spend 1k. For many a drop in the bucket but it was a weird feeling for me. I think have a better long term plan knowing we have taken some risk off the table might help me. We have a spread sheet of projected long term spending and if we delay SS a few years it would cover our essential spending.
Firecalc has an option under the "Investigate" tab that allows you to set a minimum portfolio value. It's the last item on that page
 
Yep, that's essentially the gist of it.

One thing the author only mentioned, but deliberately didn't use in his example - is that this doesn't have to be a "die with zero" strategy. At least for the stock portion, if you're amortizing using the Excel (or Googlesheets) PMT function, one of the parameters is the terminal value of the portfolio. A die with zero plan populates that with $0, whereas somebody who wants to leave a fixed legacy, or just keep some amount in reserve, can target any terminal value they wish. The withdrawal calculation then takes that into account.

(VPW, as it is currently written, doesn't have this feature)

Cheers.
Thanks for the great replies. I used the new calculator I mentioned in the original post and it generated this list of needed tbills. Not that familiar with them. Why is it recommending 47 bonds and to purchase 64K of them. Am I spending 64k to purchase 47 X 1,000+ 47K worth of bonds? Can't wrap my mind around that part.

1764083007250.png
 
Thanks for the great replies. I used the new calculator I mentioned in the original post and it generated this list of needed tbills. Not that familiar with them. Why is it recommending 47 bonds and to purchase 64K of them. Am I spending 64k to purchase 47 X 1,000+ 47K worth of bonds? Can't wrap my mind around that part.

View attachment 60107
Those are mostly TIPS with a few STRIPS in there, not T-Bills. It looks like they're using STRIPS to cover the gap years in which there are currently no TIPS maturing. I personally wouldn't do that as that does have some inflation risk.

Instead, I would recommend using tipsladder.com to create a full TIPS ladder instead. It will cover the gap years by adding an additional amount of TIPS to purchase for the last year before the gap and the first year after the gap. Then each year, you sell some of those and purchase a new 10 year TIPS. For example in Jan 2026, you will be able to purchase a 10 year TIPS that will plug the 2036 hole, by selling some 2035's and 2040's to make the purchase. Same thing each year until you're done. Less risk that way but it does require some maintenance over the new few years to plug the hole, but fairly straightforward.

I used this tool earlier this year to design two 30 year ladders and one 6 year SS bridge ladder. It was pretty easy to use. Actually making the purchases at my brokerage house took some time, however.

Bogleheads thread where the author of the tool answers questions:

Cheers.
 
Last edited:
Thanks for the great replies. I used the new calculator I mentioned in the original post and it generated this list of needed tbills. Not that familiar with them. Why is it recommending 47 bonds and to purchase 64K of them. Am I spending 64k to purchase 47 X 1,000+ 47K worth of bonds? Can't wrap my mind around that part.

View attachment 60107
Looks way too complicated for my way of doing things. YMMV
 
Looks way too complicated for my way of doing things. YMMV
The simplest I know to implement the overall strategy is as follows:
1) Build a TIPS ladder for the income part. (Or have somebody build it on your behalf). After it's built, enjoy the income from it for the next (up to) 30 years.
2) For the non-TIPS portion, it can be as hard or as easy as you want it to be. If you're spreadsheet savvy download the spreadsheet and just use it - it's fairly easy. If you want simpler, then on the VPW wiki are tables for how much to with withdraw each year. If you can calculate how much to withdraw for RMDs then you can easily multiply the percentage in the table by the value of the portfolio you're withdrawing from. It covers AA's from 30/70 to 70/30

VPW Wiki: https://www.bogleheads.org/wiki/Variable_percentage_withdrawal

Cheers.
 
Last edited:
I recall reading the 26-page Sharkansky article a few months ago when it was mentioned in some thread here. It struck me as just a more rigorous implementation of the general methodology many retirees use. Being a relative newbie at all this, I tried to absorb what I could, then filed a copy of the article away and figured someone might discuss it someday. That seems to be this thread. Got my popcorn.
 
This is Stefan Sharkansky, author of the article and creator of The Best Third. Thank you, 2HOTinPHX, for mentioning my work! With respect for community guidelines against commercial promotion I will write about The Best Third only to respond to specific questions and comments about it.

The Best Third is very much a work in progress. Your comments are appreciated and helpful for improving it.

I'll try to address here the main points about the research article and the website made in previous replies:

1) The strategy is to use TIPS (Treasury Inflation-Protected Securities) for the guaranteed income portion (not T-bills or other nominal bonds)

2) 2HPTinPHX's question about the outputs "Why is it recommending 47 bonds and to purchase 64K of them. Am I spending 64k to purchase 47 X 1,000+ 47K worth of bonds? Can't wrap my mind around that part."

You are paying 64K for 47 $1000 bonds. The $1000 for a bond is its face value, which represents the amount that you are repaid at maturity. The trading price of a bond will typically vary from the face value (higher or lower) depending on changes in interest rates. With TIPS it's more complicated as the inflation adjustments cause the trading price to rise with inflation. Also the $1000 face value for a TIPS means that the repayment at maturity will be $1000 in terms of the dollars at the time the bond was issued. The actual dollar amount that is repaid will be adjusted upwards from there with inflation.

3) big-papa's comment "It looks like they're using STRIPS to cover the gap years in which there are currently no TIPS maturing. I personally wouldn't do that as that does have some inflation risk." he instead recommends " cover the gap years by adding an additional amount of TIPS to purchase for the last year before the gap and the first year after the gap. Then each year, you sell some of those and purchase a new 10 year TIPS. For example in Jan 2026, you will be able to purchase a 10 year TIPS that will plug the 2036 hole, by selling some 2035's and 2040's to make the purchase. Same thing each year until you're done. Less risk that way but it does require some maintenance over the new few years to plug the hole, but fairly straightforward."
Both approaches for dealing with the years without maturing TIPS are imperfect work-arounds and both have some risk. I mitigate (though cannot eliminate) the inflation risk of STRIPS by building in a 3% annual increase in the nominal value at maturity. While the interest payments and principal repayments from TIPS are essentially riskless, the market yields and trading prices of TIPS are volatile as with all other bonds. big-papa's approach is sensitive to changes in the yield curve. I opted for the approach that is less demanding on the user. Implementing big-papa's approach is just a matter of programming. We can always add it to the site if enough users ask for it.

4) big-papa's comment "If you're spreadsheet savvy download the spreadsheet and just use it - it's fairly easy." My paper also mentions that its strategy can be implemented using a spreadsheet. That is true in the context of the article's stated limitation that taxes weren't considered. When you take taxes into account and want to maximize after-tax spending capacity and want to smooth annual variation in tax costs it gets way more complicated -- with different tax considerations for taxable vs tax-deferred vs tax-exempt accounts, and progressive brackets. The after-tax outcomes of withdrawals computed by a straightforward application of a VPW rule are likely to be even more variable than one initially anticipates. The Best Third addresses a lot of the tax considerations already and will be improving in tax efficiency.

I welcome questions and comments, whether on the forum, by personal message or in the website's feedback box.
 
I didn't read the article, but the comments reminded me of Bill Bernstein's recommendation in a book I read a while back. It was How to Retire, by Christine Benz, interviews with 20 economists, FAs, etc. Here is a relevant few paragraphs from the Bernstein interview:

Christine: If we were to think of a base case of a 65-year-old with a spending rate of 4% and a self-professed moderate risk tolerance, how would such an individual approach constructing a portfolio for retirement?

Bill: The single most important thing that a person in that situation should do, since they’re at risk of running out of money if they invest too aggressively, is to invest relatively conservatively. The bedrock of their portfolio should be something that approximates a TIPS ladder, a portfolio of Treasury Inflation-Protected Securities that matures in such a fashion that it pays off their liabilities every year.

Let’s say that you have a person who needs $70,000 a year to live and is getting $30,000 a year in Social Security. Social Security provides a real payoff because the amount you receive increases with inflation. So they have to pay off that $40,000 deficit every single year, ideally. I say “ideally” because it’s hard to realize perfectly in the real world. They want to have $40,000 worth of TIPS maturing every single year for the next 30 years, because that’s as far out as TIPS go. They need $1.2m worth of TIPS—30 years times the $40,000 they need per year. That should be the bedrock of their portfolio.

I’m not saying to put it all in TIPS. It’s not a problem to layer some stocks and nominal bonds on top of that as well.
 
This is Stefan Sharkansky, author of the article and creator of The Best Third. Thank you, 2HOTinPHX, for mentioning my work! With respect for community guidelines against commercial promotion I will write about The Best Third only to respond to specific questions and comments about it.


3) big-papa's comment "It looks like they're using STRIPS to cover the gap years in which there are currently no TIPS maturing. I personally wouldn't do that as that does have some inflation risk." he instead recommends " cover the gap years by adding an additional amount of TIPS to purchase for the last year before the gap and the first year after the gap. Then each year, you sell some of those and purchase a new 10 year TIPS. For example in Jan 2026, you will be able to purchase a 10 year TIPS that will plug the 2036 hole, by selling some 2035's and 2040's to make the purchase. Same thing each year until you're done. Less risk that way but it does require some maintenance over the new few years to plug the hole, but fairly straightforward."
Both approaches for dealing with the years without maturing TIPS are imperfect work-arounds and both have some risk. I mitigate (though cannot eliminate) the inflation risk of STRIPS by building in a 3% annual increase in the nominal value at maturity. While the interest payments and principal repayments from TIPS are essentially riskless, the market yields and trading prices of TIPS are volatile as with all other bonds. big-papa's approach is sensitive to changes in the yield curve. I opted for the approach that is less demanding on the user. Implementing big-papa's approach is just a matter of programming. We can always add it to the site if enough users ask for it.
I only wish it was my idea, but it wasn't. It was originally formulated by some of my fellow bogleheads and the owner of tipsladder.com, kaesler, incorporated it into the tool this year. Only my opinion, but I would suggest that users simply use tipsladder.com's capabilities rather than adding it to yours, though I suspect that yours has other things going on besides just the ladder.

If you look at the default way that tipsladder.com calculates how many extra 2035 and 2040 TIPS to purchase, the risk is, I believe, low. This method uses a linear interpolation of durations between 2035 and 2040 to determine how many extra 2035 and 2040 maturities to buy for each missing rung. Now, there is no fundamental law of nature that prevents the yield curve from bending in any manner between now and January when the first, current gap year can be filled nor for the subsequent years. But there is also no requirement that you use exactly the number of 2035 and 2040 TIPS that were originally reserved for each gap year, either. By the time you need to make the purchase, it could be that a different number of each would work better. There is risk of running out of spares by 2039, if you do that, however. No perfect solution, I'm afraid.
 
Last edited:
So great to have you here to No expert here but I just wanted to say I really like the layout of your calculator. It flows very smoothly in my opinion. Just the right amount of drop down menu options along the way.

We also subscribe to Boldin and the amount of options and charts and graphs there can be overwhelming to us newbie's.
 
I hope this thread continues (especially with Stefan Sharkansky participating!), as I think the "Safety-First" model doesn't get enough attention on these boards. Full Disclosure: DW and I use the traditional balanced-portfolio method. We might, of course, change in the future if our analysis says a different method is better for us.

FWIW, I've posted book reviews on Pfau's Retirement Planning Guidebook (2025 revision) and How Much Can I Spend In Retirement. I've read, but not posted, on Pfau's Safety-First Retirement Planning. Perhaps this will be the motivation for me to re-read and do a review...

For now, let me say that I consider the 30-year TIPS ladder's greatest strength to also be its greatest weakness - it gives a safe, inflation-adjusted (providing that your spend increases at the CPI-U rate) for 30 years maximum, but fails completely in year 31.
 
I hope this thread continues (especially with Stefan Sharkansky participating!), as I think the "Safety-First" model doesn't get enough attention on these boards. Full Disclosure: DW and I use the traditional balanced-portfolio method. We might, of course, change in the future if our analysis says a different method is better for us.

FWIW, I've posted book reviews on Pfau's Retirement Planning Guidebook (2025 revision) and How Much Can I Spend In Retirement. I've read, but not posted, on Pfau's Safety-First Retirement Planning. Perhaps this will be the motivation for me to re-read and do a review...

For now, let me say that I consider the 30-year TIPS ladder's greatest strength to also be its greatest weakness - it gives a safe, inflation-adjusted (providing that your spend increases at the CPI-U rate) for 30 years maximum, but fails completely in year 31.
The 30 year limit only comes into play if you['re 65+, at least for most of us. If younger than that, can always roll forward and add another year as the calendar turns.
 
I appreciate you
I hope this thread continues (especially with Stefan Sharkansky participating!), as I think the "Safety-First" model doesn't get enough attention on these boards. Full Disclosure: DW and I use the traditional balanced-portfolio method. We might, of course, change in the future if our analysis says a different method is better for us.

FWIW, I've posted book reviews on Pfau's Retirement Planning Guidebook (2025 revision) and How Much Can I Spend In Retirement. I've read, but not posted, on Pfau's Safety-First Retirement Planning. Perhaps this will be the motivation for me to re-read and do a review...

For now, let me say that I consider the 30-year TIPS ladder's greatest strength to also be its greatest weakness - it gives a safe, inflation-adjusted (providing that your spend increases at the CPI-U rate) for 30 years maximum, but fails completely in year 31.
i agree, hope this thread gains more traction. It's an interesting topic and I think we have gotten some good feedback already. I thought the spend down phase was supposed to be the easier part. Wrong.... protecting what you were fortunate enough to save for the great unknown future is tricky. Balancing what you can comfortably spend now with what you may need for the long run 😲.
 

Latest posts

Back
Top Bottom