@jkandell Great response as usual. Just looking for clarity, if I bought a 5 year TIP today at todays rate, and assuming inflation stays the same for the next six months, what would be the actual interest payment % in 6 month? For example, if the interest payout was 2.5%, do you get 2.5%/2 of the beginning price of the TIP after the first 6 months??
OK Kevin @boston-spam-02101gmail-com this is your thread, time to cough up your retirement plan so hopefully others will follow 😋. Don't worry, we will be gentle 🙃.
OK Kevin @boston-spam-02101gmail-com this is your thread, time to cough up your retirement plan so hopefully others will follow 😋. Don't worry, we will be gentle 🙃.
Okay, please see plan below.
All feedback would be appreciated.
This is my second attempt to post a summary of my retirement plan. First attempt was lost when I spend too long writing and the connection to the server timed out.
I'll update this message whenever I realize what I missed.
----
Demographics
- Status: Retired
- Current age: 48
- Planning age: 100
- Marital Status: Single
- Family:
- Live-in long-term partner (51F)
- Neither of us have children
- Housing: Renter
Residency & Cost/Standard of Living
- 2025 - Boston, MA ($235/yr, Pre-retirement)
- 2026 - Rochester, NY (est ~$145k/yr)
- 2028+ - Charleston, SC (est ~$165k/yr)
Investment Assets & Asset Location
- Self
- Taxable Equities = $3.7M
- Taxable Bonds (TIPS) = $1.2M
- 401k Equities = $0.7M
- HSA = $26k
- Partner Assets
- Excluded from plan.
Asset Allocation
- Cash: Hold ~1yr of planned expenses in cash for convenience
- Bonds:
- TIPS ladder (7 yrs x $145y/yr)
- A dynamically rolling bond ladder to be gradually depleted during equity bear markets and extended whenever real portfolio value exceeds original plan forecast. Expected to result in increasing equity glide path.
- TIPS ladder (7 yrs x $145y/yr)
- Equities:
- 60% US Large Cap (VOO)
- 35% International Total Mkt (VXUS)
- 5% US Small & Mid Cap (VXF)
- Annuities:
- When/if efficiently-priced annuities that are indexed to inflation return to the US market, I will seriously consider purchasing enough annuity coverage to guarantee my future essential spending (~$145k/yr). I anticipate this to be cost-optimal circa age 80.
Retirement Income
- Deferred compensation: 2026 = $735k, 2027 = $205k
- Social Security_SELF = ~$40k/yr starting age 70
- Social Security_Partner = ~$30k/yr starting at HER age 70
Spending Strategy: Actuarial / Variable Percentage Withdrawal
Legacy Goal: Zero
Spending Targets:
- Spending Floor = $165k/yr
- Target Spending = Max sustainable spending (est ~$350k/yr)
- Planned Spending Evolution By Year
- 2026-2031 - Ramp up from ~$165k/yr to max sustainable spend in 2031
- 2032+ - Max sustainable spending (est ~$350k/yr)
Roth Conversions
- Current Plan
- Pralana Optimization
- Pralana Optimization Goal= maximize net effective legacy
- Optimization performed under consumption smoothing spending strategy
- Plan result: Large conversions from 2028-2031 will take all my Pre-Tax assets (~$0.7-0.8M) to zero by 2032
- Pralana Optimization
- Future Plan
- I will revisit this plan between now and 2028 to see if there are better ways to optimize.
- Specifically, I plan to explore whether it is beneficial to maintain ~$200k (real) in 401k from 2032-2077, continuing to perform annual Roth conversions each year in zero/low tax-brackets
Long Term Care
- Self-insured/planned LTC expense of $100k/yr from age 97-99 (net incremental)
Capital Market Assumptions
- Inflation (Updated periodically based on Forecast from Cleveland Fed)
- 2026-2035 = ~2.3%
- 2036-2077 = ~2.4%
- Real Rates of Return (Updated periodically based on *TIPS & ** A. Damodaran analysis of market pricing implied equity risk premium)
- Cash = 1%
- Bonds = 1.85%*
- Stocks = 5.5%**
----
Current Outlook
- Maximum Sustainable Spending
- Planning SW - Withdrawal Strategy - Simulation Method - Capital Mkt Assumptions - Sustainable Real Spending
- Boldin - Max Spending - Deterministic - Boldin Defaults (Moderate Returns - Average) - 364,730
- Income Lab - GuardRails (Flat Income) - Traditional MC - Income Lab Defaults - 359,000
- Pralana - Actuarial - Deterministic - Custom (NYU, A. Damodaran) - 354,342
- Projection Lab - Variable Percentage Withdrawal - Deterministic - Custom (NYU, A. Damodaran) - 357,100
- Planning SW - Withdrawal Strategy - Simulation Method - Capital Mkt Assumptions - Sustainable Real Spending
- Chance of Success (Scenario: Target Spending = Max sustainable spending, with $165k floor)
- Boldin - No CoS for Max Spending strategy. For fixed spending: 99% @ $165k, 92% @ $235k, 64% @ $350k
- Income Lab - No CoS for Guardrails strategy. 59% chance of underspending, 41% chance of overspending
- Pralana - 86% (90y) -> 72% (100y) (There's only a ~10% chance I'll live past age 98)
- Projection Lab - 92.1%
----
Plan Management
- Plan refreshed annually to determine then current-year affordable spending and Roth conversions
- Bear-market plan calls for re-balancing into equities and accelerated Roth conversions during bear markets
----
Attachments - I'll attach 3 files to the upcoming messages:
- Input1 - Pralana Input file based on Actuarial Spending Strategy
- Results1 - Pralana Results file based on Actuarial spending strategy - in Monte Carlo Charts the Red Lines & Blue bands = Actuarial solution
- Results2 - Pralana Results file based on Consumption Smoothing spending strategy - in Monte Carlo Charts the Red Lines & Blue Bands = Actuarial, Yellow Lines = Consumption Smoothing
@boston-spam-02101gmail-com Kevin you are the definition of FIRE, great job!
First observations:
The average annual cost of nursing home care in the U.S. is approximately $111,325 for a shared room and about $127,750 for a private room. So you might want to up your costs for long term care.
I think assuming inflation will stay around 2.4% over the next 50 years or so is extreme wishful thinking, respectfully 😏. I know retirement planning is or should be dynamic, but this may give you a false sense of long term security.
Not sure I follow how you are going to do Roth conversion. Sounds like just from your 401K. Your going to be in a high tax bracket so how will this work??
Living in the Midwest I am always amazed by the high cost of living on the east cost.
I like your plan for an increasing equity glide path.
I know I am a lone wolf, but I think 35% AA in international equites is a little high, especially since you are looking at a 50 year retirement.
That's it for know. It looks very good!!
That is a lot to go through, I didn't make it through everything. A couple of thoughts:
I do not follow the logic behind Roth conversions at very high tax rates and then ending up with many years with zero taxes. I don't believe that's actually profitable. Is that really the plan Pralana suggested? Note that if you need long term care, LTC would be tax deductible but by pre-paying those taxes with Roth Conversions, you would be out of luck.
Why not a more tax efficient portfolio where you put your bonds preferentially in tax deferred? See the bogleheads.org wiki on tax efficient fund placement. Holding bonds preferentially in tax deferred works better because it matches the taxation of tax deferred as ordinary income with the taxation of the investment. That leaves more stocks for taxable where their primary growth would be taxed as LTCGs. You can model that with Mode 2-Advanced Portfolio Modeling in Pralana, though you should understand that direct comparison to Mode 1 is not reliable as it requires an adjustment in the quantity of bonds in tax deferred to correct for the taxes due on them and that's a very hard to pin down number that Pralana doesn't adjust for. That's also explained in the bogleheads.org wiki if you want to read up on it. A tax efficient portfolio inherently slows the growth of tax deferred and so reduces the need for Roth Conversions.
While you have no legacy goal, you have a partner that may be your heir and she would rather not inherit HSA money as the balance would be immediately taxable. Your stored up receipts become useless, or as my CPA put it, they fly off to heaven with you. So once you have to choose between HSA money and Roth money to spend, spend HSA money up to the limit of your medical receipts (which includes Medicare B and D costs).
The attempt to spend down your bonds in bad markets is not readily modeled in Pralana. Karsten Jaske (retired Fed economist) at EarlyRetirementNow.com did a study using monthly historical data and found that whether it outperforms a constant allocation is more a matter of chance than anything else. For every plan, there is a market that makes it a bad plan and a market that makes it a good one, you just don't know which you will get. Holding a constant asset allocation tries to steer you down the middle, avoiding the worst cases by not trying so hard to optimize things.
@boston-spam-02101gmail-com Kevin you are the definition of FIRE, great job!
First observations:
The average annual cost of nursing home care in the U.S. is approximately $111,325 for a shared room and about $127,750 for a private room. So you might want to up your costs for long term care.
I think assuming inflation will stay around 2.4% over the next 50 years or so is extreme wishful thinking, respectfully 😏. I know retirement planning is or should be dynamic, but this may give you a false sense of long term security.
Not sure I follow how you are going to do Roth conversion. Sounds like just from your 401K. Your going to be in a high tax bracket so how will this work??
Living in the Midwest I am always amazed by the high cost of living on the east cost.
I like your plan for an increasing equity glide path.
I know I am a lone wolf, but I think 35% AA in international equites is a little high, especially since you are looking at a 50 year retirement.
That's it for know. It looks very good!!
@pizzaman, Thank you, this is good feedback.
I appreciate the thoughtful feedback and suggestions!
- Long-Term Care
- @Pizzaman wrote "The average annual cost of nursing home care in the U.S. is approximately $111,325 for a shared room and about $127,750 for a private room. So you might want to up your costs for long term care."
- Thank you for bringing up LTC for (re)consideration.
- How much to save for LTC - The average real lifetime Long-Term-Care cost for a married couple today is ~$300k, spread out over approx 3 years. So I modeled this as $100k/yr for 3 years at the end of my plan. It gets rather complex to model LTC in Pralana based on individuals needs and years.
- Probability of needing ANY LTC is ~70%, with LTC often provided by unpaid family members
- Odds of needing PAID LTC: Men ~40%, Women ~50%
- Average duration of LTC: Men 2.2yrs, Women 3.7yrs
- Probability of needing ANY LTC is ~70%, with LTC often provided by unpaid family members
- LTC Expense v. NET LTC Expense - I think LTC cost should represent the NET EXPENSE of long-term care, and I expect that once I go into LTC my other non-LTC spending will go down significantly. (No more trips to Europe! when you need help with daily living). This is a bit difficult in Pralana when using Actuarial withdrawal strategy, which I use.
- My Approach - I currently plan using a conservative age at death of 100, because the odds that I'll actually live to age 100 is <5%.
- A Better Approach: Age of Death as a Stochastic Variable - What I really WANT to do is set a conservative age of death for linear planning purposes and define LTC expenses as starting 3 years before death while making age at death a stochastic variable in the Monte Carlo simulations. This would be a better way of answering my actual question, which is "will I run out of money before I die?"
- As Pralana currently operates, nearly all the plan failures that happen in Pralana's Monte Carlo simulations due to not being able to afford LTC actually happen after I'm already dead! Definitely not ideal, and I don't just don't trust the Pralana Monte Carlo analysis to be "accurate enough to be useful" beyond about age 90.
- As a result, I'm still experimenting with how best to model LTC costs. Ideas I'm toying around with are moving LTC expenses into my 80s (when they are statistically more likely to occur) or treating LTC expenses as a legacy target.
- Inflation
- Historical inflation:
- Recent history: ~2.5% from 2000-2025
- Long-term average: ~2.3% from 1871-2025
- Inflation Outlooks:
- Current market pricing: ~2.4% average for the next 30 years
- Fed Target: 2.0%
- My approach - I apply the current market pricing expectation (2.4%) in my plan because it's the best number I have. I'm not saying that the market prediction is right, but if I thought that long-term inflation would be 3%+ then I'd be investing my money accordingly; buying less stocks and more TIPS.
- A Better Approach - I really wish that I could model Inflation as a stochastic variable in Pralana, ideally with auto-correlation, so that we can simulate some periods of high inflation in the Monte Carlo analysis.
- Historical inflation:
- Roth Conversions
- Thank you for pressure-testing this, I agree that I need to refine this strategy over the next couple of years before my Roth conversions would start.
- Without Roth Conversions
- If a follow a traditional withdrawal strategy (Taxable, Tax Deferred, Tax Free) without any Roth conversions, then my tax deferred accounts would grow substantially and RMDs would force me into unfavorable tax positions from age ~80-100
- 35% Federal tax bracket, with Social Security taxation and IRMAA expenses
- If a follow a traditional withdrawal strategy (Taxable, Tax Deferred, Tax Free) without any Roth conversions, then my tax deferred accounts would grow substantially and RMDs would force me into unfavorable tax positions from age ~80-100
- Current Plan: Pralana-recommended Roth Conversions: (Front Loading)
- The Pralana optimizer recommends for me to front-load my Roth conversions, even up into the ~24-32% bracket depleting my Tax Deferred (401k/IRA) balance down from ~$900k at age 50 to $0 before RMDs would start. Depending on my starting conditions, capital market assumptions, and assumptions about future single v. married tax filing status the conversion period can be as short as 4 years or as long as 20.
- Super-Contribution & Tax-Drag Elimination: I believe that the main factor driving Pralana towards Front-Loading my Roth conversions is the effect of Super-Contribution & Tax-Drag Elimination that results from paying the Roth Conversion taxes from my taxable accounts.
- A hypothetical $100k Roth conversion
- $100k in ROTH is more valuable than $100k in Pre-Tax 401k
- Converting $100k from Pre-Tax 401k into Roth triggers an income tax expense (e.g., 24%)
- Paying the Roth conversion income tax expense from my taxable accounts allows me to move all $100k from my Pre-Tax 401k and into my tax preferred (ROTH) accounts
- Thus performing a $100k Roth conversion and paying the tax from my taxable accounts is better than either
- The no-conversion case where I will have $100k in my tax-free (which will eventually be taxed some future income tax rate), and $24k in my taxable account that will experience tax drag
- The conversion case where I pay the conversion taxes via withholding from the conversion, in which case I will have $76k in my tax-free Roth account and $24k in my taxable account that will experience tax drag
- By using outside cash, I have effectively "contributed" an extra $24k into a tax-free shield.
- Another way to think of this is that I have now "shielded" $24k from future tax drag on growth, and with ~50 years in my planning horizon this has a material impact
- A hypothetical $100k Roth conversion
- Future Plan: Lifetime Tax Diversification
- Under the Pralana-recommended Roth conversions plan (Front-loading) I'll have fully depleted my Tax Deferred 401k accounts by the time my taxable accounts are depleted (approx age 80), and from that point forward I'll have zero taxable income from age 80-100.
- I need to study this further, but I suspect that it may be beneficial for me to maintain on the order of ~$100-300k in my Pre-tax 401k well into my 90s just so that I can make modest withdrawals every year at very low tax rates from age 80-100 (the standard deduction for seniors is ~$18k/yr)
- Pizzaman wrote "The average annual cost of nursing home care in the U.S. is approximately $111,325 for a shared room and about $127,750 for a private room. So you might want to up your costs for long term care."
- Thank you for bringing up LTC for (re)consideration.
- How much to save for LTC - The average real lifetime Long-Term-Care cost for a married couple today is ~$300k, spread out over approx 3 years. So I modeled this as $100k/yr for 3 years at the end of my plan. It gets rather complex to model LTC in Pralana based on individuals needs and years.
- Probability of needing ANY LTC is ~70%, with LTC often provided by unpaid family members
- Odds of needing PAID LTC: Men ~40%, Women ~50%
- Average duration of LTC: Men 2.2yrs, Women 3.7yrs
- LTC Expense v. NET LTC Expense - I think LTC cost should represent the NET EXPENSE of long-term care, and I expect that once I go into LTC my other non-LTC spending will go down significantly. (No more trips to Europe! when you need help with daily living). This is a bit difficult in Pralana when using Actuarial withdrawal strategy, which I use.
@boston-spam-02101gmail-com, There is a genuine dilemma with how to handle LTC. The Genworth site gives reliable data on the current costs of different types of LTC. (In my case I use two years of assisted living + 1 year of more intense care, in AZ = ~350k pv per person total.) You can work in a projected inflation (e.g. 3% more than CPI) if you think current trends will continue.
But... when you run Vanguard's LTC calculator (which was developed in collaboration with Genworth), it gives a much lower figure (~30k total each) based on the costs of everyone, not just those needing care. The Vanguard calculator includes the top 25%th and 10%th and 5%th percentile costs. I recently have started using the top 25th costs in my plan instead of the full "average cost of those who use it" I had formerly used.
A case can be made for either approach: using the 25th or 10th percentile probabilistic costs makes sense since odds are high that's all you'll need; or, alternatively, "covering it all" also makes sense given individuals are not "insurance pools", and each of us has only one trajectory with no "do overs".
Obviously the Vanguard approach is a riskier strategy, because it it turns out you are in the top 10% percentile of users, you have not planned for it. In my case, my actuarial discretionary withdrawal is large enough I feel confident in covering the shortfall should I turn out to be in the top 10% and not the top 25% I set aside. In other words, I'd prefer a higher consumption in the many years prior compared to "full coverage" of long term care on the chance it's needed. But I don't think there is a "right" or "wrong" to this analysis. Kevin, given your detailed shared plan, I venture you might consider making the same decision as I to cover the <25th highest percentile costs, using discretionary monies to fill any gaps if needed.
One thing I really like about the actuarial method we both use is that you don't need to know exactly when you'll be needing the LTC: I put it in a bit early just to maximize the npv. As long as it's accounted for in your plan, you won't be spending it away.
- My Approach - I currently plan using a conservative age at death of 100, because the odds that I'll actually live to age 100 is <5%.
- A Better Approach: Age of Death as a Stochastic Variable - What I really WANT to do is set a conservative age of death for linear planning purposes and define LTC expenses as starting 3 years before death while making age at death a stochastic variable in the Monte Carlo simulations. This would be a better way of answering my actual question, which is "will I run out of money before I die?"
- As Pralana currently operates, nearly all the plan failures that happen in Pralana's Monte Carlo simulations due to not being able to afford LTC actually happen after I'm already dead! Definitely not ideal, and I don't just don't trust the Pralana Monte Carlo analysis to be "accurate enough to be useful" beyond about age 90.
- As a result, I'm still experimenting with how best to model LTC costs. Ideas I'm toying around with are moving LTC expenses into my 80s (when they are statistically more likely to occur) or treating LTC expenses as a legacy target.
I've long admired opensecurity's longevity weighted NPV calculations. I've been experimenting with doing these same types of calculations on my Pralana downloads, with things like LTC and Roth conversions. In theory this would give me a better NPV for LTC than any of the ways above, or how I'm doing it now.
A Better Approach - I really wish that I could model Inflation as a stochastic variable in Pralana, ideally with auto-correlation, so that we can simulate some periods of high inflation in the Monte Carlo analysis.
The best I've found is doing exactly that in Portfolio visualizer monte carlo. However, it only uses 10 years of data (in the free account) so is not very useful in this regard. I do my own monte carlo in excel sometimes but the calculation time is burdensome, trying to avoid that.
I will give our plan in two parts. This first part is the background/theory. Our withdrawal plan isn't necessarily suited to others, but based on our circumstances. In particular, my wife and I are working class without intergenerational family wealth, and have modest retirement savings. The lack of "safety net" if we're wrong has heavily influenced our accumulation and our "safety first" decumulation strategies.
Me 62, wife 55. Me retired last year, wife working a few more years, relatively lcol. One 16 year old kid, with no college plans at the moment (but we would pay in-state tuition if they did). ACA for health. 340k home paid off. We live very modestly, only spending about $30k above and beyond taxes and medical and house maintenance.
Our accumulation strategy since 2010 was basically to invest everything beyond our emergency fund at 50/50. This wasn't arrived at via some fancy analysis. At the time I didn't know much about economic history, and had no idea if stocks would do better than bonds, so just hedged my bets with the guiding principle of "one dollar kept safe for every one dollar put at risk". We would have obviously made more in retrospect during what turned out to be the biggest bull in history. No one I knew in my family or friends "invested". (If I knew then what I know now, i would have been 100% stocks when starting with nothing to lose.) I stopped working a year ago (after reaching a funded ratio of 1.6)to be master of my domain, but my wife wants to continue her job because "what else will I do". I've budgeted her in for another 3 years.
Now on the cusp of decumulation, we will use a "safety first" actuarial approach. The basic idea is to cover all our lifetime's essential expenses with SS + Pension + individual TIPs (supplemented by bond funds), and then invest "the rest" beyond that at some AA--at moment 70/30--amortizing a withdrawal each year. (Knowing our "essentials" are safely covered allows more risk with "the rest".) I use Pralana to help model all our essential expenses through my wife's death; plus I add into the "essentials" bucket the $30k "dignity floor" we spend presently on bills and lifestyle. "Essentials" also includes some for Long term care, of course, though that is a huge wildcard, as is health care in general. I am trying to cover the whole floor beyond SS with a ladder--not just "bridging to social security".
The withdrawal plan is that each year I will (A) calculate how much to reserve for (future and present) essentials, and (B) how much to take out for discretionary (amortization). Each January I will add the value of our portfolio + NPV of the future social security stream + pension stream, and subtract out the NPV of the essential expense stream + dignity floor through death as ascertained by Pralana. This tells me my total discretionary wealth, taking into account the time value of money.
Each year I will do a simple amortization PMT formula in excel on this discretionary bucket to calculate how much to withdraw for this year beyond "paying the bills":
PMT(3.1%, 45, ,-p,0,1),
where p is the value of the discretionary bucket, 45 is my wife's life expectancy to the 10th percentile, and 3.1% is the expected return of my discretionary bucket at the beginning of the year. (Note this return/discount will of course change every year, and is inversely related to the CAPE10.)
I will then run the same equation a second time but with the 30y TIP rate as the discount rate, currently 2.4%. I will withdraw discretionary money between those two results. The low boundary is based on the idea that it is unlikely over the long run I would earn less than the 30y TIPs yield.
Note the above isn't the same as VPW or Pralana's actuarial method, since it looks at my remaining future income and expenses in doing the actuarial equation, not just my current portfolio and expenses. I think it is closer to Pralana's consumption smoothing or Maxifi's approach.
to be continued part two
A little off topic but if looking at long term investing in International funds:
VXUS - Vanguard Total International Stock Market Fund ETF started in 2011. Total return since that time was 54%. S&P 500 during the same time frame returned 421%
VTRIX - Vanguard International Value started in 1983. Total return since then was 68%. For S&P 500 3,786%.
VWIGX - Vanguard Growth Fund started in 1982. Total return since then 539%. For S&P 500 about the same as above.
Just something to think about 🤔.
A little off topic but if looking at long term investing in International funds:
VXUS - Vanguard Total International Stock Market Fund ETF started in 2011. Total return since that time was 54%. S&P 500 during the same time frame returned 421%
VTRIX - Vanguard International Value started in 1983. Total return since then was 68%. For S&P 500 3,786%.
VWIGX - Vanguard Growth Fund started in 1982. Total return since then 539%. For S&P 500 about the same as above.
Just something to think about 🤔.
I remember a long thread of yours about this. For me it comes down to this: looking backwards, it's no contest. But I'm investing looking forwards. Equity returns are a mix of (1) the unique nature of the equity structure itself (with capital gain being the primary force, and a premium over bonds reflecting higher risk of that), (2) the geo-politics of where certain companies are located. The US dominated the world since WW2. Before that other countries dominated--it is just another empire. And in the remaining 50 years of my plan I am not willing to bet the US will continue will dominate over e.g. China, India, Scandinavia or who knows where. Perhaps no country will dominate. The Equity Premium is the main thing I am after when I invest in stocks; and that is somewhat stable across time (see McQueen, and see Jorda et al.). I am less interested in the second part, gambling on the country that will dominate the world economy. It is a core principle in investing to never put all your eggs in one risk basket.
That is a lot to go through, I didn't make it through everything. A couple of thoughts:
I do not follow the logic behind Roth conversions at very high tax rates and then ending up with many years with zero taxes. I don't believe that's actually profitable. Is that really the plan Pralana suggested? Note that if you need long term care, LTC would be tax deductible but by pre-paying those taxes with Roth Conversions, you would be out of luck.
Why not a more tax efficient portfolio where you put your bonds preferentially in tax deferred? See the bogleheads.org wiki on tax efficient fund placement. Holding bonds preferentially in tax deferred works better because it matches the taxation of tax deferred as ordinary income with the taxation of the investment. That leaves more stocks for taxable where their primary growth would be taxed as LTCGs. You can model that with Mode 2-Advanced Portfolio Modeling in Pralana, though you should understand that direct comparison to Mode 1 is not reliable as it requires an adjustment in the quantity of bonds in tax deferred to correct for the taxes due on them and that's a very hard to pin down number that Pralana doesn't adjust for. That's also explained in the bogleheads.org wiki if you want to read up on it. A tax efficient portfolio inherently slows the growth of tax deferred and so reduces the need for Roth Conversions.
While you have no legacy goal, you have a partner that may be your heir and she would rather not inherit HSA money as the balance would be immediately taxable. Your stored up receipts become useless, or as my CPA put it, they fly off to heaven with you. So once you have to choose between HSA money and Roth money to spend, spend HSA money up to the limit of your medical receipts (which includes Medicare B and D costs).
The attempt to spend down your bonds in bad markets is not readily modeled in Pralana. Karsten Jaske (retired Fed economist) at EarlyRetirementNow.com did a study using monthly historical data and found that whether it outperforms a constant allocation is more a matter of chance than anything else. For every plan, there is a market that makes it a bad plan and a market that makes it a good one, you just don't know which you will get. Holding a constant asset allocation tries to steer you down the middle, avoiding the worst cases by not trying so hard to optimize things.
Thank you @ricke, This is helpful feedback!
Below are a few comments about some of the topics you brought up
Paying for LTC with Pre-tax accounts - I think that you're right that holding some $$s in my pre-tax accounts might help avoid tax expenses on the portion of LTC costs that are qualified medical expenses (~100% in nursing homes, ~50% in assisted living, and ~75% in in-home care). I'm currently ear-marking HSA funds for that purpose, but I think you're right that holding some funds in my pre-tax accounts would also help.
Holding Bonds in Taxable Accounts v. Tax preferred 401k/Roth - Thank you for spotlighting this topic and providing useful suggestions and references regarding how this can be improved. I will need some time to follow up on your suggestions and perform my own analyses.
About a year ago I briefly worked with a professional financial planner and I asked him this same question, whether I should hold my TIPS ladder in my 401k or my Taxable account, and he gave me the default answer that "holding your income bolds in pre-tax is better" just like you and I intuited, but when I made him do the analysis for me and show me exactly how much better it would be he was surprised to discover that holding the bonds in the Taxable account was actually better, resulting in a slightly higher Legacy and a Chance of Success that was a couple of percent higher.
- Why my financial planner told me that holding bonds in Taxable accounts was better for me
- If the choice is to hold new, incremental bonds in Taxable or Pre-tax/Roth accounts, then the Pre-tax/Roth account will win every time due to Tax Efficiency. However, that's not really the choice I face.
- Because I have a fixed amount of dollars available in my Pre-tax/Roth accounts, every Pre-tax dollar that I hold in bonds means that I hold one less Pre-tax dollar in Stocks.
- Upon running the numbers, we discovered that it was better for me to pay the higher income taxes from holding my TIPS bond ladder in the taxable account rather than give up the higher Tax-free growth from having more Pre-tax/Roth dollars allocated to stocks.
It's something like the trade-off between the cost of tax on the TIPS interest for 7 years vs the "cost" of 7 years of lower returns in the Pre-tax account followed by having a smaller Pre-tax/Roth account from years 8 until death.
I'll admit that I'm not 100% sure of this outcome, so I need to dig into it deeper and may change my TIPS asset location if his finding was wrong.
Spending the HSA - Thank you, this is good advice. I didn't realize that medical receipts from HSA's become useless if the the HSA is inherited by a non-spouse. Wade Pfau commented in passing once that he expected his HSA heir(s) to benefit from his meticulous record-keeping of his medical receipts, and I didn't realize that benefit didn't apply to a non-spouse. I had been toying with the idea of using HSA to pay for Medicare Premiums or medical-related Long-Term-Care expenses, but your insight put greater urgency on this question.
Modeling Bond Ladders in Pralana - Thank you, I've read a decent amount of Karsten's work, but probably not on this topic. I'll go dig into it!
It's really a tragedy that NO consumer retirement planning software, including Pralana, has a good way of modeling a bond held to maturity. This makes it hard to properly test our bond related risk-hedging strategies.
Wade Pfau and Michael Kitzes have published some research on the topic of Bond Ladders v. Rebalancing fixed-allocation portfios and they concluded that DEPLETING BOND LADDERS (which create rising equity glide-paths) perform better than FIXED ASSET ALLOCATION PORTFOLIOS WITH REBALANCING. If I recall correctly, after further research Wade concluded that DYNAMICALLY ROLLING BOND LADDERS perform even better than DEPLETING BOND LADDERS and he published some optimal guidelines regarding when to role v. deplete the bond ladder such as: "If total real assets in your plan are greater than what they were expected to be at this point, then re-fund/roll your bond ladder. But, if your real total portfolio value is less than what it was expected to be, then do not re-fund/role the ladder, deplete it by 1 year." Wade also identified an approximate rule of thumb guideline that worked almost as well: "If your nominal portfolio value is above/below what it was at retirement, then Do/Do not re-fund the ladder."
Here's my version of the logic supporting supporting DYNAMICALLY ROLLING BOND LADDERS:
- 100% Equity Portfolios - Historically the highest average lifetime spending outcomes have come from 100% equity portfolios
- However 100% equity have high year-to-year volatility and taking withdrawals during bear markets can deplete the portfolio too quickly, leading to SORR and creating Chance of Success failures
- Therefore, we may choose to start from a 100% equity allocation conceptually and determine what deviations should be made to reduce probability of failure, SORR, year-to-year volatility, drops in spending, and other undesirable risks of the 100% equity portfolios
- Contextual note: 85% of all bear markets in US history have recovered within 8 years. The longest bear market + recover took 14 years
- Fixed-allocation Equity/Bond Portfolios - The traditional approach to mitigating bear-market risk is to choose a fixed allocation of equities and bonds in your portfolio (e.g. 60/40) and to re-balance annually to maintain that allocation
- However, a fixed allocation equity/bond portfolios creates a lifetime drag on expected return due to the lower expected returns from bonds than from equities creating lower expected returns in every year of retirement
- Fixed Duration, Depleting Bond Ladders - A fixed duration, depleting bond ladder can mitigate the risk of a bear-market + recovery lasting up to X years by holding a fixed duration bond ladder of X-years.
- A fixed duration bond ladder is an extreme case of the bucket approach where you never refill the bond bucket. This may be the type of bond ladder that Carsten was researching, but I'm not sure.
- If the bond ladder duration is "short enough", then a fixed duration, depleting bond ladder can create less lifetime return growth drag than a perpetual fixed equity/bond allocation while mitigate the risk of bear markets IF the recovery comes before the bond ladder fully depletes (Years_for_Recovery < Bond_Ladder_Duration).
- A typical fixed-duration bond ladder (5-12 years) requires a lesser portion of your investment portfolio to be invested in bonds than would be needed for a typical a fixed allocation portfolio (e.g., 60/40)
- A Fixed Duration, Depleting Bond Ladder creates a rising equity glide path as the the bond ladder "rung" years deplete every year and the equity portion continues to grow.
- Holding a fixed duration bond ladder for 5-10 years at the beginning of retirement can provide "good" protection against SORR when using a fixed withdrawal spending strategy, because IF a bear market occurs in the first 5-10 years after retirement (the fragile decade), then the bond ladder protects against SORR and prevents over-depleting the portfolio during those down years. Furthermore, if a bear market doe not occur in that 5-10 year period then by the time that the bond ladder is depleted then the retiree's equity portfolio should have appreciated materially and should provide additional buffer against bear markets.
- Dynamically Rolling Bond Ladders - A dynamically rolling bond ladder improves on the fixed duration bond ladder a bit in that you start with an X-year bond ladder and only deplete the ladder "rungs" during bear markets / not-yet-recovered years. Otherwise, you "roll forward" the ladder.
- A Dynamically Rolling Bond Ladder creates a rising equity glide path as the the bond ladder "rung" years deplete during during bear markets / not_yet_recovered years
- A dynamically rolling bond ladder can offer "similar" protection to fixed equity/bond asset allocations or long-term depleting bond ladders, but at a lower cost
- A dynamically rolling bond ladder allocates fewer dollars to bonds than a fixed duration bond ladder providing lifetime protection
- Bond ladder duration X (e.g., 8 years) << A fixed duration bond ladder covering the entire lifetime planning lifetime (e.g., 40 years)
- AND
- A dynamically rolling bond ladder allocates fewer dollars to bonds than a fixed allocation (higher equity allocation)
- Bond ladder $$s < the equivalent $s needed for a fixed allocation portfolio
- e.g., perhaps a 40 year retirement with a 60/40 portfolio -> bond allocation $$s equivalent to a ~16 Year bond ladder)
- Bond ladder $$s < the equivalent $s needed for a fixed allocation portfolio
- A dynamically rolling bond ladder allocates fewer dollars to bonds than a fixed duration bond ladder providing lifetime protection
- Therefore Dynamically Rolling Bond Ladders offer good, but not perfect, protection against bear markets / long recoveries at a reasonable cost.
Kevin mentioned that he was planning for a 50 year retirement and that he has an AA of 35% international stocks. I just provided some factual information that people can use however they want. That's all.
Just a quick note on HSA:
You can use your HSA to pay for Medicare premiums tax-free, including Part B, Part D, and Medicare Advantage plan premiums, but not Medicare Supplement (Medigap) premiums. You can also reimburse yourself for past Medicare premiums as long as you have receipts.
Kevin mentioned that he was planning for a 50 year retirement and that he has an AA of 35% international stocks. I just provided some factual information that people can use however they want. Thanks all.
Thank you @Pizzaman
I agree that Warren Buffet and 50+ years of historical global stock market returns support your position that holding 100% US equity is the right bet.
However, the vast majority of academic researchers, investment professionals, and retirement planning investment communities disagree with this US-only approach given the currently high US market valuations, pragmatic challenges to the US repeating it's 20th century global success again in the 21st century, and the resulting risks of lower US returns in the future, therefore I'm personally not so confident in US exceptionalism over the next decade that I'm willing to go "All-in" on the USA for my retirement portfolio, so I choose to diversify internationally.
But again, you may be right!
Thanks Kevin! It's not so much who's right or wrong, it's what you are comfortable with and work's best for you. I tend to come across a little strong with my opinions, that's just my writing style. Feel free to fire back at me 😘. Please limit your barrage to water balloons 🤣.
Part two of my plan, getting into the details. Theory doesn't quite equal reality. (See post above for the "safety first" theoretical model I am using. And, no, I will not be pulling down my pants with exact amounts.) I've listed some areas where I can use some advice.
Me 62, wife 55. Me retired last year, wife working a few more years, relatively lcol in desert southwest. One 16 year old kid, with no college plans at the moment (but we would pay in-state tuition if they did). ACA for health. 340k home paid off. We live very modestly, spending about $30k above and beyond taxes and medical and house maintenance.
As explained in my first post, I have two buckets: one to cover essentials the other for discretionary. At the moment about 70% of my assets are discretionary and about 30% are reserved for liabilities, with 46% overall in equity. (My glidepath will likely fall over time, see below.) We're in the 12% tax bracket through my death, then 22% afterwards.
Liability Bucket. Intension is to reserve my essentials in 100% safe assets, ideally a TIPs ladder. At present I have about 80% of the npv of my essentials through my age 104 (wife's 10% longevity per LongevityIllustrator) in a TIPs/ibond ladder going 2026-2043. The remaining 20% is in VGIT, VTIP, BND, VTP and money markets.
The rub:
- The duration of this ladder is only about 8 years, and there are twenty more years needed. But I am reluctant to buy more bonds maturing when elderly, so am keeping it liquid in misc bond funds. Plan is to buy an annuity, or more likely just keep it in bond funds. This is influenced by the plan to move all accounts to fund-of-funds after I die (see below). But I don't like the interest rate risk; what happens if bonds return 0% when I try to extend them? And worse, what happens if we have another "bond fund crash" like a few years ago, which defeats the purpose of the "essential" bucket?
- The other issue is that about a third of my ladder are taxable Ibonds coming due 2030-33 with a very large capital gain. My original plan was to use these to pay my son's tuition (since treasury interest is tax free if used for qualified educational expenses). But life threw a twist: son has no intention of going to college, nor do we and at 16 grandchildren are just a dream. So we may still shove them into a 529 as they mature to postpone the tax. As other have mentioned, I wish there was a way for me to model these individual bonds in Pralana. Worst case: I patriotically pay a heck of a lot of taxes celebrating the 3.6% real interest I've earned for 30 years.
Remaining discretionary bucket. Intention is to invest the remainder at 60-70% AA (tbd), and to amortize this bucket each year for a variable discretionary withdrawal. Since we live simply, we anticipate much of this being given to charity, promoting things in our city, and to causes we believe in, and helping our son.
My overall asset allocation (including the ladder) is currently 46% stocks, 23% bonds, 23% cash. Within equities, my intended AA is: 44% tsm, 31% total international, with the remaining 25% split evenly between AVDV (small value intl), US SV, Vanguard Multifactor ETF, and Emerging Markets funds.
The rub:
- I thought I was so smart in placing most of my stocks in taxable. About three quarters of all my stocks are in my Taxable account, and they have a huge unrealized capital gain. The unintended effect is I've not rebalanced at all the last few years (which has given me both too much US stock, and too many stocks in general). Since I am on ACA through 2030 when I go on Medicare, I've not wanted to raise my income even at the 0% ltcg rate. I am not sure how to model/weigh tax-gain harvesting vs roth conversions vs ACA subsidy since none of the programs do all that.
- I am not sure what AA to keep my discretionary bucket. I am exploring monte carlo (TPAW's) and the "merton share". The latter is attractive since it adjusts AA to maintain the same risk aversion as the equity premium changes, but I don't trust it.
Social Security A very small non-inflation adjusted pension starts this year. I like OpenSecurity's methodology for social security (longevity weighted npv); and following its advice we plan on wife claiming at her age 63 and me at age 70 (roughly in the same year).
Falling glide path: The amounted needed to cover the essentials bucket rises over time until my potential LTC and I die. After that point, the amount needed for essentials gradually decreases. So my overall AA will decrease over time until my death, after which it will stabilize.
Housing: We own our home, but will almost certainly want to downsize after our son leaves. And maybe even rent from that point on. This is an unknown, but I need to explore more in Pralana.
Health care: I did not count on ACA going to hell. My wife is still working as a sole proprietor, and to minimize ACA premiums we are contributing 40k each year to her solo 401k, which is more than I had planned. The impact of this is that just when our taxes are at their lowest (10% bracket) I can't really do conversions or even tax gain harvesting (see above), and am kind of trapped in limbo of aiming for a low MAGI. Pralana models some of this, but doesn't allow me to do tax gain harvest vs conversions vs ACA.
Long term care: Huge wildcard. Genworth shows the "typical" 2 years of assisted living plus another 1 year of higher intensity care comes out to about $700,000 for both of us. (PV=$400,000, and if we work in the price of selling our home, this comes to PV of about $200,000.) I'm not sure about the validity of my reasoning; but I've elected to instead utilize the Vanguard LTC estimator at their "25% percentile of all people like you", which is about a third of those figures. That way my yearly discretionary allowance won't be as impacted for something we may not need. However, on the off chance we are in the top 10%, I'll utilize monies from my discretionary bucket. (In that event I likely won't care about pleasures.) The rub of course is: with bad luck I may not have the discretionary bucket to cover.
Roth conversions: My original purpose in buying Pralana was to explore conversions, and after analysis my conclusion is that they are not worth it for me. Part of this is due to our plan to use QRCs charity to offset some of the RMD, but most of it is: it just isn't worth it in the 12% bracket. (According to Pralana after I die we will still be in the 12% bracket, despite the "widow tax.") I was surprised to see Pralana not recommending conversions even if we didn't use QRCs. In the end the arbitrage of 10% > 12% is not very valuable. However, this is not taking into account the fact taxes may rise to 15%, or other contingencies.
Simplifying when elderly: Instructions will guide my my wife to (a) shift everything in all accounts to the same single fund-of-funds (e.g. a LifeStrategy) in all accounts. By then there shouldn't be a large taxable account left, since we are spending it down first; but simplicity will trump tax efficiency in any case. (b) I will guide her to use the RMD formula once a year to withdraw an amortized amount of our portfolio for all expenses (essential + discretionary). At this point we will switch from consumption smoothed discretionary withdrawal to a simpler RMD-style "one big pot" withdrawal. Coming up with a simple two page plan for her is proving more challenging than I had anticipated.
