Thanks for the detail. Not sure I completely followed all your logic but if it works for you great. It does seem to be complex and I am curious how you feel about your cognitive ability to implement it as you age and what your plan is if you pass early and your spouse takes over the plan.
I am going to go out on a limb and ask a question that seems to be emotionally charged at times. What are your thoughts and have you modeled covering some (portion of essential, all of essential, all essential + some discretionary) expense with a single premium annuity. I am sure you have read all the research that explains both pros and cons.
If this topic has any traction we should start a new topic.
I will be out of touch for a few weeks traveling in Africa so I won’t start it. I am spending my money in early retirement - no budget just an unknown life span I doubt I will want to do this when I am in my 80s.
Although I have not purchased a SPIA, I have elected to take pensions as annuities to help overcome my behavioral problem taking withdrawals as I have explained before.
I know of an approach that does not include all income taxes as essential expenses. They include only the taxes associated with the income that covers essential expenses. Other taxes associated with income such as RMDs not required to cover essential expenses as discretionary taxes. The thinking is that if your IRA that is giving you the RMDs goes to zero, those taxes would also go to zero. You need to pause and think about this for a while but I see their point.
Interesting thread.
Like many here, I started in the ’90s with spreadsheets to track our company 401(k) holdings and performance, plus simple automation to rebalance when thresholds tripped. Choices in the 401(k) were limited—maybe 20 stock and bond funds. After our kids moved out, my wife went back to school, became an RN, and that income was a game-changer. We maxed her employer 403(b), though its menu was just as constrained. With the extra income, we finally opened and maxed out tax-free Roth IRAs at Fidelity, which gave us real investment options. As HSAs became available, we elected HDHP coverage to gain access to those tax-free accounts, invested the HSA money instead of spending it, and periodically moved it to HSAs at Fidelity. I managed the tax-deferred and tax-free accounts as one portfolio in spreadsheets. But with Roth and HSA limits stuck around $4k–$6k, the bulk of assets remained in employer plans, which made implementing our design tricky. My focus then was learning investing and accumulation—not yet retirement.
In the late 2000s I began using Sound Mind Investing (SMI) to guide portfolio design—momentum principles within a risk-tolerant allocation. As a member I also got MoneyGuidePro (MGP) access for $50. It’s an advisor tool I’d rate “medium fidelity,” but it made retirement planning approachable for my wife with clear inputs, charts, and probabilities. At 50 I added catch-up contributions to my 401(k), and my wife did the same six years later. We opened a joint taxable brokerage at Fidelity and added what we could. I shifted my learning from accumulation to decumulation and tax management. I still ran everything as a single portfolio in spreadsheets and modeled withdrawal strategies (4% rule, guardrails, etc.) with mixed success—taxes make withdrawal planning complicated fast.
In 2015, my employer was acquired, and I was thrilled to roll my 401(k) into an IRA at Fidelity—finally, broad flexibility for most of our funds. I kept maxing the new employer 401(k), and they added a small “pension” (effectively a company-funded 401(k)). As I dug into decumulation, I learned about safe withdrawal rates, risk-parity design, and tax-efficient asset location. We used MGP and Fidelity’s tools to shape our retirement plan. By 2020, after years of strong returns, I wanted to “take some chips off the table.” We defined a comfortable income floor and bought a simple FIA to cover that income floor in combination with our combined SS. Last year, I rolled that annuity into another annuity paying nearly double the original (not inflation-indexed). With the floor secured, the remaining portfolio will bridge our income needs until we start SS (my wife at 62 and me at 70) and provide for long-term care. We’re both healthy, have a strong family support system, and our sons are financially secure (one mortgage, no other debt). We don’t have any legacy goals; we’d rather make meaningful gifts to children, grandchildren, and charities while we can see the impact.
In 2020, I discovered Pralana, tested Bronze, then bought Gold. I retired in 2022 and rolled my 401(k) and company “pension” into my IRA; my wife retired the next year and rolled her 403(b) into her IRA. With everything at Fidelity, we finally had full flexibility. I shifted to a simple, risk-parity-inspired allocation: 42% stocks (21% large-cap growth, 21% small-cap value), 26% long-term bonds, 16% gold, 10% alternatives (managed futures and commodities), and 6% cash. I rebalance if an allocation drifts more than 20% from target, let dividends/interest refill the cash bucket, and only check on the 15th of each month to see if rebalancing is needed. Because we draw from cash for living expenses, cash deviations don’t trigger rebalancing until January each year. I also implemented asset location principles to manage taxes. First, I fill the tax-deferred accounts (tIRA) with all the gold, alternatives, & bonds I can. Next, I fill the tax-free accounts (Roth & HSA) with all the equities possible. Then the taxable account takes the rest. Our split between tax-deferred/tax-free/taxable is currently 50/25/25. I have a spreadsheet with VBA code to input the CSV from Fidelity with our total portfolio and calculate the variance from targets (with indicators) and amounts needed to rebalance. I'm working with ChatGPT to refine the spreadsheet so it will explicitly outline how much to change each asset in each account to maintain optimal asset location. It’s simple, mostly static, and easy to manage.
I’ve since moved to Pralana’s online version and love the cash-flow planning and reports—they help my wife see the whole picture. The forum is excellent, and Stuart & Charlie are incredibly responsive. My wish list: deeper historical data across more asset types (à la Portfolio Charts) and a fix for Georgia’s retiree tax rates (I have an open feedback). Overall, I’m happy. Each January, I still update Fidelity’s retirement planner and MGP, then compare both to Pralana as a cross-check. Over time, I can see us retiring the other tools. And that’s how I ended up here.
OK, what I saw in the MaxiFi article is just that it reduced your discretionary spending during the years of Roth Conversions, it was my guess as to why. In any case, that seems to be the opposite of what folks want - requiring less consumption up front to get more later. Worse, if I understand what it's doing, I think MaxiFi is doing bad math, mixing the effects of Roth Conversions with the effects of short term low spending. That is, your far future discretionary spending would automatically increase if you cut your spending back for a few years even if you did no Roth Conversions at all.
I hate to keep defending Maxifi, which I've never used. But according to their how-to-videos you get three choices of how it will conduct Roth Conversion Optimization:
(1) All plans
(2) Ignore plans where the annual discretionary spending in any year is below the Base Plan
(3) Ignore plans where the annual discretionary spending in any year is more than ___ % lower than the Base Plan.
The Base Plan represents your yearly consumption-smoothed discretionary "allowance" for your AA if stocks were to return 0%.
@boomdaddy3 Very interesting. We are all so different. It's sound like you're like me in not using Pralana the "heart" of our decisions but more of a supplement. I find its cashflow analysis unparalleled. Can you please explain what "risk parity" is and how you arrived at your allocation, and with with those particular assets?
Yes, but the constraint that you can always spend your mandatory spending is not a fix. It is asking you to spend less discretionary in the early years (compared to later) to make room for Roth Conversion taxes and that means that unspent discretionary money is available to invest and compound. That would be a weird plan, but mathematically defensible if Roth Conversion taxes behaved similarly, since they would balance out with the low early discretionary consumption, but Roth Conversion taxes don't behave the same as consumption.
Roth Conversion taxes are different in that if you should do the conversions at all, then by definition they don't reduce lifetime consumption, so there is no financial planning reason to force a trade-off of Roth Conversion taxes vs. discretionary spending. I think it is programming shortcut to get the LP solver to work in an integrated way with Roth Conversions since it uses the same objective function (maximize spend) as when there are no Roth Conversions. Whereas Pralana does it in a more correct, but non-integrated way by forcing you to iterate between Consumption Smoothing and the Roth Optimizer.
Forcing the trade-off in MaxiFi not only creates plans that ask folks to have low discretionary spending in early retirement, but they mix in the effects of consumption deferral with the effects of Roth Conversions and are claiming it all for Roth Conversions. Since I don't use MaxiFi, I'm not interested enough to figure out how significant it is, but I don't like the bragging claim the guy made in his paper, when what he really did was employ a not-well-thought-out shortcut.
Concept. Risk parity aims for a diversified portfolio where each asset contributes a balanced share of overall risk. In practice, that means larger weights in lower-risk assets (e.g., bonds) and smaller weights in higher-risk assets (e.g., stocks), with mechanical rebalancing to sell recent winners and buy recent laggards.
Implementation choices. I favor assets with low correlations and meaningful volatility, so rebalancing has more to harvest across most economic regimes. That’s why I use long-term Treasuries rather than a balanced bond fund: they’ve historically offered powerful capital appreciation when rates fall, plus steady interest income. On the equity side, I hold large-cap growth and small-cap value—both volatile and only loosely correlated with each other. I also keep a core gold position for diversification and inflation protection. While committing a sizable gold weight was initially hard, multi-tool backtests showed it improves safe-withdrawal rates versus gold-free mixes. It looks brilliant in today’s environment, but there were long stretches when gold did little while other assets soared—that’s the nature of diversification.
Alternatives & cash. My “alternatives” sleeve began with REITs but migrated to managed futures due to lower correlation with stocks; later, I split part of that into a broad commodities fund. The exact split of the last ~10–15% isn’t critical—I’m mainly seeking additional, uncorrelated volatility. I also added a 2% Bitcoin allocation (IBIT) within alternatives. For flexibility, I keep up to 6% in cash so I can buy without forced sales; as markets peak, cash tops out near 6%, and during major drawdowns, I’ll deploy up to ~4% into bargains.
Allocation framework. My targets follow a “Golden Ratio” (~1.618) progression:
-
Stocks (total): 42% ≈ 1.6 × Bonds: 26%
-
Bonds: 26% ≈ 1.6 × Gold: 16%
-
Gold: 16% ≈ 1.6 × Alternatives: 10%
-
Alternatives: 10% ≈ 1.6 × Cash: 6%
I first encountered this design on Risk Parity Radio (by Frank Vasquez), which launched in July 2020 with several $10k sample portfolios at Fidelity using different withdrawal rules; the Golden Ratio portfolio was one of them. For a quick primer, start with episodes 1, 3, 5, 7, and 9. There’s also an excellent Golden Ratio write-up on Portfolio Charts, and I use testfol.io alongside other tools for backtesting.
Why I like this approach. It trades stock-picking for stable, rules-driven diversification. At any moment, some holdings will lag—that’s expected. The mix, plus volatility, creates disciplined opportunities to sell high and buy low at rebalance thresholds. The result: better sleep, less effort, and only minimal tinkering around the edges.
Hedging note. Lately, I’ve added SPY puts as low-cost insurance (IV had been near historical lows). Even when those hedges don’t “pay,” gains elsewhere typically offset their cost. If the correction I've been waiting for shows up, the puts will offset most of my stock losses. I'm pretty bearish in outlook and have been for some time - my error has been a happy problem, but I fear the correction, when it comes, will really hurt.
I hope this clarifies the rationale, structure, and resources behind my risk-parity setup.
@boomdaddy3 Thanks for the explanation of risk parity and golden ratio. What happens if the AA derived from Risk Parity (the optimal AA of an asset to equalize the overall risk) conflicts with the golden ratio of 1.62? (The GR is fixed, but the risk parity varies with returns and variance.)
@jkandell I will survive. The diversity, coupled with the volatility, provides the edge. I’m not seeking perfection but a simple, workable design. The GR design is the 2nd best portfolio of all submitted to Portfolio Charts. Follow the link in my post to see how it stacks up against.
