Navigator
Navigator does two things:
- Model your retirement. It can simply project your savings and predict your future balances.
- Navigate your retirement. Enter your information and Navigator will optimize for retirement success, giving you results plus optional levers to change your outcome.
Navigator helps you face the different decisions retirement brings — stock-market performance, Social Security start ages, the widow’s trap, RMDs, Roth conversions, Medicare Part B & IRMAA impacts, spending, gifting, and more — simulating your full retirement year-by-year under the tax code and validating with a 1,000-path Monte Carlo.
Executive Summary
The core retirement decisions — and the rules of thumb behind them.
Educational rules of thumb to frame your thinking — not personalized financial or tax advice. Navigator computes your specific numbers in the tabs that follow.
1When to start Social Security?
- Single — start SS at 62.
- Married
- higher-earner at 70, lower-earner at 62, or
- both at 70 (if both are expected to live long; this is the best hedge against poor long-term stock-market performance).
2FEHB and Medicare Part B, or just FEHB?
- Generally it's a good idea to take both (about 70% of federal employees elect Part B at 65).
- Can be too expensive for federal employees in high Modified Adjusted Gross Income (MAGI) brackets. At higher MAGI there is an additional monthly surcharge — IRMAA (Income-Related Monthly Adjustment Amount). MAGI also rises in RMD years. Compare your total annual Medicare Part B cost against your total FEHB cost (including the Max Out-of-Pocket) to guide the decision.
3Roth Conversions?
- This comes down to paying taxes at the lowest tax brackets.
- If you were born in 1960 or later, RMDs start at 75; otherwise they start at 73.
- RMDs are driven by your total tax-deferred balances, which are influenced by:
- stock allocation and stock-market performance,
- Congress (future tax law),
- tax brackets — when one spouse dies, the survivor moves into the individual tax brackets (the “widow's trap”), and the tax burden rises.
- Strategy — make assumptions, project your future tax brackets, and do Roth conversions before RMDs begin.
4Spending
Rule of thumb — the 4% rule
- Withdraw 4% of the financial portfolio in year one, then increase that dollar amount each year for inflation.
- ~90%+ chance the portfolio survives 30 years.
- ~65% chance the portfolio doubles over 30 years.
- Assumes a portfolio between 50/50 and 80/20 (stock/bond).
5Stock Market Performance
- Historically, the stock market gains about 12% per year.
- Vanguard expects reversion-to-the-mean over the next few years (~6–8%).
- Lost Decades — ~10-year periods where the market gain is close to zero (the last began in the early 2000s).
6Sequence-of-Return Risk
- Withdrawing during a down market early in retirement is a major risk to the portfolio.
- Strategies:
- Lower the spending rate.
- Spend from the non-stock allocation. This gradually raises the stock percentage — effectively “buying low” — though it can start to feel like timing the market. This is a bucket strategy.
- Recognize this is the risk of starting retirement with an aggressive portfolio, so you might begin more moderate. A 60/40 portfolio is a common starting point — but those with pensions and high-value portfolios can often handle more risk and a more aggressive mix. A personal decision with potentially significant consequences (recall the 2008 housing crash).
The Rules — in Plain English
The tax rules that actually drive the math — and precisely how Navigator applies each one.
An educational summary of current federal rules — not personalized tax or financial advice. Rules change; confirm against IRS / SSA publications before acting.
1Social Security
- You may claim any time from 62 to 70. Your Full Retirement Age (FRA) is 67 if you were born in 1960 or later.
- Claiming early permanently shrinks every check; waiting grows it. At 62 you receive 70% of your full benefit; at FRA 100%; at 70 you receive 124% — an 8% per year delayed-retirement credit that stops at 70. There is no reward for waiting past 70.
- Benefits are taxed by an unusual formula. Your provisional income = other income + half your Social Security. Above $32,000 married / $25,000 single, part of the benefit becomes taxable; above $44,000 / $34,000, up to 85% is taxable.
- Those thresholds have never been indexed to inflation — frozen since 1984 — so each year more retirees are pulled in. This is the “tax torpedo.”
How Navigator applies it: the exact SSA factors (62 → 70%, 67 → 100%, 70 → 124%, 8%/yr after FRA, capped at 70), and the real provisional-income test with the frozen $32K/$44K (MFJ) and $25K/$34K (single) thresholds. Calculate searches claim ages 62–70 for each spouse and picks the best pair. Note: the engine still uses the pre-OBBBA taxation formula.
2RMDs (Required Minimum Distributions)
- Traditional TSP / 401(k) / IRA money grows untaxed — but the IRS eventually forces you to withdraw it and pay ordinary income tax. That is an RMD.
- Once you reach RMD age you must withdraw at least: prior year-end balance ÷ a divisor from the IRS Uniform Lifetime Table. The divisor shrinks every year you age, so the forced withdrawal keeps climbing — often pushing you into higher brackets and higher IRMAA tiers exactly when you least want it.
- SECURE 2.0 set the age at 73, rising to 75 for those born in 1960 or later.
- Roth IRAs have no RMDs — nobody forces you to touch that money.
How Navigator applies it: RMDs begin at your SECURE 2.0 age, derived per spouse from birth year — 75 if born in 1960 or later, 73 if born earlier — so spouses who straddle the 1960 line each get their own. It divides your Traditional balance (including the G-Fund buffer) by the Uniform Lifetime Table divisor, computed on the pre-conversion balance as the IRS requires. Roth is excluded entirely, and each spouse’s Roth conversion window closes automatically when their own RMDs begin.
3Roth Conversions
- A conversion moves money from Traditional to Roth. You pay ordinary income tax on the converted amount this year.
- In exchange, that money grows tax-free, is never subject to RMDs, and passes to your heirs tax-free.
- The entire game is paying tax at a lower rate now than you would later — filling up the cheap brackets in the window after you retire (income drops) but before RMDs start and before the widow’s trap compresses your brackets.
- The cost: tax paid up front, plus the growth that money would have earned. Convert too much and you pay a high rate now to avoid a lower one later.
How Navigator applies it: two modes — bracket fill (convert just enough each year to reach the top of a chosen bracket, e.g. 24%, adjusting to that year’s actual income) or a fixed dollar amount per spouse per year. The optimizer searches amounts and timing; conversion windows close at 75, when RMDs begin. Conversions raise MAGI, so they raise your IRMAA tier two years later — which the engine models.
4Medicare Part B & IRMAA
- If your income is high, Medicare Part B costs more. The surcharge is IRMAA (Income-Related Monthly Adjustment Amount).
- It is based on your MAGI from two years ago — so what you do today sets your premium in two years.
- It is a cliff, not a ramp: one dollar over a threshold moves you into the next tier and you pay that entire step. Thresholds for a single filer are half the married ones.
How Navigator applies it: the real 2-year MAGI lookback, so a Roth conversion (or a big RMD) raises your tier two years later. Cliff brackets are enforced to the dollar; thresholds, per-tier surcharges, and the Part B base premium all inflate each year. IRMAA applies only when Part B is elected.
5The Widow’s Trap
- When the first spouse dies, the survivor files as single from the following year on.
- Income barely falls — but the tax on it jumps. Brackets compress to roughly half the width, the standard deduction halves, and IRMAA thresholds halve.
- The household also loses the smaller of the two Social Security checks; the survivor keeps only the larger.
- Net effect: nearly the same money, taxed considerably harder — often for decades. It is the single most under-planned event in a married retirement, and the main reason to convert to Roth early.
How Navigator applies it: at the first death the engine automatically switches to single-filer brackets, the single standard deduction, and halved IRMAA thresholds, and drops the smaller SS benefit. The Net to Spouse objective scores precisely this risk, and an optional lever reduces spending after the first death (defaults to 80% of the couple’s spending).
6Other Rules the Engine Honors
- NIIT — a 3.8% Net Investment Income Tax on investment income above $250,000 (married) / $200,000 (single).
- Step-up in basis at death — appreciated assets reset to their date-of-death value, erasing the capital gain. In Texas community property, 100% steps up on the first death; in most other states only 50% does.
- State income tax — Texas has none, so the engine models none. If you live elsewhere, read the projections as pre-state-tax.
- Capital gains & the home sale exclusion — up to $500,000 of gain on a primary residence (married) is excluded; other property is fully taxable at long-term rates.
How Navigator applies it: all of the above on the deterministic path, with the Texas 100% step-up as the default (toggleable in inputs). Brackets, the standard deduction, and SS PIA all inflate at your assumed inflation rate; healthcare costs inflate at their own, higher rate.
How Navigator works
Five steps from your inputs to a final plan.
1
📝
Inputs
Your financials, spending, real estate, healthcare, and settings — market forecast, portfolio mix, plus your SS & Roth choices.
2
🧭
Analyze
Navigator finds the optimal Roth conversion strategy and Social Security claim ages, runs Monte Carlo, and builds your year-by-year projection. Analyze with your current settings, or Navigator’s recommendation.
3
🔍
Results
- · Roth Optimizer
- · Monte Carlo (chance of success)
- · Year-by-year projection
- · Healthcare & IRMAA
4
📊
Dashboard
Your headline results: the objective, key decision points, the optimized solution, and widow’s-trap protection.
5
⚖️
Strategy
Explore lever decisions to grow wealth or chance of success: lever rankings, two-lever heat map, stress test across forecasts, real-estate keep-vs-sell, and a final consolidated plan (My Plan).
Navigator · Personal retirement planning tool · For methodology details, see the How It Works tab.
Spending
Spending
Base living expenses only — food, travel, utilities, transportation, hobbies, discretionary. Don’t include mortgage, property tax, insurance, or maintenance (→ Real Estate), healthcare premiums or out-of-pocket (→ Healthcare), life-insurance premiums, gifts, or one-time purchases (→ Gifts, Purchases & Recurring). Those are entered in their own tabs and added on top automatically — including them here double-counts.
Base subtotal: $0
From your other tabs (read-only — first full-retirement-year estimate; Run to populate)
| Mortgage P&I edit | -- |
| Property tax / insurance / maintenance edit | -- |
| Healthcare (premiums, OOP, IRMAA) edit | -- |
| Life-insurance premiums edit | -- |
| Gifts, purchases & recurring | -- |
Total annual spending--
%
%
Excludes federal tax & health insurance — engine adds those automatically. Pension COLA = annual cost-of-living adjustment applied to pension benefits (default 2%; CSRS ≈ 3%, FERS ≈ 2%, private pension typically 0%).
Gifts, Purchases & Recurring Payments
Annual Gifting
$
Recurring outflow from Taxable each year (today's $; inflates 3%/yr). For example, $19K × 3 children = $57,000/yr to use the IRS annual exclusion. Under the annual exclusion no gift tax applies and no Form 709 is needed.
One-time Purchases
Lump-sum events (car, wedding, gift) in today's $. Engine inflates 3%/yr to trigger year.
Trigger by calendar year (e.g. 2030) or Spouse 1's age (e.g. 75). If both set, year wins.
Recurring Payments
Annual outflows that recur between two S1 ages (e.g. alimony, club dues, charitable pledge).
Amount is today's $; engine inflates 3%/yr through the active years.
1. Plan Basics
Plan Setup
Beneficiaries
2. Spouse 1 — Spouse 1
Identity & Timeline
Note: Retire age, SS claim, and death age are annual resolution — the engine treats retirement as starting January 1 of the calendar year the spouse reaches that age. No sub-year (mid-year) timing.
Income & Pension
$
$
%
$
$
SS PIA = 12 × monthly PIA at FRA (67).
Retirement Accounts
$
$
$
$
Contributions are during working years only.
HSA (Health Savings Account)
$
$
📋 How HSAs are modeled (tax + spending treatment)
Triple-tax-advantaged bucket. Contributions are pre-tax (reduce ordinary income that year, same as Traditional). Growth is tax-free. Withdrawals for qualified medical expenses are tax-free.
Contributions stop at Medicare age. Once you're enrolled in Medicare (default age 65), HSA contributions are no longer allowed by the IRS. The engine zeroes future contributions automatically from that age on. Existing balance keeps growing tax-free.
Spending treatment. Each year, the household's combined HSA balance pays for that year's medical costs FIRST (Medicare B + plan premium + OOP), up to the available HSA balance. The remainder is funded from the normal waterfall (Taxable → Traditional → Roth). The Medicare B / Plan premium / OOP columns in the Yr-to-Yr table show GROSS costs; the HSA column shows the balance after each year's medical draw.
At first death (MFJ). Surviving spouse inherits the HSA tax-free (federal spousal rollover). The combined household balance continues to pay for medical.
At end of plan (both deceased). For "Net to Spouse," the surviving spouse's HSA is counted at full value (assumes continued medical use). For "Net to Beneficiaries," the HSA is treated like Traditional — non-spousal beneficiaries pay ordinary income tax on the full balance.
Simplifications. Doesn't model: pre-65 non-medical withdrawal penalty (20%), "save receipts and reimburse decades later" strategy, HDHP eligibility validation, separation of Medigap (non-qualified) from Part B/D (qualified) premiums.
3. Spouse 2 — Spouse 2
Identity & Timeline
Note: Retire age, SS claim, and death age are annual resolution — the engine treats retirement as starting January 1 of the calendar year the spouse reaches that age. No sub-year (mid-year) timing.
Income & Pension
$
$
%
$
$
Retirement Accounts
$
$
$
$
HSA (Health Savings Account)
$
$
Tax/spending treatment: see the expandable note on Spouse 1's HSA card.
4. Total Taxable Account (these are non-retirement accounts)
Taxable Account
$
$
$
$
Annual contribution: working years only. Reserve floor: min cash buffer engine won't draw below unless Trad/Roth depleted.
5. Cash Flow Events
Windfalls
Lump-sum cash in (inheritance, settlement, bonus). Today's $; inflates 3%/yr to trigger year.
Trigger by calendar year or Spouse 1's age. Adds to Taxable as after-tax money (no LTCG drag on the windfall itself).
Term Life Insurance
Add as many policies as you have (laddered terms, group term, individual). Each pays the face amount tax-free if the insured dies during the coverage period.
If insured outlives coverage, premiums are sunk. Face amount and premium are nominal $ (level term — not inflation-indexed).
Real Estate
Add properties below. Each is modeled with its carrying costs, mortgage (if any), and tax treatment at sale or step-up at death.
📋 Tax-treatment reference (click to expand)
Property types:
- Primary residence — $500K MFJ / $250K Single capital-gains exclusion at sale (§121)
- Vacation home — full LTCG at sale, no exclusion
- Other — land, inheritance, etc. (same tax as vacation home)
0 for "never sell during plan" (held to death — gets full step-up in basis, beneficiaries inherit tax-free).
Healthcare Coverage
Determines health insurance cost in retirement. Medicare uses base premium + IRMAA tiers based on income.
FEHB / Other use the custom premium + OOP fields below. To compare these options visually, see the Healthcare Results tab.
Plan
Medicare uses IRMAA tiers automatically. ~70% of FEHB retirees enroll in Part B as secondary.
HSA usage
When reserved, the HSA is not drawn for routine healthcare — it keeps growing tax-free and stays in your assets, available later (e.g. for long-term care).
📋 Inputs Snapshot
Everything you have entered, at a glance — so you can check it against the settings below.
Open the full Dashboard
for your objective, decision points and the optimized solution.
▶ Settings — please review below & make selections
▸ 🎯 Set your scenario assumptions — what Navigate does (click to expand)
These key decisions drive your analysis — they mirror the Inputs pages, so changes here propagate everywhere. Adjust, then Analyze (here or from the sidebar).
Primary Objective
%
All three are always computed in the Roth Optimizer; this only sets which one the recommendation targets.
Optimization Horizon
End of plan scores at S2’s death (default — favors Roth conversions, which avoid years of widow-bracket tax). S1 death scores at the widow transition (favors what the survivor inherits; no-conversion baselines often lead here). Single households fall back to end-of-plan.
Household
Market Assumptions
Conversion Strategy
Healthcare & IRMAA
Spouse 1 — Retirement & SS Decisions
Spouse 2 — Retirement & SS Decisions
Run an analysis (the buttons above or below) to see your initial recommendations here.
Roth Conversion Optimizer
Three optimal Roth conversion strategies are shown below, each maximizing a different objective:
Max Net to Spouse @ end of plan, Max Gross to Spouse @ end of plan (pre-tax — useful
for seeing how much the conversion strategy gives up to taxes), and Max Net to Beneficiaries.
Two no-conversion baselines (SS at 62/62 and at 70/70) are also shown for comparison. Each strategy
assumes you keep your current Real Estate, Social Security, portfolio, and spending settings —
only the Roth conversion strategy varies.
Click Calculate to run.
Distribution of Outcomes (Fan Chart)
Distribution of outcomes:(Click Recalc to populate)
|
MC sampling method:Switching modes will automatically re-run MC if results exist.
Note: Both "Total portfolio" and "Net assets" reflect financial accounts only (Trad + Roth + Taxable). Real estate held to end of plan is not included in these lines. Real estate that is sold during the plan flows into the Taxable account from the sale year forward and IS included from that point on.
Sampling method:Normal — Normal(mean, vol) per year, independent across years, fixed 3% inflation.
Bootstrap mode samples random 5-year blocks from actual 1928-2024 history — preserving fat tails (1929/1931/2008-style crashes), sequence-of-returns clustering, and inflation regimes (1970s-style 8% vs 1990s-style 2%). Toggle above to switch; results may differ noticeably for marginal plans.
Bootstrap mode samples random 5-year blocks from actual 1928-2024 history — preserving fat tails (1929/1931/2008-style crashes), sequence-of-returns clustering, and inflation regimes (1970s-style 8% vs 1990s-style 2%). Toggle above to switch; results may differ noticeably for marginal plans.
📋 Forecast modes — what they mean and why these numbers (click to expand)
Navigator supports four forecast frames you can toggle in the Assumptions bar at the top of the page (most pessimistic to most optimistic):
Lost Decade: Years 1-10 use stocks -2% / 17% vol (sequence-of-returns stress test calibrated to actual U.S. lost decades). Years 11+ revert to Baseline 11.5%. Captures the realistic threat to retirees: bad early decade, then normal recovery — but with less capital.
Vanguard 10yr: Years 1-10 use stocks 5.5% / 17% vol (Vanguard VCMM Q1 2026 midpoint of 4.5-6.5%). Years 11+ revert to Baseline 11.5%. Applies institutional 10-year forecast for the horizon Vanguard actually forecasts, then historical for the longer period.
Baseline (default): stocks 11.5% / 17% vol, bonds 4.5% / 7% vol, S/B corr 0.10. Stock CAGR ~10.06% nominal — matches historical.
Historical: stocks 11.79% / 19.4% vol, bonds 4.79% / 7.9% vol, S/B corr 0.017 — pure Damodaran 1928-2024.
Why Baseline = 11.5% nominal arithmetic mean for stocks?
The defensible long-horizon number for U.S. equities, grounded in 96 years of data:
- S&P 500 long-run geometric mean (CAGR): 9.8-10.2% nominal (1928-2024, dividends reinvested) — Damodaran NYU Stern
- S&P 500 long-run arithmetic mean: 11.7-12.0% nominal (same source)
- Inflation 1928-2024 average: ~3.0%
- Real CAGR: ~6.9-7% (nominal CAGR minus inflation)
30-year rolling window data — most relevant for retirement planning:Four Pillar Freedom analysis
- Worst 30-year period (1965-1994): 4.3% real (~7.3% nominal) — stagflation era
- Median 30-year period: 7.1% real (~10.1% nominal)
- Best 30-year period (1932-1961): 10.1% real (~13.1% nominal) — post-Depression boom
- The S&P 500 has NEVER had a negative 30-year period since 1928
How Baseline 11.5% nominal arithmetic translates:
- 11.5% arithmetic − (17% vol)²/2 = ~10.06% nominal CAGR
- ~7% real CAGR (after 3% inflation)
- This matches the historical 9.8-10.2% nominal CAGR — Baseline is calibrated to "next 33 years average out to historical experience"
- Vol of 17% is slightly below the long-run 19.4% to reflect modern post-1980 stability; using historical vol gives CAGR ~9.65% (still within historical range)
- Above the worst-30-year-period CAGR of 7.3% nominal — meaning even in the most adverse 30-year window in 96 years, real returns were positive in real terms
- If you believe today's elevated valuations (S&P P/E ~28× vs ~17× long-run average) warrant a haircut, dial down stock weight or accept a more conservative figure manually
Why not just use the institutional 10-year forecasts?
Vanguard, BlackRock, Morningstar etc. forecast 3.3-5.5% nominal for U.S. equities over the next 10 years. These are 10-year forecasts, not 33-year. Saying "Vanguard predicts 4.5% for 33 years" misrepresents what they actually publish. Over a longer horizon, mean-reversion has more time to work; the long-run productivity-driven floor of equity returns reasserts. Vanguard VCMM
Why Lost Decade as the stress test instead of Conservative?
"Conservative" (extrapolating 10-year forecasts for 33 years) was misrepresenting what institutions actually predict. The realistic failure mode for a retiree isn't "average returns are 4.5% forever" — that has never happened in U.S. history over 30 years. The realistic failure mode is sequence-of-returns risk: a bad first decade depletes capital that the recovery can't fully replace.
- 1929-1939: S&P CAGR ≈ -1% nominal
- 1965-1974: S&P CAGR ≈ +1% nominal (real -3% with high inflation)
- 2000-2009: S&P CAGR ≈ -1% nominal — rolling-returns analysis
Lost Decade mode (-2% nominal mean for years 1-10, then 10% from year 11) is calibrated to be slightly worse than the worst rolling 10-year actuals, then recover. This stress-tests whether the plan survives the realistic worst-case sequence.
How to use the modes:
- Plan against Baseline. Your central planning assumption.
- Stress-test with Lost Decade. If your plan still works under a 2000-2009-style early decade, you have meaningful resilience.
- Sanity-check with Historical. If history repeats, you'll likely have substantially more wealth than Baseline projects.
⚠️ Scheduled review: revisit annually. The 9.8-10.2% historical CAGR is anchored in 96 years of data, but valuations and current conditions warrant calibration. Watch CAPE ratio (currently ~33), 10-year Treasury yield (currently ~4.3%), and rolling-30-year actuals as they update.
Five percentile lines: 95th (teal, upper tail), 75th (blue), median (black, thicker), 25th (orange), 5th (red, lower tail).
"Total portfolio" = combined Trad + Roth + Taxable balances.
"Net assets" = portfolio minus estimated tax on Trad (per-year effective rate) and LTCG on taxable gains (proper bracket math). Same method as the MC table's "Net Assets to Spouse" — values match.
Monte Carlo — Distribution of Outcomes
Each scenario runs 1000 paths with year-by-year returns sampled from
Normal(mean, volatility)
for the chosen stock/bond allocation. P10 / P50 / P90 show the 10th, 50th, and 90th
percentile of ending net assets and beneficiary net across the 1000 paths.
Takes ~10 seconds. Results are stochastic — re-run for slightly different numbers.
Takes ~10 seconds. Results are stochastic — re-run for slightly different numbers.
⚠️ Longevity is NOT modeled in MC. All paths assume you and your spouse live exactly to the death ages you entered. Returns and (under bootstrap) inflation vary across paths, but lifespans do not. This means the "Chance of success" metric does not capture longevity risk. If you live longer than expected, your plan needs more years of coverage. If a spouse dies early, the survivor faces the widow trap (single-filer brackets, halved IRMAA thresholds). To stress-test these scenarios, manually adjust the death ages on the inputs tab and Recalc — or wait for the planned Survivor Risk tab to do this automatically.
Click Run Monte Carlo to compute distribution.
Healthcare Cost Comparison
Compare lifetime healthcare costs across plan choices (Medicare, FEHB, FEHB+B, etc.).
Costs are in today's dollars and account for: Medicare Part B (with IRMAA tiers based on income), Part D,
Medigap supplemental insurance, and the custom premium / OOP / inflation values from the Healthcare Inputs tab.
The currently-selected plan from the Healthcare Inputs tab is shown in bold.
The currently-selected plan from the Healthcare Inputs tab is shown in bold.
□ Real estate as a bond sleeve.Enter your balances (or apply a demo) to compute. · Target total stock exposure:
%
Illiquid ballast: raises total net worth & legacy, but cannot cushion a bad-sequence spending year without a sale. Keep real liquid bonds/G-Fund for that.
Considering a future purchase? Add it here, then fill in the form on Real Estate inputs:
💡 Why does adding a vacation home have such a big retirement-wealth impact?
A vacation home looks on paper like swapping one asset (cash or portfolio) for another (real estate). But over a 25–30 year retirement, three compounding financial drags accumulate. The Roth conversion optimizer, Monte Carlo, and Net to Spouse / Beneficiaries figures all reflect these — which is why “sell” or “never bought” scenarios can be $1M–$3M ahead of “hold forever” in present-value dollars.
- Carrying costs eat 2–4% of home value every year, forever.
Property tax + insurance + maintenance + repairs typically run 2–4% of home value annually. For a $900K vacation home, that's $18K–$36K/yr leaving the portfolio. Over 25 years with 3% inflation, the present-value cost is roughly 18–22× the annual figure — about $400K–$800K just to hold the asset.PV(carrying) ≈ annual cost × ((1 + g)N − 1) / g, g = real escalation rate - Equity locked in real estate compounds at ∼0% real.
Real estate appreciates roughly with inflation (∼3% nominal, ∼0% real). A 70/30 stock/bond portfolio compounds at ∼6% real. The opportunity cost of locking equity in the home is the differential compounding:opportunity cost ≈ equity × ((1.06)N − (1 + real RE growth)N)
Example: $400K of locked equity over 25 years. Portfolio FV = $400K × (1.06)25 = $1,716K. RE equity FV (real, at 0% real growth) = $400K. Wealth gap = $1,316K. This is usually the largest drag. - Selling triggers capital gains tax.
Federal LTCG at 15% (sometimes 20% at high income) applies to the appreciation since purchase. A $900K vacation home with $525K basis, sold after 25 years of 3% nominal growth, sells for ∼$1,884K nominal → $1,359K of nominal gain → ∼$204K in federal LTCG. (If held until death, basis steps up and this disappears — but the carrying + opportunity costs already happened.)
Worked example — a $900K vacation home with $400K equity, held 25 years, then sold:
• Carrying drag (PV): ∼$600K
• Opportunity cost on equity: ∼$1,300K
• LTCG at sale: ∼$200K
• Total impact: ∼$2.1M vs. holding the same capital in a 70/30 portfolio.
The trade-off this analysis doesn't capture: the consumption value of actually using the home — vacations you'd otherwise pay for, time with family, the option to host visitors. If you'd otherwise spend $30K/yr on vacation rentals, that's ~$750K of avoided expense over 25 years, which offsets some of the financial drag above. The net financial impact is the drag minus avoided consumption — and the decision is whether the remaining net cost is worth the consumption value to you.
For your specific properties with actual sale ages, growth rates, and mortgage terms, see the “Why is the impact so large?” details inside the RE Sell Decision Analysis card below — it computes carrying / opportunity / LTCG with your numbers.
Real Estate Sell Decision Analysis
Compares the financial impact of holding non-primary properties vs.
selling at various ages vs. never having bought. Each scenario runs
the full optimizer to find the best Roth conversion strategy given that real estate
plan, then reports net assets and beneficiary net at end of plan (today's $).
Takes ~15–25 seconds per non-primary property + ~6 seconds for combination scenarios. Other RE inputs in the Joint card drive this analysis — adjust them, click Recalc, then re-run this card.
Takes ~15–25 seconds per non-primary property + ~6 seconds for combination scenarios. Other RE inputs in the Joint card drive this analysis — adjust them, click Recalc, then re-run this card.
Click "Run Analysis" to compute. (This card is hidden when no non-primary real estate is configured.)
Mortgage Early Payoff Analysis
For each property with a mortgage, tests payoff at every other age from 62 to 90 plus
"no early payoff" baseline. Each scenario runs the full optimizer to find the best Roth
conversion strategy given that payoff plan. Reports net assets and net to beneficiaries
at end of plan (today's $).
Takes ~12 seconds per property. For typical retiree scenarios with low-rate fixed mortgages and equity-heavy portfolios, holding the mortgage usually beats paying it off — pulling Trad to fund the payoff triggers immediate ordinary income tax, while the Trad balance left in place keeps compounding tax-deferred. Paying off can win for high-rate mortgages (6%+), fixed-income-heavy portfolios, or when funded from after-tax cash. Click ⭐ Run Analysis to see the optimal answer for your specific rates, balances, and tax situation.
Takes ~12 seconds per property. For typical retiree scenarios with low-rate fixed mortgages and equity-heavy portfolios, holding the mortgage usually beats paying it off — pulling Trad to fund the payoff triggers immediate ordinary income tax, while the Trad balance left in place keeps compounding tax-deferred. Paying off can win for high-rate mortgages (6%+), fixed-income-heavy portfolios, or when funded from after-tax cash. Click ⭐ Run Analysis to see the optimal answer for your specific rates, balances, and tax situation.
Click "Run Analysis" to compute. (This card only shows results for properties with active mortgages.)
Year-by-Year Projection
Calculate first.
Healthcare columns:Medicare B = Part B premium + IRMAA surcharge (only when Medicare B is elected); Plan premium = FEHB/Other plan premium (when not on Medicare-only); Health OOP = out-of-pocket costs on the custom plan. Empty cells (—) mean that component doesn't apply to your current health configuration.
FEHB + Part B coordination: when Part B is elected with FEHB/Other and all covered spouses are 65+, OOP is set to $0 — most major FEHB plans (BCBS Standard, GEHA Standard, etc.) waive cost-sharing when Medicare is primary. If your specific plan doesn't fully coordinate, manually set OOP to a partial value.
FEHB + Part B coordination: when Part B is elected with FEHB/Other and all covered spouses are 65+, OOP is set to $0 — most major FEHB plans (BCBS Standard, GEHA Standard, etc.) waive cost-sharing when Medicare is primary. If your specific plan doesn't fully coordinate, manually set OOP to a partial value.
Exports the scenario shown above in nominal dollars, with live formulas for Total spending, Total draw, Total assets, and a lifetime-totals row.
Charts (for selected scenario above)
Calculate first.
Scenarios used
⭐ Net to Spouse / Beneficiaries — Observations
Based on your inputs, the optimizer found two strategies — one optimized for
your wealth at end of plan, one for what your beneficiaries receive.
When the same strategy wins both, that's a strong recommendation. When they diverge,
the differences highlight a meaningful trade-off.
Updates automatically when you click Recalc.
Updates automatically when you click Recalc.
Click Recalc to populate observations.
🎯 Levers to Improve Chance of Success
Tests real-world adjustments — spending cuts, portfolio mix changes, real estate
sales, mortgage payoff, working longer — and ranks each by how much it would raise
your Monte Carlo chance of success. Useful when your current chance is below your
comfort threshold, or just to see which levers carry the most weight in your plan.
Takes ~60–120 seconds (runs ~12 scenarios with full optimizer + 1000 MC paths each). Levers are filtered to those applicable to your situation. Each row shows the new success rate, the change in pp, and the wealth cost so you can see the trade-off.
Takes ~60–120 seconds (runs ~12 scenarios with full optimizer + 1000 MC paths each). Levers are filtered to those applicable to your situation. Each row shows the new success rate, the change in pp, and the wealth cost so you can see the trade-off.
Display:
Click "Run Analysis" to compute. The target rate determines which levers get highlighted as "meets target" in the table.
🌡️ Heat Map — Chance of Success vs. Two Levers
Pick two levers and see how their combinations affect your Monte Carlo chance of
success. Each cell shows the success rate for one (X, Y) combination. The grid
reveals the frontier — for each value of one lever, the smallest change in
the other that meets your target.
What each cell shows. Every cell runs 1,000 Monte Carlo market simulations. Success is the share of those paths in which the portfolio lasts through the end of the plan — the last surviving spouse’s assumed death age (set by the dies at controls below) — without running out. The large dollar figure is Net to Spouse: the after-tax wealth remaining at that end-of-plan point, in today’s dollars, valued as the surviving spouse would hold it (tax-deferred balances still owe income tax on withdrawal). B is Net to Beneficiaries — what beneficiaries receive after the second death, once the step-up in cost basis resets capital gains.
v3.142: grid expanded to up to 10×10 = 100 cells (~3 minutes). Cells fill in progressively as they compute. Switch between Absolute coloring (fixed thresholds: <80% / 80–94% / ≥95%) and Relative (best/middle/worst within this grid) using the toggle below.
What each cell shows. Every cell runs 1,000 Monte Carlo market simulations. Success is the share of those paths in which the portfolio lasts through the end of the plan — the last surviving spouse’s assumed death age (set by the dies at controls below) — without running out. The large dollar figure is Net to Spouse: the after-tax wealth remaining at that end-of-plan point, in today’s dollars, valued as the surviving spouse would hold it (tax-deferred balances still owe income tax on withdrawal). B is Net to Beneficiaries — what beneficiaries receive after the second death, once the step-up in cost basis resets capital gains.
v3.142: grid expanded to up to 10×10 = 100 cells (~3 minutes). Cells fill in progressively as they compute. Switch between Absolute coloring (fixed thresholds: <80% / 80–94% / ≥95%) and Relative (best/middle/worst within this grid) using the toggle below.
Color:
finds the best stock/bond mix under the Forecast above (~30–60s)
The grid computes automatically when you open this page. After changing any control, click "Click to update heat map" to refresh.
📋 What does the Heat Map assume? (click to expand)
For each cell of the grid, the engine:
- Clones your baseline household.
- Applies the X-axis lever value and the Y-axis lever value.
- Runs the Roth optimizer (~350 candidate strategies: 10 fixed amounts × 5 start ages × 5 end ages = 250 fixed scenarios, plus 4 bracket-fill targets × 5 starts × 5 ends = 100 bracket-fill scenarios).
- Picks the strategy with Max Net Assets to Spouse @ end of plan (the same winner shown as the headline scenario elsewhere).
- Runs a 1000-path Monte Carlo under that strategy to compute the cell's success rate.
- Runs one deterministic projection to compute Net to Spouse and Net to Beneficiaries (today's $, RE-aware) shown beneath the success rate.
What's optimized per cell:
- Roth conversion amount, start age, end age, or bracket-fill target (unless Roth is one of the axes — then it's pinned to the cell's value)
What's fixed per cell (taken from your inputs):
- Spouse 1 / Spouse 2 SS claim ages (from the Spouse cards — the heat map optimizer does NOT search over SS unless SS is an axis)
- Spending mode, withdrawal rate, healthcare assumptions, beneficiaries, gifting, one-time purchases
- All your real estate (carrying costs, sale ages, mortgage balances/rates/payoffs) — unless overridden by an axis lever
- Portfolio forecast mode (unless Forecast is an axis — then it's varied per cell)
Conversion-tax sourcing: always waterfall (tax comes from Taxable → Trad gross-up → Roth, in that order).
Why this matters:
- Each cell shows the best achievable success rate and wealth under that combination of inputs — assuming you'll also adjust your Roth strategy optimally. This is realistic for planning since you control the Roth choice.
- The chosen Roth strategy can differ between cells. The heat map doesn't show which Roth strategy each cell picked — only the resulting metrics. To see the impact of a specific Roth strategy, set Roth as one of the axes.
- Heat-map results are not directly comparable to the Lever Rankings tab, which only varies the lever in question and uses your existing default Roth strategy.
🧪 Stress Test — Robust Plan Across the Uncontrollables
Finds the strategy that holds up best across the things you can't control — the market regime and each spouse's date of death. It commits to one plan (portfolio mix, SS claim ages, Roth conversion policy) and scores that single plan across every scenario. Unlike the rest of the tool, conversions are not re-optimized per scenario — this answers “will my actual plan survive?”, not “what's best with hindsight?”
1. Scenarios to stress
Market regimes (return environments):
Longevity cases (each spouse's death age, relative to your current inputs):
Run to populate.
Within each regime, the 1,000-path Monte Carlo already captures market luck; these regimes vary the environment, and the longevity cases vary how long each of you lives — the two big things outside your control.
2. Search the committed plans
Sweeps the full mix (11) × SS combos (floor-earner 62/64/66/68/70 × other 62/67/70) × conversion policy (none / 12% / 22% / 24% / 32%) — roughly 825 committed plans against the scenarios above.
Heads-up: hundreds–thousands of simulations — a couple of minutes, with progress below.🎯 Withdrawal Rate Test
Tests portfolio survival under a fixed withdrawal-rate rule (4%, 6%, 8%, 10% — classic Trinity/Bengen
stress tests) and validates that the main engine reproduces the textbook result. This page contains
three tools, each more rigorous than the last:
What this isn't. A retirement plan. These tools are validation, not forecast. The standalone test ignores SS / pensions / taxes / etc.; a real plan typically does better (guaranteed income reduces portfolio reliance) or worse (taxes eat draws). For your actual plan's success rate, see Monte Carlo.
- Standalone withdrawal-rate test — textbook portfolio-only MC, no engine logic. Pick a rate, see if the portfolio survives.
- Engine validation — runs the main engine on a synthetic single-filer 4% household. Should match (1) within sampling noise.
- Engine validation ladder — runs the engine on 5 progressively complex households (taxes, draw order, SS, pension) and grades each. Final verdict: does the engine work?
What this isn't. A retirement plan. These tools are validation, not forecast. The standalone test ignores SS / pensions / taxes / etc.; a real plan typically does better (guaranteed income reduces portfolio reliance) or worse (taxes eat draws). For your actual plan's success rate, see Monte Carlo.
One-click full validation
Sets paths to 10,000 and runs all three tests in sequence at your chosen withdrawal rate. Takes ~30–60 seconds for the ladder.
Setup
$
Withdrawal Rule
Fixed real: withdraw the initial rate × starting portfolio in year 1, then
inflate that dollar amount by the inflation rate annually thereafter. The classic Bengen
rule. Portfolio CAN run out — "failure" = balance hits zero before the horizon ends.
Click Run withdrawal rate test to compute. Default inputs (60/40, 30 years, 4.0%) should
produce ~92–95% success — if so, the MC machinery is calibrated correctly.
🧪 Test the Engine with a Single Scenario
Validates the main engine against the standalone test above. Builds a synthetic single-filer
Roth-only household (no form changes): age 65 onwards,
spending_mode=four_percent_rule,
no other income. Roth-only neutralizes taxes so withdrawal == spending. Uses the same starting
portfolio, withdrawal rate, and horizon as the standalone test above, plus your current global
Forecast + Portfolio mix. Engine result within ±3 pp of standalone = calibrated correctly.
🪜 Engine Validation Ladder
Runs the full engine MC on 5 progressively-complex synthetic households — all using the chosen withdrawal rate.
Each step adds one engine subsystem (taxes, draw order, SS, pension). The delta from the previous step
tells you what that subsystem costs or earns in MC success rate. Uses the starting portfolio, withdrawal rate,
and horizon from the standalone test above, plus the current global Forecast + Portfolio mix.
Stress-testing: change the standalone test\'s Initial rate (%) to e.g. 8%, re-run it (sets a new anchor), then re-run the ladder. The engine should agree with the standalone at any rate — that\'s a stronger calibration test than 4% alone.
Stress-testing: change the standalone test\'s Initial rate (%) to e.g. 8%, re-run it (sets a new anchor), then re-run the ladder. The engine should agree with the standalone at any rate — that\'s a stronger calibration test than 4% alone.
📋 Dashboard
Synthesizes everything you've entered + analyzed into one page:
your inputs at a glance, your chosen objective, the key decision points you control,
and the engine's recommended optimal strategy for your objective.
🎯 My Plan — One Plan Per Objective
Generates two recommended plans by trying levers in the order below, stopping at the first combination that reaches the ≥90% target. Reorder them with the arrows to set your own preference — Roth is always first, portfolio stays manual:
For each objective (Net to Spouse, Net to Beneficiaries) the search finds the smallest combination
of levers (preferring earlier ones) that reaches the ≥90% success target — then uses the chosen wealth metric (Net to Spouse or Net to Beneficiaries) only to break ties among plans that all clear 90%.
Two plans total: one per objective (Net to Spouse, Net to Beneficiaries) at ≥90% chance of success.
Takes ~5–10 minutes (1000 MC paths × dozens of scenarios per target). Results cache so re-runs are fast if you tweak nothing.
Takes ~5–10 minutes (1000 MC paths × dozens of scenarios per target). Results cache so re-runs are fast if you tweak nothing.
□ Survivor spending: a surviving spouse typically needs about 75–80% of the couple's combined spending to keep the same lifestyle — roughly a 20–25% reduction, not half. The “reduce spending after the first death” lever models this and defaults to 80% (a 20% cut) when you turn it on.
Click "Calculate" to generate your plan.
📖 How It Works — Methodology Reference
Deterministic vs. Stochastic Analyses
Navigator runs two engines under the hood. Deterministic projects a single path year-by-year using the configured average return (the volatility-drag-adjusted CAGR). Stochastic (Monte Carlo) runs 1000 paths sampling annual returns from a distribution, capturing the range of possible outcomes.
Each analysis uses one or both, depending on what question it answers.
📐 Pure Deterministic — single path, fast, reproducible
Used when the question is "given my plan, what does the year-by-year projection look like?" Same inputs always produce the same output.
| Analysis | Where | Why deterministic |
|---|---|---|
| Roth Conversion Optimizer 6-scenario comparison | Roth Conversion Optimizer tab (top of page) | Picks "best" scenarios on a single representative path |
| Year-by-Year projection | Year-by-Year tab | Single row-by-row breakdown — must be one specific path |
| Slider live updates | Slider card | Real-time response to bracket-fill / SS / age changes |
| SS optimum check | SS optimum hint | Searches across SS claim ages on deterministic path |
| Healthcare Plan Comparison | Healthcare Results tab | Compares 3 plans on deterministic path (informative, not predictive) |
| Audit balance check | Year-by-Year audit column | Reconciliation — needs single deterministic flow |
🎲 Pure Stochastic — Monte Carlo only
Used when the question is "given uncertainty, what's the range of outcomes?" 1000 paths × multiple scenarios.
| Analysis | Where | What it shows |
|---|---|---|
| Monte Carlo summary table | After Recalc, MC card | Success rate + P10 / P50 / Mean / P90 for each scenario |
| MC fan chart | After Recalc, fan chart | Net assets percentile bands over time |
⚖️ Hybrid — Deterministic for ranking, Stochastic for safety
Used when the question is "which plan should I choose?" Each candidate gets both a wealth metric (deterministic) and a success rate (stochastic), so you can compare wealth-at-target-safety trade-offs.
| Analysis | Deterministic part | Stochastic part |
|---|---|---|
| Real Estate Sell Analysis | Per-scenario optimizer + lifetime metrics | Success rate per scenario |
| Mortgage Payoff Analysis | Per-scenario lifetime cost + interest saved | Success rate per scenario |
| Plan Safety Levers | Each lever's wealth metrics | Each lever's success rate |
| Heat Map | 2D grid of lever combos — wealth metric per cell | Success rate per cell |
| Portfolio Allocation (📊) | 11 allocations × optimizer wealth | Success rate per allocation |
| Wealth vs Risk | All ~25 levers' wealth values | All levers' success rates |
| Final Plan | Plan composition + wealth lift | Plan's success rate (the threshold criterion) |
When to Use Which Number
- Wealth / dollar projections — use deterministic. The optimizer "Net to Spouse" numbers, the lever "wealth lift," the bracket-fill percentage are all single-path projections.
- "Is this plan safe?" decisions — use stochastic. Success rate, percentile bands, the "Net to Spouse, ≥95%" threshold in Final Plan all come from MC.
- Real-time tweaks (slider, age changes) — deterministic only. MC is too slow for sub-second feedback.
Why Deterministic and MC Don't Match Exactly
You'll see the deterministic "Net financial @ end (excl. RE)" sit roughly 1% above the MC P50 (median) at 70/30, growing to ~3-4% at 100/0. Two reasons:
- Sampling noise — 1000 paths is finite. Standard error of P50 is ~1-2% of the median value. Re-running MC will give slightly different results each time.
- Sequence-of-returns risk (the more interesting one) — when there are withdrawals, paths with bad early returns shrink the portfolio before it can recover, and forced withdrawals consume a larger fraction of the depleted balance, locking in losses. The good-early-return case doesn't generate a symmetric gain. The deterministic CAGR projection compounds at
mean − σ²/2(geometric mean), but withdrawals make the realized geometric mean even lower than that. The asymmetry is real, not a bug, and it's one of the more important findings in retirement-planning research (Bengen et al.).
This applies roughly equally to all candidate strategies, so the optimizer's ranking is preserved — but absolute dollar projections are slightly optimistic.
Forecast Modes
The "Forecast" toggle (above the Recalc button) changes the deterministic growth rate AND the mean of the MC distribution. Both engines respect the same setting.
| Mode | Stock μ | When to use |
|---|---|---|
| Baseline | 11.5% all years | Long-run S&P historical assumption — default planning view |
| Lost Decade | −2% × 10 yrs → 11.5% | Stress test: what if returns are flat or negative for the first decade of retirement? |
| Vanguard 10yr | 5.5% × 10 yrs → 11.5% | Vanguard VCMM Q1 2026 midpoint forecast — moderate near-term suppression |
| Historical (Damodaran) | 11.79% all years | NYU Stern long-run including dividends — slightly more optimistic than Baseline |
Key Engine Assumptions
- NIIT (3.8% Net Investment Income Tax) — modeled on the deterministic side; respects MFJ ($250K) / Single ($200K) thresholds.
- State income tax — not modeled (Texas has none; if you live elsewhere, treat the projections as pre-state-tax).
- Step-up in basis at death — Texas community property: 100% step-up on first death. Other states: 50%. Toggleable in inputs.
- RMDs — current SECURE 2.0 ages (75 for those born 1960+; older ages for older births).
- SS benefits — taxed under current rules (up to 85% taxable above the second threshold). Note: SS taxation rules were modified by the 2025 OBBBA "no tax on SS" provision; this engine still uses the pre-OBBBA formula.
- IRMAA thresholds — start at 2026 cliff levels and inflate each year at the general inflation rate (3%), along with the per-tier surcharge amounts and the Medicare Part B base premium. Each cliff $1 above triggers the next bracket.
- Real estate — appreciates at user-specified rate (default 3% nominal). Held until specified sale age, then proceeds (net of $500K MFJ exclusion for primary home, full LTCG for other property) flow into Taxable.
- Inflation — single rate applied uniformly (default 2.5% nominal). Brackets, standard deduction, SS PIA all inflate at this rate.
- Healthcare cost inflation — separate rate (default 6% nominal) applied to plan premiums and OOP.
Roth Conversion Strategy
Conversions are optimized on the deterministic path — the optimizer doesn't directly factor in volatility when picking conversion amounts. This is fine because the relative ranking of strategies is preserved across stochastic paths. But it means the optimizer's "best" strategy is best-on-average, not best-in-the-worst-case. For worst-case considerations, look at MC P10 (the 10th percentile, bad market sequence).
Two conversion modes:
- Bracket fill — each year, convert just enough to fill up to a specified marginal bracket top (e.g., 24% bracket → ~$394K MAGI for MFJ). Dynamically adjusts to actual fixed income that year.
- Fixed amount — convert a specified dollar amount per spouse per year, regardless of bracket position.
This page is a living reference — additional sections will be added as new analyses are introduced. If a tab or analysis isn't documented here yet, check the footnotes on its output.