How the planner turns your life into a stress-tested retirement plan.

How the planner handles each retirement year

Every expense uses today's buying power. The planner handles inflation behind the scenes and reports results in today's dollars. A displayed future value of $100 means it could buy roughly what $100 buys today—not that the future account statement would literally show $100.

Expenses can stay the same in today's buying power, change across life periods, or be entered year by year. The planner first collects repeating expenses and income, then collects one-time expenses, one-time income, and life events before asking which assets will fund the plan. Repeating income is included only between its entered start and end ages. One-time entries affect the calculation only when marked as money spent or received; life-event-only entries label charts without changing the math.

The connected children's education planner creates a separate college schedule for each child. It projects dedicated education savings and planned contributions using the entered return above general inflation, estimates college costs using the entered cost growth above inflation, applies dedicated savings first, and adds only the remaining household payment in college years that overlap retirement. College years before retirement remain outside this analysis and should be reflected in the account balances expected at retirement.

Testing many possible market returns

This test—often called a Monte Carlo simulation—creates 1,000, 5,000, or 10,000 possible sequences of yearly market returns. Investments that often rise or fall together are modeled together, and the test allows for unusually strong or weak years. Every sequence uses the same expenses, income, investment rules, and inheritance target you entered.

These central-view values are assumptions, not forecasts. The planner also offers lower and higher/long-run views so the result does not depend on one hidden forecast. The model assumes 2.5% general inflation and a 1.5 percentage-point healthcare inflation premium. Investment returns are before fund fees; estimated taxes are calculated separately inside each tested path. Changing an ETF ticker within the same category does not create a different modeled return.

Testing real historical retirement starting years

This test replays every complete retirement period available from 1928–2024 yearly data. It uses the actual order of U.S. stock, 10-year U.S. Treasury, 3-month Treasury bill, gold, and inflation results. It highlights difficult retirement starting years such as 1929, 1966, 1973, and 2000.

A starting year counts only when the data contains enough later years to cover your full retirement plan. The result is a count of what happened in the available history—not the probability of a future result and not a promise that markets will repeat the past.

Testing history remixed in five-year chapters

This test—technically called a five-year block bootstrap—joins real five-year chapters of market history into new retirement timelines. Stocks, U.S. Treasuries, Treasury bills, gold, and inflation remain paired as they actually occurred inside each chapter. That keeps more of history's real multi-year relationships than shuffling every investment or year separately.

The result is still limited to the years and investment categories in the source data. It is another way to test whether the order of returns could hurt the plan; it cannot represent every future economy.

How each simulated year is calculated

1. Record the year's starting investable portfolio.
2. Apply that year's return separately to U.S. stocks, international stocks, intermediate Treasuries, TIPS, the near-term reserve, and gold.
3. Add every active income source and one-time receipt, then calculate required expenses, discretionary expenses, and one-time spending.
4. Calculate required traditional-account withdrawals from the applicable prior year-end balances.
5. Separate total return into price movement and estimated distributions. In regular taxable accounts, estimate bank and reserve interest, Treasury and TIPS income, qualified dividends, ordinary dividends, realized gains, and—when applicable—the net investment income tax. The payout remains part of investment return, so it is not added to the portfolio a second time.
6. Estimate taxable Social Security, traditional-account distributions, federal tax, and the entered state-tax percentage. Treasury interest is removed from the simplified state-tax base. Additional withdrawals needed to pay tax are included.
7. When Marketplace coverage is turned on, use the same year's household income to estimate the 2026 premium tax credit and net premium. Withdrawals needed to pay that premium can change income and the credit, so the calculation repeats until those amounts agree.
8. Use income from two years earlier to estimate any income-related Medicare Part B and Part D addition, then include that amount in the applicable later year.
9. Fund the remaining need using cash, taxable, traditional, HSA, Roth, and other accounts in the disclosed starting order. At the investment-category level, the near-term reserve is used first.
10. Record investment income, gains, household income used for ACA and Medicare, withdrawals, estimated taxes, required distributions, healthcare additions, any unfunded amount, and the ending portfolio before advancing one year.

The tax-strategy comparison reruns the same life plan and the same seeded market paths under four educational Roth-conversion rules. It changes only the conversion rule, so differences in taxes, Medicare income tiers, Marketplace credits, portfolio survival, and ending balances can be compared fairly. The rules are deliberately simple reference strategies—not a tax optimizer or a recommendation.

Entered pre-Medicare coverage periods are evaluated year by year. Employer, spouse, COBRA, and private-plan costs are added as entered. For a Marketplace period, the model uses that year's estimated household income—including interest, dividends, gains, Social Security, conversions, and retirement-account withdrawals—to estimate the premium credit and net premium. Entered healthcare costs rise using the healthcare-cost assumption. The coverage review identifies missing and overlapping ages before the analysis is rerun; an age with no separate cost entered still contributes $0 and is therefore labeled as incomplete rather than assumed to be free.

An optional healthcare stress test adds the user-entered net cost as a person-specific healthcare expense for the selected starting age and number of years. The net cost equals the entered yearly cost minus any entered long-term-care insurance benefit and household spending that would stop. It grows using the healthcare-cost assumption and is included in every tested market path. The test changes expenses; it does not estimate the probability of illness or care.

The stable-reserve builder is a planning view, not a separate investment forecast. For each selected early year, it uses entered spending and recurring income plus the middle estimated tax and pre-Medicare coverage cost from the latest full analysis. It discounts the remaining yearly needs using the user-entered expected return after inflation, adds them, and compares that amount with the cash-like reserve, Treasuries, and TIPS in the recommended portfolio tested by the model. It does not buy securities, alter the modeled portfolio, quote live yields, or model individual bond prices, taxes, bid-ask spreads, or reinvestment.

The Social Security start-age planner estimates each person's worker retirement benefit for an integer start age from 62 through 70. For a two-person household, it also estimates whether either person may receive a spouse-benefit top-up after both people file. The top-up is based on the difference between that person's own full-retirement-age benefit and one-half of the other person's full-retirement-age benefit, with a separate reduction when spouse benefits begin early. If one modeled life ends, the analysis keeps the larger eligible worker benefit rather than adding both checks. Applied estimates become the Social Security income used in the plan.

The fixed Conservative and Growth benchmarks keep 50% and 70% in stocks, respectively. The personalized changing portfolio is different: it first estimates ten years of required-expense gaps after entered Social Security, pensions, annuities, QLAC income, and relevant one-time amounts, then compares that amount with the near-term reserve need. Its starting stock share is whatever remains after the plan-specific stable portion and a 5% gold diversifier, subject to stated limits.

The personalized portfolio permits a gradual increase in stocks only when entered Social Security, pension, annuity, or QLAC income covers at least 70% of required expenses across up to five reviewed years, no reviewed year falls below 50%, and at least ten modeled years remain. The increase takes five years and stops at 80% stocks. The reserve and gold percentages stay fixed while Treasuries and TIPS become smaller. A custom investment mix stays fixed. After a negative stock year, the planner does not automatically sell stocks solely to refill the reserve. After a non-negative stock year, it resets the categories toward their intended percentages.

The success calculation follows a strict order. First, it applies the selected portfolio percentages to the entire investable balance. Second, it estimates where those investment categories could sit across the cash, taxable, traditional, Roth, and HSA accounts entered. Only then does it test market returns. The tax-aware placement generally keeps the selected near-term reserve in cash or a regular account, places interest-heavy investments in traditional accounts, and gives more long-horizon stock exposure to Roth or HSA accounts when capacity permits. Cash / HYSA above the selected reserve percentage is treated as money moved into a regular taxable investment account; it does not silently make the tested portfolio more conservative. This is a modeling convention—not an individualized recommendation. Contribution, rollover, eligibility, and trading rules can limit real-world implementation. Rebalancing sales in taxable accounts can create estimated gains. Dedicated 529 and education savings are excluded from the general retirement portfolio and are used only inside the connected education plan.

To estimate taxable investment payouts without counting return twice, the model uses transparent annual nominal-yield assumptions inside the total-return model: 1.3% for U.S. stock dividends (95% estimated qualified), 2.8% for international stock dividends (75% estimated qualified), 3.8% for Treasuries, 1.5% plus inflation for TIPS, 3.3% for the near-term reserve, and no annual payout for gold. These are category-level tax assumptions—not promised yields or predictions for a ticker. Actual distributions, qualified-dividend percentages, bond discounts, TIPS tax treatment, fund turnover, and account holdings can differ.

What “success” means

A modeled future succeeds when the investable portfolio funds every scheduled expense through the entered planning age. When an inheritance goal is enabled, the combined goal succeeds only when the plan also ends with at least that real-dollar amount.

The planner reports a range of outcomes and tradeoffs rather than treating one percentage as certain. Read the many-possible-returns estimate, the count of real historical periods that worked, and the remixed-history estimate together because they ask related but different questions.

Important exclusions and conservative boundaries

• Primary-residence value is shown only as context unless usable net proceeds are entered deliberately.
• 529 and other dedicated education savings are excluded from general retirement spending to prevent the same dollars from funding both education and retirement.
• The current success rate includes an educational federal-tax estimate using 2026 brackets and standard deductions, taxable Social Security, estimated interest and dividends, dynamically changing taxable-account cost basis, realized gains, NIIT, required withdrawals, and income-related Medicare premium additions using the usual two-year lookback. Tax laws and future Medicare amounts can change.
• When the user supplies Marketplace premiums and household details, the model estimates the 2026 ACA premium tax credit using the restored 100%–400% federal-poverty range and current contribution percentages. It does not determine eligibility for Medicaid, employer coverage, immigration exceptions, family-member filing requirements, or a specific Marketplace plan. Premiums must not also be entered in Step 2.
• Detailed state rules, itemized deductions, AMT, tax-loss carryforwards, foreign tax credits, municipal-bond income, inherited-account rules, tax-lot selection, short-term versus long-term sales, and tax withholding or estimated-payment penalties are not calculated.
• The primary result pays every planned expense. When selected, separate guardrail comparisons use the same market paths and reduce only discretionary spending by 10%, 20%, or 30% in the year after a portfolio return of −10% after inflation or worse. Full spending returns after a year at least 5% above inflation. Essential expenses and one-time amounts are unchanged.
• The What-if Lab reruns the same calculation after changing one selected idea. It does not overwrite the saved plan. The first-year stock-decline test replaces the normal first-year U.S. and international stock return with the selected loss; cash-like reserves, bonds, and gold retain their independently modeled returns. Withdrawals then occur, and every remaining retirement year returns to the normal market model. The TIPS-bridge comparison is simplified: it sets aside the sum of selected early essential needs, assumes a 0% real yield, and adds scheduled real payments; it does not model individual bond prices, taxes, spreads, or trading costs.
• The tax pathway is educational. It does not prepare a tax return or claim that an account placement, withdrawal, ACA-income target, or Roth-conversion amount is optimal.
• The tax-strategy comparison uses simplified annual conversion targets. It does not model every deduction, filing choice, conversion deadline, withholding requirement, state-specific rule, surviving-spouse tax change, or future-law change.
• The pre-Medicare healthcare planner depends on the premiums and coverage periods entered. It does not determine plan eligibility, provider networks, drug coverage, Medicaid eligibility, employer affordability, or every Marketplace reconciliation rule.
• The healthcare stress test uses the amount, duration, insurance benefit, and spending offset entered by the user. It does not estimate medical diagnoses, utilization, local care prices, long-term-care eligibility, policy exclusions, caregiver availability, or the likelihood that a chosen event will occur.
• The stable-reserve builder is an educational cash-flow map. Maturity-year ETF tickers are current issuer examples—not recommendations or model inputs. The builder does not construct executable trades or guarantee that a security can be purchased at the displayed amount, yield, or maturity value.
• The Social Security comparison does not calculate either person's earnings record, future earnings, exact SSA rounding, benefits withheld while working before full retirement age, disability, child-in-care or family-maximum rules, divorced-spouse rules, government-pension interactions, or eligibility. Spouse and survivor amounts are planning estimates—not an SSA claim calculation. Each person's current SSA estimate remains the source of truth.
• Unentered inheritances, bonuses, RSUs, future earnings, and other windfalls are excluded.
• Historical international-stock performance is represented by the available U.S. stock series. Intermediate Treasuries and TIPS both use the available 10-year Treasury series because a comparable TIPS record does not extend to 1928. These are disclosed stand-ins, not exact histories for those categories.
• Representative funds explain investment categories; the calculator does not predict or recommend a specific security.