Retirement Planner

• Nine primary spending rules, with eight comparison-ready withdrawal strategies for side-by-side analysis
• Social Security and pension income stream modeling
• Spending adjustments for healthcare and lifestyle changes
• Historical and Monte Carlo simulation modes
• Progressive 2026 tax mode with filing-status-aware brackets, taxable Social Security, and optional IRMAA modeling
• Tax-aware account breakdown with explicit withdrawal sequencing and Roth conversion modeling
• Allocation glidepaths with age-based transition milestones
• Saved comparison snapshots that can round-trip through save and share flows when present in the request payload
• Failure-year analysis showing which conditions cause plan depletion
• Annual account-flow summaries for tax-aware plans
• Cash flow and tax decomposition chart with an Engine cash residual node that surfaces the engine accounting gap (up to roughly 10 percent in mixed RMD and discretionary years) and a separate Tax median reconciliation node for independently medianed bucket residuals

• Test whether a portfolio can fund a target spending plan over a fixed retirement horizon.
• Compare spending rules and see how much stability each one trades for a higher or lower success rate.
• Stress-test a plan against poor historical sequences and Monte Carlo variability before changing allocation or spending.
• Add Social Security, pensions, annuities, and one-time or recurring spending changes that materially affect withdrawal needs.
• Compare several named plans side by side when deciding between different retirement ages, spending levels, or allocations.

The planner simulates retirement one year at a time. In each year, spending is determined by the selected spending rule, income streams offset withdrawals, spending adjustments are applied, and the remaining portfolio grows by that year's real return. Values are shown in constant purchasing-power dollars.

The most effective workflow is:

• Start with a realistic allocation and horizon.
• Add recurring income and any known spending changes.
• Start with the historical result, then enable Monte Carlo if you want a stochastic stress test alongside it.
• Use the failure diagnostics, sensitivity sweeps, and plan comparison table to identify the smallest change that improves the plan.
• Keep comparison snapshots with the plan if you want to reopen the same side-by-side context from a saved analysis or share link.

Historical analysis is always available. Monte Carlo can be enabled to add a second stochastic result:

• Initial portfolio value: starting balance in today's dollars.
• Asset allocation: percentage weights across supported retirement asset classes. Weights must sum to 100%.
• Allocation glidepath: optional milestones that replace the base allocation from a chosen age onward.
• Time horizon: retirement start age and end age. This defines the full decumulation window.
• Annual spending: either a fixed real dollar target or, for percent-of-portfolio mode, an initial withdrawal rate.
• Spending rule: how spending adjusts over time.
• Income streams: recurring income that reduces portfolio withdrawals during the years it is active.
• Spending adjustments: one-time or recurring spending changes such as travel, health care, or a mortgage payoff.
• Tax-aware account breakdown: split the starting value across taxable, traditional, and Roth buckets. This enables the decumulation policy controls.
• Decumulation policy: choose withdrawal sequencing, set Roth and taxable balance floors, control surplus destination, and optionally model Roth conversions.
• Tax model: use the flat approximation or the progressive 2026 federal model with filing status and optional IRMAA estimation.

Enable the account breakdown when the source of withdrawals matters. The planner can model required minimum distributions, different tax treatment across taxable / traditional / Roth balances, explicit withdrawal ordering, and optional Roth conversions.

The sequencing controls support taxable-first, traditional-first, Roth-last, proportional, bracket-aware, and custom ordering. You can also preserve a Roth floor, keep a minimum taxable cash buffer, and direct modeled surplus back into taxable or Roth.

In progressive mode, the planner uses filing-status-aware 2026 federal brackets, models taxable Social Security benefits, and can add a simplified Medicare IRMAA surcharge estimate. This is still a planning approximation rather than a full household tax return.

The comparison rail supports the eight withdrawal strategies below. Percent-of-portfolio spending is also available as a primary plan rule, but it is not exposed in the comparison picker.

Every comparison strategy is calibrated so year 0 spending equals the configured base annual spending. Strategy-native dynamics apply from year 1 onward.

Any table entry can also drive the primary plan. The primary spending rule picker exposes nine total rules, including percent of portfolio, and selecting a dynamic rule reveals an inline parameter panel for its configuration. If you want to compare alternatives against a baseline instead of picking one up front, use the comparison rail described below.

The "Compare withdrawal strategies" toggle on the form runs multiple spending rules side by side on the same simulation. Each entry sees the same return sequence per cohort (historical) or per path (Monte Carlo); only the spending logic differs. Use this view to compare rules such as Guyton-Klinger and VPW for the same plan assumptions.

Comparison is opt-in. Default runs return only the primary strategy and stay byte-identical to single-strategy behavior. Enabling the toggle populates two default entries (current spending plus Guyton-Klinger) and unlocks an "Add strategy" picker covering the eight comparison strategies listed above. Up to four entries per request.

Year-0 calibration applies across the rail: every entry starts at your plan's base annual spending in year 1 of display (year 0 internal). Strategy-native dynamics, including Guyton-Klinger cuts and raises, VPW age-banded percentages, the Blanchett smile, and the ABW / TPAW amortization formulas, take over from year 2 onward. This makes the year-by-year comparison apples-to-apples regardless of which strategies you picked.

The results surface shows a decision-metrics table (year-1 spending, median and P10 lifetime spending, cut frequency and size, success rate, median and P10 terminal value), a spending-paths overlay with a focused-entry P10–P90 band and a real/nominal toggle, an event timeline showing where each entry's rules fired, and a terminal wealth distribution.

Tax-aware comparison: when the account breakdown is enabled, every entry runs through the same per-account tax-aware loop the primary plan uses, with shared withdrawal sequencing and Roth conversion policies. RMD-based comparison entries set only the discretionary spending target; mandatory required minimum distributions still flow through the tax engine without double-counting. Multi-entry tax-aware Monte Carlo caps simulations at 3,000 paths to bound runtime and surfaces the cap as an assumption note in the results.

Income streams (Social Security, pensions, annuities, or custom recurring income) offset portfolio withdrawals during the years they are active. Each stream has a start age, optional end age, and optional cost-of-living adjustment (COLA).

Spending adjustments model one-time or recurring changes to baseline spending (e.g., a travel budget from age 65 to 75, or a one-time home repair at age 70). Positive amounts increase spending; negative amounts reduce it.

• Start with a clean base case. Use your current planned allocation, retirement age, and spending target before trying optimizations.
• Enter material cash flow offsets. A plan without Social Security, pensions, or large recurring expenses is usually not decision-useful.
• Run the historical result first. Then enable Monte Carlo if you want to compare sequence-specific history with a broader distribution of random paths.
• Use the sensitivity analysis to change one lever at a time. This is the fastest way to see whether spending, retirement age, or allocation matters most.
• Save candidate plans and use plan comparison instead of relying on memory. This is especially useful when comparing a spending cut against a later retirement date.

• Verdict: pass (90%+ success), caution (75-90%), or fail (below 75%).
• Success rate: fraction of scenarios where the portfolio lasted the full plan duration.
• Annual flow summary: for tax-aware plans, shows median guaranteed income, withdrawals by account type, taxes, conversions, reinvested surplus, and ending balances by year. Median fields are calculated independently, so the tax bucket total can differ from median total tax in aggregated result rows.
• Cash flow and tax decomposition: shows the selected year's income sources, withdrawals, taxes, spending, and surplus. The Engine cash residual node represents an engine accounting gap between sources and uses (up to roughly 10 percent in mixed RMD and discretionary years) and is not solely a median aggregation artifact. A separate Tax median reconciliation node appears when independently medianed tax buckets do not sum to median total tax.
• Median terminal value: 50th percentile portfolio balance at plan end.
• Worst-case terminal value: 5th percentile portfolio balance at plan end.
• Spending variability: coefficient of variation of annual spending (shown for dynamic spending rules).
• Failure analysis: table of historical cohorts that failed, showing which start years led to depletion and how many years the portfolio lasted.
• Portfolio fan chart: percentile bands of portfolio value over the plan duration.
• Spending path chart: percentile bands of annual spending over the plan duration.
• Terminal value distribution: histogram of ending portfolio values across all scenarios.
• Failure diagnostics: for weaker plans, the tool estimates a lower spending target, a safer withdrawal rate, and ranked next-step guidance for the smallest changes that are most likely to help.
• Sensitivity analysis: shows how success rate changes when you sweep spending, retirement age, or stock allocation.
• Comparison helpers: lets you run common plan variants directly from the results view and add them to the comparison table without rebuilding the form.
• Scenario comparison: compares the current result against saved plans so you can see which change improved the plan most.

• Treating a high success rate as a guarantee. The tool is a stress test, not a forecast.
• Comparing plans with missing income or missing expense changes. Those omissions usually matter more than small allocation changes.
• Overreacting to one metric. Use success rate, terminal wealth, spending variability, and failure timing together.
• Ignoring the shape of the path. Two plans can have similar success rates but very different worst cases and spending volatility.

• Returns are annual real (after-inflation) total returns compiled from standard academic datasets (1926-2024).
• Asset classes beyond US Stocks and US Bonds use simplified approximations based on the primary series.
• Beginning-of-period withdrawal convention: spending is taken at the start of each year, then the portfolio grows by that year's return.
• Monte Carlo draws random years with replacement (bootstrap), which preserves the empirical return distribution but not serial correlation.
• When tax-aware mode is off, taxes, fees, and transaction costs are not modeled. Flat tax-aware mode applies simplified federal rates to traditional and taxable withdrawals. Progressive tax-aware mode uses 2026 federal brackets, taxable Social Security, LTCG, NIIT, AMT, and optional IRMAA estimates; it is still not a full household tax return.
• Annual flow rows are median summaries across cohorts or simulation paths. They are useful for interpreting the typical path, but they should not be read as one exact household ledger unless the run has a single underlying path.
• The result panel makes these assumptions explicit at the top of the output so you can see the modeling shortcuts before evaluating the charts and tables.