What is covered well
FERS timing, household cash flow, taxes, TSP drawdown, FEHB, Medicare, and state taxes are all in scope today.
Assumptions, Coverage & Limits
The output is only as trustworthy as the scope and assumptions behind it.
This page is the canonical description of what FERSCalc models today, what it does not, and where results should be treated as directional planning context rather than final truth.
FERSCalc is not affiliated with OPM or any federal agency, and it does not replace official estimates or professional tax, legal, or financial advice.
What still needs caution
Edge cases such as part-time service prorating, post-retirement earned income for the FERS supplement earnings test, Roth basis tracking, and some state-specific rules remain simplified or excluded.
How to use the result
Use the output to compare scenarios and sharpen questions, then confirm major decisions with official sources and qualified advisors.
Modeled Today
Retirement income and timing
FERS pension timing, deferred retirement handling, the FERS supplement, survivor elections, sick leave credit, and Social Security claiming-age scenarios are modeled in the main projection.
Taxes and paycheck bridge
Federal taxes, current supported state income taxes, working-side FERS contribution drag, and tax-adjusted retirement cash flow are modeled in the year-by-year output. Local city, county, and municipal income taxes are not modeled.
TSP drawdown and stress testing
Deterministic TSP withdrawals are modeled in the core projection, with Traditional and Roth balances tracked separately. Four withdrawal-source modes are supported: Traditional-first, Roth-first, Proportional, and a threshold-aware guardrail mode. Fixed annual in-plan Roth conversions after retirement are also modeled. When more than one withdrawal-source mode applies, results rank them on projected lifetime federal tax plus Medicare drag. The threshold-aware mode is a year-by-year guardrail heuristic, not a full lifetime optimizer. Optional historical replay (Results → Stress) runs additional deterministic projections from bundled CPI and TSP fund-return data on your timeline; it is illustrative and may truncate if history ends before your horizon. See the Monte Carlo and Historical Replay scope notes below.
Health and household context
FEHB premiums, Medicare Part B, IRMAA (including tax-exempt interest in MAGI when provided), spouse or partner scenarios, survivor pension and SS survivor benefits, filing-status changes on death including the IRS Qualifying Surviving Spouse window, TSP spousal transfer, and exportable reporting are part of the current product.
Coverage Today
FERS tiers and 6C coverage
FERS Classic, FERS-RAE, FERS-FRAE, and 6C special-category retirement coverage are supported. Standard contribution tier can auto-detect from service start date or be manually overridden for edge cases. The standard tier changes paycheck deductions and take-home comparisons while working, while 6C changes retirement eligibility and pension formula.
State-tax coverage
State income tax is modeled for all 50 states and DC. Detailed bracket and deduction rules cover the 42 jurisdictions with income tax; the 9 no-income-tax states are correctly handled as $0.
See per-state FERS, TSP & SS treatment →Calculator assumptions you can change
Inflation, salary growth, FEHB premium inflation, TSP returns, COLA, mortality years, projection horizon, and historical stress-test presets are exposed as scenario assumptions.
See the Field Guide for input mechanics →Known Gaps The important missing pieces should be easy to spot. 9 current gaps are tracked here, including part-time service prorating, Roth basis handling, local-tax exclusions, and survivor-edge simplifications.
- CSRS and CSRS-Offset are not modeled.
- Part-time federal service prorating is not yet modeled.
- The FERS supplement assumes no post-retirement earned income, so the SRS earnings-test reduction is not modeled for bridge jobs or contracting income.
- Roth TSP basis is not tracked — when the 5-year requirement is not met, the entire Roth withdrawal is treated as taxable rather than only the earnings portion. This is conservative but may overstate taxable income.
- Some state-specific edge rules are intentionally simplified where the current household-level input model does not support per-person precision.
- Local city, county, and municipal income taxes are intentionally excluded from the model.
- Survivor Social Security uses a simplified age-based reduction factor rather than the full SSA survivor benefit computation.
- FEHB continuation rules after a spouse's death are not modeled — premiums remain unchanged in the projection.
- Remarriage effects on survivor pension eligibility and filing status are not modeled.
Baked-In Assumptions The projection starts from a clear rule set. Federal tax rules, state-tax baselines, FEHB growth, TSP return assumptions, COLA handling, and Roth qualification rules are all defined here.
- Federal tax brackets and standard deductions are inflation-indexed forward from the 2025 baseline using the scenario inflation assumption.
- Pre-retirement salary uses the scenario annual salary growth rate and compounds year-over-year until retirement; growth set to 0% preserves flat salary behavior.
- State-tax rules use a 2025-law baseline, but most state brackets are held constant rather than inflation-indexed year by year.
- The Social Security wage base is indexed forward from the 2025 base ($176,100) using a fixed 3.5%/yr average-wage-index (AWI) growth assumption and simple rounding rather than SSA’s official wage-base rounding convention. Federal tax and Medicare baseline constants are reviewed against annual IRS, SSA, and CMS releases before each tax year.
- FEHB premiums grow at the scenario FEHB premium inflation rate.
- TSP growth uses separate pre-retirement and post-retirement return assumptions in the deterministic projection.
- Roth TSP withdrawals apply the IRS 5-year qualification rule when a first contribution year is provided. When the year is left blank or the account is already qualified, earnings are treated as tax-free. Basis is not tracked — unqualified withdrawals conservatively treat the entire amount as taxable.
- Social Security benefit amounts are calculated using whole-year claiming ages. Benefits claimed in a month other than your birthday month may differ from the calculator estimate by roughly 1–2%.
- FERS COLA follows the standard FERS rule structure: full CPI up to 2%, capped at 2% for 2% to 3%, and CPI minus 1% above 3%.
- Projection length is based on the chosen simulate-through age, with a minimum window so short scenarios still produce usable output.
- When historical stress replay is enabled, CPI and TSP fund returns for that run come from the app’s bundled annual dataset (not from your static TSP return assumptions). The primary projection row in the calculator still uses those static rates unless you are looking at the Stress tab outputs.
Monte Carlo Scope Useful stress test, not full-world uncertainty modeling Monte Carlo randomizes TSP investment returns only. The rest of the household cash-flow model stays deterministic.
- Monte Carlo randomizes annual TSP investment returns only.
- Pension, Social Security, FEHB, Medicare, taxes, and other cash-flow rows stay fixed from the deterministic scenario.
- The default portfolio statistics represent a blended TSP allocation assumption rather than your exact holdings unless you interpret them that way yourself.
Historical Replay Scope Bundled history on your timeline — illustrative, not predictive Results → Stress runs optional extra projections using annual CPI and fund-return sequences from the app’s history file, separate from Monte Carlo and from your long-run static return assumptions.
- Historical replay uses the bundled annual dataset in the app (CPI and per-fund TSP returns). It is not your personal account history and not a guarantee of future results.
- Replay aligns the chosen history sequence to your projection timeline; calendar labels follow your plan while economics follow the selected slice.
- L Fund glide during replay is approximated from published glide-path weights.
- If the dataset ends before your full simulate-through horizon, stress paths stop early and the UI should indicate truncation.
Use Directionally Some scenarios should sharpen your questions, not settle the decision. Treat the result more cautiously when future law, unsupported retirement systems, or timing-sensitive edge cases are driving the answer.
- When your situation depends on future law, tax, Medicare, or policy changes that are not known yet.
- When you are in a retirement system or employee category the calculator does not yet support directly.
- When small timing details matter enough that integer-age Social Security claiming or simplified state-tax handling could move the answer materially.
- When you are using Monte Carlo output as if it were a forecast of total household uncertainty rather than a TSP-return stress test.
- When you are treating historical replay output as a prediction of future markets or exact fund performance rather than a directional illustration with bundled, imperfect proxies.
Evergreen Note
This page changes whenever coverage changes, so what you read here should match what the calculator actually does today.
Use FERSCalc to compare options, then confirm major decisions the right way.
The right workflow is to use the calculator for scenario planning, then verify final retirement timing, tax treatment, and agency-specific questions with official sources and qualified advisors.