Methodology

How we compute the numbers — and what we leave out.

Every result on this site is calculated from publicly cited tax tables. This page lists each tool’s method, every source we use, the dates we last verified them, and the things we deliberately don’t compute so you know where the math stops.

Last full refresh: May 22, 2026 · Calc engine v1.3.0 · Next scheduled refresh: November 2026
01 — Calculation engine

Three tools, each computed server-side from versioned, auditable code.

Every calculation runs on our server in TypeScript, not in the browser. The reasons: compliance (one auditable place for the math), trust (we can cite the exact source and verification date in every result), and integrity(the business-critical logic doesn’t ship in fork-able client-side JavaScript). Each tool exports a CALC_VERSION that gets stored with every result, so months later you can tell exactly which version of the math produced a given number.

The Paycheck Snapshot

CALC_VERSION 1.3.0

Real take-home pay after federal income tax, FICA (Social Security + Medicare), state income tax, and state disability where applicable. Uses the IRS percentage method (Pub 15-T) for federal, and each state's official withholding schedule for state.

Federal method
IRS Pub 15-T percentage method (W-4 2020+)
State method
Per state — see §3
Tests
Property-based vs. IRS reference cases

The Cash Flow Lens

CALC_VERSION 1.1.0

Pure subtraction across user-entered income and expense categories. Surfaces a margin-of-comfort figure and where the largest single pressure sits. No tax math here — this tool consumes the Paycheck Snapshot's net figure rather than recomputing it.

Inputs
Net income · fixed costs · variable costs
Output
Monthly delta + pressure analysis
Tests
Unit tests on edge cases (zero income, negative delta)

The Debt Clock

CALC_VERSION 1.2.0

Amortization schedule for a single fixed-rate debt at a given monthly payment, then a second schedule at payment + extra. Reports both payoff months and total interest paid. Uses standard monthly compounding (APR / 12).

Formula
Monthly amortization · APR/12
Inputs
Balance · APR · monthly payment · optional extra
Tests
Property-based — payoff months bounded by analytic formula

Multi-Debt Avalanche

CALC_VERSION 0.1.0

Month-by-month simulation across up to 5 debts. Every debt pays its minimum each month; the extra-payment budget is routed to the highest-APR debt first, then cascades to the next as each debt clears. A parallel snowball (smallest-balance-first) simulation runs on the same inputs to show the head-to-head difference. Mortgages are excluded — they're tax-advantaged and modeled separately in Refi Break-Even.

Method
Avalanche primary · snowball comparison · minimums-only baseline
Compounding
Monthly · APR/12 applied before each month's payment
Inputs
1–5 debts (balance · APR · min payment · kind) + extra budget
Cascade
Freed minimums roll into the next debt in priority order
Tests
Property-based, 19 cases — provenance, ordering, edge cases, invariants

Emergency Fund Sizing

CALC_VERSION 0.1.0

Months-of-fixed-costs target range based on convention from the CFPB Emergency Savings guidance and the FINRA Investor Education Foundation. Base of 3 months plus risk additions for variable income, single-earner households, and dependents. Lower bound is base + boosts; upper bound is the lower bound plus 3 months, capped at 12. Compares against the user's current balance and projects time-to-fill at their savings rate.

Base
3 months — CFPB + FINRA convention
Risk additions
Income type (0–2) · single-earner (+1) · dependents (+1)
Range width
3 months between lower and upper bound, upper capped at 12
Inputs
Monthly fixed costs · current balance · income type · household · dependents · (optional) monthly savings
Tests
20 cases — recommendation model, gap analysis, time-to-fill, notes, edge cases

Savings Rate Reality

CALC_VERSION 0.1.0

Closed-form years-to-FI projection. The user's savings rate (take-home minus spending, divided by take-home) becomes the input to a derivation of the future-value-of-annuity formula set equal to the 25× annual-spending target. Assumes a 5% real return on invested assets and a 4% safe withdrawal rate (Trinity Study + Bengen + MMM convention). Supports an optional existing-balance input and emits a sensitivity table at +5pp / +10pp / +15pp savings-rate bumps.

Real return
5% (after inflation) · MMM / Bogleheads convention
Withdrawal rate
4% · Trinity Study + Bengen 1994
FI multiple
25× annual spending (1 / withdrawal rate)
Closed form
n = log((target + S/r) / (B₀ + S/r)) / log(1+r)
Tests
23 cases — canonical MMM-table match, derived values, edge cases, sensitivity monotonicity

Refi Break-Even

CALC_VERSION 0.1.0

Standard mortgage amortization on the new loan to derive the new monthly payment, then two break-even computations. Simple break-even is closing costs divided by monthly savings — the number every refi calculator surfaces. Total-cost crossover is a month-by-month walk of cumulative current-loan payments versus cumulative new-loan payments plus closing costs; the crossover month is when the new path becomes cheaper. Optional planned-holding-period drives a structural verdict (clear-win / marginal / underwater / refi-costs-more).

Amortization
Monthly compounding · annual rate / 12
Simple break-even
closing_costs ÷ monthly_savings
Crossover
Month-by-month walk to MAX_CROSSOVER_MONTHS = 480
Verdict tiers
Break-even ≤ 50% of plan → clear-win · ≤ 100% → marginal · > 100% → underwater
Not modeled
Cash-out · points · PMI threshold · tax deduction · ARM
Tests
21 cases — happy path, verdict tiers, notes, same-term refi, edge cases
02 — Federal sources

Three federal publications, re-verified per tax year.

The federal portion of every tax calculation comes from one of these three sources. We re-fetch and re-parse each one when the IRS or SSA publishes a new annual update (typically late October / early November), and again on a quarterly spot-check to catch mid-year errata.

Federal · 3 sources
Federal ITIRS Publication 15-T (2026) — percentage method2026
FICASSA OASDI & Medicare rates 20262026
FICA wage capSSA contribution & benefit base 20262026
03 — State sources

All 50 states + the District of Columbia.

Each state’s source is the official Department of Revenue (or equivalent) withholding publication for tax year 2026. Italicized entries note states where the calculation has a known caveat — typically because the state uses a flat tax, no income tax, or a method we can’t fully replicate without information we don’t ask for.

States with graduated income tax + withholding tables · 32 states
AlabamaAL DoR Form A-4 (2026)May 17
ArkansasAR DFA AR4EC (2026)May 17
CaliforniaCA EDD DE 44 (2026), Method BMay 17
ConnecticutCT DRS IP 2026(1)May 17
DelawareDE DoR Withholding 2026May 17
GeorgiaGA DoR Pub IT-560 (2026)May 17
HawaiiHI DoT Booklet A (2026)May 17
IowaIA DoR Withholding Tables (2026)May 17
KansasKS DoR KW-100 (2026)May 17
LouisianaLA DoR R-1306 (2026)May 17
MaineME MRS Withholding Tables 2026May 17
MarylandMD CoT Withholding Guide 2026May 17
MinnesotaMN DoR Withholding Tables 2026May 17
MississippiMS DoR Form 89-350 (2026)May 17
MissouriMO DoR Pub 4282 (2026)May 17
MontanaMT DoR Withholding Tables 2026May 17
NebraskaNE DoR Circular EN (2026)May 17
New JerseyNJ DoT NJ-WT (2026)May 17
New MexicoNM TRD FYI-104 (2026)May 17
New YorkNY DTF NYS-50-T-NYS (2026)May 17
North DakotaND OST Income Tax Withholding 2026May 17
OhioOH DoT Employer Withholding Tables 2026May 17
OklahomaOK TC Packet OW-2 (2026)May 17
OregonOR DoR Withholding Tax Tables 2026May 17
Rhode IslandRI DoR Withholding Tables 2026May 17
South CarolinaSC DoR SC1040ES (2026)May 17
VermontVT DoT GB-1210 (2026)May 17
VirginiaVA TAX Pub 2026-1May 17
West VirginiaWV TaxDiv Withholding Tables 2026May 17
WisconsinWI DoR Pub W-166 (2026)May 17
D.C.DC OTR Withholding Tables 2026May 17
IdahoID STC EPB00006 (2026)May 17
Flat-tax states · 10 states
ArizonaAZ DoR — 2.5% flat (uniform method)May 17
ColoradoCO DoR — 4.4% flat (DR 1098)May 17
IllinoisIL DoR — 4.95% flat (IL-700-T)May 17
IndianaIN DoR — 3.05% flat (Departmental Notice #1)May 17
KentuckyKY DoR — 4.0% flat (2026 Withholding Tables)May 17
MassachusettsMA DoR — 5.0% flat (Circular M 2026)May 17
MichiganMI Treasury — 4.25% flat (446-T 2026)May 17
North CarolinaNC DoR — 4.25% flat (NC-30 2026)May 17
PennsylvaniaPA DoR — 3.07% flat (REV-415)May 17
UtahUT TC — 4.55% flat (Pub 14, 2026)May 17
No state income tax · 9 states (caveat: federal-only result)
AlaskaNo state income tax · federal-only calculation
FloridaNo state income tax · federal-only calculation
NevadaNo state income tax · federal-only calculation
New HampshireNo tax on wage income (interest/dividends only)
South DakotaNo state income tax · federal-only calculation
TennesseeNo state income tax · federal-only calculation
TexasNo state income tax · federal-only calculation
WashingtonNo state income tax on wages (capital gains only)
WyomingNo state income tax · federal-only calculation
04 — What we don’t compute

Things we deliberately don’t include — so you know where the math stops.

Most paycheck calculators tell you what they cover. Few tell you what they don’t. These are the categories we know our calculation excludes — if any of them apply to your situation, your actual paycheck will diverge from our estimate. For decisions that depend on a precise figure, confirm with your payroll team or a licensed tax professional.

  • Local taxes

    San Francisco, NYC, Portland, Cleveland, Detroit, Philadelphia, Birmingham, Louisville, Kansas City and others. State portion only.

  • Multiple jobs / spousal income

    We treat each calculation as if it's a single W-2 source. The W-4 Multiple Jobs Worksheet adjustment is not applied.

  • Bonuses & supplemental wages

    Bonus withholding (22% flat or aggregate method) and other supplemental wage rules are not modeled. Enter base wages only.

  • Equity compensation

    RSU vesting, ISO/NSO exercises, ESPP purchases, and Section 83(b) elections are out of scope. These often dominate paycheck variance for tech employees.

  • Itemized deduction modeling

    We assume the standard deduction. Mortgage interest, SALT, charitable contributions, and other itemized deductions aren't modeled in the paycheck-level withholding calc.

  • YTD wage progression

    Each result treats the paycheck as if it's mid-year (steady state). Year-to-date wage caps like the Social Security cap aren't applied until you enter YTD wages, which we don't ask for.

  • Employer-specific deductions

    Beyond 401(k)/HSA/health premiums you enter, we don't model garnishments, union dues, transit/parking, dependent-care FSA, or commuter benefits.

  • Personalized tax advice

    Nothing on this site recommends a course of action for your specific situation. We compute figures from public tables — interpretation and decisions are yours.

05 — Annual refresh process

When and how the tax tables get updated.

Tax tables change every year — usually announced October–December for the following tax year. Here’s the schedule we run on so the numbers stay current without quietly drifting:

OCT

Watch for IRS Pub 15-T draft

Auto-filed Linear ticket reminds us to begin tracking federal changes the moment the draft drops.

NOV

State DoR publications

Each state's publication is re-parsed into a versioned data/YYYY.ts file. The diff is reviewed in PR.

DEC

Run reference cases

Calc engine runs against published reference paychecks (IRS Pub 15-T examples, state DoR examples) to confirm parity.

JAN 1

New tax year live

New tables ship behind a date-gated switch. Old-year scenarios are still computable with their original calc_version.

06 — Disclosure & legal

What this site is, and what it isn’t.

Aplomia is not a financial advisor, planner, lender, broker, or tax professional. Every result is a planning estimate generated from publicly cited tax tables. Nothing on this site constitutes personalized tax or financial advice for your specific situation.

We don’t sell leads, don’t run display advertising, and don’t make money from your decision after a calculation — no incentive changes a tool’s result. Aplomia is free today; planned revenue comes from advisors who license our white-label embed, plus clearly-labeled affiliate or offer placements that never affect the math — see How we make money for the full picture.

If you believe a number on this site is wrong — different from the published source, or missing a relevant edge case — please report it at methodology-brendan@aplomia.com. We fix bugs in the calc engine within 5 business days of confirmation and post a changelog entry.