How we calculate

We apply 2026 federal tax brackets based on your filing status, then estimate state taxes using your state. FICA (Social Security 6.2% + Medicare 1.45%) is deducted from gross. The result is your net monthly income — the number that actually matters for affordability.

Lenders measure housing cost against gross income. SyntheFi measures it against take-home pay, and counts upkeep and PMI in the cost. The bands are 28%, 36% and 43%. They take their shape from the familiar lender ratios, but they are SyntheFi's own and stricter by design.

A household earning $120k gross in a 25% effective tax bracket has $7,500 a month in take-home pay. Our 28% band allows $2,100 a month for housing. A lender's gross-based 28% would allow $2,800. We think the smaller number is the honest one.

We add up monthly debt (housing, car payments, student loans, minimum card payments and other recurring debt) and divide by gross monthly income. Between 36% and 43% lowers a Safe verdict to Stretch. Above 43% lowers Safe or Stretch to Risky. It never raises a verdict.

Test 1 is the ratio. Housing cost against take-home pay is Safe up to 28%, Stretch to 36%, Risky to 43% and Not Viable above 43%. Debt-to-income can then lower it.

Test 2 is flex: what is left each month after housing, living costs and debt payments, before saving. Safe needs 17.5% of gross income, the midpoint of the 15–20% that advisors recommend saving. Stretch needs one month of take-home pay per year, which is take-home divided by 12. CFPB research (2022) finds that one month of income in savings separates households that fall behind on bills from those that do not. Risky is anything from $0 up. Below $0, or with no take-home pay, is Not Viable.

Your debt payments count in flex. We judge flex before saving, so raising your savings line never lowers your verdict.

Then a cash check can cap the verdict by the cash left after closing, if you tell us your cash on hand. See Cash after closing below.

If you tell us your cash on hand, we subtract your down payment and closing costs and count how many months of housing costs are left. Closing costs are assumed to be 3% of the home price, in the lower middle of the 2–5% range commonly cited. The JPMorgan Chase Institute (2019) found that borrowers with under one month of payments in reserve were 20% of borrowers but 54% of defaults. HUD 4000.1, the handbook for federally insured loans, treats three months of reserves as a compensating factor.

With three or more months left, nothing changes. With one to three months, the verdict is Stretch at most. With under one month, or cash that falls short of the down payment and closing costs, it is Risky at most. Cash never raises a verdict, and it never rules a home Not Viable on its own, because you may hold savings we cannot see, such as retirement accounts or gifts.

Monthly rent recommendations for your ZIP come from HUD's published Small Area Fair Market Rents (the same numbers HUD uses for Section 8 housing). We carry the 2BR FMR for every ZIP HUD covers. Where HUD's Small Area FMR doesn't cover a ZIP, we fall back to Census ACS median gross rent at the ZCTA level. HUD FMR is intentionally a median-renter figure (typically 10–15% below market in expensive metros), which we view as the right anchor for affordability — what most people in that ZIP actually pay, not what the most expensive listings show.

Buy-scenario home prices are anchored to the US Census American Community Survey 5-Year median owner-occupied home value, by ZCTA. When you start a buy scenario, the form pre-fills with the Census median for your ZIP as a starting point. You can change it; the verdict updates live.

Utilities are state electricity (from EIA Form 861 average monthly residential bills) plus a national-average bundle for gas + water + sewer + residential internet + 2 mobile phone lines. Electricity varies dramatically by state (HI is $202/mo, NM is $84/mo); the other components vary much less, so we hold them at national averages. The combined figure is what a typical 2-bedroom household pays for all utilities.

Non-housing spending categories (food, transport, healthcare, entertainment, etc.) are adjusted by the BEA's Regional Price Parities at the metro level. A user in San Francisco metro sees the same category at 109% of US average; a user in Hanford-Corcoran CA sees 99%. This replaces the old single-state-multiplier approach which over-estimated SF non-housing by ~35%. Metro-level coverage today is the top 42 US metros; outside those, we fall back to state-level RPP.

Per-category spending recommendations (food, transport, healthcare, entertainment, subscriptions, personal care) start from the US Bureau of Labor Statistics Consumer Expenditure Survey 2024, bucketed by income quintile. We then calibrate the figures to a specific persona — urban dual-income professional couple, 28–38, renting, no kids, saving for a first home — because raw BLS quintile averages include retirees, suburban families, and multi-car commuters whose spending shape is fundamentally different. The calibration sources are USDA Moderate Cost Food Plan (food), KFF Employer Health Benefits Survey (healthcare), JCHS Rental Housing report (transport), Lemonade 2026 (renter's insurance), and Perplexity Sonar Deep Research against Federal Reserve SCF microdata.

Unlike spending categories — which show what comparable households actually spend — the SAVINGS line shows what advisors recommend you save. The number is 17.5% of your bracket's average gross income, the midpoint of the 15–20% range broadly recommended by financial planners. This is the only line in the recommendations that is prescriptive: every other line answers 'how much do households like mine spend?', the SAVINGS line answers 'how much should I save?'

Persona calibration is applied to Q3, Q4, and Q5 income brackets ($56k–$220k+), covering the urban-professional life-stage range. Q1 and Q2 still use raw BLS averages because their demographics are structurally different (entry-level wages, single-earner households) and require their own persona work. Geographic data coverage: full-US Census ACS (~33,000 ZCTAs) for home value and median rent; full-US county property tax (~3,200 counties); HUD FMR currently top-50 metros while we finalize bulk-access setup. ZIPs without coverage fall back gracefully to state-level data with a clear citation.