Where you stand
| Range | Tier |
|---|
Front-end (housing only) and back-end (everything). Lenders care about both. We tell you exactly where you stand.
| Range | Tier |
|---|
Underwriters compute DTI two ways. Front-end isolates housing burden — useful for asking "can I afford this PITI on this income?" Back-end is the full debt picture — what lenders actually approve against. The 28/36 rule is the textbook benchmark; the CFPB Qualified Mortgage rule (12 CFR 1026.43) sets 43% as the de-facto regulatory ceiling for the best pricing.
PITI — Principal + Interest + Taxes + Insurance + HOAGross income — pre-tax monthly (verifiable W-2/1099/tax return)Minimum debt — minimum payments on the credit report, not actual| Loan type | Front-end | Back-end (typical / max) | Authority |
|---|---|---|---|
| Conventional (Fannie/Freddie) | 28% | 43% / 50% | Fannie Mae Selling Guide |
| FHA | 31% | 43% / 56.99% | HUD Handbook 4000.1 |
| VA | — | 41% (residual-income primary) | VA Lender's Handbook M26-7 |
| USDA Rural Development | 29% | 41% | USDA RD Handbook 1980-D |
| Jumbo (non-conforming) | 28% | 43% | Lender overlay (Fannie not eligible) |
"Compensating factors" that let lenders push above the standard ceiling: 12+ months of reserves after closing, FICO 740+, down payment 20%+, residual income materially above county thresholds (VA), low housing-payment increase from current rent.
Lenders use gross income (pre-tax). Most household budgets think in net income (after-tax take-home). A 36% DTI on gross can feel like 50%+ of net — perfectly approvable on paper but tight in lived budget. Run both numbers before signing: PITI ÷ net (target <30%) is a useful "can I actually live with this" sanity check.
Lenders re-pull credit days before closing. A new auto loan, a furniture financing line, or a charged-up card can push DTI over the qualification threshold and kill the deal. From application to closing: no new credit, no big purchases, no employment changes. The CFPB explicitly warns about this in its Know Before You Owe guidance.
If your federal student loans are on Income-Driven Repayment, some lenders count the actual IDR payment (sometimes $0) for DTI; others impute 0.5-1% of balance. Conventional Fannie/Freddie accept the IDR payment if documented; FHA traditionally imputed 1% but updated guidance allows actual payment. Ask your loan officer which treatment applies — it can change DTI by 5-10 points on a typical loan balance.
Sources: CFPB Ability-to-Repay/QM Rule (12 CFR 1026.43); Fannie Mae Selling Guide §B3-6; Freddie Mac Single-Family Seller/Servicer Guide §5401; HUD Handbook 4000.1 (FHA Single-Family); VA Lender's Handbook M26-7; USDA Rural Development Handbook HB-1-3555; CFPB "Know Before You Owe" mortgage disclosure guidance.
Front-end DTI = housing payment (PITI) ÷ gross income. Measures housing burden alone. Back-end DTI = (housing + all minimum debt) ÷ gross income. Lenders evaluate both. The 28/36 rule sets front ≤28%, back ≤36%. The CFPB Qualified Mortgage rule (12 CFR 1026.43) caps back-end at 43% for QM-eligible loans, though many lenders extend to 45-50% with strong compensating factors.
Under 36% back-end qualifies for the lowest mortgage rates. 36-43% is acceptable for most loans. 43-50% requires compensating factors and limits to FHA or non-QM. Above 50% is rarely approvable. For non-housing borrowing, keep under 36% to maintain rate access.
Lenders count: PITI, auto, student loan (or 1% of balance for deferred), credit card minimums, personal loans, child support, alimony, and any debt with ≥10 months remaining. Not counted: utilities, groceries, insurance premiums (unless escrowed), 401(k) loan payments, tax withholding. Some lenders also exclude debts paid off before closing.
Future PITI replaces current rent. Rent at $1,800 + future PITI of $2,400 = back-end DTI rises by $600/month. Lenders use full PITI including escrowed taxes/insurance/HOA, not P&I only. This is why a "rent < mortgage" situation often pushes DTI just over the qualification threshold.
The Ability-to-Repay/QM rule (12 CFR 1026.43, effective 2014) gives lenders legal safe harbor on QM loans — historically a 43% back-end DTI cap, replaced in 2021 by an APR-based pricing test. Most lenders still treat 43% as the practical ceiling for best pricing. Above, loans are non-QM — legal but typically 0.5-1.5 points higher. FHA, VA, USDA have separate exemption pathways.
Conventional: 28/43-50%. FHA: 31/43-56.99% (HUD Handbook 4000.1). VA: no front-end, 41% back-end but residual-income primary metric. USDA: 29/41%. Jumbo: 28/43% with stricter overlays. The calculator above flags which thresholds you clear.
Yes, twice over. Eliminating a $100/month minimum drops back-end debt by $1,200/year. It also lowers utilization, lifting FICO 30-100 points — and a higher FICO unlocks better mortgage rates. The double-win is why mortgage planners recommend paying cards down to ~10% utilization (not zero) 60-90 days before applying.
Usually no. Student loans carry 4-7% rates — below mortgage rates — and qualify for IDR, deferment, and forgiveness pathways personal loans don't. Aggressive paydown burns liquidity. Better tactics: enroll in IDR to lower the reported minimum (some lenders use the actual IDR payment, even if $0), or refinance to extend term. Confirm the lender's policy first.
Sometimes. A non-occupant co-borrower's income is added, lowering combined DTI — but their debts are added too. Only helps if their income/debt ratio is materially better. Co-signer takes equal liability — late payments hurt both reports, and the loan counts against the co-signer's own DTI. FHA and VA have specific co-borrower rules; conventional Fannie loans allow non-occupant co-borrowers with full documentation.
Faster than improving FICO. DTI updates as soon as a debt is paid off and the lender pulls fresh credit. Three highest-leverage moves: (1) pay off any debt with fewer than 10 months remaining — FHA can exclude these entirely; (2) pay down cards to lower minimums and utilization; (3) avoid new debt — a new auto loan during the 60 days before closing can disqualify an approved mortgage.