Your Home Loan Debt-to-Income Ratio Just Got Approved.
Your Cash Flow Didn’t.

Lenders will approve a mortgage that puts up to half of your gross income toward debt. Convert that same approval to what actually lands in your bank account, and it can leave a household with less than $800 a month for food, gas, and everything else that is not a bill. That is not a hypothetical. It is the real math behind an approval most people assume means they can afford the loan.

What Debt-to-Income Ratio Actually Measures

Debt-to-income ratio, DTI, is the percentage of gross monthly income that goes toward debt payments. Lenders use two versions.

Front-end DTI counts only housing: mortgage principal, interest, property taxes, homeowners insurance, PMI if it applies, and HOA dues, divided by gross monthly income.

Back-end DTI counts everything: housing plus every other required monthly debt payment, car loans, student loans, credit card minimums, personal loans, divided by that same gross figure.

Back-end is the number that actually decides mortgage approvals. Both ratios are built on gross income, by design. Gross is the number a lender can verify in seconds from a W-2 or a pay stub.

What Lenders Actually Allow in 2026

The decades-old guideline is the 28/36 rule: housing at or under 28% of gross income, total debt at or under 36%. Almost nobody actually lends at those numbers anymore.

Conventional loans routinely approve back-end DTI up to 45%, and automated underwriting will stretch to 50% with compensating factors like cash reserves or a strong credit score. FHA loans target 31% front-end and 43% back-end on paper, but automated underwriting commonly approves up to 50%, and in some cases as high as 57%. VA loans can go as high as 60% in certain cases.

The Consumer Financial Protection Bureau’s old Qualified Mortgage rule used to cap DTI at 43% as a hard legal line. That cap was replaced in 2021 with a rate-based test instead, and 43% stuck around as an industry rule of thumb anyway. A Federal Reserve Bank of St. Louis analysis of more than 30 million mortgage applications from 2018 to 2024 found that 43% is not where the real risk boundary sits. A DTI of 45% gets treated about the same as a DTI of 35%. The real cliff is 50%, where denial rates jump 15 to 17 percentage points almost immediately.

All of it, the 36%, the 45%, the 50% cliff, is measured against gross income. Nobody in that chain is looking at the net income that is deposited in your account each pay period.

Why Gross Is the Wrong Denominator

We have written before about how the 30% housing rule was set in an era when gross and net pay were nearly the same number. Federal income tax barely touched most wage earners, Social Security did not exist until 1937, and Medicare would not exist for another 35 years after that. The gap between gross and net was a rounding error.

It is not anymore. Depending on income, 38 to 42 cents of every gross dollar disappears before a single spending decision gets made. DTI inherited the exact same gross-income convention, for the exact same reason housing rules did: it was never adjusted for a gap that did not exist yet when the convention formed.

So what does a 50% gross approval actually look like once it is measured against the paycheck a household actually takes home?

What a 50% Gross Approval Actually Costs You

Take the two households from that earlier analysis: one earning $70,000 a year, one earning $150,000. Both were shown, on a full realistic monthly budget, to already run a deficit against net pay, $854 a month short at $70K, $1,045 a month short at $150K, once every ordinary category (food, transportation, healthcare, utilities, clothing, entertainment) is included at national averages.

Now run the same households at the industry’s real ceiling, a 50% gross back-end DTI.

Both households were approved. Both were told, in effect, that this payment was affordable. Neither number came anywhere close to covering the cost of an ordinary month.

House Rich, Cash Poor, and Why the Bank Doesn’t Mind

There is a name for a mortgage that looks affordable on the approval letter but consumes so much of a household’s real cash flow that almost nothing is left: house rich, cash poor. The table above is exactly that condition, quantified.

It happens because a bank’s profit scales with loan size. A bigger mortgage means more interest paid over 30 years and a larger origination fee, and in most cases the loan is sold or serviced by someone else within months of closing anyway, so the originating lender’s exposure to what happens to a borrower’s quality of life over the next three decades is limited. The underwriting question being answered is whether this person will very likely make the payment. It is not whether this person will have anything left over for their kid’s soccer cleats. Those are two different questions, and only one of them gets asked.

This is not a claim that lenders are acting in bad faith. The bank is solving its own risk equation correctly. It was never solving the borrower’s.

The approval letter feels like validation, the bigger, better home, the number that says yes. Then the first mortgage payment clears, and the excitement meets the calendar. Groceries, gas, the kids’ activities, one night out, all of it now competing for a share of the paycheck much smaller than the approval letter implied.

The number that got a household approved was never the number that keeps it solvent. Here is the one that actually is.

What a Sustainable Net Budget Actually Requires

Using the same published figures, the real cost of everything that is not housing or debt, food, transportation, healthcare, utilities, clothing, personal care, entertainment, travel, a baseline savings contribution, runs about 70% of net income at both income levels tested (69.9% at $70K, 69.0% at $150K).

That leaves roughly 30% of net income as the real ceiling for housing and debt combined, at both levels, a striking convergence given an $80,000 difference in income.

The Real Number, Converted Back to What Your Lender Understands

Restate that 30%-of-net ceiling in the language a lender actually speaks, gross income, and it comes out to:

• $70K household: 18.8% gross back-end DTI
• $150K household: 18.0% gross back-end DTI

Roughly 18 to 19% gross back-end DTI is the real target. Not the 36% textbook guideline. Nowhere near the 45 to 50% a lender will actually let a borrower reach.

Worth being direct about the caveat: this is a break-even number, zero cushion, no extra retirement contribution beyond what is already assumed, no house maintenance reserve, no adjustment for a high-cost metro or an added dependent. In most of those situations, the real target is lower, not higher.

Where Can You Actually Buy This?

A target percentage only matters if the price it produces exists somewhere in the real market. At a 6.58% rate (the Freddie Mac national average for the week this was written), current national property tax and insurance averages, and PMI where it applies, here is what the real target actually buys, at three down payment levels.

At $150,000, every one of these is a real market. $217,716 sits close to Detroit’s metro median of $210,000. $233,397 clears Memphis at $215,000. And $272,853 at 20% down comfortably clears Pittsburgh, the cheapest large metro in the country this year at $250,000. This income level, held to the sustainable target, can actually buy a house.

At $70,000, the picture is tighter. The 5% and 10% down figures, $80,139 and $85,911, fall short of every real market found, including Akron, Ohio, one of the least expensive housing markets in the country, where prices often run under $101,000. Saving for 20% down changes that meaningfully: removing PMI entirely frees up enough of the budget to reach $100,435, just under Akron’s number, the first of the three scenarios that gets genuinely close to a real market.

Notice what is missing from every market cited so far: nothing is a West Coast or East Coast metro. Detroit, Memphis, Pittsburgh, and Akron are interior and Rust Belt markets. San Francisco’s three-month median sits at $1.7 million. Boston is at $852,000. Both are multiples of even the $150,000 household’s most generous scenario, and the national median itself, $398,771, already sits above what a $70,000 household can sustainably reach at any down payment tested here. On the sustainable target, coastal homeownership is not a stretch. It is a different order of magnitude entirely, one that a bigger down payment does not close.

One honest caveat: a $70,000 household is most plausibly a single income. Add a second earner, and total gross income rises, which directly raises the dollar amount available for housing at the same or a similar sustainable percentage. Working out exactly how much, and which real markets that opens up, requires modeling a complete second income tier the same way this analysis modeled $70,000 and $150,000, not just scaling one number. That is a deeper dive for its own piece.

Expect the Upsell, and Resist It

Here is the part almost nobody warns you about. 18 to 19% gross DTI sits so far below the 36% textbook guideline, and further still below the 45 to 50% real-world ceiling, that a borrower targeting the real number gets approved easily, often instantly.

That gap, between what someone is targeting and what they technically qualify for, is exactly the gap a loan officer or a realtor is trained to notice and fill. Expect some version of: you actually qualify for a lot more house.

It is the same incentive from the section above, showing up at the exact moment it is hardest to resist. A bigger loan means more interest paid over 30 years and a larger origination fee. Someone else’s bottom line, not the buyer’s quality of life.

Getting pre-approved is useful. Spending up to the pre-approval ceiling is optional. Nothing obligates a borrower to take on the full amount a lender is willing to offer.

The hard part was never the math. It is holding the number once someone with a financial incentive says a borrower can do better.

Find Your Own Number

The 18 to 19% figure came from two specific households at national-average costs. It is not a universal constant, it is evidence of a pattern, and the real number for any household depends on its own cost of living. Three steps get to a personal version:

1. Add up the realistic monthly cost of everything that is not housing or debt: food, transportation costs, healthcare, utilities, personal care, entertainment, and a baseline savings contribution.
2. Subtract that total from net, take-home, monthly pay. What is left is the most a household can safely put toward housing and debt combined.
3. Divide that number by gross monthly income. That percentage, not 36%, not 50%, is the real sustainable back-end DTI target.

Why Your Bank Will Never Hand You This Number

No lender is going to run this calculation for a borrower. It is not the question their underwriting model is built to answer, and there is less profit in a smaller loan.

aiSmartBudget already runs the calculation a lender never will: modeling real spending against actual take-home pay from real transactions, not a national average or a gross-income assumption. It is the same forward-looking budget engine behind the Annual Budget and Register, applied to the exact question this article raises: not what a household is approved for, but what it can actually live on once the loan closes.

Approved and Affordable Are Different Words

They are different words for a reason. A lender’s approval answers one question. A household’s actual month answers a different one. See both numbers before signing anything, not after the first payment clears.