Reading room

How much home can I afford, and which ratio decides it

Lenders apply two debt-to-income tests and only one binds you. Which one it is changes what you should do: clearing a debt can beat half a point of rate.

Every affordability conversation runs into the same wall: the lender’s answer and the household’s answer are different numbers, and the lender’s is bigger. That gap is not a trick. It comes from a specific piece of arithmetic that is worth understanding, because once you can see it you can see which lever actually moves it.

The two tests

Underwriting applies two debt-to-income ratios, both against gross monthly income.

Front-end (housing) ratio. The whole housing payment over gross income. “Whole” means principal, interest, property tax, homeowner’s insurance, HOA dues and mortgage insurance — the figure the industry calls PITIA, not the principal-and-interest number a quote leads with.

Back-end (total debt) ratio. The housing payment plus every other recurring monthly obligation — car finance, student loans, minimum card payments, personal loans, child support, alimony — over the same gross income.

The classic conventional guideline is 28% and 36%. FHA is more commonly cited at 31% and 43%, and automated underwriting will approve outside all of these numbers on compensating factors. The point is not the exact percentages. It is that there are two of them, they produce two different housing budgets, and the smaller one wins.

Which one binds you, worked

A household earns $118,000 a year — $9,833.33 a month gross — has $40,000 for a down payment, and is looking at a 30-year loan at a placeholder 6.5%, with property tax at 1.25% of value, insurance at $1,800 a year, and mortgage insurance at 0.58% of the loan where it applies.

With $640 a month of other debt, at 28% and 36%:

  • Front-end budget: 28% of $9,833.33 = $2,753.33 for housing.
  • Back-end budget: 36% of $9,833.33 = $3,540.00, less $640 of debt = $2,900.00 for housing.

The front-end test is the tighter one. The housing budget is $2,753.33 and the maximum price is $366,506 — a $326,506 loan at 89.1% LTV, with $2,063.74 of principal and interest, $157.81 of PMI, and the rest going to tax and insurance.

Here the other debts are not the constraint at all. Paying off the car would not raise this household’s ceiling by a dollar, because the back-end test has $146.67 a month of slack in it.

Now raise the other debt to $940 a month and nothing else changes:

  • Front-end budget: still $2,753.33.
  • Back-end budget: $3,540.00 less $940 = $2,600.00.

The back-end test now binds. The housing budget falls to $2,600 and the maximum price to $346,962 — nineteen and a half thousand dollars of house lost to $300 a month of debt.

Same income, same down payment, same rate. Different binding constraint, and therefore a completely different thing to do about it.

The affordability calculator prints which constraint bound the answer, in a sentence, next to the number. That is the field worth reading first.

Why the distinction changes your plan

For the $940-a-month household, the back-end ratio is binding, so retiring debt is the highest- return action available. Clearing $500 a month of it — paying off the car — takes the housing budget from $2,600 back to $2,753.33 and the price ceiling from $346,962 to $366,506. That is $19,544 more house.

Compare that with chasing the rate. Leaving the $940 of debt in place and finding a loan half a point cheaper — 6.0% instead of 6.5% — lifts the ceiling to $360,235, a gain of $13,273.

Clearing one car payment beat half a point of interest rate, and it is entirely within the borrower’s control on a timescale of months. For the $640-a-month household the picture inverts: the debt is irrelevant to the ceiling and the rate is the only lever, because the front-end test does not look at debts at all.

This is the whole practical value of knowing which test binds. The advice “pay down your debts before you apply” is right roughly half the time and a waste of effort the other half.

What raises the ceiling, in order of effect

  • Retire the debt with the worst payment-to-balance ratio — but only if the back-end test is the binding one. A $4,000 balance costing $340 a month buys more ceiling than a $20,000 balance costing $200.
  • More down payment, which helps twice: less loan, and above 20% no mortgage insurance competing for room inside the same housing budget.
  • A longer term. Thirty years instead of fifteen lowers the payment and raises the ceiling — and roughly doubles the interest. The amortisation schedule shows what that trade costs.
  • A lower rate, which is the lever everyone reaches for first and is usually neither the largest nor the one you control.
  • Documentable income you forgot to include — consistent overtime, a bonus with a two-year history, rental income. This is administrative rather than financial and it is free.

The costs that sit outside the ratio

The ratio is computed on the mortgage payment. Owning a house costs more than the mortgage payment, and none of the following is in the calculation:

  • Maintenance. A working planning figure is around 1% of the home’s value a year, lumpy and unpredictable — nothing for four years and then a roof.
  • Closing costs. Commonly 2–5% of the price, due in cash and on top of the down payment. On a $310,000 purchase that is roughly $6,200 to $15,500.
  • Utilities, which usually rise when you move out of a flat and into a house.
  • Furnishing the extra rooms, which is the one that surprises people in month two.
  • Income tax, which the gross-income convention has quietly ignored throughout.

A more useful question than the maximum

Work the problem backwards. Decide the monthly housing payment you would be content with in a bad month — a month with a vet bill and a broken boiler — and find the price that produces it. The monthly payment calculator runs in that direction as readily as the other one, and the answer is normally somewhere between 10% and 20% below the ratio ceiling.

That number is not a compromise. It is the one that survives contact with the next five years.

Also asked

Questions this raises

What is the difference between the front-end and back-end ratio?

The front-end ratio is your total housing payment — principal, interest, taxes, insurance, HOA and mortgage insurance — divided by gross monthly income. The back-end ratio is that same housing payment plus every other recurring debt payment, over the same income. The classic conventional guideline is 28% front-end and 36% back-end, and government programmes and automated underwriting both run wider. The lower of the two budgets is the one that actually limits you.

Why do lenders use gross income rather than what I take home?

Because tax withholding varies with filing status, state, deductions and elections in ways a lender cannot verify or standardise, while gross income is on the pay stub and the W-2. It is a convention for comparability, not a claim about what you can afford. It is also why a ratio-derived maximum tends to overshoot what a household can comfortably carry.

Does a bigger down payment raise the price I qualify for?

Yes, and by more than the down payment itself, because it does two things at once: it removes that much from the loan, and above 20% it removes the mortgage insurance from the monthly payment, which frees room inside the same housing budget for a larger loan. Below 20% the effect is smaller but still more than dollar for dollar.

Should I borrow the maximum I qualify for?

The ratio is a lending limit, not a budget. It is computed before income tax, before retirement contributions, before childcare and before maintenance, and it assumes today's income continues. Most of the trouble comes from buying at the ceiling and then meeting a roof, a boiler or a redundancy.

Floor four

Put your own numbers through it

Every figure in this article came out of the calculators on this site, run on the inputs printed beside it. Change the inputs and the answer changes — that is the point of showing the working rather than the conclusion.

Storey Home Lending is a fictional company built to demonstrate a website. It does not lend money, take applications, or hold any licence. Every figure, person and scenario on this site is illustrative.