ARM vs Fixed Mortgage in 2026: Pricing the Adjustment Risk

Should you take an ARM or a fixed rate in 2026?

Take the ARM only if you have a concrete reason to believe you will be out of the loan, by sale or refinance, before the fixed period ends. In 2026 the discount an adjustable rate buys you is small, and a single adjustment can erase years of savings in under two years. The fixed rate costs a little more each month and hands you certainty for the whole loan. The honest question is not which rate is lower today; it is what the adjustment risk is worth to you.

As of August 6, 2026, Freddie Mac put the average 30-year fixed at 6.69% and the 15-year fixed at 6.01%. Adjustable rates are no longer part of Freddie Mac's weekly survey, and in practice the starting rate on a 5/6 or 7/6 ARM in 2026 has been running only modestly below the 30-year fixed, often a quarter to half a point. That thin discount is the whole story.

What an ARM actually is in 2026

An adjustable-rate mortgage has a fixed period, then a floating one. The name tells you the schedule. A 5/6 ARM is fixed for five years, then adjusts every six months after that. A 7/6 ARM is fixed for seven years, then every six months. The old 5/1 naming (adjust once a year) is mostly gone; today the adjustments are usually every six months.

When it floats, your new rate is an index plus a margin. The index moves with the market. The margin is a fixed number set in your contract, and it does not change. So your adjusted rate is roughly the index on the reset date plus that margin, subject to caps.

Caps are the safety rails, and they come in three numbers, per the Consumer Financial Protection Bureau. The initial cap limits the first jump when the fixed period ends, commonly 2% or 5%. The periodic cap limits each later change, commonly 1% or 2%. The lifetime cap limits the total increase over the life of the loan, usually 5% above your start rate. Those three numbers are printed on your Loan Estimate. Read them before anything else, because they define your worst case.

The honest comparison at today's rates

At an illustrative half-point discount, the monthly savings on an ARM are real but small, and the fixed rate buys certainty for the price of one dinner out a month. Take a $400,000 loan (a $500,000 home with 20% down). Compare a 30-year fixed at 6.69% to a 5/6 ARM starting at 6.19%, half a point lower.

The 30-year fixed principal and interest is $2,578 a month. The ARM at 6.19% is $2,447. That is a difference of $131 a month during the fixed years. Over the full five-year fixed period, the ARM saves you about $7,871, assuming the discount holds and nothing else changes.

That $7,871 is the entire prize. It is what you are being paid to take the adjustment risk. Whether that is a good trade depends completely on what happens in year six, which no one selling you the loan can promise.

If you want to run your own two rate quotes side by side and see the monthly gap, the mortgage points break-even calculator computes the difference between any two rates and the month one catches the other, and the full Offer Buildr calculator folds the payment into your true monthly cost with taxes and insurance.

What the adjustment can actually cost

One ordinary adjustment can wipe out the entire five-year cushion in less than two years. Stay with the same $400,000 loan. After five years of paying the 6.19% ARM, the balance is about $373,081. Now the fixed period ends and the rate adjusts.

Assume a common 2% initial cap. Your rate can jump from 6.19% to 8.19% in one step. The lender recasts the remaining balance over the remaining 25 years, so the new payment on $373,081 at 8.19% is about $2,927 a month. That is $479 more than your old ARM payment, and $348 more than the 30-year fixed would have cost you the whole time.

At $348 a month above the fixed, the five-year savings of $7,871 are gone in about 23 months. Everything after that is pure loss compared with having simply taken the fixed rate. And 8.19% is not the worst case. With a 5% lifetime cap, the rate could reach 11.19%, which on that balance is about $3,708 a month, roughly $1,261 above where you started.

The point is not that rates will do this. It is that they can, and the contract lets them. The fixed rate removes that entire branch of outcomes for about $131 a month.

When an ARM is a defensible bet

An ARM makes sense when you have a specific, near-certain exit before the fixed period ends, not a vague hope that rates will fall. The clean cases are real. You know the job relocates in four years. You are buying a starter home you plan to sell inside five. You have a large, scheduled liquidity event, a vesting cliff or a maturing asset, that will pay the loan off. In those cases you capture the discount and are gone before the risk arrives.

The weak case is the common one: taking the ARM because the payment is lower and assuming you will refinance before it adjusts. Refinancing depends on rates being lower then, on your income and credit holding, and on the home still appraising. None of those are promises. If you are betting on a refinance you do not control, you are not buying a discount, you are buying a hope and pricing it as a certainty.

There is also a discipline version of the same trade, and it is more honest. If the ARM's lower payment is only affordable because it is lower, that is a warning, not a feature. The fixed rate is the version that still works if nothing goes your way.

If your real plan is to pay it off fast

If your plan is to attack the balance aggressively, compare the ARM against a 15-year fixed, not just the 30-year, because the 15-year locks the low rate and the fast payoff together. On the same $400,000 loan, the 15-year fixed at 6.01% is about $3,378 a month. That is $799 more than the 30-year fixed and $931 more than the ARM, but the rate is locked and the loan is gone in half the time.

A frequent plan is to take the ARM for its low rate and prepay hard, aiming to kill the balance before it adjusts. It can work, but notice what it depends on: your willpower every month for five years. The 15-year fixed enforces the same discipline by contract. If you genuinely have the cash flow to prepay an ARM down fast, you usually have the cash flow to carry the 15-year, and the 15-year removes the risk that a bad year derails the plan right before the reset. Contract-enforced beats willpower-enforced when the downside is a payment shock.

For a related tradeoff on paying to lower your rate up front, see whether paying mortgage points is worth it, and if a builder or seller is offering rate help instead, temporary vs permanent buydowns covers where that money does the most good.

Pressure-test it before you sign

Do not compare the two starting payments. Compare the fixed payment against the ARM's worst realistic payment, and ask whether you could carry that number. Pull the initial cap, periodic cap, and lifetime cap off the Loan Estimate and compute the payment at the start rate plus the initial cap, then at the lifetime cap. If either of those numbers would break your budget and you do not have a certain exit, the fixed rate is the answer.

Remember too that the rate is only one line of the payment. Taxes and insurance move on their own schedule regardless of which loan you pick, which is a separate reason the monthly number can climb. That mechanic is worth understanding on its own, and it is covered in why your mortgage payment goes up in year two.

The ARM is not a trap and it is not a trick. It is a discount in exchange for a risk. In 2026 the discount is thin and the risk is real, so the burden of proof sits on the ARM: it should win only when you can name the exact reason you will be gone before the rate can move.

Estimates, not appraisals · not legal or financial advice.