Educational analysis only, not mortgage advice; all rate figures reflect August 2026 survey data from Freddie Mac and the Mortgage Bankers Association and change weekly.
TL;DR — Quick Verdict
- The 5/1 ARM discount is 0.68 percentage points as of mid-August 2026 — 5.99% versus 6.67% on a 30-year fixed, per MBA and Freddie Mac survey data.
- On a $400,000 loan, that gap is worth $178 per month, or $10,652 over the five-year fixed period.
- Add faster principal paydown and the ARM’s true five-year advantage reaches $13,648 — the payment savings alone understate it by 28%.
- A first reset to 7.99% under a 2/2/5 cap structure erases that entire advantage in 46 months, meaning the ARM wins only if you exit or refinance before roughly year nine.
- ARMs made up 7.9% of applications in the week ending August 7, 2026 — the market is still largely passing at this spread.
- Take the ARM if your documented exit horizon is under seven years and you can absorb the worst-case capped payment; otherwise take the fixed.
Sixty-eight basis points. That is the entire distance between a 30-year fixed mortgage and a 5/1 adjustable-rate mortgage in August 2026 — Freddie Mac’s Primary Mortgage Market Survey put the fixed at 6.67% for the week ending August 13, while the Mortgage Bankers Association recorded 5.99% on 5/1 ARMs for the week ending August 7. Historically, that spread has been wide enough to justify the gamble. Today it buys you $178 a month on a $400,000 loan, and the market knows it: ARMs accounted for only 7.9% of applications in the same MBA survey week.
Borrowers at Rocket Mortgage, Chase, and Navy Federal are being quoted both products side by side, usually with a monthly payment comparison and nothing else. That comparison is incomplete. It ignores the principal-paydown advantage that flows from a lower rate, it ignores the 0.83 points the MBA reports ARM borrowers are paying at origination, and it ignores what happens at month 61 under the cap structure written into the note. This analysis models all four variables on real August 2026 rates, shows the exact month at which a reset erases every dollar the ARM saved, and identifies which borrower profiles actually come out ahead.
The 2026 Rate Gap: What Fixed and Adjustable Actually Cost Right Now
Two federal-adjacent surveys anchor this comparison, and they measure slightly different things. Freddie Mac’s PMMS pulls from loan applications submitted through Loan Product Advisor and reflects conventional, conforming, fully amortizing purchase loans for borrowers putting 20% down with excellent credit. The MBA Weekly Applications Survey covers closed-end residential applications through retail and consumer-direct channels and is the only remaining weekly national source for ARM contract rates — Freddie Mac stopped publishing its 5/1 ARM series, so the two figures below come from different instruments and are one week apart.
Sources: Freddie Mac Primary Mortgage Market Survey and MBA Weekly Applications Survey. PMMS does not report points under current Loan Product Advisor requirements.
Notice the 0.83 points on the ARM. On a $400,000 loan that is $3,320 paid at closing, and it is included in the MBA’s origination-fee-inclusive figure. Any honest comparison has to net that against the monthly savings — which is exactly what most lender-supplied comparison sheets omit. Understanding mortgage points and rate buydown math matters more on an ARM than a fixed, because you are prepaying for a rate that only lasts 60 months.
The spread also moves independently of what most borrowers assume drives it. Short-term rates set ARM pricing while long-dated yields set fixed pricing, which is why the ARM discount compresses when the curve flattens — and with short rates anchored by a Fed that has held since December 2025, the discount has narrowed rather than widened, with the ARM share of applications holding at 7.9% in the latest MBA survey week. The relationship between the 10-year Treasury and mortgage rates explains why a Fed cut does not automatically widen the ARM advantage.
How an ARM Reset Actually Works — And What Caps Really Protect
Every 5/1 ARM contains four numbers that determine your post-reset payment: the index, the margin, the periodic caps, and the lifetime cap. Miss any one and the payment projection is meaningless.
Most conforming ARMs originated since the LIBOR transition reference the 30-day Average Secured Overnight Financing Rate published by the Federal Reserve Bank of New York, which sat near 3.64% in mid-August 2026. Your lender adds a fixed margin — commonly 2.75 percentage points on agency-eligible loans, though this is written into your note and varies by lender. Index plus margin gives the fully indexed rate. At August 2026 levels, 3.64% plus a 2.75% margin produces a fully indexed rate of roughly 6.39% — already above the 5.99% teaser.
Caps then constrain how fast you get there. A 2/2/5 structure means the first adjustment cannot exceed 2 percentage points, each subsequent annual adjustment cannot exceed 2 points, and the rate can never exceed the start rate by more than 5 points. Starting at 5.99%, that puts the worst-case first reset at 7.99% and the absolute lifetime ceiling at 10.99%.
Here is the scenario that catches borrowers. A buyer closes a $400,000 5/1 ARM at 5.99% in August 2026 with a $2,396 payment. Sixty payments later the balance is $372,171 and the rate resets to the capped 7.99% over a remaining 300-month term. The new payment is $2,870 — a $474 jump, and $297 more than the fixed-rate payment they declined. That is not a tail risk. That is the contractual maximum happening on schedule.
Caps limit the rate, not the payment shock. Because the reset amortizes a large balance over a shortened remaining term, a 2-point rate increase produces a payment increase closer to 20%. Anyone weighing this should first model the specific ARM versus fixed break-even analysis using their own margin and cap disclosure, not a generic calculator.
5/1 ARM vs 30-Year Fixed: Which Is Better for a Five-to-Seven-Year Holding Period?
Run the numbers on a $400,000 loan at August 2026 rates and the ARM’s five-year advantage is larger than the payment difference implies. The fixed payment is $2,573 against the ARM’s $2,396 — a $178 monthly gap worth $10,652 across 60 months. But the lower rate also routes more of each payment to principal. After five years the fixed-rate borrower owes $375,252 while the ARM borrower owes $372,171, a $2,996 equity edge. Combined advantage: $13,648.
Original amortization modeling by Real Cost Report using contract rates from Freddie Mac PMMS (August 13, 2026) and MBA Weekly Applications Survey (week ending August 7, 2026). Top row loan amount reflects the 2026 baseline conforming loan limit of $832,750 per the Federal Housing Finance Agency. Figures exclude points, taxes, and insurance.
Now extend the timeline. If the ARM resets to the capped 7.99% in month 61 and holds there, the borrower pays $297 more per month than the fixed borrower. Dividing the $13,648 accumulated advantage by $297 gives 46 months. The ARM stays ahead until roughly month 106 — just under year nine.
Under the lifetime cap of 10.99%, the picture collapses fast. The payment becomes $3,645, a $1,072 monthly disadvantage that burns through the entire five-year advantage in 12.7 months. That outcome requires short-rate increases well beyond the additional tightening markets currently price for 2026, but it is contractually permitted.
Verdict
For a documented holding period of five to seven years, the 5/1 ARM wins on cost — $13,648 on a $400,000 loan at five years, and still ahead at seven even after a maximum first reset. Beyond nine years the fixed wins under the capped-reset scenario. The decision is therefore not about rate forecasting; it is about whether your exit date is a plan or a hope. If you cannot name the year you will sell or refinance, the 0.68-point discount is not compensation for the risk.
What Most Borrowers Get Wrong About ARM Cost Comparisons
Four errors show up repeatedly, and each one distorts the comparison in the ARM’s favor.
Mistake 1: Comparing monthly payments only
The consequence is understating the ARM’s advantage by 28% — $10,652 in payment savings versus $13,648 in true five-year benefit on a $400,000 loan. Ironically this error favors the fixed. The correct action is to compare remaining principal balance at your expected exit date, not just cumulative payments, since the balance is what you settle at sale.
Mistake 2: Ignoring points paid at origination
The MBA recorded 0.83 points on 5/1 ARMs for the week ending August 7, 2026, which is $3,320 on a $400,000 loan. That consumes 31% of the first five years’ payment savings. Correct action: demand a zero-point quote on both products before comparing, or convert points to an equivalent rate. This is the same discipline required when comparing lenders by APR and total borrowing cost.
Mistake 3: Assuming you can simply refinance before the reset
Refinancing requires qualifying credit, sufficient equity, and a rate environment that makes it worthwhile — none guaranteed in month 60. Borrowers who planned to refinance out of 2021-vintage ARMs found the exit closed. Correct action: treat the capped reset payment of $2,870 as the payment you are actually underwriting yourself for, and only proceed if it fits your budget today.
Mistake 4: Treating the ARM decision as independent of loan size
Freddie Mac notes that ARMs remain most popular for nonconforming loan sizes, and the dollar math explains why: the same 0.68-point spread produces $133 per month at $300,000 and $370 at $832,750. Below roughly $300,000 the absolute savings rarely justify the reset exposure. Correct action: if your balance exceeds the conforming limit, evaluate the ARM alongside jumbo versus conforming rate differences, where ARM pricing is often most aggressive.
Who Should Take the ARM — and Who Should Not
Conditional logic, not preference, should decide this.
Take the 5/1 ARM if you have a contractually or professionally determined exit inside seven years — a military relocation cycle, a residency or fellowship term, a vesting schedule you intend to liquidate against — and you can service the $2,870 capped-reset payment on current income without relying on future raises. A second qualifying condition: your loan balance exceeds $400,000, where the monthly gap of $178 or more compounds into a meaningful sum.
Take the 30-year fixed if any of the following apply. You are buying a home you intend to hold through retirement, where a payment that cannot rise is worth more than $178 a month. Your debt-to-income ratio leaves no room for a 20% payment increase. Your income is variable or commission-based. Or you are financing under $300,000, where the $133 monthly gap does not compensate for the modeling burden and reset exposure.
Retirees and pre-retirees face a specific asymmetry here. A fixed payment is an inflation hedge against a largely fixed income; an ARM inverts that protection at precisely the life stage when income flexibility disappears. The 0.68-point discount is not priced for that risk.
One structural alternative deserves consideration before defaulting to either. A 15-year fixed at 5.96% now prices three basis points below the ARM’s 5.99% with zero reset risk — the trade is a substantially higher payment against a shorter term, with no rate premium at all. For borrowers who can absorb the payment, the 15-year versus 30-year total interest comparison now dominates the ARM outright rather than merely on a risk-adjusted basis. Borrowers should also confirm they are not solving a pricing problem that better execution would solve: credit score impact on mortgage rates by tier and online lender versus bank and credit union pricing can move a fixed-rate quote by more than the entire ARM discount. Reviewing controllable factors for securing the lowest mortgage rate is the cheaper first move.
Frequently Asked Questions
How much lower is a 5/1 ARM than a 30-year fixed in 2026?
The 5/1 ARM contract rate averaged 5.99% in the MBA Weekly Applications Survey for the week ending August 7, 2026, against 6.67% for the 30-year fixed in Freddie Mac’s PMMS for August 13, 2026 — a spread of 0.68 percentage points. On a $400,000 loan that translates to $178 per month. Note the ARM figure included 0.83 points, roughly $3,320 at that loan size.
What is the maximum my ARM payment can increase at the first reset?
Under a common 2/2/5 cap structure, the rate cannot rise more than 2 percentage points at the first adjustment. Starting at 5.99%, that caps the reset at 7.99%. On a $400,000 loan that moves the payment from $2,396 to $2,870 — a 20% increase, larger than the 2-point rate change suggests, because the balance re-amortizes over 300 remaining months. Your specific caps appear in your note and adjustable-rate disclosure.
Why are so few borrowers choosing ARMs right now?
ARMs represented 7.9% of applications in the MBA survey week ending August 7, 2026, unchanged from the prior week and up from 7.1% in mid-July. That is still a small share by historical standards, and the reason is the size of the discount: at 0.68 percentage points, borrowers are being asked to accept reset risk for progressively less compensation than a wider spread would offer.
Which index do most new ARMs use?
Most conforming ARMs originated after the LIBOR transition reference the 30-day Average Secured Overnight Financing Rate published by the Federal Reserve Bank of New York, which stood near 3.64% in mid-August 2026. Your fully indexed rate equals that index plus your contractual margin. Confirm both the index name and margin in your loan documents, as margins vary by lender and program.
How We Researched This Article
Rate inputs came from two weekly primary surveys. The 30-year and 15-year fixed contract rates were taken from Freddie Mac’s Primary Mortgage Market Survey for the week ending August 13, 2026. PMMS results derive from mortgage rates collected across loan applications submitted to Freddie Mac through Loan Product Advisor and cover conventional, conforming, fully amortizing home purchase loans for borrowers making a 20% down payment with excellent credit. Freddie Mac no longer publishes average fees and points, and it discontinued its 5/1 hybrid ARM series — a material limitation for this comparison.
Because of that discontinuation, the 5/1 ARM contract rate, points, and application share were sourced from the Mortgage Bankers Association Weekly Applications Survey for the week ending August 7, 2026, covering 80% LTV loans with points inclusive of the origination fee. Readers should note the two surveys measure different loan populations one week apart; the true spread available to any individual borrower will differ. Index data came from the 30-day Average SOFR series published by the Federal Reserve Bank of New York and mirrored on FRED. The 2026 baseline conforming loan limit of $832,750 was verified against the Federal Housing Finance Agency announcement of November 25, 2025.
All payment, balance, and break-even figures are modeled, not measured. We used standard fixed-payment amortization on the stated contract rates with no points financed, no mortgage insurance, and no taxes or insurance escrow. The 2/2/5 cap structure and 2.75-point margin used in the reset scenario are illustrative conventions common on agency-eligible ARMs, not universal terms — actual margins and caps are lender-specific and appear in the borrower’s note. We did not model negative amortization, interest-only structures, or lender credits. Break-even months were computed by dividing the accumulated five-year advantage by the post-reset monthly payment differential, which assumes the reset rate holds constant thereafter; in practice rates adjust annually. Research was last conducted August 2026. All figures were verified against named primary sources before publication.