ARMs carry benefits over fixed-rate mortgages beyond the lower monthly payment:

• ARM borrowers automatically benefit when interest rates fall after the fixed period. During recessions and other economic slowdowns, this feature acts as a built-in stabilizer, while fixed-rate borrowers often pay a higher initial rate and fail to take advantage of refinancing options.
• ARMs also alleviate interest rate lock-in by repricing to the current interest rates post-reset if rates increase, as borrowers leave considerably less money on the table if they move or refinance in the first seven years and are not locked into a low rate after the initial period.  

But ARMs carry a cost as well. If interest rates are higher after the end of the fixed period, borrowers make higher monthly payments. Borrowers likely have increased earnings, home values, and home equity when this happens, but higher future rates will inevitably lead to borrowers paying more on their mortgage, whether because of the reset rate or because they move and take out a new, higher-rate mortgage.

Why do so few borrowers opt for ARMs?

Today’s ARMs are not the ARMs of the period before the financial crisis. Those earlier products often had initial rates teased for two years or less that would then rise substantially when the initial period expired. The earlier products also frequently contained dangerous features, such as negative amortization, and were subject to sloppy underwriting standards. Many borrowers and real estate market participants may continue to associate today’s ARMs with those prior products.

But today’s ARMs adjust to the SOFR plus a margin after the initial period, with caps to ensure the payment change is manageable. The cap structure is characterized by three numbers: the cap at the first reset, at each subsequent reset, and over the loan’s lifetime. Most ARMs have a cap of 2 percent at the first reset, 1 percent at subsequent resets, and 5 percent over the life of the loan (“2/1/5,” as set by the GSEs). The cap structure has symmetric floors as well, meaning the rate can’t increase quickly or decrease quickly. If interest rates increase substantially, like they did postpandemic, the caps allow borrowers a year or more of monthly payments lower than the full ARM margin so the borrower can adjust to the new higher-rate environment.

ARM margins are generally 275 to 300 basis points over the SOFR, leaving the post-reset mortgage rate high. In January 2026, the SOFR sat around 3.7 percent, making the post-reset rate at least 6.45 percent (3.7 percent + 2.75 percent), higher than the fixed rate. As a result, many borrowers feel that lower interest rates for the initial period are not worth the uncertainty that they may face a much higher rate later. Even when borrowers do take out ARMs, they often refinance when the initial period ends.

The GSEs set an explicit limit of 300 basis points on ARM margins, and 275 to 300 basis points has been the industry standard for the past two decades. In comparison, the difference between fixed-rate mortgages and 10-year Treasury bonds (a spread that indicates the incremental cost a mortgage borrower must pay) is around 200 basis points, with the historical average around 188 basis points. ARMs shouldn’t have a higher spread than fixed-rate mortgages, as the investor’s interest rate risk is much lower with ARMs and prepayment risk is far less costly, which partially offsets the fact that ARMs are less liquid and have a limited cap risk.  

How federal housing agencies can promote affordable housing with ARMs

Current ARMs are almost balloon loans. Because of the high margin reset, most borrowers refinance if they can, meaning investors get little benefit from the higher rate, as few loans remain outstanding. Instead, we recommend that the GSEs and the Federal Housing Administration require the ARM margin to be 175 basis points, comparable with the historical fixed-rate spread. This would make ARMs more appealing to borrowers while retaining investor upside.

Improving borrower take-up (and hence market liquidity) should also lower rates during the initial period. Currently, the initial rate for an ARM with a seven-year fixed period is around 5.5 percent, while the five-year Treasury bonds are 3.75 percent. The market should theoretically set the initial ARM rate at 3.75 percent plus 90 basis points for fees and servicing and an added premium reflecting the remaining investor prepayment and interest rate risk—resulting in an overall rate decrease.

To homogenize ARMs and increase liquidity, we recommend the GSEs and the Federal Housing Administration coalesce around a set of standards for ARMs, potentially using current practices. These standards could include

• a seven-year initial period, followed by six-month resets;
• a 2/1/5 cap structure;
• the SOFR index; and
• a mandated 175 basis-point margin.

Further, the GSEs’ ARMs currently contain an option to convert into a fixed-rate mortgage at the initial reset date if the borrower is current on their existing mortgage, typically without the need for appraisal or underwriting. But borrowers are generally unaware of this option. To increase awareness, we suggest servicers include the conversion option and expected conversion rate in the note they are required to send to the borrower 30 days before their ARM rate reset.

Ultimately, ARMs can be an effective affordability solution if a few simple adjustments are made. Right now, borrowers are reluctant to take out ARMs because the post-reset margin is too high, meaning ARMs have limited liquidity, which keeps even the initial-period interest rates high. Lowering the margin and increasing standardization across ARMs would increase the popularity of and liquidity of ARMs, benefiting prospective buyers, especially first-time homeowners.