The quick read: It depends on whether your exit can survive September 2026 long-term rates, not on which coupon starts lower. The floating base is cheaper: SOFR printed 3.87% on September 23, 2026 per FRED, while the U.S. Treasury's par yield curve put the 5-year at 5.03% on September 24, a 116-basis-point gap that ran 42 basis points the other way a year earlier. But a floating bridge loan's rate risk sits at the refinance: in the illustrative exit test below, the same property needs 11.3% more net operating income to refinance than at year-earlier rates. Take floating only if the business plan clears that exit with room to spare, and count the rate cap in the floating column.
This page makes no rate forecast: every scenario is stated arithmetic, and any spread no citable page states is labelled HYPOTHETICAL. For when a bridge loan beats permanent debt in general, see bridge loan vs. permanent financing; this article is about rate structure on the September 23–24, 2026 prints.
What are the floating and fixed base rates as of September 24, 2026?
As of September 24, 2026, the floating base was well below the fixed base: SOFR printed 3.87% on September 23 per FRED, while the Treasury's par yield curve put the 5-year at 5.03% and the 10-year at 5.18% on September 24. A year earlier, SOFR sat above the 5-year.
A floating coupon is an index plus a spread. Freddie Mac's multifamily floating-rate term sheet (4/26) names its pricing index as "30-day Average SOFR," which FRED shows at 3.69764% on September 24, 2026, below spot SOFR because the average still includes days before the September 17 step-up. Freddie Mac's fixed-rate term sheet calls "the Treasury index" the "most volatile part of the coupon." This article uses the 5-year for bridge-length comparisons and the 10-year for a takeout.
Table 1: The rate prints behind this article (each one dated)
| Series and source | Print used here | Year-earlier print | Change |
|---|---|---|---|
| SOFR (FRED) | 3.87% (Sept 23, 2026) | 4.12% (Sept 23, 2025) | −25 bp |
| 30-day Average SOFR (FRED) | 3.69764% (Sept 24, 2026) | 4.348% (Sept 24, 2025) | −65 bp |
| 5-year Treasury (par yield curve) | 5.03% (Sept 24, 2026) | 3.70% (Sept 24, 2025) | +133 bp |
| 10-year Treasury (par yield curve) | 5.18% (Sept 24, 2026) | 4.16% (Sept 24, 2025) | +102 bp |
| 5-year Treasury minus SOFR | +116 bp | −42 bp | curve shape flipped |
Floating base: SOFR 3.87% as of September 23, 2026 (FRED).
Fixed base: 5-year 5.03% and 10-year 5.18% as of September 24, 2026 (U.S. Treasury par yield curve).
Recheck these benchmarks on the rates page before you sign.
What exactly did the Fed say when it raised rates on September 16, 2026?
The Federal Reserve's September 16, 2026 statement says the Committee "decided to raise the target range for the federal funds rate by 1/4 percentage point to 3-3/4 to 4 percent, in support of the Federal Reserve's dual mandate," approved by a 12–0 vote. It adds: "Inflation remains elevated."
The next sentence reads: "Today's policy action will support a timelier return to the Committee's 2 percent goal." The statement does not announce further increases, so do not underwrite one from it; run scenarios instead. FRED shows SOFR at 3.62% on September 16 and 3.85% on September 17, and the bank prime rate at 6.75% and then 7.00% on the same two days. The long end moved on its own path: the Treasury par yield curve shows the 10-year at 5.01% on September 16, 4.94% on September 17 and 5.18% on September 24. How the hike passes through a floating coupon's index, floor and reset is covered in what the September 2026 Fed decision means for bridge loan pricing.
Is floating-rate carry actually cheaper than fixed on the September 23–24, 2026 prints?
Only the base is reliably cheaper on the September 23–24, 2026 prints: SOFR sat 116 basis points below the 5-year Treasury after sitting 42 basis points above it a year earlier. Whether your coupon is cheaper depends on the spread gap between your two quotes, and no public 2026 bridge-spread data we could cite settles that.
Loan: $5 million, interest-only (illustrative).
Floating spread: 4.00% over SOFR, HYPOTHETICAL; substitute the spread in your own quote.
Formula: monthly interest = loan × (SOFR + spread) ÷ 12.
At the September 23, 2026 print, $5 million × (3.87% + 4.00%) ÷ 12 = $32,792 a month. At the year-earlier print it was $5 million × (4.12% + 4.00%) ÷ 12 = $33,833, so the index alone makes the same loan $1,042 a month cheaper to carry.
Against fixed, use a break-even rather than the base gap: break-even average SOFR = fixed coupon − floating spread. Take an illustrative fixed coupon of 6.80%: the 10-year's 5.18% plus a HYPOTHETICAL 1.62% fixed-rate spread; substitute the spread in your own fixed quote. Against the HYPOTHETICAL 4.00% floating spread, the break-even is 2.80%, so SOFR would have to average 107 basis points below its September 23 print for the floating coupon to match. A lower base does not guarantee a lower coupon; the spread gap decides it.
How much does a SOFR move change the payment on a floating-rate bridge loan?
On a $5 million interest-only bridge loan, every 100-basis-point move in SOFR changes the interest bill by $50,000 a year, or about $4,167 a month, in either direction, and a rate cap, if the loan has one, is what limits the upside. The table shows the arithmetic for moves of 50 to 200 basis points.
Table 2: Illustrative monthly interest on a $5 million interest-only loan at SOFR + 4.00% (HYPOTHETICAL spread). Scenarios, not forecasts.
| SOFR scenario | SOFR | Coupon | Monthly interest | Change vs. Sept 23 print |
|---|---|---|---|---|
| Print −200 bp | 1.87% | 5.87% | $24,458 | −$8,333 |
| Print −100 bp | 2.87% | 6.87% | $28,625 | −$4,167 |
| Print −50 bp | 3.37% | 7.37% | $30,708 | −$2,083 |
| Sept 23, 2026 print | 3.87% | 7.87% | $32,792 | no change |
| Print +50 bp | 4.37% | 8.37% | $34,875 | +$2,083 |
| Print +100 bp | 4.87% | 8.87% | $36,958 | +$4,167 |
| Print +200 bp | 5.87% | 9.87% | $41,125 | +$8,333 |
An index floor stops the down rows at the floor. A loan priced off an average, as Freddie Mac's floater is on 30-day Average SOFR, trails a move by weeks: on September 24, 2026 the average was still 3.69764%, against a 3.87% spot print for September 23. To scale the table, multiply the change column by your balance divided by $5 million.
How does a rate cap change the floating-versus-fixed decision?
A rate cap turns an open-ended floating payment into a bounded one: the borrower pays an upfront premium to a third-party provider, and the provider covers index increases above an agreed strike for an agreed term. It changes the decision because its cost belongs in the floating column before you compare coupons.
Caps are not universal, even inside one agency program. Freddie Mac's floating-rate term sheet refers to "standard, capped or uncapped floating-rate loans" and says the "Borrower may obtain its own cap coverage from a third-party provider." On cost, no public page we could cite states a 2026 cap price, strike or term for bridge loans, so this article prints none.
Worst-case coupon while the cap lasts: strike + spread.
Monthly cost of the cap: premium ÷ months of coverage, added to the floating payment before you compare it with fixed.
Whichever Table 2 row matches your strike is the most the payment can reach while the cap lasts. Ask the lender:
- Is a cap required, and at what strike and for what term?
- Must it come from an approved counterparty, and who receives the payouts?
- If you extend, must you buy a new cap, and do you escrow for it monthly?
Where does the real rate risk sit on a floating-rate bridge loan?
The real rate risk on a floating-rate bridge loan sits at the refinance, because a fixed-rate takeout is sized on long-term rates, and on September 24, 2026 the 10-year Treasury, at 5.18%, was 102 basis points above its year-earlier 4.16%. A lower SOFR helps the carry; it does nothing for the exit.
Bridge payoff: $12 million (hypothetical, above the $10 million minimum on Freddie Mac's fixed-rate term sheet).
Takeout test: 1.25x minimum amortizing debt coverage, 80% maximum loan-to-value on a 7-year term, 30-year maximum amortization, per Freddie Mac's fixed-rate term sheet (4/26).
Takeout rate: 10-year Treasury + 1.62%, a HYPOTHETICAL fixed-rate spread held constant in every row; substitute the spread in your own quote.
Formula: NOI needed = 1.25 × annual debt service on the payoff; maximum loan = (NOI ÷ 1.25) ÷ annual loan constant.
Table 3: Illustrative exit test on a $12 million bridge payoff (10-year + 1.62% HYPOTHETICAL spread, 1.25x, 30-year amortization). Scenarios, not forecasts.
| 10-year Treasury scenario | Exit rate | NOI needed to refinance $12 million | Maximum loan at a HYPOTHETICAL $1.1 million NOI | Versus the $12 million payoff |
|---|---|---|---|---|
| Year-earlier print, 4.16% (Sept 24, 2025) | 5.78% | $1,053,864 | $12,525,335 | +$525,335 surplus |
| Sept 24, 2026 print −100 bp, 4.18% | 5.80% | $1,056,155 | $12,498,160 | +$498,160 surplus |
| Sept 24, 2026 print, 5.18% | 6.80% | $1,173,465 | $11,248,734 | −$751,266 shortfall |
| Sept 24, 2026 print +100 bp, 6.18% | 7.80% | $1,295,767 | $10,187,017 | −$1,812,983 shortfall |
At the September 24, 2026 print, the property needs $119,601 more NOI, 11.3% more, than at the year-earlier print to clear the same exit. With a $1.1 million NOI it comes up $751,266 short, and the sponsor writes that check at the refinance. Compare the carry saving: on a $12 million floating balance, SOFR's 25-basis-point drop from its year-earlier print saves $2,500 a month, or $60,000 over a HYPOTHETICAL 24-month hold. If you already hold a bridge loan facing this test, see why you can't refinance out of a bridge loan in 2026.
If you lock a fixed rate now, what are you giving up?
Locking a fixed rate off the September 24, 2026 curve fixes your base near a 5.03% 5-year or 5.18% 10-year Treasury, which ends payment risk but trades it for prepayment cost: if long rates fall or you sell early, getting out of the loan can be expensive. It also sizes the loan on in-place income.
Freddie Mac's fixed-rate term sheet sets prepayment as "Yield maintenance until securitized followed by 2-year lock-out; defeasance thereafter." Its floating-rate sheet pitches that product as "ideal for borrowers who want to take advantage of lower, short-term rates with prepayment flexibility," with four prepayment options, one of which is a year-one lockout followed by 1% premiums. Those are agency products, not bridge loans; ask your lender for its schedule and read up on yield maintenance and defeasance before you lock.
Sizing is the other trade. On Table 3's assumptions, $1.1 million of NOI supports at most $11,248,734 of fixed-rate debt at 6.80%, the same number the bridge exit faces if the curve is unchanged at maturity. A floating bridge is mainly a bet that the business plan raises NOI before the refinance, with rate exposure layered on top; a fixed loan sizes you on the income you already have.
Which structure fits your deal on September 2026 rates?
A floating-rate bridge fits a deal whose value comes from a business plan that raises net operating income before the refinance, with enough cushion to clear a fixed-rate exit at September 2026 long rates. A fixed-rate loan fits a property whose income already supports the proceeds you need, held long enough to live with prepayment terms.
Table 4: Floating-rate bridge versus fixed-rate loan on the September 23–24, 2026 prints
| Test | Floating-rate bridge loan | Fixed-rate loan |
|---|---|---|
| Base rate (dated) | SOFR 3.87% (Sept 23, 2026) | 5-year 5.03% or 10-year 5.18% Treasury (Sept 24, 2026) |
| What moves the payment during the hold | Every index reset, up to any cap strike | Nothing after the rate lock |
| Main rate exposure | The refinance, sized on long rates at maturity | Prepayment cost if you exit early |
| Rate cap | Required or optional by program; no public 2026 price to cite | Not applicable |
| Fits when | Projected NOI clears the exit test with cushion | In-place income supports the loan and the hold matches the prepayment schedule |
Carry test: monthly interest at SOFR + your quoted spread, plus the cap premium divided by months of coverage.
Exit test: NOI needed = 1.25 × annual debt service on the payoff at the 10-year plus your expected fixed spread, run at the print and 100 basis points either side.
Cushion: projected stabilized NOI minus NOI needed; a negative number is cash you bring to the refinance.
For all-in bridge costs, see how much a short-term bridge loan costs, and for terms worth negotiating, see what good bridge loan terms look like.
How do you get lenders to quote floating and fixed on the same deal?
You get floating and fixed quotes on the same deal by submitting one complete file, with in-place and projected net operating income, the business plan, the hold period and the exit you expect, and asking every lender for its index, spread, cap terms and prepayment terms side by side.
YieldStack is a commercial mortgage brokerage, not a lender. Your file is matched against 20,000+ loan programs, so floating and fixed options are compared on the same numbers. Every credit decision is made by the lender; YieldStack arranges the introductions and negotiates on the borrower's side, and no loan, rate, or closing is guaranteed. It costs Zero upfront to submit a deal and review offers; YieldStack's broker fee is 0.50–1.00% of the loan amount and is paid only at closing. Submit your deal to see floating and fixed quotes side by side.
The bottom line
On the September 23–24, 2026 prints, floating debt starts from the cheaper base: SOFR at 3.87% sat 116 basis points below the 5-year Treasury at 5.03%. The Fed's September 16 statement announced no further moves, so underwrite scenarios, not a path. A bridge loan's rate risk sits at the refinance, where the 10-year at 5.18% was 102 basis points above its year-earlier print. Choose floating when the plan clears that exit with cushion and the cap is in the carry math; choose fixed when in-place income already supports the loan.