Rate snapshot
| Fed funds target | 3.75% to 4.00% | raised 25 bp on September 16, 2026 |
| SOFR, daily | 3.88% | October 2, 2026 |
| 30-day average SOFR | 3.76116% | October 2, 2026, still 11.9 bp below spot |
| 90-day average SOFR | 3.68732% | October 2, 2026 |
| 5-year Treasury CMT | 5.01% | October 1, 2026, the refinance exit |
The September hike has not finished arriving. SOFR printed 3.88% on October 2, but the 30-day average that most bridge loans actually index to printed 3.76116%, because the average still carries two weeks of pre-hike days. If you are testing an extension option this month, you are testing it on a rate that is 11.9 basis points stale and rising on its own.
That matters more than it sounds. On a $14 million interest-only bridge loan, the convergence is worth only $16,638 of additional annual interest. The same convergence raises the principal paydown required to clear a 1.20x DSCR extension test from $526,667 to $684,260. A 12 basis point drift in an index costs $157,592 in cash, because the paydown is solving a covenant, not an interest bill.
This post works a first extension option on a $14 million small-bay industrial bridge loan at its initial maturity. The finding worth taking away: the sponsor’s own debt service coverage reads 1.303x and passes, the coverage the loan agreement actually tests reads 1.155x and fails, and the quarter-point extension fee in the term sheet is 4% of what exercising the option really costs.
What an extension option actually is
Bridge loans are written with a short initial term and one or more extension options because the business plan needs time it cannot promise. An extension option is not a right to more time. It is a conditional right, and the conditions are performance tests measured on a specific date against a specific definition. Fail one and the option disappears, which means the loan matures and you are refinancing or selling into whatever market exists that month.
Four conditions show up in nearly every bridge loan of this size. A minimum debt service coverage ratio. A minimum debt yield. A fee, usually 25 to 50 basis points of the balance. And a requirement to have an interest rate cap in place for the extension period, which means buying a new one because the original expires with the initial term. There is always a no-default condition and often a minimum occupancy or leasing threshold as well.
Sponsors read these as a list of hurdles roughly equal in difficulty. They are not. One of them will bind, the others will be decorative, and which one binds is determined by where rates are, not by how the property is performing.
The worked example
A sponsor closed a $14,000,000 bridge loan on a small-bay industrial park in October 2024. Twenty-four month initial term, interest only, 30-day average SOFR plus 350 basis points, with two 12-month extension options. The plan was to roll below-market tenants to current rents as leases expired. The plan worked, just not all the way, and the initial maturity is now.
| Loan and test terms | Value |
|---|---|
| Original loan amount | $14,000,000 |
| Index, 30-day average SOFR | 3.76116% |
| Spread | 350 bp |
| All-in rate | 7.26116% |
| Test rate floor in the covenant | 7.00%, not binding |
| Minimum DSCR to extend | 1.20x |
| DSCR tested on | hypothetical 30-year amortizing payment |
| Minimum debt yield to extend | 9.00% |
| Extension fee | 0.25% of balance |
| In-place NOI, trailing twelve annualized | $1,325,000 |
Now run the two ratios. This is where most sponsors discover they were measuring the wrong thing.
| Test | Result | Required | Outcome |
|---|---|---|---|
| Debt yield, NOI divided by loan | 9.46% | 9.00% | Pass |
| DSCR on actual interest-only debt service of $1,016,562 | 1.303x | 1.20x | Pass, but not the test |
| DSCR on the covenant’s 30-year amortizing debt service of $1,147,328 | 1.155x | 1.20x | Fail |
| Definitional gap between the two coverage figures | 0.149x | n/a | The whole problem |
The property is covering its actual debt service at 1.30x. The sponsor’s asset management report says so and it is correct. The loan agreement does not test actual debt service. It tests a constructed payment that includes principal the borrower is not paying, and on that basis coverage is 1.155x and the extension option is not available.
Note also that debt yield passed with room to spare, and would have passed even at a balance of $14,722,222. Debt yield ignores the interest rate entirely, which is exactly why it is the wrong test to worry about in a 7% world and the right one to worry about when rates are low.
What it costs to cure
The lender’s remedy is almost always a paydown. Solve each test for the largest balance that clears it and the asymmetry is stark.
| Test | Maximum passing balance | Paydown required |
|---|---|---|
| Debt yield at 9.00% | $14,722,222 | $0 |
| DSCR at 1.20x, amortizing basis | $13,473,333 | $526,667 |
| Binding constraint | DSCR | $526,667, or 3.76% of the loan |
Then add the rest of the exercise cost, which is the number that should have been in the sponsor’s reserve model two years ago and usually was not.
| Cash required to exercise the first 12-month extension | Amount |
|---|---|
| Principal paydown to cure the DSCR test | $526,667 |
| Extension fee, 0.25% of the post-paydown balance of $13,473,333 | $33,683 |
| Replacement interest rate cap, 12 months (assumed) | $185,000 |
| Lender legal, appraisal update, title endorsement (assumed) | $45,000 |
| Total | $790,351 |
| As a percentage of the original loan | 5.65% |
The extension fee was 25 basis points. Exercising the extension costs 565 basis points. The fee is 4.3% of the true cost, and the two line items nobody budgets, the paydown and the replacement cap, are 90% of it.
Simplifications: the DSCR covenant is modeled as a hypothetical 30-year amortizing payment at the greater of the in-place all-in rate or a 7.00% floor, which is one common formulation among several; your document may test actual debt service, use a different amortization period, apply a higher assumed constant, or underwrite NOI with lender-imposed management fees, reserves and vacancy that differ from in-place. Loan terms, the test thresholds, the fee, the rate cap premium of $185,000 and the $45,000 of third-party costs are illustrative term sheet and modeling assumptions, not quotes and not QuadBlock terms. NOI is held at trailing twelve annualized with no credit for signed but unoccupied leases, which many lenders would partially allow. Rate cap pricing depends on strike, term and the forward curve and is not a figure sourced from this run.
Where the cliff is steepest
Two variables drive the outcome: the index and NOI. Holding the loan at $14,000,000, here is the tested DSCR and the paydown it implies.
| Index | All-in | NOI down 5% | NOI flat | NOI up 5% |
|---|---|---|---|---|
| 30-day avg SOFR, 3.761% | 7.261% | 1.097x, $1,200k | 1.155x, $527k | 1.213x, $0 |
| Spot SOFR, 3.880% | 7.380% | 1.084x, $1,350k | 1.141x, $684k | 1.198x, $18k |
| 4.250% | 7.750% | 1.046x, $1,798k | 1.101x, $1,156k | 1.156x, $514k |
| 4.500% | 8.000% | 1.021x, $2,087k | 1.075x, $1,460k | 1.129x, $833k |
| 4.750% | 8.250% | 0.997x, $2,365k | 1.050x, $1,752k | 1.102x, $1,140k |
Read the top row across. A 5% swing in NOI, which is one anchor tenant going dark or one year of insurance renewals, moves the required paydown from zero to $1,200,000. The covenant converts small operating variance into large capital calls, because DSCR is a ratio and the cure is a principal balance.
Put the same point in NOI terms. At the current all-in rate the property needs $1,376,794 of NOI to clear 1.20x with no paydown. It has $1,325,000. The shortfall is $51,794, under 4%, and it costs $526,667 to fix. That is roughly ten dollars of paydown for every dollar of missing NOI, which is the single most useful rule of thumb in this post.
| All-in rate | NOI needed for 1.20x | Shortfall vs. $1,325,000 | Required increase |
|---|---|---|---|
| 7.26116%, today’s average | $1,376,794 | $51,794 | 3.9% |
| 7.38000%, spot | $1,393,088 | $68,088 | 5.1% |
| 7.75000% | $1,444,287 | $119,287 | 9.0% |
| 8.00000% | $1,479,269 | $154,269 | 11.6% |
The test date is a decision, not a fact
Most extension provisions set the measurement period as the trailing three, six or twelve months ending on a date tied to the notice deadline. Notice is typically due 30 to 60 days before maturity. That gives the borrower a narrow but real amount of control, and in an index that is drifting upward it is worth money.
Here the index is doing the drifting. The 30-day average SOFR of 3.76116% contains pre-hike observations that will roll out of the window over the coming weeks. If the average converges to the 3.88% spot rate with no further Fed action, the tested DSCR falls from 1.155x to 1.141x and the paydown grows by $157,592. Nothing about the property changed. The borrower who delivered notice and locked the test in September is $157,592 better off than the one who waits for a better rent roll.
The inverse applies when the curve is falling: waiting is cheap and the trailing average works for you. The point is that the direction of the index tells you whether to test early or late, and almost nobody checks before choosing a notice date.
Checklist before your extension notice deadline
- Find the extension section in the loan agreement, not the term sheet, and write down the notice deadline and whether notice is revocable. Many are not.
- Copy out the exact definition of DSCR. Confirm whether debt service means actual payments or a hypothetical constant, and at what rate and amortization.
- Copy out the exact definition of NOI. Lender-underwritten NOI usually imposes a management fee, replacement reserves and a vacancy factor whether or not you incur them.
- Identify the measurement period and its end date, then compute the ratio on that basis rather than on your own.
- Compute both ratios and identify which binds. If the rate is high, expect DSCR; if the rate is low, expect debt yield.
- Solve for the paydown. Do not wait for the lender’s number, because by then you are negotiating against a deadline.
- Get a rate cap indication early. Premiums move with the forward curve and the strike your lender requires may not be the one you want.
- Total the full exercise cost and compare it against refinancing today and against selling. Extending is one of three options and it is not automatically the cheapest.
- If you expect to fail, open the conversation 90 days out. A lender asked early will consider a lower strike, a modified test, a partial cure with a reserve, or a short extension to a new test date. A lender asked at maturity has far less room.
- Check whether a failed extension triggers cash management or a springing lockbox. Losing control of cash flow is often worse than the paydown.
Red flags in extension language
Four provisions do most of the damage. A DSCR tested on a hypothetical amortizing payment rather than actual debt service, which is the 0.149x gap in this example. A test rate defined as the greater of the in-place rate and a stated floor, which means a rate cut does not help you. No partial cure mechanic, so a miss of any size voids the option outright rather than letting you pay down to compliance. And a requirement that the replacement cap carry a strike at or below the original, which can be expensive or unavailable depending on where the forward curve sits.
How QuadBlock can help
We write bridge, transitional and permanent loans from $250,000 to $30,000,000 across multifamily, industrial, retail, self-storage, specialty use and ground-up SFR and townhome communities, in every state except Oregon and Nevada. Across 30-plus closings we have refinanced plenty of loans whose extension tests were going to fail, and we would rather look at it 90 days out than 10. Send the loan agreement, the current rent roll and trailing twelve with your deal at quadblockcapital.com and we will tell you which test binds before your notice deadline does.
This article is for general educational purposes and is not tax, legal, or investment advice. Extension rights, covenant definitions, cure mechanics and cash management triggers depend entirely on the language of your own loan documents, and the formulations described here are common examples rather than standards. The worked example uses illustrative loan terms and assumed costs for the rate cap and third-party expenses, is not an offer, quote or commitment, and does not reflect QuadBlock Capital rate, fee or leverage terms. Interest rate caps and other derivatives carry their own risks and costs. Consult your own counsel, tax advisor and lender before acting on any extension, refinance or hedging decision.
Sources
- Federal Reserve Bank of New York, SOFR daily rate, October 2, 2026
- Federal Reserve Bank of New York, SOFR Averages and Index, 30-day and 90-day averages as of October 2, 2026
- Federal Reserve Board, H.15 Selected Interest Rates, release of October 2, 2026
- FOMC statement, September 16, 2026, raising the target range to 3-3/4 to 4 percent
- Federal Reserve Board, Open Market Operations, federal funds target range history
- FRED, 5-Year Treasury Constant Maturity Rate (DGS5)
- Bureau of Labor Statistics, Employment Situation Summary for September 2026, released October 2, 2026
