Rate snapshot
| Fed funds target | 3.75% to 4.00% | Last changed Sept. 17, 2026 (+25 bp) |
| SOFR | 3.90% | Oct. 6, 2026; 3.88% on Oct. 2 |
| 1-month Treasury | 4.06% | Oct. 6, 2026 close |
| 1-year Treasury | 4.46% | Oct. 6, 2026 close |
| 2-year Treasury | 4.79% | Oct. 6, 2026 close, down 5 bp from Oct. 5 |
| Implied short-rate path | +136 bp | Over 24 months, derived from the Oct. 6 curve |
The front end of the curve closed Tuesday with the two-year at 4.79%, five basis points lower than Monday, while overnight SOFR ticked up to 3.90%. Nothing is scheduled to land today, and the next FOMC meeting is three weeks out on Oct. 27 and 28. That gap is useful, because it is a good moment to look at a budget line that gets set once at closing and then quietly decides whether your bridge loan makes it to the exit.
Here is the part of the curve that matters for an interest reserve. The one-month bill sits at 4.06% and the one-year at 4.46%, which means the market expects the average short rate over the next twelve months to be 40 basis points above today’s one-month rate. Push it out to the two-year at 4.79% and the implied average for the second year is 5.12%. Run that as a path and the market is telling you the short rate ends up roughly 136 basis points higher in 24 months than it is now.
So the obvious question is how much reserve you need to absorb 136 basis points of rate increase. We modeled it on the fact pattern below, and the answer is that the rate path is almost irrelevant: 150 basis points of rate increase burned $2,579 of extra reserve, while a six-month lease-up delay burned $227,216. The reserve is not a rate hedge. It is a bet on your leasing schedule, leveraged about 88 to 1 in that direction.
The deal
Small-bay industrial in the Southeast, 86,400 square feet across 24 units, 42% leased at closing at $8.40 per square foot NNN. The business plan is demising walls, a roof, truck court repair, and electrical upgrades, then lease the balance at market. National industrial asking rents averaged $10.45 per square foot in the second quarter of 2026 with vacancy at 6.8%, and small-bay product trades at a premium to that average, so $11.75 per square foot is the stabilized assumption here.
| Item | Amount |
|---|---|
| Purchase price | $9,200,000 |
| Price per square foot | $106.48 |
| In-place NOI at closing | $228,000 |
| Going-in cap rate on purchase | 2.48% |
| Stabilized NOI (93% leased, $11.75 NNN) | $900,000 |
| Hard and soft capex budget | $2,150,000 |
| Loan rate | SOFR + 325 bp |
| Rate at closing | 7.15% |
| SOFR floor | 3.50% |
| Term | 24 months, interest only |
| Leverage | 70% of total cost |
A 2.48% going-in cap looks indefensible until you remember you are buying 50,100 square feet of vacancy. The whole deal is the spread between a $106 basis and a 7.54% stabilized yield on cost. Getting there requires 24 months of carrying a loan the property cannot service, and the reserve is what pays for that.
The reserve is circular in three directions
Most sponsors treat the reserve as an input. It is actually a solve, because it appears on both sides of its own equation three separate times.
First, the reserve is a line in the project budget. Second, the loan is sized as a percentage of that budget, so a larger reserve produces a larger loan. Third, the larger loan accrues more interest, which the reserve has to pay. And because the reserve is funded by draws that add to the outstanding balance, the reserve pays interest on its own prior draws.
On this deal that last loop is not a rounding error. Of the $245,000 reserve, $28,430 exists only to service the reserve’s own draws. That is 11.6% of the reserve, or roughly one dollar in nine, doing nothing but paying interest on itself.
Sources and uses at the correct reserve
| Uses | Amount | Funded at |
|---|---|---|
| Purchase price | $9,200,000 | Closing |
| Capex budget | $2,150,000 | Drawn, months 1 to 12 |
| Closing costs and third-party reports | $265,000 | Closing |
| Origination fee (1.00%) | $83,605 | Closing |
| Interest reserve | $245,000 | Drawn monthly as needed |
| Total project cost | $11,943,605 |
| Sources | Amount | Share |
|---|---|---|
| Bridge loan | $8,360,524 | 70.0% |
| Sponsor equity | $3,583,082 | 30.0% |
| Initial advance at closing | $5,965,524 | |
| Holdbacks (capex plus reserve) | $2,395,000 | |
| All-in basis per square foot | $138.24 | |
| Stabilized yield on cost | 7.54% |
Four ways to size it, and what each one actually does
We ran four sizing methods against the same $8,360,524 loan and the same market-implied rate path, then asked a single question of each: does the reserve make it to month 24?
| Method | Reserve | Basis | Outcome over 24 months |
|---|---|---|---|
| A | $600,000 | 12 months of gross interest at the closing rate | Survives, $366,910 never drawn |
| B | $1,195,000 | 24 months of gross interest at the closing rate | Survives, $977,988 never drawn |
| C | $210,000 | Net of projected NOI, SOFR held flat at 3.90% | Runs dry in month 11, short $33,643 |
| D | $245,000 | Net of NOI, market-implied path, self-funding | Survives, $1,750 left |
Method A is the rule of thumb you hear most often, and on a deal with any in-place income it overfunds by 145%. That is not harmless. The reserve sits in the cost basis, so the sponsor funds 30% of it, which means $355,000 of excess reserve ties up $106,500 of equity that earns nothing and inflates the denominator on every return metric in the model.
Method C is the more dangerous error, because it looks rigorous. It nets the reserve against projected NOI, which is correct, but holds SOFR flat at today’s print. Against the curve the market is actually quoting, a flat-rate assumption underfunds by $35,000, or 14.3%, and the reserve runs out in month 11 of a 24-month deal.
How the reserve actually burns
Method D is the sized reserve. Note where the draws cluster, because this is the shape that makes the reserve fragile.
| Period | Avg rate | Avg balance | Interest | NOI | Net draw | Reserve left |
|---|---|---|---|---|---|---|
| Months 1 to 6 | 7.30% | $6,663,079 | $242,521 | $114,000 | $128,521 | $116,479 |
| Months 7 to 12 | 7.65% | $7,862,710 | $300,369 | $198,000 | $102,369 | $14,111 |
| Months 13 to 18 | 8.01% | $8,357,897 | $334,544 | $342,000 | $12,361 | $1,750 |
| Months 19 to 24 | 8.36% | $8,358,774 | $349,488 | $446,000 | $0 | $1,750 |
94% of the reserve is spent in the first twelve months, and the final draw lands in month 15. That is not a sizing flaw, it is the correct answer: the reserve exists to bridge the window between closing and the month the property can service its own debt. On this deal the crossover is month 16, the first month NOI exceeds interest, and the reserve is designed to run out just before it.
The problem is what that shape does to your margin for error. By month 12 you have $14,111 of reserve left against a $300,000 semiannual interest bill, and at month 24 the leftover is $1,750 on a $245,000 reserve. A correctly sized reserve has, by construction, almost no cushion, so it has no capacity left at exactly the point in the deal where bad news tends to arrive.
The sensitivity that matters
Here is the grid. Reserve held fixed at $245,000. Columns are months of lease-up delay, rows are how much SOFR rises over the 24 months. Each cell shows either the unused reserve at month 24 or the month the reserve runs dry and the resulting cash shortfall.
| SOFR path | On plan | 3 months late | 6 months late | 9 months late |
|---|---|---|---|---|
| Flat at 3.90% | $36,974 left | Dry m11, short $56,436 | Dry m11, short $148,904 | Dry m11, short $241,348 |
| +50 bp | $25,137 left | Dry m11, short $75,173 | Dry m11, short $175,821 | Dry m11, short $277,807 |
| +100 bp | $11,876 left | Dry m11, short $95,102 | Dry m11, short $204,803 | Dry m11, short $317,240 |
| +150 bp | Dry m14, short $2,579 | Dry m11, short $117,577 | Dry m10, short $236,283 | Dry m10, short $358,359 |
Read it across and down and the asymmetry is hard to miss. Moving down the on-plan column, from a flat SOFR to 150 basis points of increase, costs $39,553 of reserve capacity across the whole range. Moving right along the flat-SOFR row, from on plan to nine months late, costs $278,322. The grid is far steeper horizontally than vertically.
Decomposed
| Scenario | Cash shortfall over 24 months |
|---|---|
| Base case: market-implied path, lease-up on plan | $0 |
| Rate shock only: +150 bp, lease-up on plan | $2,579 |
| Schedule slip only: market path, 6 months late | $227,216 |
| Both | $236,283 |
A six-month slip is 25% of a 24-month term, and it consumes 93% of the entire reserve. The rate shock, nearly 50% larger than what the curve is pricing, consumes 1%. Two reasons for the gap, and they compound each other.
The first is that a rate increase applies to a declining number of reserve-funded months, because the reserve is nearly spent by the time the rate path has risen much. The second is more important: delaying lease-up does not postpone six average months of NOI, it postpones the six highest-NOI months and replaces them with six of the lowest. A six-month slip removes $332,000 of NOI from the 24-month window, roughly $55,333 for each month of delay, against average monthly interest of $51,122. Every month of delay therefore costs slightly more than a full month of interest, which is why only $227,216 of that $332,000 shows up as a shortfall: the reserve absorbed the rest before it ran out.
One qualifier. If your lender requires a purchased rate cap, which most bridge lenders do on floating-rate paper, the rate exposure in that left column is bounded by the strike and the already-small rate sensitivity gets smaller. The cap premium is a real cost and belongs in the budget, but buying it does nothing about the risk that is actually 88 times larger.
What it costs to fix
If the sensitivity that breaks the reserve is schedule, the answer is to size the reserve against a schedule you actually believe rather than the one in the offering memorandum. We re-solved for a reserve that survives a six-month delay on the market-implied rate path.
| Item | Sized on plan | Sized to a 6-month slip | Change |
|---|---|---|---|
| Interest reserve | $245,000 | $475,000 | +$230,000 |
| Total project cost | $11,943,605 | $12,175,227 | +$231,622 |
| Loan amount | $8,360,524 | $8,522,659 | +$162,135 |
| Sponsor equity | $3,583,082 | $3,652,568 | +$69,486 |
| Stabilized yield on cost | 7.54% | 7.39% | -15 bp |
| Month 24 balance | $8,358,774 | $8,521,144 | +$162,370 |
| Debt yield at month 24 | 10.77% | 10.56% | -21 bp |
Doubling the reserve costs the sponsor $69,486 of additional equity, because the reserve sits in the cost basis and the lender funds 70% of it. Every dollar of extra reserve costs the sponsor 30 cents. Set against a $227,216 cash exposure that would otherwise arrive as a capital call in month 10, that is cheap insurance.
The exit still works. At month 24 the larger loan leaves a $8,521,144 balance against $900,000 of stabilized NOI, a 10.56% debt yield, which clears a 10.00% takeout threshold with $478,856 of room. The 15 basis points of yield-on-cost dilution is the real price, and it is a number you can show an investor.
Simplifications: interest computed 30/360 on the outstanding balance, capex drawn straight-line over 12 months, NOI ramped linearly from month 7 to month 20, and the SOFR path modeled as a straight line to the stated endpoint rather than the actual forward curve. The implied path is derived from Treasury par yields as a proxy for expected short rates and does not strip out term premium, so treat 136 bp as an order of magnitude and not a forecast. Excluded: rate cap premium, extension fees and tests, lender legal, reserve replenishment covenants, and any cash trap that would divert property NOI away from debt service. Rate, fee, and leverage figures are illustrative modeling inputs, not a QuadBlock quote.
What to check before you sign the budget
- Confirm whether the reserve is sized gross or net of projected property NOI. These produce answers that differ by a factor of two or more, and the term sheet often does not say which was used.
- Ask what rate assumption sits behind the number. If it is the closing rate held flat, re-solve it on the forward curve yourself.
- Check that the model accrues interest on the reserve’s own draws. If the reserve is funded from loan proceeds and the sizing is a simple balance times rate times term, that loop is missing.
- Rebuild the sizing on a lease-up schedule that is three to six months behind the business plan, and ask what the shortfall is in dollars and which month it lands in.
- Find out what happens when the reserve is exhausted. A reserve replenishment covenant backed by a sponsor guaranty, a springing cash trap, and a straight capital call are three very different outcomes.
- Compare the reserve exhaustion month to the month the property first services its own debt. If the reserve runs out before the crossover, the sizing is wrong regardless of how the math was done.
- Confirm whether unused reserve is released to the sponsor, applied to principal, or simply not advanced. On Method A that distinction is worth $366,910.
Red flags
- A reserve expressed as a round number of months rather than as a solved dollar amount.
- A business plan where the reserve runs dry in the same quarter the first renewal or lease commencement is scheduled.
- No stated assumption about when construction completion allows leasing to begin. On a demising-wall project, leasing cannot start before the work is substantially done, and that lag is where reserve sizings usually break.
- A reserve that was sized at application and never re-solved after the loan amount, spread, or capex budget changed in committee.
- Interest reserve and capex contingency drawn from the same bucket with no sub-limit.
How QuadBlock can help
We underwrite bridge and transitional loans from $250,000 to $30 million across multifamily, industrial, retail, self-storage, specialty, and ground-up residential, and we have closed more than 30 deals in every state except Oregon and Nevada. On transitional deals we size the interest reserve against the lease-up schedule we believe rather than the one that makes the proceeds look best, and we would rather have that conversation before the term sheet than during month 10. Send us the rent roll, the capex budget, and your leasing assumptions at quadblockcapital.com and we will show you the solve.
This article is for general educational purposes and is not tax, legal, or investment advice. The deal shown is a hypothetical model built from the market data cited below, and the rate, fee, and leverage inputs are illustrative rather than quoted terms. Loan structures, pricing, and underwriting standards vary by lender, asset, sponsor, and market, and actual results will differ. Consult your own tax advisor, counsel, and lender before acting on any of the structures discussed.
Sources
- Federal Reserve Bank of New York, Secured Overnight Financing Rate, observations through Oct. 6, 2026
- U.S. Department of the Treasury, Daily Treasury Par Yield Curve Rates, October 2026
- Board of Governors of the Federal Reserve System, Open Market Operations, federal funds target range history
- Board of Governors of the Federal Reserve System, FOMC Meeting Calendar, 2026
- JLL, U.S. Industrial Market Dynamics, Q2 2026
