Floating Rate Loan Risks and Interest Rate Caps
Floating-rate commercial real estate loans carry four main risks: debt service that rises with the benchmark, caps that expire before the loan does, replacement hedges that cost far more at extension, and refinancing proceeds that shrink when rates and spreads reset higher. A cap limits the benchmark component of interest for the term and balance it covers. The spread, the premium and everything after expiry stay with the borrower. Coverage that looks adequate while a cap is paying can therefore sit beside a maturity that needs fresh equity.
What a cap covers and what stays on the borrower
A floating-rate loan prices at a benchmark such as SOFR plus a fixed contractual spread. A cap is a payment owed to the borrower by a third party whenever the benchmark finishes above the strike. In single-asset floating-rate structures the borrower usually writes the cap into the financing: if SOFR rises above 4%, the borrower pays 4% on the benchmark leg and the cap counterparty pays the difference.
The spread is untouched by any of this.
Two boundaries decide how much protection a cap delivers. The first is notional: if the cap covers less principal than the loan balance, the uncovered slice floats freely. The second is term: once the cap contract ends, the full benchmark exposure returns on whatever balance is still outstanding, even if the cap performed exactly as written until its last day.
A reconciliation catches most modeling errors. Gross interest equals balance multiplied by benchmark plus spread. The cap payment equals covered notional multiplied by the excess of the benchmark over the strike. Net interest is the difference. A model that substitutes the strike for the whole note rate has dropped the spread and ignored any gap between the loan balance and the covered notional.
Cap expiry ahead of loan maturity
A hedge can end before the debt does. In one sample of CRE CLO floating-rate loans, 40% of the rate-cap agreements with expiration dates in the next four years expire before their respective loans mature.
The same sample puts a size on what a cap is doing. Current note rates on loans without a cap agreement averaged approximately 200 bps higher than the effective note rates on capped loans, so expiry reopens an interest gap of that general order.
That gap lands on debt service coverage first and on refinancing second. Cap expirations pressure coverage during the remaining term and add refinancing risk at maturity, because the lender sizing a takeout loan sees the uncapped cost of carrying the debt.
Office floaters show the pressure plainly. Many carry deteriorating asset fundamentals and a high debt service burden, and that burden grows when the caps expire.
A hedge schedule laid next to the maturity schedule is the working test. A loan with time left on its term and no cap behind it is a different credit from the same loan with protection in force.
Extension options and the replacement hedge bill
Extension options give a borrower time, and time on a floating-rate loan has to be hedged again. A substantial portion of debt used in the post-pandemic transaction market was two-year floating-rate securitized paper, often with three one-year extensions. Each extension is a separate decision with a separate cap to buy.
In CRE CLO loans the lender requires a cap for the initial three-year term, and borrowers can face higher hedging costs when they exercise either one-year extension option. The original premium is a sunk cost and says little about the price of the next contract.
Option pricing explains the direction. A cap becomes more expensive as its strike moves toward or below the prevailing benchmark, because the odds that it pays rise. A strike that was far out of the money at closing can be close to the money, or already in it, when the extension comes due.
Multifamily borrowers can find the reset costly, and in the experience of one fund manager new caps on maturing high-octane debt can run 10 or 20 times the original price.
That cash comes due at the moment the borrower is also testing whether the extension conditions can be met. An extension option supplies time only if the property clears the lender’s tests and the sponsor can fund the new cap.
“Floating-rate borrowers in the multifamily space may have to reset interest rate hedges at costly premiums to extend maturing loans, and rising rates may put downward pressure on refinancing proceeds.”
Why a passing coverage ratio can sit beside a refinancing shortfall
A working cap lowers the interest on the loan it hedges. It has no bearing on how much a new lender will advance, because the takeout loan is sized on its own rate and its own coverage requirement.
For an interest-only loan, DSCR is net operating income divided by annual interest. A lender sizing to a required DSCR divides NOI by that ratio to get allowable debt service, then divides allowable debt service by the underwriting rate to get the loan amount. With NOI and the required ratio fixed, a higher rate produces a smaller loan.
Both components of the rate moved against borrowers. SOFR rose from 0.05% to 5.31% over about a year and a half, which raises the cost of debt underwritten into deals, and lenders also widened spreads as rates rose, which added to all-in interest cost.
Banks reduce proceeds to protect against debt service coverage ratios, and over $900 billion of commercial real estate loans were forecast to mature in a single year.
The refinancing gap is the existing balance plus transaction costs, minus gross replacement proceeds. Payment capacity and payoff capacity are separate tests, and a loan can pass the first while failing the second. Coverage thresholds vary by lender, so the replacement lender’s assumed rate and required ratio belong in the analysis explicitly.
Testing coverage at the uncapped rate
Reported coverage on a capped loan measures the cap as much as the property. The cleaner decision test recomputes DSCR with the cap removed and the benchmark at a stressed level.
In one estimate of the single-asset floating-rate market, 20% of it could not meet its debt service coverage ratio if SOFR were allowed to rise to 4.8%. Caps are what hold those loans inside their covenants.
That changes the question a credit committee or a limited partner asks. The issue is how much of the reported coverage depends on a contract with an end date, and what coverage looks like on the day after it ends.
A cap that is in the money and paying also carries a warning: the benchmark has already moved past the strike, so the replacement contract will price from that higher level.
How property type changes which risk binds first
Office and multifamily floaters can reach trouble through different doors. On office loans the asset side is weak and the debt service burden is high, so cap expiry adds to a coverage problem that already exists.
In multifamily, the pressure arrives at extension and takeout: costly hedge resets to extend a maturing loan, and rising rates putting downward pressure on refinancing proceeds.
A cap payment offsets covered interest. It does not replace lost income. DSCR carries income in the numerator and debt service in the denominator, so falling NOI can erode coverage while effective interest stays flat.
Rate sensitivities that hold NOI constant therefore describe only the interest half of the exposure.
Worked example: a $20,000,000 floating-rate loan with and without its cap
This hypothetical compares one $20,000,000 interest-only loan over the same 12 months in two cases. Property NOI is $2,520,000 in both. The loan pays a benchmark plus a 3% spread, the cap strike is 4%, and the benchmark sits at 6% throughout. Case A has the cap in force on the full balance. Case B has the cap expired and no replacement bought.
The note rate is 6% plus 3%, which is 9%. Gross annual interest is $20,000,000 multiplied by 9%, which is $1,800,000. That is the full cost in Case B.
In Case A the benchmark exceeds the strike by 6% minus 4%, which is 2%. The cap pays $20,000,000 multiplied by 2%, which is $400,000. Effective debt service is $1,800,000 minus $400,000, which is $1,400,000. That equals the 4% strike plus the 3% spread, an effective 7% on the balance.
Case A coverage is $2,520,000 divided by $1,400,000, which is 1.8x. Cash after interest is $2,520,000 minus $1,400,000, which is $1,120,000.
Case B coverage is $2,520,000 divided by $1,800,000, which is 1.4x. Cash after interest is $2,520,000 minus $1,800,000, which is $720,000.
The cash difference is $1,120,000 minus $720,000, which is $400,000, exactly the cap payment.
Now add the extension. Suppose a replacement cap at the same 4% strike costs a hypothetical $600,000 for the extension year. It restores interest of $1,400,000. Cash after interest and premium is then $1,120,000 minus $600,000, which is $520,000.
Against the $720,000 left in Case B, buying the cap leaves $720,000 minus $520,000, which is $200,000 less cash in that year. The premium of $600,000 exceeds the $400,000 the cap pays at a 6% benchmark. The cap earns its cost only if the benchmark runs higher or the lender requires the hedge as a condition of the extension.
Mistakes that distort floating-rate loan analysis
Reading the cap strike as the all-in rate drops the spread. The cap offsets only the benchmark excess over the strike, so the correct build is gross interest first, then the cap payment subtracted. In the example above the strike alone suggests 4%, while the effective rate is 7%.
Carrying a capped interest assumption through the whole modeled hold overstates protection. In one sample, 40% of cap agreements expiring in the next four years end before the loan they hedge. Correct analysis ties each period’s interest to the cap actually in force and prices the exposed periods at the uncapped rate.
Using the original premium as the extension budget substitutes a sunk cost for an unresolved one. Replacement caps on maturing high-octane debt can cost 10 or 20 times the original price. The right line item is a fresh quote for each extension year.
Treating a passing DSCR as proof of a funded takeout confuses payment capacity with payoff capacity. The new lender sizes on its own rate and its own coverage requirement, and higher borrowing costs shrink the principal the same income supports. The refinancing gap is what remains after net proceeds meet the payoff balance.
Holding NOI constant in a rate sensitivity covers half the exposure. On assets where income is weakening, a cap can leave interest flat while coverage still falls, so rate shocks and income shocks belong in the same table.
Floating-rate loan review questions
What share of the note rate does the cap cover? The benchmark and the spread reconcile separately to total interest, and only the benchmark leg is protected.
Does the covered notional match the outstanding balance? Any difference leaves principal floating freely even while the cap is paying on the rest.
When does the cap end relative to the final maturity and each extension? The gap between those dates marks every period when the loan runs unhedged.
What will a replacement cap cost at the strike and forward curve available at the extension date? The sunk premium on the original contract carries no information about that number.
Does reported coverage use gross interest or interest net of cap receipts? The convention decides whether the ratio measures the property or the hedge.
What is DSCR with the cap removed and the benchmark stressed? That figure shows how much of the covenant cushion depends on a contract with an end date.
Whichever underwriting rate and coverage requirement the takeout lender applies will set the proceeds, so a borrower should get both in writing before assuming the new loan will repay the existing one. Proceeds follow NOI divided by the required ratio, divided again by the underwriting rate.
Who funds the combined cash demand of the replacement hedge and the refinancing gap? Both can arrive in the same period while the capped payment still looks manageable.