Worked example: rates rise on a floating-rate loan

When the benchmark on a floating-rate loan rises, interest expense grows on the full balance while property income stays where it was, so coverage and cash to equity fall. The arithmetic runs from benchmark plus spread to annual interest, then to debt-service coverage and cash after debt service. A rate cap interrupts that chain only while it is in force, and the refinancing date adds a separate test that depends on the replacement lender’s own thresholds.

How a benchmark move reaches interest expense and equity cash

On a floating-rate loan the first number to move is interest expense, and NOI plays no part in it.

For an interest-only loan, annual interest equals the balance multiplied by the benchmark plus the spread. The spread is fixed by contract, so every point the benchmark rises adds one percent of the full balance to the annual interest bill. Floating-rate borrowers pay more interest as benchmark rates rise, and their cash flow goes down accordingly.

The size of the move can be large. Over one two-year span, floating-rate borrowers saw loan rates climb 400 to 500 basis points, while borrowers refinancing into fixed-rate debt generally priced off the 10-year Treasury. The floating borrower carries the reset on every payment date, and the fixed borrower carries it once, at the refinancing.

Two ratios respond differently. Debt-service coverage divides NOI by debt service for the same period, so a larger payment lowers it even when NOI is flat. Debt yield divides NOI by the loan balance and contains no rate term, so it reads the same before and after the reset.

Rising rates also work on the asset side. They can lift capitalization rates, reduce transaction volume and raise borrowing costs across regions, while increasing repayment costs on floating-rate loans. The payment problem and the exit problem can therefore arrive together.

The same pattern holds inside securitized structures. Rising rates can increase payment difficulty and delinquencies, and they raise the interest owed on floating-rate CRE CLO notes, which reduces the cash available to investors.

Worked example: a hypothetical $20 million loan after a 2.5 point benchmark rise

The hypothetical loan is an interest-only floating-rate loan of $20,000,000 on a stabilized property with annual NOI of $2,100,000 and $10,000,000 of equity. The contract rate is the benchmark plus a 3% spread. Both cases hold the balance, NOI, fees (none) and the twelve-month period identical, so the benchmark is the only variable.

Base case: the benchmark is 2%, so the loan rate is 2% plus 3%, or 5%. Annual interest is $20,000,000 multiplied by 5% divided by 100, which is $1,000,000. DSCR is $2,100,000 divided by $1,000,000, or 2.1x. Cash after debt service is $2,100,000 minus $1,000,000, or $1,100,000, which is 11% on the $10,000,000 of equity ($1,100,000 divided by $10,000,000, multiplied by 100).

Stressed case: the benchmark is 4.5%, a rise of 4.5 minus 2, or 2.5 points. The loan rate is 4.5% plus 3%, or 7.5%. Annual interest is $20,000,000 multiplied by 7.5% divided by 100, which is $1,500,000. DSCR is $2,100,000 divided by $1,500,000, or 1.4x. Cash after debt service is $2,100,000 minus $1,500,000, or $600,000, which is 6% on equity.

Interest rises by $1,500,000 minus $1,000,000, or $500,000, and cash after debt service falls by the same $500,000 ($1,100,000 minus $600,000). The cash yield on equity drops 5 points (11 minus 6). NOI never moved. Debt yield is $2,100,000 divided by $20,000,000, or 10.5%, in both cases.

Cap layer: assume the stressed case carries a cap with a 3% strike on the benchmark, with notional equal to the full $20,000,000. While it is in force the cap pays 4.5% minus 3%, or 1.5%, on the balance: $20,000,000 multiplied by 1.5% divided by 100, or $300,000. Net interest is $1,500,000 minus $300,000, or $1,200,000. DSCR is $2,100,000 divided by $1,200,000, or 1.75x, and cash after debt service is $2,100,000 minus $1,200,000, or $900,000.

At expiry, with the benchmark still at 4.5%, the offset disappears. Net interest returns to $1,500,000, DSCR to 1.4x and cash to $600,000. The $300,000 swing happens with no further rate move.

Refinance sizing on the same NOI: the replacement lender’s tests are assumed at 1.25x coverage and a 10% debt yield. Both thresholds are lender-specific and chosen only for illustration. Pricing is assumed at the same all-in rate as the existing loan. Allowable debt service is $2,100,000 divided by 1.25, or $1,680,000. At 5% that supports $1,680,000 multiplied by 100 divided by 5, or $33,600,000. At 7.5% it supports $1,680,000 multiplied by 100 divided by 7.5, or $22,400,000. The rate move removes $11,200,000 ($33,600,000 minus $22,400,000) of coverage-based capacity.

The debt yield test gives $2,100,000 multiplied by 100 divided by 10, or $21,000,000, at either rate. The lower of the two tests controls, so proceeds are $21,000,000 in both cases, which clears the $20,000,000 payoff by $1,000,000 ($21,000,000 minus $20,000,000). Coverage headroom over the payoff shrank from $13,600,000 ($33,600,000 minus $20,000,000) to $2,400,000 ($22,400,000 minus $20,000,000), yet the binding test never changed. Coverage becomes the binding test once the loan rate passes $1,680,000 divided by $21,000,000, multiplied by 100, or 8%. A lender with different thresholds would produce a different answer.

Why caps, hedge terms and extensions put a clock on the loan

A cap protects a stated exposure for a stated period, and the end of that period moves the borrower’s cost without any change in the market.

In a single-asset floating-rate structure, the borrower usually writes an interest rate cap on the underlying loan. If SOFR rises above 4%, the borrower pays 4% on the benchmark component and the cap provider pays the difference. The spread remains payable on top of the strike, so the strike is a ceiling on the benchmark and never the all-in loan rate.

Term mismatch creates the clock. Floating-rate debt often carries a five-year term but rate hedges for only the first two years, which pulls asset decisions forward as hedge renewals approach. The post-pandemic U.S. transaction market leaned heavily on two-year floating-rate securitized loans, often with three one-year extensions, and each extension date is another point where protection has to be priced again.

Property type changes how the clock bites. Office floaters can pair deteriorating asset fundamentals with a high debt service burden that becomes heavier when caps expire, so the liability side worsens just as the asset side is weakest.

For floating-rate multifamily, the pressure comes from the cost of the replacement hedge. Borrowers may have to reset interest rate hedges at costly premiums to extend maturing loans, and that premium is cash paid up front. Later cap receipts do not fund the purchase of the protection that produces them.

“Many office floaters are acutely asset-liability mismatched with deteriorating asset side fundamentals paired against high debt service burden that will only become more burdensome when interest rate caps expire.”

When the payment test passes and the payoff test fails

Positive cash after debt service answers whether the existing loan can be paid this period. It says nothing about whether a replacement loan covers the payoff.

Refinancing proceeds are sized from NOI at the replacement rate. A higher rate lowers the loan that a coverage test supports, and rising rates may put downward pressure on refinancing proceeds that then require new capital. In the hypothetical, coverage-based capacity fell by $11,200,000 while the existing loan stayed fully serviceable, and a lender whose binding test sat closer to the payoff would have shown a gap.

Lenders have met this problem from the other side. Floating-rate loans underwritten at debt yields that proved unsustainable after rate increases have led to workouts, extensions or sponsor equity injections. Each outcome brings new capital or new terms at the maturity or extension date.

The scale of the exposure shows up in structured data. In one snapshot of the single-asset floating-rate market, 20% of loans could not meet their coverage ratio if the benchmark were allowed to move to 4.8%. That is one sample at one time, and it shows how much of a book can sit near its threshold.

Value compounds the sizing problem. The same rate environment that lowers coverage-based proceeds can raise capitalization rates and thin transaction volume, so a sale as the alternative exit faces its own headwind.

Floating-rate underwriting mistakes that hide the cash gap

Reading the cap strike as the borrowing rate understates interest. A 3% strike covers the benchmark component only, and the 3% spread stays payable, so the net rate with the cap in force is 3 plus 3, or 6%. A model that stops at the strike shows interest of $20,000,000 multiplied by 3%, or $600,000, against a correct net figure of $1,200,000.

Treating a steady debt yield as steady payment safety confuses two ratios. Debt yield is 10.5% before and after the reset because NOI and balance did not change, while DSCR fell from 2.1x to 1.4x. A ratio with no rate term cannot register a rate shock.

Carrying the cap through the whole projection removes the expiry step. With the cap in force, the stressed case shows $900,000 of cash after debt service. At expiry the figure drops to $600,000 with the benchmark unchanged, so a model that holds the cap to maturity overstates cash for every period after expiry.

Equating an extension option with a funded extension skips the premium. Extending may require buying new protection, and that premium is a day-one cash outlay. The lower net interest the new cap produces arrives after the premium is paid.

Sizing refinancing from one test misses the controlling constraint. In the hypothetical, coverage supports $22,400,000 while debt yield supports $21,000,000, so the lower figure governs. Both thresholds are lender-specific, and the existing loan’s covenant level does not transfer to a replacement lender.

Underwriting on the in-place benchmark with no rate stress leaves the threshold unseen. Some lenders have found floating-rate loans were underwritten at debt yields that proved unsustainable once rates rose. A table of coverage at several benchmark levels shows where the break-even sits before the market finds it.

Questions for testing a floating-rate loan against a rate rise

At what benchmark level does DSCR reach the loan’s covenant?

Does the model show interest-only and amortizing debt service separately?

Is the cap strike applied to the benchmark and then added to the spread?

When does the cap expire relative to the loan maturity and each extension date?

Does the cap notional match the outstanding balance?

What does the replacement hedge cost, and where is that cash coming from?

Which refinancing test controls, coverage or debt yield?

How much cushion remains between supported proceeds and the payoff?

Does the stress case move NOI as well as the benchmark?