# Modeling cash flow through the interest rate cycle
A retail center throwing off $2 million in net operating income can look bulletproof on a spreadsheet and still fail. The trigger is rarely the tenants. It is the debt. When a floating rate resets 300 basis points higher, or a loan matures into a market that will only refinance at half the old leverage, the same cash flow that covered debt comfortably turns thin, then negative. This lesson shows you where that happens by building the model line by line.
Start with the property, then layer on the debt.
Net operating income (NOI) is rental income plus recoveries, minus operating expenses. It excludes debt service and capital costs. For a stabilized retail center it is the top of your waterfall.
Debt service is the payment to the lender: interest, plus principal if the loan amortizes. Many commercial loans are interest-only (IO) for a period, meaning no principal is repaid until maturity.
Levered cash flow is NOI minus debt service minus capital items (tenant improvements, leasing commissions, reserves). This is what the equity actually keeps.
The single number that predicts distress is the debt service coverage ratio (DSCR): NOI divided by annual debt service. A DSCR of 1.25x means NOI is 125% of the payment. Lenders often set covenants near 1.20x. Below 1.0x, the property cannot pay its own debt from operations.
Build ten columns, one per year. Each column carries NOI up top, debt service in the middle, and levered cash flow at the bottom. The two events that matter most sit at the edges: the rate on the debt (which can float) and the refinancing at year 5.
If the loan is fixed-rate, debt service is a known constant. Model risk is low until maturity.
If the loan is floating-rate, the rate is a benchmark plus a spread. The common benchmark today is SOFR (the Secured Overnight Financing Rate, which replaced LIBOR). A loan might price at "SOFR plus 250 basis points." When SOFR moves, your payment moves. That is where the stress begins.
You can track SOFR yourself through the New York Fed's reference rate page, which publishes the rate daily for free.
Assume SOFR sits at 3.5% and the spread is 2.5%, so the all-in rate is 6.0%. On $18 million interest-only, annual debt service is $1.08 million.
That looks safe. Comfortable coverage, healthy cash to equity. This is exactly the point where inexperienced modelers stop. The base case flatters you.
Now hold everything constant except the benchmark. SOFR climbs from 3.5% to 6.5% over three years, a swing well within what markets saw between 2021 and 2023.
Your all-in rate goes from 6.0% to 9.0%. On $18 million, debt service jumps from $1.08 million to $1.62 million.
Meanwhile NOI grows only 2.5% per year, reaching about $2.15 million by year 3.
Still above covenant, but coverage fell by more than a quarter and cash to equity dropped sharply. Notice the asymmetry: NOI crawls up 2.5%, while debt service can leap 50% in the same window. Floating-rate borrowers are short the rate, and the model shows exactly how fast that erodes equity cash flow.
Many floating-rate borrowers buy an interest rate cap, a contract that pays them if the benchmark rises above a strike. If our borrower capped SOFR at 5.0%, the all-in rate stops at 7.5% no matter how high SOFR goes. Model the cap as a ceiling on the rate line. It costs an upfront premium, but it converts an open-ended risk into a known worst case. In your model, the difference between capped and uncapped can be the difference between 1.33x and 1.0x coverage.
🎬 [VIDEO: "How Interest Rate Caps Work in Commercial Real Estate" — youtube.com — a clear walkthrough of caps, strikes, and premiums for CRE loans]
The floating-rate risk is painful but gradual. The refinancing risk at year 5 is a cliff.
Your original loan was $18 million at 60% loan-to-value on a $30 million property. At maturity you must repay it, usually by taking a new loan. Two things may have changed:
1. Cap rates widened. If buyers now demand a 7.5% cap rate instead of 6.7%, and your NOI is $2.25 million, the property is worth 2.25M / 0.075 = $30 million. Flat value, despite five years of rent growth, because the market repriced.
2. Lenders tightened. A lender that offered 60% leverage in a low-rate market may now offer only 55%, and may size the loan to a minimum 1.25x DSCR at today's higher rates.
Run the DSCR-constrained sizing. At a 7.5% refinance rate and a required 1.25x coverage, the maximum interest-only loan is:
NOI / (rate x DSCR) = 2.25M / (0.075 x 1.25) = $24 million
That is more than enough to repay $18 million. Good. But flip the inputs. If the refinance rate is 9.0% and the lender wants 1.30x:
2.25M / (0.09 x 1.30) = $19.2 million
Now you can just barely refinance. If NOI had slipped because an anchor tenant left, or if the required coverage were higher, the new loan would not cover the old balance. The gap is called a refinancing shortfall, and the borrower must write an equity check to close it or lose the property.
Layer the two stresses and the model tells a clean story:
The lesson: a property can have perfectly good tenants and still default. Distress lives in the financing assumptions, and only a model that stresses the rate line and the refinance sizing will surface it before it happens.
Knowledge check
1. According to the lesson, why can a retail center with strong NOI still fail?
2. A property has a DSCR of 0.95x. What does this indicate?
3. How does levered cash flow differ from NOI?
4. Select ALL correct answers about what is included or excluded in Net Operating Income (NOI).
Select all the correct answers.
5. Select ALL correct answers about the two 'edge' events in a levered cash flow model that most affect distress risk.
Select all the correct answers.
A few modeling habits separate a decision-grade model from a decorative one.
Make the rate an input, not a constant. Put SOFR in one cell and let every debt line reference it. Then you can flex it in seconds.
Model the refinance explicitly. Add a row at maturity that sizes the new loan on both a leverage test and a DSCR test, and takes the lower. Compare that to the outstanding balance. A negative result is your shortfall.
Show DSCR every year, not just year 1. The whole point is watching coverage decay across the cycle.
Run scenarios, not a single line. Build at least a base, a rising-rate, and a hard-refinance case. For a deeper reference on structure, the CFI guide to real estate financial modeling is a solid free starting point.
None of this is investment advice. It is a framework for seeing risk clearly before you commit capital.