DSCR Sensitivity Design for Smaller Lending Pools

Most investment committee decks contain a DSCR sensitivity table: coverage at a range of interest rates, everything else held constant. It looks like risk analysis and it is not, because the scenario it describes has never happened.

Rates do not rise while occupancy, expenses and cap rates politely stay still. They move together, and in smaller markets they move together harder — which is the whole reason to design the test differently.

Why single-variable tables mislead

A table showing DSCR at 6%, 7% and 8% coupons implicitly claims that a 200 basis point rate move leaves the rest of the world unchanged. In practice, the conditions that push rates up also:

• Raise cap rates, cutting your exit value and your refinance LTV.
• Slow leasing, because tenant demand softens for the same macro reasons.
• Increase concessions, reducing effective rent below the asking rent in your model.
• Tighten lender appetite, raising debt yield floors and DSCR minimums at exactly the wrong moment.

Each of those is a separate row in a naive sensitivity table. In reality they are one event.

The correlated scenario approach

Replace the grid with three named scenarios in which every variable moves together and consistently.

Base

Your underwritten case. State it, then stop using it for decision-making — it is the marketing case, not the risk case.

Downside

A recession-shaped scenario, all variables moving at once:

• Effective rent 5–10% below base, through concessions rather than asking rent.
• Occupancy 300–500 bps below base.
• Operating expenses 5–8% above base, weighted toward insurance and payroll.
• Refinance coupon 150–250 bps above base.
• Exit cap 50–100 bps above entry.
• Debt yield floor 100 bps above your quote.

Severe

Not a forecast — a survivability test. Roughly double the downside deltas, and add the market-specific failure that would actually hurt you: the largest employer contracting, a supply wave delivering into your lease-up, or an insurance renewal at double.

The adjustments smaller markets require

This is where a generic stress template stops being adequate.

Wider dispersion, not just a worse mean. Small markets do not simply perform worse; they perform less predictably. A single 400-unit delivery can move a submarket's vacancy by hundreds of basis points. Your downside case should be wider than one built for a diversified metro, because the range of plausible outcomes genuinely is.

Employment concentration is a real variable. If one employer, one base, one campus or one plant drives a meaningful share of demand, model its contraction explicitly. County-level employment concentration is visible in BLS data and belongs in the underwriting file, not in the narrative section.

Time-to-transact is a variable. In a market with a handful of comparable trades a year, disposition can take two or three quarters. Extend the hold in your downside case and carry the debt service for the extra period. A deal that survives a 15% value decline but not a nine-month delay in selling is a deal with a liquidity problem, not a value problem.

Lender exit risk. Model the refinance as though your current lender is unavailable. In markets where two or three community banks provide most of the debt, this is not a tail scenario — a number of regional banks reduced commercial real estate exposure between 2023 and 2025.

Insurance as its own scenario. In coastal, Gulf and wildfire-exposed markets, model a renewal at 1.5x and 2x independently of everything else. It has been the single largest NOI surprise of the last several years, and it does not correlate with the rate cycle.

What to measure, and where the thresholds sit

For each scenario, produce these five and compare them to a threshold set before diligence:

1. DSCR at the trough, not the average. Averages hide the quarter you breach a covenant.
2. Debt yield at maturity, against the floor you expect to face rather than the one you were quoted. See what debt yield floor works in tertiary markets.
3. Break-even occupancy — the point at which the property covers debt service and nothing else.
4. Refinance proceeds versus maturity balance — the gap, in dollars.
5. Months of reserve at the downside burn rate.

A useful discipline: write down what you will do if each threshold is breached, before you close. "Fund from reserve for six months, then market for sale" is a plan. "Reassess" is not.

Presenting it without hiding it

Three habits that separate credible committee material from optimistic material:

• Lead with the downside. If the recommendation only survives when the base case leads, the recommendation is weak.
• Show the correlation explicitly. State that the variables move together and why, so nobody reads the downside as improbably compounded pessimism.
• Name the breach. Identify which threshold fails first in the severe case and at what point. That single sentence is more useful than the whole table.

What to do next

• Run the numbers with the DSCR calculator.
• Extend the test to value with the cap rate, debt yield and exit cap stress test.
• Size the maturity gap in how to underwrite refinance risk in non-core markets.
• Choose the structure with floating vs fixed rate structures for thin-liquidity CRE.

Sources

• Bureau of Labor Statistics
• FRED: Federal Reserve Economic Data
• Census Building Permits Survey
• Federal Reserve Senior Loan Officer Opinion Survey