Stress Testing a Commercial Real Estate Acquisition

A stress test finds the specific conditions under which an acquisition runs out of cash, trips a loan covenant, cannot refinance or loses its equity, and then sizes the capital each failure would require. Stress tests commonly examine occupancy, net operating income, interest rates and construction costs together, because the combined case is where reserves get consumed. Each failure sits at a different threshold, so a deal can pass one test and fail the next. The output is a list of break points with a dated cash schedule behind each.

Four break points a stress test has to locate

Cash break-even is the first. It is the point where collected revenue covers operating expenses and debt service. Distributions to investors depend on property revenue clearing that bar, and below it the shortfall has to be funded from reserves or new capital.

The covenant break comes second. DSCR is net operating income divided by debt service, and a lender sets its own minimum, which varies by loan and by lender. Any minimum above 1.0 is crossed while the property still pays its bills, so a deal can be in default on coverage with cash still positive.

The refinance break comes third. At maturity the replacement loan is sized by the replacement lender’s constraints, and the resulting amount can fall short of the balance owed even when cash flow is positive.

The equity break is the last. Equity is whatever value remains after the debt, so the thinner the equity, the smaller the value decline that eliminates it. With only 5% equity in the capital stack, a 5% drop in property value takes the equity to zero.

A deal can clear the first test and fail the other three.

Reverse stress testing finds each break directly. It solves for the occupancy at which NOI equals debt service, the occupancy at which coverage hits the lender minimum, and the exit cap rate at which sale value equals the loan balance. Stress tests commonly examine occupancy, NOI, interest rates and construction costs, and the reverse method reports how far each can move before something breaks.

Building the downside from the operating statement

A flat haircut applied to NOI produces a lower number without showing how the property got there. Stress belongs on the inputs that make NOI: occupancy, rent per unit of space, concessions, collection losses and each expense line.

Physical occupancy and collected revenue are different quantities. Occupied space that carries free rent or unpaid balances produces less cash than occupied space paying full rent, so an occupancy sensitivity that skips those adjustments overstates receipts.

Expense behavior changes the answer as much as revenue does. A fixed expense stays payable when occupancy falls, and a variable expense moves with its driver. Treating every cost as a percentage of revenue gives the property automatic relief whenever income drops, and that relief is a modeling choice that needs a stated basis.

Insurance premiums and property taxes can carry their own multipliers, set by market, so that income falls and a large fixed cost rises in the same scenario. That is the pattern in which NOI drops from both sides of the statement at once, with income down and expenses up.

Property type changes the unit being stressed. The method differs for hotels, Class A offices and open-air retail centers, because each earns revenue through a different operating unit. A room-night assumption and a leased-area assumption each need their own price and collection assumption before either produces cash.

Location and competing properties shape outcomes, as do tenant industries and market volatility, and these conditions are harder to measure and model than loan metrics. A stressed input is credible when it connects to one of them, and a uniform occupancy cut applied to unrelated assets leaves their different exposures untested.

“It starts with the foundation, the fundamental granular analysis of the actual real estate itself. That’s what we all grew up on in the business, underwriting.”

What flat rents do to a plan built on momentum

Underwriting built on low vacancy and fast rent growth carries an embedded assumption that those trends continue. When conditions shifted, deals bought on that assumption saw income fall while expenses rose, and NOI fell with them.

A downside case that keeps rent growth positive but smaller still contains improvement. Flat rents and falling rents are the cases that show what the capital structure does when the improvement does not arrive.

A projected rent dip followed by a recovery has a cash cost in the dip. The recovery only helps if reserves or committed capital last through the low point, so the schedule has to show the lowest cumulative cash balance and when it occurs.

Rate resets, terminal caps and the maturity date

Floating-rate debt is stressed with the payment the loan documents produce. Where the loan has an index floor, the applicable index is the greater of the observed index and the floor, and the contractual spread is added to it. A purchased rate cap is modeled under its own strike and term, and protection assumed past its expiration understates the exposure.

Fixed-rate debt behaves differently. Payments follow the contract until maturity, so a rate shock leaves current cash flow unchanged and raises the cost of the replacement loan, which is where the risk shows up.

On the exit side, stress tests commonly examine terminal cap rates of 6% or 7%, applied to the NOI in the year of sale. A cap rate is one year of NOI divided by price, so a higher terminal cap rate applied to a lower stressed NOI compounds both effects on sale value.

Maturity timing links these pieces. If stabilization slips past the maturity date, the plan faces a repayment event before the property produces the income the replacement loan would be sized on.

Refinance sizing: the smallest ceiling governs

Positive equity in the property does not prove the existing loan can be repaid from new debt. A replacement loan is sized by several constraints at once, and the smallest result governs.

A value-based ceiling multiplies appraised value by the maximum debt-to-value ratio. A coverage-based ceiling divides stressed NOI by the required DSCR to find permitted debt service, then converts that payment into a loan amount. For an interest-only loan the conversion divides permitted debt service by the rate. An amortizing loan needs its full payment factor in that step, and applying the interest-only shortcut overstates the amount.

A break-even refinance test asks whether the property could refinance all invested money and still break even with a buffer for capital and operating expenses. Applied to a stressed case, it turns the refinance question into a number that can be compared with the balance due.

Lender minimums, maximum ratios and rates differ by lender, so every figure in the replacement test is an assumption the model must state.

Worked example: a $40,000,000 acquisition with occupancy loss and a rate reset

All figures are hypothetical and both cases use the same capital and the same one-year period. Price is $40,000,000 with $25,000,000 of interest-only debt, so closing equity is $40,000,000 minus $25,000,000, or $15,000,000, and debt is 62.5% of price. Equity also funds a $1,750,000 capital budget and a $175,000 contingency. A further $700,000 sits in operating reserves. Fees and selling costs are set to zero for clarity.

Potential rent is $5,000,000 a year, fixed expenses are $1,500,000 and variable expenses are 20% of collected rent. The base case runs at 90% occupancy and a 6% debt rate. The stress case runs at 75% occupancy and a 7.5% debt rate, with $150,000 of added fixed expense for insurance and property tax.

Base collected rent is $5,000,000 multiplied by 0.9, or $4,500,000. Variable expenses are $4,500,000 multiplied by 0.2, or $900,000. NOI is $4,500,000 minus $1,500,000 minus $900,000, or $2,100,000, which is a going-in cap rate of 5.25% ($2,100,000 divided by $40,000,000).

Debt service is $25,000,000 multiplied by 0.06, or $1,500,000. DSCR is $2,100,000 divided by $1,500,000, or 1.4, and cash after debt service is $2,100,000 minus $1,500,000, or $600,000.

Break-even occupancy comes next. Rent left after variable expenses at full occupancy is $5,000,000 multiplied by 0.8, or $4,000,000. That amount has to cover debt service plus fixed expenses, which is $1,500,000 plus $1,500,000, or $3,000,000. Cash break-even occupancy is $3,000,000 divided by $4,000,000, or 75%.

An illustrative covenant of 1.2 DSCR sets the next threshold. Required NOI is $1,500,000 multiplied by 1.2, or $1,800,000. Fixed expenses bring the required rent contribution to $1,800,000 plus $1,500,000, or $3,300,000, and $3,300,000 divided by $4,000,000 gives 82.5% occupancy.

Between 75% and 82.5% occupancy the property pays every bill and still fails the covenant.

In the stress case, collected rent is $5,000,000 multiplied by 0.75, or $3,750,000. Variable expenses are $3,750,000 multiplied by 0.2, or $750,000. Fixed expenses are $1,500,000 plus $150,000, or $1,650,000. NOI is $3,750,000 minus $1,650,000 minus $750,000, or $1,350,000.

Debt service is $25,000,000 multiplied by 0.075, or $1,875,000. DSCR is $1,350,000 divided by $1,875,000, or 0.72, and the annual deficit is $1,875,000 minus $1,350,000, or $525,000. That is $525,000 divided by 12, or $43,750 a month. Reserves of $700,000 last $700,000 divided by $43,750, or 16 months.

The capital budget competes for the same cash. A 20% overrun on $1,750,000 is $350,000. The $175,000 contingency absorbs part of it, and the remaining $350,000 minus $175,000, or $175,000, comes from reserves. Reserves fall to $700,000 minus $175,000, or $525,000, and now last $525,000 divided by $43,750, or 12 months.

At exit, the base case values the property at $2,100,000 divided by 0.0525, or $40,000,000, and proceeds after the $25,000,000 loan are $15,000,000. The stress case applies a 6% terminal cap rate to $1,350,000 of NOI, which gives $22,500,000. Debt exceeds value by $25,000,000 minus $22,500,000, or $2,500,000, so nothing reaches equity.

Equity measures the same thing from the other side. Purchase equity of $15,000,000 is 37.5% of the $40,000,000 price. The stressed value is down $17,500,000, a 43.75% decline. The exit cap rate that leaves zero proceeds is NOI divided by debt: 8.4% in the base case ($2,100,000 divided by $25,000,000) and 5.4% in the stress case ($1,350,000 divided by $25,000,000).

Refinancing the stress case at maturity uses assumed replacement terms of 60% maximum debt-to-value, 1.25 DSCR and 7.5% interest-only. The value ceiling is $22,500,000 multiplied by 0.6, or $13,500,000. Permitted debt service is $1,350,000 divided by 1.25, or $1,080,000, and $1,080,000 divided by 0.075 gives $14,400,000. The value ceiling governs at $13,500,000, so the repayment gap is $25,000,000 minus $13,500,000, or $11,500,000.

Mistakes that make a stress test look sturdier than the deal is

Reading the covenant as cash break-even is an arithmetic confusion. Both involve debt service, but DSCR is NOI divided by debt service, so any minimum above 1.0 is crossed while NOI still exceeds debt service. In the example the covenant fails at 82.5% occupancy and cash fails at 75%, and the gap between them is the range where the property pays its bills while in default on coverage.

A single expense ratio gives automatic relief. If every cost is a percentage of revenue, a fall in revenue lowers costs by the same percentage. Insurance and property tax do not fall with occupancy, so a model with relief built in understates the deficit, and separate fixed and variable lines expose the relief.

Reserves get counted twice when separate schedules each draw on the same balance. In the example, the capital overrun takes $175,000 from reserves, and the operating deficit then has $525,000 left out of the original $700,000. Funding capacity shrinks from 16 months to 12, and a single dated cash reconciliation catches this.

Positive property equity gets confused with refinancing capacity. At the stressed value, equity is already gone, and even a property with value above its loan faces a replacement loan sized by a lower debt-to-value ratio and a coverage test. The binding ceiling decides the repayment gap, and the gap is cash the sponsor or investors must find.

Extending recent conditions forward is an assumption, and the arithmetic gives it no support. Underwriting that projects fast rent growth because rents have been rising has no basis for continuation, and a downside case that keeps growth positive still embeds improvement. Flat and negative rent growth show what the debt and reserves can carry.

A terminal sale can hide an earlier shortfall when the model reports only totals. Sale proceeds arrive at the sale date and cannot fund a payment due before it. The lowest cumulative cash balance, by date, shows the contribution needed before any recovery.

Questions to settle before capital is committed

Which stressed inputs trace to property-specific conditions, including location and competing properties that a uniform cut leaves untested?

At what occupancy does NOI equal debt service, and at what higher occupancy does coverage reach a lender minimum above 1.0?

What stressed payment do the loan’s contract terms produce after the index moves, any floor applies and any purchased cap expires?

How do separate multipliers stress insurance premiums and property taxes so that a revenue decline does not hide a cost increase?

What cash remains after committed capital spending and stressed operating losses when every schedule draws on the same reserve balance?

What happens to the plan if stabilization slips past maturity, and when do the added carry payments and repayment event occur?

What does common equity receive at terminal cap rates of 6% or 7%, and which cap rate leaves nothing when calculated from stressed NOI divided by the debt balance, adjusted for selling costs and senior claims?

After computing both the value-based and coverage-based ceilings, which replacement loan constraint produces the smallest amount, and what is the resulting gap?

Can every assumption in the failure case be investigated before commitment, and what unpriced cash consequence remains for any assumption that cannot be examined?