Stress-testing: the base case is a hope

A base-case model is a hope with decimal places. It shows what happens if the plan works, the market cooperates, and the assumptions hold, which is the one scenario least worth studying, because it is the one nobody needs to prepare for. The deal worth evaluating is the one where things go wrong, and specifically the one where several things go wrong at the same time, because that is what a downturn actually is.

The variables to stress

A handful of assumptions carry the deal, and each should be pushed against. Interest rates, which matter enormously when the debt floats, because a rate rise lands directly on cash flow. Vacancy, which climbs when the market softens. Rent growth, which sponsors often assume marches upward and which can flatten or reverse. The exit cap rate, covered in the underwriting article, which can expand and gut the sale price. And the refinance, the assumption that the property can be refinanced on schedule and on acceptable terms, which is the linchpin of many bridge-financed deals.

Testing one bad thing at a time is the mistake. The real world does not send its problems one at a time.

Why single-variable tables lie

The standard sensitivity table moves one variable and holds the rest fixed: vacancy up two points, everything else unchanged. It is comforting and misleading, because the variables are correlated. A downturn does not politely raise vacancy while leaving rents, interest rates, and cap rates where they were. It raises vacancy and softens rents and, often, expands cap rates and tightens credit, all at once. A deal that survives each stress alone can die when they arrive together, which is exactly when they arrive.

Walk a combined case. Take a deal underwritten to lift net operating income to $700,000 and sell at a 6.0 percent exit cap for about $11.67 million, comfortably above the debt, returning investors a healthy multiple. Now run the correlated stress. Softer demand holds net operating income near $560,000 instead of $700,000. The exit cap expands to 7.0 percent. The sale price falls to about $8.0 million. If the debt on the property is near $6.5 million and there are selling costs and accrued obligations, the equity that looked like a strong multiple in the base case is now a fraction of what went in, and if the debt floated higher along the way, the deal may have been calling investors for more capital before it ever reached the sale. Same property, same sponsor, one honest scenario instead of the hopeful one.

The refinance case deserves its own attention, because it is the specific killer for short-term, floating-rate deals. When the plan depends on refinancing to pay off a bridge loan, three of the stresses converge on that moment: rates are higher, the property is worth less at an expanded cap, and lenders are more conservative. A refinance that was assumed to return capital instead fails to close, and the sponsor is choosing between a forced sale at the bottom and a capital call.

The structuring consequence

Evaluate the combined-stress case, not the single-variable table, because the combined case is the one the market will actually deliver if it delivers a bad one. A deal that only survives its base case has not been underwritten; it has been hoped for. And any deal built on short-term or floating-rate debt must have its refinance scenario stressed explicitly, with rates up and value down at the same time, because that is the scenario that turns a projected return into a capital call.