Stress testing a deal before you sign

Most underwriting produces one set of numbers. That set describes the world in which every assumption holds.

It has never described any actual holding period.

The variables worth stressing

Rent. Test at ten percent below your assumption, and at zero growth for the entire hold.

Zero growth is not pessimistic. Plenty of markets have gone three or four years without real rent growth, and some have gone longer.

Vacancy. Test at double your assumption. If you modeled five percent, run it at ten.

Operating expenses. Test at fifteen percent above. Insurance and property taxes have both moved much more than that in recent years in some markets.

Capital expenditure. Test a major unplanned item in year two — the roof, the HVAC system, the sewer line.

Interest rate at refinance. Test three points above today's rate.

Exit cap rate. Test a full point above entry.

Time to lease or stabilize. Test double your assumption on any value-add or lease-up plan.

The three scenarios worth building

Base case. Your realistic expectation. Not optimistic — realistic, which for most people requires deliberately talking themselves down.

Downside case. Several of the stresses above applied together, because they correlate. A recession produces soft rents, higher vacancy and tighter credit simultaneously, not one at a time.

The common modeling error is testing one variable at a time. Real bad years arrive as a package.

Break-even case. Not a probability estimate but a threshold: how far do things have to deteriorate before the deal fails?

Specifically — at what occupancy does cash flow reach zero? At what rent level? At what expense level?

The break-even occupancy is the most useful single figure in the whole exercise. If it is ninety-two percent, you have almost no room. If it is seventy-five, you have a great deal.

The questions the stress test should answer

Can I fund a negative year? If the downside case produces negative cash flow for eighteen months, do you have the reserves? From where?

Do I breach a loan covenant? Many commercial loans include ongoing coverage tests. Breaching one can restrict distributions or trigger cash management even while payments are current.

Can I refinance at maturity? Model the balloon at a stressed rate and check whether the supportable loan covers the balance. If not, quantify the shortfall.

Am I a forced seller? The critical question. A deal that produces poor returns in a downturn is disappointing. A deal that forces a sale at the bottom is ruinous, because the loss becomes permanent.

The difference between the two is entirely about reserves and debt structure.

Correlation, stated plainly

The events that hurt real estate arrive together.

A local employer closes. Employment falls, rents soften, vacancy rises, and the value of every property in the submarket declines at the same moment — including the ones you would sell to raise cash.

Credit tightens in exactly the conditions where you would want to refinance.

Which means a portfolio of similar properties in the same submarket is one position, not a diversified set of them, and it should be stressed as one.

What passing looks like

The deal survives the downside case without requiring capital you do not have.

The break-even occupancy is comfortably below realistic worst-case occupancy for the submarket.

The refinance works at stressed rates, or you have long enough fixed-rate debt that the question does not arise during the hold.

You are never a forced seller.

Note that none of these ask for a good return in the downside case. Surviving is sufficient. The upside takes care of itself when the base case is even roughly right.

The behavioral part

The stress test is usually built after the investor has decided to buy, which makes it decorative.

Building it first, before emotional commitment, produces a different result. The natural tendency once committed is to adjust assumptions until the model agrees with the decision.

A useful discipline is to write down the assumptions before looking at the asking price, and to note explicitly any assumption you later revise and why.

Assumptions that improve only after you see the price are not analysis.

The simplest version

For investors who will not build a model, one question captures most of the value.

If rent fell ten percent, one unit sat empty for four months, and the furnace failed — could I carry this property for a year without selling?

If the honest answer is no, the leverage is too high or the reserves are too thin, whatever the spreadsheet says.

General information about real estate underwriting, not investment advice. Scenario analysis does not predict outcomes. Consult qualified professionals about your own circumstances.