Deal metrics and sensitivity thinking

A one-page deal read: price and $/SF; in-place and stabilized NOI; cap rate on each; yield-on-cost for any project (stabilized NOI ÷ all-in cost) versus market cap — the development spread; DSCR/LTV/debt yield on the proposed debt; cash-on-cash year one; IRR and equity multiple over the hold; break-even occupancy = (OpEx + debt service) ÷ gross potential income.

Then flex it. One variable at a time: cap rate ±50 bps, market rents ±10%, downtime ±6 months, rate ±100 bps. The output is not one number but a shape — where the deal breaks, which lever dominates, how much cushion the structure leaves. Two deals with the same base IRR and different shapes are not the same deal.

Finally, honesty about precision: inputs are estimates, so outputs are ranges. Present base / downside / upside with the driving assumptions attached — decisions made on ranges survive contact with reality better than decisions made on decimals.