Probability-Weighted NOI for an Uncertain Cap Rate

The common mistake is to pick a single occupancy number, usually the seller's pro-forma, and treat it as fact. That gives you one NOI, one value, and a false sense of precision. The probability-weighted NOI for an uncertain cap rate method instead assigns probabilities to multiple occupancy scenarios and combines them into a single planning estimate. You keep the uncertainty visible in the math rather than burying it in one guess.

You start by listing mutually exclusive occupancy scenarios. For a small multifamily building, those might be 90% occupied (stable market), 80% occupied (moderate vacancy), and 70% occupied (high vacancy, lease rollover risk). Each scenario produces a different NOI because vacancy directly reduces effective gross income. The operating expense ratio stays roughly constant per unit, but the total NOI drops as vacancy rises.

Assign a probability to each scenario based on market evidence, historical vacancy data for the submarket, current lease expiration schedules, and comparable property performance. The probabilities must sum to exactly 1.0. If they don't, the expected NOI will be systematically biased.

Work a concrete example using inputs from the research. For a 12-unit apartment building in a secondary market, you define three occupancy scenarios based on the building's lease rollover schedule and local vacancy trends from the CBRE survey data.

Scenario A: 90% occupancy, NOI = $100,000, probability 60%. This is the stabilized case, with market rents collected and a 10% vacancy allowance built in. Scenario B: 80% occupancy, NOI = $80,000, probability 30%. This reflects a moderate vacancy period during a lease rollover. Scenario C: 70% occupancy, NOI = $60,000, probability 10%. This is a worst case, perhaps after a major tenant departure or during a market downturn.

The expected NOI calculation: ($100,000 × 0.6) + ($80,000 × 0.3) + ($60,000 × 0.1) = $60,000 + $24,000 + $6,000 = $90,000.

That $90,000 is your single planning estimate. You then apply the cap rate, say 7.5%, based on the latest CBRE survey for this property type and market tier, to get a value: $90,000 / 0.075 = $1,200,000.

If you had used the seller's pro-forma 95% occupancy and $105,000 NOI, the value would have been $1,400,000. The probability-weighted method shaved $200,000 off the price, reflecting the real vacancy risk.

Run the calculation step by step so you can reproduce it for any property. You need three inputs: the NOI for each scenario, the probability of each scenario, and a cap rate from a defensible market source.

Step 1: Define the scenarios. Make them exhaustive and mutually exclusive. For a commercial property with a single NNN lease expiring in 18 months, your scenarios might be: lease renewed at market rent (NOI $150,000, 70% probability), lease renewed at 10% below market (NOI $135,000, 20% probability), and new tenant at 15% below market after 6 months vacancy (NOI $110,000, 10% probability). The probabilities sum to 1.0.

Step 2: Compute the expected NOI. Multiply each NOI by its probability and sum the products. For the commercial example: ($150,000 × 0.7) + ($135,000 × 0.2) + ($110,000 × 0.1) = $105,000 + $27,000 + $11,000 = $143,000.

Step 3: Apply the cap rate. Use a transaction-based cap rate from the RCA/MSCI Capital Trends Report, not a survey asking price. If the market cap rate for that property type is 6.5%, the value is $143,000 / 0.065 = $2,200,000.

At this point you would normally reach for an expected value calculator to test how changes in the probability weights affect the final number. A simple spreadsheet data table showing value at 5.5%, 6.5%, and 7.5% cap rates across all three scenarios gives you a sensitivity band, which is far more useful than a single point estimate.

The method works because it makes the uncertainty explicit. Instead of one NOI that hides the risk of vacancy, you have a distribution of outcomes weighted by their likelihood. The resulting expected NOI is not a prediction, it is a planning estimate that accounts for the range of possible futures.

Geltner et al. in Commercial Real Estate Analysis and Investments treat this as standard risk analysis. The Appraisal Institute's The Appraisal of Real Estate covers probability and statistics in the context of income capitalization. Both sources agree that the cap rate applied to that expected NOI must come from comparable transactions, not from a generic survey number.

The interaction between the probability-weighted NOI and the cap rate is critical. A 50 basis point change in the cap rate can swing value by 7-8% for a stabilized property. When you combine that with an occupancy range of 70-90%, the total value band can easily be 20-25% wide. The probability-weighted method narrows that band by anchoring it to evidence, but it does not eliminate the band entirely.

The most common failure is when the sum of probabilities does not equal 1.0. If you assign 50% to Scenario A and 30% to Scenario B, with no Scenario C, the expected NOI is understated because 20% of the probability mass is missing. The same error in reverse, assigning probabilities that sum to 1.1, overstates the expected NOI.

Another failure: using trailing NOI instead of stabilized NOI. If the building had a one-time repair that lowered last year's NOI, that trailing number does not represent normal operations. The probability-weighted method amplifies that error because each scenario is built on a flawed base.

The method also fails if the probabilities are not calibrated to market evidence. A 60% chance of 90% occupancy looks reasonable, but if the local vacancy rate has averaged 12% over five years, that 90% scenario is optimistic. The CBRE survey and RCA/MSCI transaction data exist precisely to prevent that error, use them.