Sensitivity analysis is a modeling technique that measures how much an investment's return changes when a single underwriting assumption is moved while everything else is held constant. It isolates one variable, such as exit cap rate or rent growth, and reports the resulting IRR or equity multiple across a range of values.
How Sensitivity Analysis Works
Sensitivity analysis is run by selecting an output, selecting one or two inputs, and recalculating the model across a defined range of input values. The output is a grid: input values on the axes, the return metric in the cells. The method answers a single question, which is how far an assumption can move before the deal stops working.
The inputs chosen matter more than the mechanics. An underwriter picks the assumptions that are both uncertain and high-leverage on the outcome. Exit cap rate, rent growth, lease-up timing, and interest rate are the standard four in most commercial models, because each is forecast rather than observed, and each compounds across a hold period.
Element | Definition |
|---|---|
Output metric | The result being measured, typically IRR, equity multiple, or cash-on-cash |
Input variable | The single assumption being flexed, held apart from all others |
Range | The band tested, expressed in basis points, percentage points, or months |
Step | The increment between tested values, commonly 25bps for cap rates |
Breakeven | The input value at which the output hits a defined floor, such as a 1.0x multiple |
Convention sets the range. Institutional underwriting conventionally presents a cap-rate sensitivity table that shows IRR and equity multiple at exit caps from -50bps to +100bps in 25bps increments. The range is asymmetric on purpose: the downside band is wider because the downside is what the analysis exists to price.
Why Sensitivity Analysis Matters
Sensitivity analysis matters because a single-point pro forma reports one number and hides the distribution around it. A deal that shows a 13.8% IRR at the underwritten exit cap and a 6.7% IRR 100bps wider is a different deal from one that holds above 12% across the same band, even though both print the same headline.
The exit cap is usually the dominant term, and the reason is arithmetic. Value is NOI divided by a cap rate, so a 50bps widening on a 6.00% exit cap cuts the sale price by roughly 7.7%. That loss lands entirely on equity, because the debt balance does not shrink when the valuation does. On a 65% loan-to-value deal, a 7.7% hit to value is a 25.5% hit to invested equity.
A deal is not the IRR on the cover of the memo. A deal is the range of IRRs the assumptions can produce, and sensitivity analysis is the only step that shows the range.
The second reason is discipline. Sensitivity output separates deals that work because the asset works from deals that work because the model assumed the market cooperates. If a transaction only clears its hurdle when the exit cap is tighter than the going-in cap, the underwriter is forecasting cap compression, not underwriting an asset.
Example
Sensitivity analysis is easiest to see on one deal with stated inputs. A $10,000,000 asset is acquired at a 6.00% going-in cap on $600,000 of NOI, funded with $6,500,000 of interest-only debt at 6.00% and $3,500,000 of equity. NOI grows 3% annually. The hold is five years, and the exit is priced off forward NOI of $695,564.
Annual cash flow to equity is NOI less $390,000 of debt service, totaling $1,235,481 over the five years. Only the exit cap is flexed. Everything else is held constant.
Exit cap | Sale price | Net proceeds after debt | Equity multiple | IRR |
|---|---|---|---|---|
5.75% | $12,096,773 | $5,596,773 | 1.95x | 15.7% |
6.00% | $11,592,741 | $5,092,741 | 1.81x | 13.8% |
6.25% | $11,129,031 | $4,629,031 | 1.68x | 12.0% |
6.50% | $10,700,991 | $4,200,991 | 1.55x | 10.2% |
7.00% | $9,936,635 | $3,436,635 | 1.33x | 6.7% |
The table reads in three numbers. First, the underwritten case is 13.8%. Second, each 25bps of cap widening costs roughly 180bps of IRR, so the deal is highly cap-sensitive. Third, the breakeven exit cap, the cap at which the deal returns exactly 100% of equity, is 7.94%, which sits 194bps above the underwritten exit. That gap is the margin of safety, and it is the single most useful figure the grid produces.
Variations and Edge Cases
Sensitivity analysis is a family of techniques, not one procedure, and the variant chosen changes what the output means. A one-way table flexes a single input. A two-way table flexes two and produces a grid. A tornado chart ranks every input by its effect on the output. Each answers a different question about the same model.
Variant | What it does | When to use it |
|---|---|---|
One-way table | Flexes a single input across a range | Isolating the effect of one assumption |
Two-way table | Flexes two inputs on perpendicular axes | Testing correlated inputs such as exit cap and rent growth |
Tornado chart | Ranks all inputs by output impact | Finding which assumptions deserve diligence |
Breakeven solve | Solves for the input value at a target output | Quantifying margin of safety |
Joint stress | Moves two inputs adversely at once | Modeling a recession, where caps widen as NOI softens |
Two edge cases distort the output. The first is correlated inputs treated as independent: exit caps widen in the same conditions that soften rents, so a one-way cap table understates real downside. The concurrent case, exit cap +50bps while NOI underperforms by 5%, is the test most operator models skip. The second is flexing an assumption that does not drive the answer. A grid on a variable the model barely responds to produces a reassuring table and no information.
Sensitivity Analysis vs Scenario Analysis
Sensitivity analysis is often confused with scenario analysis. Sensitivity analysis changes one variable at a time and holds the rest constant, isolating that variable's effect. Scenario analysis changes a coherent set of variables together to describe a state of the world, such as a recession in which rents fall, caps widen, and lease-up slows at once.
The distinction is what each is for. Sensitivity analysis pinpoints specific risks and isolates individual variable changes, while scenario analysis examines how multiple interconnected variables affect outcomes together. Sensitivity tells you which assumption to argue about. Scenario tells you what happens when the argument is settled badly on several fronts at once.
Sensitivity analysis | Scenario analysis | |
|---|---|---|
Variables moved | One, or two on a grid | A correlated set |
Question answered | How much does this input matter | What happens in this state of the world |
Output | A table or grid of returns | A small number of named cases |
Primary use | Ranking risk, finding breakevens | Stress testing, committee presentation |
Frequently Asked Questions
What variables should a real estate sensitivity analysis test? The standard set is exit cap rate, rent growth, interest rate, and lease-up or vacancy timing, because each is forecast rather than observed and each compounds across the hold. Exit cap rate is usually tested first, since value is NOI divided by a cap rate and the entire valuation move lands on equity.
How wide should the sensitivity range be? Cap-rate tables conventionally run from -50bps to +100bps in 25bps increments, deliberately asymmetric toward the downside. The range should be wide enough to contain the breakeven value; if the grid never reaches the point where the deal fails, it has not been tested.
What is the difference between sensitivity analysis and stress testing? Sensitivity analysis maps how an output responds across a range of input values, including favorable ones. Stress testing applies a specific adverse case and asks whether the deal survives it. Sensitivity produces the map; a stress test picks one point on it and checks the outcome.
Related Terms
Internal Rate of Return