Most firms run every deal through a commercial real estate underwriting template that someone built years ago and nobody has opened with a critical eye since. That file is not a neutral container. Every hard-coded default inside it is a silent opinion: a vacancy floor, a management fee percentage, a reserve per unit, an expense escalator, an exit cap spread. Those numbers were calibrated to a market that no longer exists, and they set the answer before anyone reads the offering memorandum. The template does not measure the deal. It votes on it.
Key Takeaways
An underwriting template is not a container, it is an opinion. Every constant buried in a formula is an assumption someone made in a different market and never revisited.
The most dangerous assumptions are the ones nobody types. A default that never appears on an input tab never gets challenged in an investment committee.
Two stale defaults, a vacancy floor and a management fee, compounded and capitalized, can move an underwritten value by more than 7% before anyone disagrees about the deal.
Separating inputs, calculations, and outputs is not spreadsheet hygiene. It is what makes assumptions visible, and visibility is what makes them auditable.
Every assumption cell should carry a value, a date, and a source. An assumption without a date has no expiration.
What Belongs in a Commercial Real Estate Underwriting Template?
A commercial real estate underwriting template should hold six things in strict order: a single assumptions tab, a rent roll and lease schedule, a normalized historical operating statement, a projected cash flow, a debt and capital stack, and a returns output with a sensitivity grid. Nothing else. Every number should live in exactly one of those layers.
The order matters more than the contents. Information flows one direction: inputs feed calculations, calculations feed outputs, outputs never feed back. A model where an output cell holds a typed override is not a model. It is a picture of a conclusion.
Section | What it holds | The assumption hiding in it |
|---|---|---|
Assumptions | Every rate, fee, spread, growth factor, reserve, labeled | That the values were set for this market, not the last one |
Rent roll and leases | Unit-level rent, term, escalations, options, recoveries | That in-place rent is market rent |
Historical operating statement | Trailing 12 and trailing 3, normalized | That the seller's expense line survives closing |
Projected cash flow | Year-by-year revenue, expenses, net operating income | That the expense escalator matches real cost inflation |
Debt and capital stack | Loan sizing, rate, amortization, tests, equity structure | That today's proceeds and rate hold to closing |
Returns and sensitivity | Unlevered and levered return, exit value, downside grid | That the base case is expected, not optimistic |
The right-hand column is the point of the exercise. A tab list is a filing system. The opinion sits in the constants.
Why Do Hard-Coded Defaults Set the Answer Before Anyone Looks at the Deal?
Hard-coded defaults set the answer because they never enter the conversation. An underwriter argues about rent growth because it appears on a tab with a label. Nobody argues about the 2.5% management fee sitting inside a formula on the cash flow tab, visible only by clicking the cell. Invisible assumptions are unfalsifiable assumptions.
The distortion compounds. A stale revenue-side default inflates effective gross income, which shifts every percentage-driven expense calculated against it, and the resulting overstatement of net operating income gets capitalized into a valuation error many multiples of its own size.
A worked example: two stale defaults, one valuation gap
Take a 200-unit multifamily asset with gross potential rent of $3,600,000. Two defaults come inherited.
The first is a vacancy and credit loss floor hard-coded at 5%. Submarket data supports 8%. That gap is 3% of $3,600,000, or $108,000 of overstated effective gross income.
The second is a management fee hard-coded at 2.5% of effective gross income. Third-party contracts for assets of this size typically run in the range of 3.0% to 3.5%, so call the audited figure 3.25%. Under the template, effective gross income is $3,420,000 and the fee is $85,500. Under audited inputs, it is $3,312,000 and the fee is $107,640, understated by $22,140.
Combined, net operating income is overstated by $130,140. Capitalized at a 5.75% exit cap, that is $2,263,304 of value. If audited net operating income is $1,800,000, the honest value is $31,304,348 and the template value is $33,567,652.
Line | Template default | Audited input | Effect on NOI |
|---|---|---|---|
Vacancy and credit loss | 5.0% | 8.0% | ($108,000) |
Management fee | 2.5% of EGI | 3.25% of EGI | ($22,140) |
Combined | ($130,140) | ||
Value at 5.75% cap | $33,567,652 | $31,304,348 | ($2,263,304) |
The gap is 7.2% of value. No one lied, disagreed, or made an arithmetic error. Two numbers nobody typed produced a seven-figure difference in what the asset is worth. That is the case for the audit.
How Should Inputs, Calculations, and Outputs Be Separated in an Underwriting Model?
Separation means one rule: an assumption appears exactly once, on the assumptions tab, as a labeled cell, and every formula references that cell rather than restating its value. Calculation tabs contain no typed numbers. Output tabs create no new information. The discipline is mechanical, and the payoff is a model readable by someone who did not build it.
Two tests reveal whether a model observes the rule. Count the typed constants outside the assumptions tab: in a clean model that count is close to zero. Then hand the file to someone who has never seen it and ask them to list the ten assumptions driving the return. If they cannot do it in fifteen minutes, the assumptions are not visible, whatever the tab structure claims. A model that cannot be read by a stranger cannot be reviewed by an investment committee, only ratified by one.
The second discipline is metadata. Every assumption cell carries a value, a date it was last set, and a source. A 3.25% management fee sourced to two signed contracts in the same submarket in March 2026 is an assumption. The same 3.25% with no date and no source is a habit.
How Do You Audit the Default Assumptions in an Underwriting Template?
Run the audit on a schedule, not when a deal breaks. Once a year, open the template with no live deal in it, list every constant, and force each one to be re-justified against current evidence or removed. Survivors get a fresh date stamp. The rest get replaced or converted into a deal-level input the underwriter must supply.
The case for an annual cadence is that the conditions surrounding these defaults move fast. CBRE's H1 2026 U.S. Cap Rate Survey, built on roughly 3,600 cap rate estimates across more than 50 U.S. markets from over 200 professionals, reported that the 10-year Treasury fell below 4% in late February 2026, peaked at 4.67% in mid-May, and sat near 4.6% by mid-July. Respondents put the median 10-year yield needed to trigger a notable uptick in sales volume at 3.75%. A template calibrated before that swing describes a financing market that was not available in any month of 2026.
Default | Audit question | Acceptable evidence |
|---|---|---|
Vacancy floor | Does this match current submarket performance for this class? | Submarket data for the specific class and vintage |
Management fee | What do current contracts charge for this asset size? | Signed contracts or executed proposals, dated |
Replacement reserve | Does the per-unit figure reflect the asset's age and condition? | Property condition assessment or dated capital history |
Expense escalator | Does this match trailing inflation for these line items? | Trailing operating statements plus published cost indices |
Exit cap spread | Does the spread reflect the asset's age at exit? | Transaction comparables and published cap rate surveys |
Debt terms | Are these proceeds and this rate available today? | Current lender quotes, dated |
The output of the audit is not a better template. It is a shorter one. Files accumulate defaults because a default is faster than a decision, and that speed buys nothing that survives closing. The same discipline applies downstream, to net operating income, the exit cap rate assumption, and the sensitivity grid.
A fast screening template needs this audit more than a full model does. It runs on fewer inputs, touches more deals, and gets less scrutiny per pass, so a stale default there kills good deals quietly.
Frequently Asked Questions
What should be on the assumptions tab of a CRE underwriting model? Every rate, growth factor, fee, reserve, spread, and timing input, each as a single labeled cell with a value, a date, and a source. If a number appears in a formula elsewhere in the file, it belongs on the assumptions tab instead.
How often should an underwriting template be audited? At least annually, run with no live deal open, plus after any material shift in financing conditions. Every constant is re-justified against dated evidence or removed.
Why separate inputs from calculations in an underwriting spreadsheet? Because an assumption buried inside a formula cannot be challenged by anyone who did not build the model. Separation makes the opinion visible, and a visible opinion is one a committee can debate before capital is committed.
Conclusion
A commercial real estate underwriting template is an argument about how deals work, written once and applied silently to every deal that follows. The arithmetic is reliable. The constants are not: they do not age gracefully, and nobody notices when they stop being true. A vacancy floor and a management fee, both inherited, both invisible, produced a 7.2% valuation gap without a single error in the math.
The remedy is unglamorous and cheap. Pull every hard-coded number onto one tab. Give each a date and a source. Once a year, open the file with no deal in it and make each assumption earn its place. Firms that do this underwrite the deal in front of them. Firms that do not underwrite a market that ended some time ago, and keep getting an answer that feels right.