12 Mistakes That Break Deal Models
The most common underwriting mistakes that break real estate deal models, plus practical controls to improve decision quality and downside protection.
Introduction
Most bad underwriting outcomes are process failures, not formula failures. Teams usually have the right model architecture, but assumptions are stale, controls are inconsistent, and downside scenarios are too soft.
TL;DR: Deal models break when assumptions are not versioned, downside logic is weak, and financing constraints are treated as secondary. With policy still restrictive in early 2026 (Federal Reserve, 2026), long rates above prior-cycle norms (FRED DGS10, 2026), and refinancing pressure still relevant after major maturities (MBA, 2025), modeling discipline matters more than ever.
Mistake one to three: weak assumption discipline
The first cluster of mistakes is assumption drift. Teams often update one or two headline inputs and leave linked assumptions untouched.
Three common failures:
- Debt pricing is stale by the time a deal reaches committee.
- Rent and lease-up assumptions are not tied to current market evidence.
- Expense assumptions do not reflect current inflation and local operating pressure.
CPI moderated in recent data releases (BLS CPI, 2026), but property-level costs can still diverge from headline inflation. If your model does not account for that spread, downside coverage is overstated.
Mistake four to six: incomplete downside testing
A single-variable downside case creates false confidence. Real stress tends to hit multiple lines at once: effective rent, occupancy, operating costs, and refinancing terms.
Three more failures: 4. Downside scenarios only adjust revenue, not costs and debt terms. 5. Severe cases are not severe enough to challenge structure. 6. Refinance tests are treated as optional.
A robust downside package should connect operating stress and debt stress in the same run. That is especially important when benchmark yields remain elevated (FRED DGS10, 2026).
Mistake seven to nine: bad decision governance
Governance errors can make a technically correct model unusable. If decision rules are vague, teams can produce different recommendations from identical data.
Three governance failures: 7. Override decisions are undocumented. 8. Threshold exceptions are frequent but never reviewed. 9. Final memos hide assumption changes made late in the process.
These are preventable. Committees should require a short exception log with each recommendation and force re-approval when assumptions shift materially.
Mistake ten to twelve: portfolio-blind underwriting
Deal-level metrics can look strong while portfolio risk worsens. A good model must show interaction with existing concentration, maturity clustering, and liquidity needs.
Three portfolio failures: 10. No portfolio concentration check before approval. 11. No maturity-stack check on refinance timing. 12. No liquidity reserve check for downside execution.
The maturity context from recent cycles shows why this matters (MBA, 2025). A deal can be acceptable in isolation and risky in aggregate.
What controls fix these twelve mistakes fastest
The fastest improvements come from repeatable controls, not extra model complexity. If your process is inconsistent, adding more tabs usually makes mistakes harder to detect.
Apply this control stack:
- Lock assumptions and require source support for changes.
- Run base, downside, and severe cases every time.
- Include a refinance viability module in every memo.
- Log every override and review exceptions monthly.
- Add portfolio context checks before final approval.
If you already use the DSCR underwriting guide, tie these controls to that workflow and update debt assumptions with current policy context from the Fed (Federal Reserve, 2026).
How to build a model review workflow that scales
Model quality improves when review cadence is structured. Most teams should run a two-level review: analyst self-review first, then independent deal-team challenge before committee.
A scalable workflow:
- Analyst checklist before submission.
- Independent downside challenge review.
- Investment committee pass/fail memo.
- Post-close backtest of key assumptions.
Use portfolio reporting templates LPs actually read to keep definitions consistent between underwriting and investor reporting while using public macro context for baseline calibration (U.S. Census Bureau, 2026).
How to know whether your underwriting quality is improving
You need objective indicators, not sentiment. Quality programs fail when teams cannot show whether controls are working.
Track these indicators monthly:
- frequency of late assumption changes,
- count of override decisions,
- downside variance versus actual operations,
- refinance outcomes versus underwritten expectations.
When those indicators improve over time, model reliability is improving. When they worsen, tighten thresholds and retrain assumptions policy.
How to run a pre-committee model integrity check
Before investment committee, run a short integrity check that focuses on model behavior rather than presentation quality. This catches fragile logic before it becomes a capital decision.
A practical integrity check should verify:
- all scenario tabs reconcile to the same baseline assumptions,
- downside cases include linked revenue, expense, and financing stress,
- refinance module reflects current term expectations,
- override decisions are logged with owner and rationale,
- portfolio concentration checks are included in the recommendation.
Teams that use this check regularly usually reduce late-stage rework and improve committee confidence. It also creates cleaner post-close analysis because the decision logic is visible from day one.
Use cap rate, debt yield, and exit cap stress test during this check so valuation and financing risk are reviewed together in the same workflow, while using current market-rate context from public data (FRED DGS10, 2026).
How to train teams to avoid repeated model errors
Most repeated mistakes are training failures, not intelligence failures. Analysts often inherit legacy models and copy assumptions without seeing why those assumptions were chosen.
A stronger training loop includes:
- quarterly case reviews of both accepted and rejected deals,
- error taxonomy tracking by root cause,
- scenario-design exercises using current market conditions,
- policy refresh sessions tied to debt and liquidity shifts.
This approach improves judgment and consistency at the same time. When analysts understand how assumptions connect to decision thresholds, model quality improves and override frequency usually declines.
The important point is repetition. One workshop does not change model behavior. A recurring workflow with documented lessons and updated policy language is what turns underwriting from person-dependent craft into a scalable team capability.
Frequently Asked Questions
Are these mistakes mostly technical or organizational?
Mostly organizational. Technical model errors happen, but repeated underwriting misses usually come from weak process controls, inconsistent assumptions, and unclear decision governance.
Should small teams use the same controls as larger platforms?
Yes, but at simpler scale. Use short checklists, fixed decision rules, and explicit exception logs. Consistency matters more than complexity.
How often should a team update underwriting policy?
At least semiannually, and sooner when financing conditions shift materially. Policy should adapt to market structure, not remain frozen out of convenience.
What is the highest-impact fix to start with?
Standardize downside testing and require documented exception decisions. Those two changes usually improve decision quality quickly and reduce avoidable model variance.
Conclusion
Deal models break when teams let assumptions drift and treat governance as optional. Fixing that does not require new software. It requires repeatable controls, disciplined downside testing, and portfolio-aware decisions. Teams that enforce this process make fewer preventable errors and protect capital more effectively.
Sources
Related Resources
DSCR Loan Underwriting by Asset Type
How to underwrite DSCR loans by asset type in 2026 with tighter assumptions, refinance stress tests, and clear pass/fail controls.
Office-to-Residential Conversion Underwriting
A practical underwriting framework for office-to-residential conversions in 2026, with execution, financing, and lease-up risk controls.
Cap Rate, Debt Yield, and Exit Cap Stress Test
How to stress-test cap rates, debt yield, and exit assumptions in 2026 with a practical model framework grounded in current rates and liquidity conditions.
Get Real Estate Insights
Join other investors receiving actionable strategies and market analysis