A Strategic Guide for Sponsors, Developers, CRE Brokers, Mortgage Brokers, Credit Officers and Bankers.

The exit debt yield, not the as-stabilized LTV, sets the facility, and it is the test sponsors leave out of the model.

Reading time: 10 minutes

Two numbers decide a bridge loan, and most sponsor models contain only one. The model has the as-stabilized value: net operating income at completion divided by an exit cap rate, times the lender’s advance rate. The number usually missing is the debt yield the takeout lender will require on that same NOI. When the second produces a smaller loan, the facility is sized by the takeout, not by the appraisal.

The 10-year Treasury was 5.10% on 15 October 2026 (Federal Reserve H.15 via FRED). In CBRE’s H1 2026 Cap Rate Survey, published 12 August 2026 when the 10-year sat near 4.6%, the median response put the level needed to lift sales volume at 3.75%.

By the end of this issue you will be able to run the four tests a bridge lender applies, identify which one binds your deal, and calculate the equity it requires before you sign a purchase contract.

1. Four tests, applied in order, and only one of them binds

Purpose: Establish the sizing hierarchy so the reader stops quoting a single advance rate.

A bridge lender does not size to one ratio. It runs four and funds the lowest result. The first governs what funds at closing; the other three cap the whole facility, including future funding.

  • Test 1 — As-is LTV. An advance rate against the lower of as-is appraised value and purchase price. A day-one funding constraint.
  • Test 2 — Loan to total cost. An advance rate against purchase price plus hard capex.
  • Test 3 — Loan to as-stabilized value. An advance rate against the appraiser’s as-stabilized conclusion.
  • Test 4 — Exit, or takeout, test. Stabilized NOI divided by the takeout lender’s required debt yield.

Sponsors quote Test 2 or Test 3 because those produce the biggest number. Credit committees calculate all four.

One caveat before the arithmetic. No public source publishes bridge advance rates or takeout debt yield floors. Every percentage below is an illustrative parameter chosen to show how the tests interact, not a quoted or surveyed term. Substitute your own term sheet’s numbers.

If you cannot name which of the four tests binds your deal, you do not yet know your loan amount.

2. As-stabilized value is a conditioned conclusion, not a number

Purpose: Explain why the appraisal figure a sponsor celebrates is not the figure credit uses.

An as-stabilized value is issued subject to an extraordinary assumption, a hypothetical condition, or both: that the renovation is complete, that the rent roll reaches the assumed level, and that the market supports the exit cap rate selected. The report says so on its face, which is why credit haircuts the conclusion rather than lending against it.

The exit cap rate does most of the damage. On $2,100,000 of stabilized NOI, a 5.25% cap produces $40,000,000 of value and a 5.75% cap produces $36,521,739. Fifty basis points of appraiser judgment moves value by $3,478,261 and, at a 65% advance, the loan by $2,260,870.

CBRE’s H1 2026 survey, published 12 August 2026 from 3,600 estimates across more than 50 markets, found roughly 60% of responses indicating no cap rate change over the following six months, but a larger share than in December 2025 expecting a rise, with infill multifamily the most bearish subtype.

Ask for the exit cap rate the appraiser used before you ask for the value, because the value is a quotient and you already know the numerator.

3. The exit test is a debt yield, and that is deliberate

Purpose: Show why the binding constraint is expressed in a ratio that ignores both cap rates and interest rates.

Debt yield contains no cap rate and no interest rate, which is exactly why a bridge lender uses it to test its own repayment. An LTV test asks the appraiser a question. A debt yield test asks the property a question.

The relationship between the two is arithmetic: implied LTV equals cap rate divided by debt yield. At a 9.50% takeout debt yield and a 5.75% exit cap, the implied LTV is 60.5%. At the same debt yield and the sponsor’s 5.25% cap, it is 55.3%. A sponsor modeling 65% leverage against a stabilized value has already failed a test nobody showed him.

Coverage follows from the same number. On a 30-year amortizing takeout at 6.50%, the mortgage constant is 7.58%, so a 9.50% debt yield implies 1.25x coverage. Illustrative inputs, but they show the mechanism: a debt yield floor is a coverage requirement restated so it does not move when the appraiser moves.

A specialist will object that debt funds quote 80% loan-to-cost, so none of this binds. The LTC quote caps the facility; it does not override a lower exit test when return on cost is thin. And 80% LTC is a total-facility number sitting on a far lower initial advance.

Run your own debt yield at the loan amount you are requesting. If it starts with an 8, expect the facility to come back smaller.

4. The initial advance and the facility are different numbers

Purpose: Separate what funds at closing from what the term sheet headline says.

The headline loan amount on a bridge term sheet is the committed facility. What wires at closing is Test 1 applied to the purchase price. The difference is a future funding holdback: not money sitting in escrow, but an unfunded commitment the lender advances only when conditions are met.

Two consequences. The closing equity check is larger than the model shows, and capex competes with the interest reserve for the same holdback dollars.

Build your sources and uses with two columns, day one and future funding, or you will discover the gap at the closing table.

5. The release test looks backward, and the “and” matters

Purpose: Give the reader the actual mechanics of getting holdback money released.

Holdback release turns on a stabilization definition in the loan agreement, and that definition is conjunctive more often than sponsors expect. Take a formulation of the kind these agreements use: occupancy at or above 90% for 90 consecutive days and coverage at or above a stated level on a hypothetical takeout constant. Hitting one test releases nothing.

Two details decide outcomes. The coverage test uses trailing three-month NOI annualized, not pro forma, so a lease-up finishing in month 20 qualifies in month 23. And most agreements let the lender order a new appraisal at the sponsor’s cost before the final tranche, reopening the exit cap rate question at the worst moment.

Negotiate the stabilization definition at term sheet, not at loan documents. By documents, credit has already approved the language.

6. The interest reserve is funded by the facility it is sizing

Purpose: Expose the circularity that breaks renovation budgets.

An interest reserve is not a separate pot. It is part of the loan, it accrues interest itself, and it consumes the same holdback the capex budget needs. Lenders size it to the projected shortfall between debt service and property cash flow, so the reserve is only as good as the absorption schedule underneath it.

Priced at an index plus 350 basis points, with overnight SOFR at 3.89% on 15 October 2026 (Federal Reserve Bank of New York via FRED), an illustrative coupon is 7.39%. Note the substitution: a bridge loan references 1-month Term SOFR, published under license, and I could not verify its current level from a free primary source. Overnight SOFR stands in as a proxy, nothing more.

The direction of the risk is verified. The FOMC raised its target range to 3.75% to 4.00% on 16 September 2026, and most participants assessed that another increase would likely be appropriate by year end (FOMC minutes, released 7 October 2026). A reserve sized on a spot index assumes a path the Committee has signaled against.

Most agreements also carry a replenishment covenant: if the lender projects the reserve running dry, the sponsor funds the shortfall out of pocket within a stated period. Failure to do so is often a springing recourse carve-out, converting a non-recourse loan into a personal one.

Read the replenishment clause and the carve-out schedule together. They are usually in different sections and they are the same risk.

7. Where it goes wrong, ranked by frequency

Purpose: Give the failure modes in the order a credit officer actually sees them.

  1. The model shows the facility as day-one proceeds. The commonest error. It costs the sponsor the full holdback in extra closing equity, $3,205,000 in the example below.
  2. The sponsor’s exit cap rate is below the appraiser’s. Fifty basis points on $2.1M of NOI is $3.5M of value and $2.3M of loan at 65%.
  3. No debt yield test was run. The deal clears every LTV and LTC test and fails the only one that governs repayment.
  4. The interest reserve was sized on an absorption schedule that slipped. A three-month delay on the $22,105,000 facility below adds roughly $408,000 of interest at 7.39%, and the reserve has no slack.
  5. The release test was read as “or.” Occupancy arrives, coverage does not, and the final tranche sits unfunded through the last months of capex.
  6. The extension option was treated as automatic. Extensions turn on the same coverage test that is failing, plus a fee.

Each of these is discoverable in week one of underwriting. Most are discovered in month six.

8. Current conditions as of 17 October 2026

Purpose: State what is true right now, with every figure dated.

The index is rising, not falling. The FOMC raised the target range a quarter point to 3.75% to 4.00% on 16 September 2026, 12–0, and said inflation remains elevated (Federal Reserve Board, FOMC statement, 16 September 2026). Overnight SOFR printed 3.89% on 15 October 2026 (FRBNY via FRED). The 10-year Treasury was 5.10% on 15 October, down from 5.24% on 9 October (Federal Reserve H.15 via FRED).

Bank CRE standards are easing at the margin from a tight level. In the Federal Reserve’s July 2026 Senior Loan Officer Opinion Survey, released 3 August 2026 and covering the second quarter, moderate and modest net shares of banks eased standards on nonfarm nonresidential and multifamily loans respectively, while construction and land development standards were basically unchanged and a moderate net share reported weaker demand for them. Its special questions still placed construction and land development standards at the tighter end of their range relative to the midpoint since 2005. That survey remains the current one.

The distress is concentrated at maturity, not in current pay. Trepp’s CMBS delinquency rate rose to 8.02% in September 2026, its highest since late 2020, and office reached 12.2% (Trepp, reported by Scotsman Guide, 6 October 2026, which puts office at 12.16%; Commercial Real Estate Direct, 2 October 2026; and Colliers, 14 October 2026). In August, non-performing matured balloon loans accounted for 81% of newly delinquent balances (Trepp, reported by MBA NewsLink, 8 September 2026, and Multi-Housing News, 28 September 2026). The MBA put $875 billion, 17% of $5.0 trillion outstanding, as scheduled to mature in 2026 (MBA, 9 February 2026).

Loans are not failing because borrowers stopped paying. They are failing because the takeout did not appear, which is the same risk a bridge loan asks you to underwrite today.

9. Case study: the equity gap, calculated

Purpose: Show the arithmetic end to end on a worked example.

This is an illustrative composite, not a transaction. Every figure is an assumption, and the arithmetic is shown so you can substitute your own.

A 180-unit garden apartment property. Purchase price $27,000,000, or $150,000 per unit. As-is NOI $1,485,000, a 5.50% going-in cap. Hard capex $3,600,000, or $20,000 per unit. Total cost before financing $30,600,000. Projected stabilized NOI $2,100,000, a 6.86% return on cost. The sponsor’s model uses a 5.25% exit cap for a $40,000,000 stabilized value; the appraiser concludes 5.75%, or $36,521,739.

The four tests:

  • Test 1, as-is LTV at 70%: $27,000,000 × 0.70 = $18,900,000 (initial advance)
  • Test 2, loan to cost at 75%: $30,600,000 × 0.75 = $22,950,000
  • Test 3, as-stabilized LTV at 65%: $36,521,739 × 0.65 = $23,739,130
  • Test 4, takeout debt yield at 9.50%: $2,100,000 ÷ 0.095 = $22,105,263

Test 4 binds. The facility is $22,105,000: an initial advance of $18,900,000 and a holdback of $3,205,000. Post-closing the deal still needs capex of $3,600,000 plus a $1,900,000 interest reserve, or $5,500,000 against that $3,205,000 holdback.

Total uses are $33,100,000: price $27,000,000, capex $3,600,000, interest reserve $1,900,000, closing costs and fees $600,000. Equity at the binding test is $33,100,000 − $22,105,000 = $10,995,000, split $8,700,000 at closing and $2,295,000 during the renovation.

The sponsor’s model showed 65% of $40,000,000, a $26,000,000 loan and $7,100,000 of equity. The gap is $3,895,000, a 54.9% increase in the check. The sponsor’s loan implies a debt yield of $2,100,000 ÷ $26,000,000 = 8.08%, below the 9.50% floor.

The deal did not get worse. It was never the deal in the model.

10. What to do about it

Purpose: Convert the analysis into sequence.

  1. Calculate your debt yield at the loan amount you want, before you ask for a quote. Stabilized NOI divided by requested loan. Below 9%, reprice the equity.
  2. Ask every lender for all four tests in writing, with the advance rate and the dollar amount for each. A term sheet quoting one ratio is quoting the one that flatters.
  3. Get the appraiser’s exit cap rate and the comparable sales supporting it, and model the gap to your own in dollars.
  4. Build sources and uses in two columns, day one and future funding, and reconcile the total to the facility.
  5. Stress the interest reserve for a three-month absorption delay and a 100 basis point index move, then check whether the replenishment covenant triggers.
  6. Read the stabilization definition. Confirm whether it is conjunctive and which NOI period the coverage test uses.
  7. Negotiate that definition, the release mechanics and the extension conditions at term sheet stage, not at loan documents.

Do the first step this week. It takes one division, and it tells you whether the rest of the list matters on your deal.

Exhibit 1 is the four-test sizing table. Exhibit 2 is the worked equity bridge from the case study. Both carry their assumptions on the face.

Julio Ortiz, CVA is Founder and Executive Director of Prime Wealth Development, a Miami-based corporate finance advisory firm and commercial finance intermediary. PWD is not a lender.

This newsletter is general information, not legal, tax, accounting or investment advice, and does not create an advisory relationship. Figures are as of the dates stated.

www.primewealthdevelopment.com | jortiz@primewealthdevelopment.com

Sources