This is the sequel to our Newark bridge piece. That post walked a real $5.55M bridge closing and what it says about where the market is. This one is about what happens when your deal comes due — and the math behind why the agency takeout you penciled in 2022 doesn't hand you back the loan you're carrying.

Picture a sponsor we'll call composite, because that's what this is — a stand-in built from the kind of deal we see weekly, not a real borrower. In 2022 they bought a 40-unit property in a secondary Southeast market, value-add, below-market rents, a clear plan. They financed it on a $4.0 million floating-rate bridge, interest-only, two-to-three-year term, with the exit written on the first page of the model: stabilize the rent roll, then refinance into a cheap agency permanent loan. The business plan worked. The rents are where they projected. Stabilized NOI is $360,000 a year. The property did exactly what it was supposed to do.

And the takeout still doesn't give them their $4.0 million back.

That's the whole story, and it's not a story about a bad deal. It's a story about a rate environment that resets the takeout math on a good one. Below, we trace it all the way through — with verified program parameters and a current takeout rate — so you can run your own.

Why small-balance is its own story

Everyone is writing the maturity-wall piece right now. Almost all of it is about the $50 million CRE-CLO tranche, the rate-cap that expired on an institutional floater, the big-shop research desk's view of 2026 refinancing volume. The numbers are real: roughly $162.1 billion in multifamily loans mature in 2026 and about $167.7 billion in 2027, up from roughly $104.1 billion in 2025 (MMG Real Estate Advisors). That's a two-year peak.

What almost nobody is writing is the version of this story for the operator holding one to a handful of $1M–$5M value-add deals. You get the identical rate reset the institutional sponsor gets. You get tighter coverage tests. But you get a fraction of the advisory coverage, because the correspondent shops that quote the $30 million agency deal for free aren't focused on your loan size, and the research operations aren't writing the math for your tier. The squeeze is the same. The support is thinner. That gap — in coverage, not just in proceeds — is the one we built Blue Sky to sit in.

The takeout product — Freddie SBL, plainly

The natural takeout for a stabilized small-balance multifamily deal is the Freddie Mac Small Balance Loan, with the Fannie Mae Small Loan program as the parallel option. You know these programs; here are the parameters that actually govern the deal, confirmed against Freddie's own SBL materials and current lender sheets:

  • Loan size: $1M–$7.5M, capped at $6M in smaller markets. 5–50 units. 5/7/10-year fixed and hybrid ARM. Up to 30-year amortization.
  • LTV and DSCR by market tier:
    • Top markets — 1.20x DSCR, up to 80% LTV
    • Standard markets — 1.25x, up to 80%
    • Small markets — 1.30x, 75% LTV on acquisition, 70% on refinance
    • Very small markets — 1.40x
  • Structure: non-recourse with standard bad-boy carve-outs; carve-out waiver available at 65% LTV or lower and 1.40x DSCR or higher; sponsor net worth around 100% of the loan amount with roughly 10% liquidity; 90% physical occupancy for the trailing 90 days.
  • Indicative rate: SBL fixed starting around 5.93% earlier this year. We'll use 6.0% as the takeout rate below.

The single most important line there is the tier table. The market your property sits in flips both your LTV ceiling and your coverage floor — and as you're about to see, the coverage floor is what does the damage. If you don't know your tier before you model, you're guessing at your proceeds.

The worked example

Everything here is illustrative. The program parameters and the rate are verified; the operating inputs are a realistic composite, and the deal is hypothetical. The driving assumption is stabilized NOI — every other number derives from it.

The setup. A 40-unit value-add deal in a secondary Southeast market that prices to Freddie's "Small" tier. Value-add plan complete. Stabilized NOI: $360,000 per year. Existing debt: a $4.0M floating-rate bridge, 2022 vintage, interest-only, so the payoff is approximately $4.0M.

The takeout is a Freddie SBL refinance in a small-tier market: max LTV 70%, minimum DSCR 1.30x, 6.0% fixed, 30-year amortization. Two constraints. The lower of the two sets your loan.

Constraint 1 — the DSCR cap.

  • Maximum annual debt service the NOI supports = $360,000 ÷ 1.30 = $276,923
  • At 6.0% over a 30-year amortization, the annual mortgage constant is about 0.0719 (roughly $71,900 of P&I per year per $1M borrowed)
  • Maximum loan = $276,923 ÷ 0.0719 ≈ $3.84M
  • Check: debt service at $3.84M = $3.84M × 0.0719 ≈ $276,300, so DSCR = $360,000 ÷ $276,300 = 1.30x ✓

Constraint 2 — the LTV cap.

  • At a roughly 6.0% cap rate, value ≈ $360,000 ÷ 0.06 = $6.0M
  • A 70% LTV refinance = $4.2M

The result. The DSCR cap of $3.84M binds below the LTV cap of $4.2M. The debt-service test, not the leverage test, sets the loan.

The gap. $4.0M payoff − $3.84M of takeout proceeds = a roughly $160,000 principal shortfall — and that's before transaction costs. A real SBL closing also carries origination (on the order of 1% of the loan), third-party reports (appraisal, PCA, Phase I), and legal. Those vary by deal and lender; treat them as additional to the $160,000, not folded into it. The point isn't the exact closing number. The point is that a deal that did everything right still comes up short on principal.

The insight: DSCR binds before LTV

Look at the same deal through debt yield — NOI divided by loan amount — and the reason jumps out.

  • At the DSCR-constrained loan of $3.84M, debt yield = $360,000 ÷ $3.84M = 9.4%
  • At the max-LTV loan of $4.2M, debt yield = $360,000 ÷ $4.2M = 8.6%

The lender isn't willing to go to the 8.6% debt yield that the LTV line would allow, because at 6.0% money over a 30-year amortization, holding the loan to 1.30x coverage requires the higher 9.4% debt yield. In a 4% world, the LTV line bound first and proceeds were generous. In a 6% world, coverage binds first and proceeds shrink — even when value and NOI are exactly what you projected.

This is the part worth internalizing: at today's rates, your takeout proceeds are a function of your NOI and the coverage requirement, not your LTV. Sponsors who model to 70% LTV and assume that's their loan are modeling the wrong constraint. The cap rate could compress and the LTV-implied loan could climb, and it wouldn't matter — the DSCR test would still hand you $3.84M.

What to actually do

The gap is solvable, but only with lead time. If your bridge matures in the next 18 months, here's the work.

Start 6–9 months ahead. Agency takeouts aren't 30-day closings, and the worst position is sourcing your permanent loan while your bridge lender is asking about the maturity. The exit should drive the bridge, not the other way around.

Know your tier before you model. It's the difference between a 70% LTV / 1.30x test and an 80% / 1.25x test — a different loan entirely. Confirm where your specific market sits in the SBL tiering before you build a single proceeds number.

Model the gap early, with the DSCR test as the binding one. Run the coverage-constrained loan first, the LTV-constrained loan second, and take the lower. Then size the shortfall — principal plus closing costs — so you know the number before a lender tells you.

Line up the resolution. The shortfall gets covered one of three ways: fresh equity into the refinance; accepting a lower-leverage takeout and bringing cash to close; or gap capital — preferred equity or mezzanine — on top of the senior loan. Be candid about that last one: it's the institutional move, and at $1M–$5M deal size it's harder to source cleanly and priced accordingly. For most small-balance sponsors, the realistic answer is some equity in plus a right-sized senior loan, planned for early rather than scrambled for at maturity.

Let's model your gap

If you have a value-add deal financed on a 2021–2023 bridge and a maturity coming into view, the useful thing isn't a rate quote — it's an honest proceeds number, run against the right binding constraint, with enough runway to act on it. That's the conversation we have every week with the 100+ banks and institutional private lenders we talk to. Bring us the NOI and the market, and we'll trace the gap with you.

FAQ

What is the refinancing gap? It's the shortfall between what you owe on your maturing bridge loan and the proceeds an agency takeout will actually give you today. A 2021–2022 bridge was sized to one set of assumptions; a 2026 agency takeout at roughly 6% money and tighter coverage tests sizes to a smaller loan. The difference is the gap — principal you have to cover with fresh equity, lower-leverage debt, or gap capital, plus closing costs. In the hypothetical in this piece, a $4.0M bridge payoff against a DSCR-constrained $3.84M takeout leaves a roughly $160,000 principal shortfall before transaction costs.

Can I refinance a maturing bridge into an agency SBL? Often, yes — that's exactly what the Freddie Mac Small Balance Loan and the Fannie Mae Small Loan programs are built to do. SBL covers $1M–$7.5M loans (capped at $6M in smaller markets), 5–50 units, with terms up to 30-year amortization and 5/7/10-year fixed or hybrid ARM structures. The property generally needs to show 90% physical occupancy over the trailing 90 days, and the sponsor is expected to have net worth around 100% of the loan amount with roughly 10% liquidity. The constraint isn't whether you can refinance — it's how much proceeds the deal supports at today's rates.

What DSCR and LTV does Freddie SBL require? It depends on the market tier. Top markets price to a 1.20x DSCR minimum at up to 80% LTV; standard markets to 1.25x at up to 80%; small markets to 1.30x with 75% LTV on acquisitions and 70% on refinances; and very small markets to 1.40x. The tier your property sits in changes the answer materially, so confirm it before you model proceeds. SBL is non-recourse with standard bad-boy carve-outs, and the carve-out can be waived at 65% LTV or lower and 1.40x DSCR or higher.

Why are my proceeds lower than my old loan? Because at roughly 6% money the debt-service coverage test usually binds before the LTV test does. In the hypothetical here, a property throwing off $360,000 of stabilized NOI in a small-tier market supports a $4.2M loan on the 70% LTV line, but only $3.84M once you hold it to the 1.30x DSCR minimum at 6.0% over a 30-year amortization. The lower number wins. Coverage, not leverage, is what caps the takeout — which is the opposite of how many of these deals were underwritten in 2021.

Loans are for business purposes only and are not subject to TILA, RESPA, or HOEPA. Not for primary residences. Equal Housing Opportunity. All loans subject to underwriting approval. Rates and terms shown for illustration; actual rates depend on deal specifics. We do not lend to borrowers with credit below 600 or on owner-occupied properties.