SCG 2026-MFP2: a $609.75M single-loan, nine-property Sunbelt multifamily conduit — 75% LTV, 1.18x DSCR, 7.1% debt yield, and a structure that's floating-rate bridge risk, not agency-eligible paper

This isn't a diversified conduit pool — it's one $609.75MM floating-rate mortgage loan secured by nine cross-collateralized garden-style apartment communities (3,441 units), carved into eight rated/unrated certificate classes plus the HRR risk-retention piece 1. Borrower is Starwood REIT (SREIT), refinancing a portfolio it's owned since 2021; Highmark Residential manages 2.

Collateral geography. Two states, four markets, eight submarkets — but concentrated: North Carolina carries 66.6% of units / 53.9% of UW NCF (six Raleigh/Durham assets), Florida the other 33.4% of units / 46.1% of NCF (three Palm Beach/Fort Lauderdale assets, which run bigger and richer per unit) 3. Single-largest exposure is Arium Centro at Davie at 18.5% of UW NCF — real name risk if any one asset underperforms.

LTV / debt yield / DSCR — the headline credit metrics. Against an As-Is portfolio appraised value of $813.0MM (a ~2.0% portfolio premium over the sum of individual appraisals of $796.9MM), the $609.75MM loan prices at 75.0% LTV, a 7.4% UW NOI debt yield and 7.1% UW NCF debt yield, and 1.18x UW NCF DSCR (1.11x at the SOFR cap strike) 4. That implies a going-in cap rate of only ~5.5% on NOI against the appraised value 5 — full pricing for a value-add-recovered, institutionally-owned Sunbelt portfolio, not a wide margin of safety. $203MM of appraised equity sits behind the loan (25% cushion), and the loan runs $177,201/unit against $236,269/unit of appraised value 5.

The certificate structure spreads that risk across eight classes by subordination, not diversification — this is one loan, so every class is exposed to the same nine buildings; the tranching only allocates loss priority.

Why this wouldn't finance in agency (Fannie/Freddie) CMBS execution:

• It's floating-rate, interest-only, and structured as a two-year bridge with three 1-year extensions to a 5-year fully-extended maturity 6 — that's the profile of transitional financing meant to be refinanced out, not the fixed-rate, amortizing 5–10 year permanent paper Fannie DUS and Freddie CME conventional programs are built for. No amortization ever occurs; the balance at maturity equals the balance at close.
• 1.18x DSCR and a 7.1–7.4% debt yield sit below where the GSEs typically draw the line on conventional executions (Freddie/Fannie standard conventional loans generally target 1.25x+ DSCR and debt yield floors in the high-single-digits to ~9%+ depending on market tier); at 75% LTV with sub-1.20x coverage, this loan is priced for a rated, tranched buyer base that can price the subordinate risk piece-by-piece — the GSEs don't have that mechanism, they hold or guarantee the whole loan.
• It's a nine-property, cross-collateralized, cross-defaulted blanket loan across nine separate SPE borrowers, with individual partial-release pricing (105%/110% of allocated loan amount) 7 baked in as a structural feature. Agency execution is built around single-asset (or much simpler, same-sponsor-pool) loans; a blanket loan with release mechanics, cross-collateralization, and portfolio-premium appraisal math is a bespoke structure the GSEs' standard underwriting box doesn't accommodate.
• Future mezzanine debt is expressly permitted 8 — the GSEs place tight restrictions on subordinate financing and typically require it to be underwritten and approved at closing, not left open as a future option; a structure that pre-clears future mezz signals this loan is meant to support additional leverage down the road, inconsistent with GSE credit boxes.
• The rating-agency-driven tranched structure (AAA through unrated HRR) is itself the financing tool — it lets the sponsor monetize a loan that's too aggressively leveraged/thin-covering for a single balance-sheet or agency-guaranteed execution by distributing the risk to buyers across the credit spectrum. That's a conduit/CMBS-native solution; there's no equivalent in agency execution, where the GSE (not the capital markets) absorbs the credit risk and therefore dictates the leverage and coverage terms directly.

Net: SREIT chose conduit CMBS because the deal itself — leverage, coverage, floating short-term structure, and multi-borrower/release complexity — doesn't fit inside what Fannie Mae or Freddie Mac would underwrite; the private-label market was the only place this loan, at this balance and this leverage, could get done.