The lender's checklist for California multifamily deals

A California multifamily credit memo has to underwrite the government alongside the sponsor. On the way in, entitlement stage and vesting quality decide whether the approved unit count and the frozen fee schedule are real; mid-hold, rent regulation such as AB 1482 and local ordinances govern how fast NOI can grow; on the way out, recorded covenants and special taxes decide what a buyer will pay. None of this appears on a rent roll, and most of it is verifiable from statutes and recorded documents in an afternoon.

Two federal layers deserve their own lines. HUD-insured deals carry 24 CFR Part 246, which preempts local rent regulation that would prevent an insured project from operating on a financially sound basis — a genuine, if procedural, protection worth understanding before assuming a local cap fully binds an FHA asset. And LIHTC collateral carries the IRC § 42 extended-use agreement, which by statute survives ordinary transfers and persists through foreclosure in modified form — three years of tenant protections at minimum, and full restrictions if the foreclosure is found collusive. Both change recovery math.

Place the collateral on the ladder — raw, application-filed, entitled, RTI — and then grade the vesting instrument, because "entitled" without vesting can still be moved by a new ordinance. The three instruments differ: an SB 330 preliminary application freezes the ordinances, standards and fee schedules in effect at filing for the project's processing (see SB 330 vesting); a vesting tentative map confers a right to proceed under the rules in effect when the map application was complete, on the map's statutory life; a development agreement is a negotiated contract that can lock rules (and fees, if drafted to) for a decade or more — see vesting maps and development agreements. Condition funding on the instrument, its date, and evidence that appeal, referendum and CEQA-litigation windows have closed.

• Underwriting watch-outs:
• Ask which fee schedule is locked and as of what date — SB 330's fee freeze runs from the preliminary application, and the delta to current fees is a real number on multi-year projects.
• Approvals expire: tentative maps, entitlement terms and conditions of approval all carry clocks; confirm remaining life against the construction schedule plus a downside.
• CEQA litigation risk doesn't end at approval — verify the notice-of-determination date and that the statute-of-limitations window has run, or price pending litigation explicitly.

Scrub the sources-and-uses against the city's adopted fee schedule as of the vesting date — impact fees, school fees and utility capacity charges are commonly the largest underbudgeted line on small-sponsor deals. Then pull title for Mello-Roos: a CFD special tax (Mello-Roos Act, verbatim) is effectively senior to the mortgage at the tax collector, escalates by formula, and permanently loads operating expenses — it belongs in the DSCR math, not a footnote (see Mello-Roos and special taxes).

On redevelopment collateral — anything with tenants now or in recent years — underwrite the replacement and relocation stack: Housing Crisis Act replacement units and right-of-return (protected units), density-bonus replacement rules reaching back five years (guide), and relocation payments under state law or, with federal money, the URA. These obligations change the buildable program and the schedule; a budget with no line for them on a formerly occupied site is a red flag, not a savings.

• Underwriting watch-outs:
• Require the recorded rate-and-method for any CFD and model the special tax at its escalated values through the hold — not the current-year bill.
• Replacement-unit obligations produce covenanted affordable units inside the collateral — the stabilized rent roll must reflect them, and the exit cap should too.
• Delivery-vacant assumptions on occupied buildings import just-cause, buyout-regulation and relocation timelines into your construction-start covenant — stress the start date.

NOI growth and exit cap are regime questions. Build the unit-level regime table — local RSO, AB 1482, or exempt-until-when (the new-construction exemption rolls at 15 years, so a mid-hold conversion belongs in the model) — and cap escalations accordingly; the framework is in AB 1482 and Costa-Hawkins underwriting. Exit pricing should assume the buyer's lender runs the same table. On HUD-insured or Secretary-held projects, 24 CFR Part 246 preempts local rent regulation that prevents financially sound operation — on preemption-eligible assets, the HUD-regulated rent process, not the local board alone, sets the achievable schedule; agency and FHA/HUD debt teams should underwrite which regime actually governs.

Last, covenant seniority. A LIHTC extended-use agreement is designed to survive: statute requires the restrictions to run with the land, and foreclosure terminates them only subject to a three-year tenant-protection period (no eviction without cause, no over-limit rent increases for existing tenants) — and not at all if the foreclosure is collusive. HOME, bond and local regulatory agreements each have their own subordination and survival terms. Collect every recorded agreement, map the waterfall of what survives your foreclosure, and check the qualified-contract and Year-15 posture on anything with credits (see LIHTC compliance and extended use).

• Underwriting watch-outs:
• Do not size takeout debt on a "restrictions burn off" story without reading the recorded documents — California extended-use terms typically run 55 years and the state has curtailed qualified-contract exits on newer deals.
• Subordination agreements with public lenders often trade lien priority for covenant survival — the covenant surviving your foreclosure is the negotiated norm, so model recovery on restricted rents.
• Part 246 preemption is procedural, not automatic — it attaches to specific HUD-regulated categories and processes; confirm the asset's insurance status and the current preemption scope before crediting it in the model.