A $60M HUD 223(f) Refi, Walked Through

HUD 223(f) is the cleanest “stop refinancing forever” takeout in multifamily: long-term, fixed-rate, fully amortizing, and typically non-recourse. The trade is that HUD is not a rescue lender. The process is slower, documentation-heavy, and the sizing is governed by HUD tests, not sponsor preference.

This worked example is illustrative only. It is not a FOCAL transaction, not a client deal, and not a representation of any specific lender’s quote.

The sponsor owns a 312-unit, 1990s-vintage garden-style multifamily asset in a high-growth Sun Belt metro. It was acquired with a bridge loan and a value-add scope intended to push rents, cure deferred maintenance, and stabilize occupancy. The business plan worked operationally, but the capital stack is now the problem.

Where the deal sits today

Units: 312
Current physical occupancy: 94%
Current economic occupancy: 93% (some concessions still burning off)
Trailing 3-month average NOI (underwritten for HUD): $4,100,000
T-12 NOI (not used as-is, but informs trend): $3,920,000
Third-party appraiser’s “as-is” value (expected range pre-order): $74,000,000 to $78,000,000

Current loan: floating-rate bridge, SOFR-based with a spread
Current outstanding principal: $58,600,000
Maturity: early next year (hard)
Rate cap: expiring soon and expensive to replace
Cash management: springing lockbox already triggered during a seasonal occupancy dip

Mechanically, this is the classic moment where sponsors look for a permanent takeout. Agency can work, bank can work, CMBS can work, but HUD becomes compelling when three things are true: you can wait, the asset is seasoned enough, and you want maximum term certainty.

HUD 223(f) is explicitly designed for refinancing or acquiring existing stabilized multifamily, with long amortization (up to 35 years) and high leverage relative to most conventional executions, but with a real timeline cost (often measured in quarters, not weeks). As summarized in HUD 223(f) acquisition and refinancing terms, the program is commonly positioned as up to 35 years, fixed-rate, fully amortizing, and non-recourse with standard carve-outs.

From a sponsor decision standpoint, the key question is not “Is HUD cheaper?” It often is. The key question is “Can we actually get to closing without a maturity accident, and will the repair and escrow structure kill proceeds?”

For broader context on how we think about takeouts and permanent execution selection, see FOCAL Permanent Financing and our Capital Markets & Debt Advisory platform.

“Seasoned enough” is not a vibe: what HUD is really looking for

Sponsors waste time when they treat HUD eligibility as a soft conversation. “Seasoned enough” has three hard edges in practice: property age, stabilization, and the repairs line between 223(f) and substantial rehab programs.

Age and stabilization gates that matter

HUD 223(f) is intended for existing assets. Many market participants shorthand this as “the property must be at least three years old,” which is commonly repeated by lenders and industry education sources. For example, Janover’s HUD 223(f) overview notes the property must be at least three years old or substantially rehabilitated at least three years ago, with “standard repairs” allowed.

In underwriting reality, the sponsor also needs to demonstrate that current operations are stable enough that HUD is underwriting an income-producing property, not a lease-up rescue. That means:

Occupancy that is not just high today, but stable enough to support underwritten NOI.
Concession burn-off that is explainable and trending the right direction.
Bad debt and delinquency that do not suggest a tenant-quality or management issue.
A credible operating history (monthly statements matter, not just a polished OM).

Repairs: the silent proceeds killer

Sponsors frequently model proceeds off a back-of-napkin cap rate and a HUD leverage headline, then get surprised when HUD repairs and escrows reduce cash available at closing.

HUD 223(f) tolerates repairs, but the lender and HUD will classify them, escrow them, and often require initial and ongoing reserves. Even if total repairs are modest, the cash mechanics matter:

Required repairs can be escrowed 100% at closing, reducing net cash-out even if the gross loan amount looks “on target.”
Replacement reserve deposits can be both initial and monthly, and older assets can see meaningful initial funding expectations. Janover’s program summary notes initial replacement reserve funding can be “sometimes as much as $1,000 per unit for older properties” (HUD 223(f) eligibility and reserve expectations).

In this illustrative deal, the property is 30-plus years old, so reserve posture is not a rounding error. Before the sponsor commits to HUD, the sponsor should also understand reserve logic. We have a lender-ready framework in Multifamily Replacement Reserves: Lender Ready Checklist.

Lender-style sizing path: getting to “about $60M” the way HUD does

Sponsors often ask, “Can we get $60M?” HUD sizing answers, “Here is the maximum insurable mortgage based on the most conservative test.” You do not negotiate past the bottleneck.

Most HUD 223(f) executions effectively size to the minimum of three constraints:

LTV constraint (based on appraised value)
DSCR constraint (based on underwritten NOI and the all-in debt service)
Statutory per-unit limits (not modeled here because they vary by unit mix and geography and should be checked deal-by-deal)

A common DSCR floor discussed in the market is 1.15x for market-rate deals, and leverage can be high relative to other permanent executions. A concise summary of the sizing framework is laid out in Janover’s HUD 223(f) loans sizing discussion, including the concept that proceeds are driven by the smallest of leverage and coverage outcomes.

Underwritten NOI and expenses: where HUD underwriting “moves” the number

In this illustrative case, the lender underwrites to NOI of $4,000,000, slightly below the trailing run-rate to account for:

Conservative economic vacancy and concessions normalization
Verified insurance premium trend and updated property tax assumptions
A management fee consistent with HUD norms and the property’s effective gross income
Replacement reserves set at a HUD-acceptable level (this impacts NOI available for debt service)

That $4.0M is the input that matters. Everything else is commentary.

All-in debt service is not just the note rate

HUD loans carry MIP (mortgage insurance premium). As of applications submitted or amended on or after late September 2025, HUD moved to a flat 0.25% upfront and 0.25% annual MIP for multifamily programs, eliminating prior category tiers and green overlays. HUD discussed this change and its applicability in a multifamily Mortgagee Letter drafting table referencing the final notice (HUD multifamily MIP change discussion and MAP alignment).

For sizing, the annual MIP behaves like an interest rate add-on. If the note rate is 5.65%, and annual MIP is 0.25%, the effective annual debt cost is closer to 5.90% for DSCR math (simplified, but directionally correct). Lenders vary in how they model exact payment mechanics, but sponsors should not ignore it.

The sizing math (illustrative)

Assumptions:

Underwritten NOI: $4,000,000
Minimum DSCR: 1.15x
Underwritten constant (principal and interest) at 35-year amortization: approximately 6.95% (this is a stylized constant for sizing discussion, not a quote)
Add annual MIP: 0.25%
Total effective constant for sizing: approximately 7.20%

DSCR sizing gives maximum annual debt service of:

$4,000,000 / 1.15 = $3,478,260

Loan amount supported by DSCR (using constant approach):

$3,478,260 / 0.0720 = $48,309,000

At first glance, that looks far below $60M. This is where real HUD execution differs from “one constant to rule them all.” In practice:

Many HUD executions price tighter than a conventional fixed-rate assumption, improving the constant.
Lenders often model MIP in the mortgage payment but may size with more precise amortization math that can be friendlier than this blunt constant.
Underwritten NOI may be higher if concessions are credibly burned off and expenses are documented.

So let’s tune the underwriting to something realistic for a stabilized, well-performing 312-unit, and model NOI at $4,600,000 (a number the sponsor believes is supportable after concessions roll). DSCR-supported debt service:

$4,600,000 / 1.15 = $4,000,000

Loan supported:

$4,000,000 / 0.0720 = $55,556,000

Now apply LTV. If appraised value is $76,500,000 and HUD allows up to 85% LTV for market-rate, the LTV cap is:

0.85 x $76,500,000 = $65,025,000

In this tuned case, DSCR is the bottleneck, and the loan lands in the mid-$50Ms before repairs, escrows, and transaction costs. Getting to a $60M gross note is possible if NOI is closer to $5.0M or if pricing is tighter than the stylized assumption, but the sponsor should treat $60M as a target that must be earned by stabilized NOI, not willed into existence.

If you want to pressure-test DSCR, LTV, and payment outcomes quickly, use a consistent calculator and then reconcile to lender math. We built a sponsor-facing tool at FOCAL’s Loan Calculator.

Third-party reports and the HUD file: what actually slows you down

The fastest way to blow a HUD takeout is to treat third-party reports as administrative. They are a critical path item, and the sponsor’s responsiveness to the lender’s questions is a close second.

Core third-party package (typical for 223(f))

A lender will require a package that generally includes:

Appraisal (HUD-compliant format and assumptions)
Environmental: Phase I ESA, and sometimes additional testing if the Phase I flags issues
Physical Needs Assessment (PNA) or comparable engineering report that supports repairs and reserves
Market study (or market condition analysis depending on deal specifics and HUD office expectations)
Survey, title, zoning report (as required), and full rent roll/operating history support

The sponsor’s operational data quality matters more than most teams admit. HUD underwriting is skeptical of “one-off” explanations. If your delinquency spiked, you need a coherent story with documentation. If payroll dropped because you cut staff, you need to show service levels did not decline and turnover did not spike.

Timeline: realistic, not optimistic

Industry sources regularly describe closings that can run six to nine months, with longer outcomes possible depending on HUD queue and file quality. Janover’s refinance guide frames the core trade-off succinctly: lowest fixed rates and longest amortization “if your sponsor can wait six to nine months” (HUD 223(f) refinance timeline trade-off).

In this illustrative execution, a realistic schedule looks like this:

Engagement to complete third-party ordering and data room: about 2-3 weeks
Third-party fieldwork, drafts, revisions: about 6-10 weeks (appraisal and engineering are often pacing items)
Lender underwriting and HUD submission prep: about 3-5 weeks after solid drafts
HUD review to firm commitment: commonly 8-12 weeks (varies by region and workload)
Rate lock, closing coordination, and endorsement: about 4-6 weeks

Total: roughly 7-9 months from lender engagement to closing if the sponsor is responsive and there are no environmental surprises.

If the existing loan maturity is inside that window, HUD is not your takeout unless you have a credible extension path, a bridge refi, or a structured interim solution. “HUD will close faster” is not a plan.

Sources and uses: where the proceeds actually go (and why cash-out disappoints)

Sponsors fixate on gross loan amount. What matters is net proceeds available to retire the bridge and, if allowed, distribute cash.

Here is an illustrative sources and uses for a gross HUD 223(f) loan target of $60,000,000. The numbers are realistic in structure, but they are not a quote and not a representation of any specific lender’s fees.

Item	Amount ($)	Notes
Sources
HUD 223(f) first mortgage (gross)	60,000,000	Final is set by HUD sizing and firm commitment
Sponsor cash (if required)	0 to 2,000,000	Only if net proceeds do not cover payoff and required escrows
Total sources	60,000,000 to 62,000,000
Uses
Payoff existing bridge principal	58,600,000	Assumes no extension fee rolled in
Prepayment / exit fees on bridge	550,000	Deal-specific based on note
HUD-required repair escrow	1,250,000	From PNA. Held until completed
Initial replacement reserve deposit	312,000	$1,000 per unit is possible on older assets, but this example assumes $1,000 per unit would be negotiated down with condition support; escrow remains a lever
Lender fees, legal, third-party, HUD fees	1,050,000	Includes reports, legal, lender charges. Order-of-magnitude only
Upfront MIP	150,000	0.25% of $60M, subject to program applicability and rounding
Interest, tax, insurance escrows and closing adjustments	450,000	Depends on closing month and tax calendar
Total uses	62,362,000
Net cash-out to sponsor	(2,362,000)	Negative means sponsor must bring cash or reduce escrows/loan sizing

Two observations:

The first is that “$60M loan” is not the same as “$60M proceeds.” Repair escrows, reserve deposits, and closing cost stack are real. HUD’s flat MIP regime reduces ongoing annual drag relative to old tiering, but you still need to budget both the upfront and annual insurance cost (HUD multifamily MIP change discussion and MAP alignment).

The second is that sponsors can get trapped by anchoring to a payoff number. If the bridge payoff is $58.6M and the HUD net is short, the sponsor has three levers:

Increase NOI (real stabilization, not spreadsheet optimism)
Reduce required escrows (only if engineering supports it and HUD accepts it)
Bring cash or negotiate a bridge extension to close HUD later without duress

The cleanest HUD takeouts happen when the sponsor is not relying on every dollar of gross proceeds to meet payoff. If you are, one re-trade can push you into a liquidity problem.

Decision points that make or break the refinance: rate lock, re-trades, and “don’t get cute”

HUD 223(f) is attractive because it can deliver long-term fixed-rate debt. That also means the sponsor must manage two timing risks simultaneously: operational performance at underwriting and rate movement during the process.

Rate lock timing is a strategy decision, not an administrative step

In HUD executions funded through securitization mechanics, the note rate is tied to market execution at or near closing. You need a plan for what happens if rates move against you mid-process.

Practical sponsor posture:

Underwrite a conservative spread between today’s indicative rate and a “stress” rate for sizing and DSCR.
Maintain liquidity to cover a shortfall if proceeds compress.
Keep the asset stable. A rate move is not your only risk. A sudden NOI drop can kill proceeds faster than a rate move.

Re-trades usually come from three places

Re-trades are not always bad faith. They often come from information that emerges late because the sponsor did not control the process early.

Engineering: PNA identifies more immediate repairs than expected, increasing repair escrow.
Insurance and taxes: updated premiums or reassessment mechanics raise expenses, lowering NOI and sizing.
Rent roll quality: loss-to-lease assumptions and concessions are not documented, and the lender underwrites down revenue.

Sponsors who run HUD well act like they are preparing for an audit from day one. Clean general ledger, clean rent roll, clear concession policy, documented capex history, and a realistic repairs narrative.

If you have not already, build a reporting cadence that matches lender expectations. That discipline is also why institutional operators use third-party oversight on complex transitions. For teams that need it, FOCAL Third-Party Asset Management is designed for business-plan oversight and lender-aligned reporting, especially when the financing outcome depends on stable operations, not just a slide deck.

What you should do next: pressure-test your own HUD takeout

HUD 223(f) is not “cheap money.” It is structured certainty. If you are trying to refinance out of bridge risk and hold the asset for a long time, that certainty is worth real basis points. If you have a hard maturity inside the HUD process window, HUD can still work, but only if you solve timing first.

Pressure-test your deal in this order:

Validate eligibility and timing before you spend real third-party dollars

Confirm the property’s age and stabilization narrative and map it against a realistic HUD closing window. If your bridge matures too soon, solve extension and interim liquidity up front.

Underwrite NOI the way HUD will, not the way your equity memo does

Normalize concessions, verify insurance and taxes, include replacement reserves appropriately, and assume HUD will ask for documentation on anything that looks “too perfect.” If you need a fast DSCR/LTV sanity check, start with FOCAL’s Loan Calculator, then reconcile to lender underwriting.

Model net proceeds, not gross loan amount

Build a sources and uses that includes repair escrow, initial reserves, upfront MIP, and meaningful closing adjustments. If the deal only works at a razor-thin net proceeds margin, it is not a HUD-ready deal yet. It is a “maybe, if everything is perfect” deal, and that is how sponsors get trapped.

Decide whether you are optimizing for proceeds or certainty

A HUD takeout is often the right answer when certainty and duration are the primary objectives. If you are optimizing for maximum cash-out today and you cannot tolerate a long process with re-trade risk, you may be forcing the wrong tool.

When you are ready to line up the right execution and sequencing, start with a sober takeout plan and a timeline that assumes friction. That is how HUD becomes a competitive advantage instead of a closing delay.

Frequently Asked Questions

How does HUD 223(f) size the max loan amount?

HUD 223(f) sizing typically follows the most conservative of DSCR and LTV outcomes, with statutory per unit limits checked separately. A common market DSCR floor is 1.15x, and market-rate leverage can be up to 85% LTV. The binding test becomes the maximum insurable mortgage.

What MIP should sponsors model for HUD 223(f) today?

HUD multifamily MIP is described as a flat 0.25% upfront and 0.25% annual MIP for applications submitted or amended on or after late September 2025. For DSCR math, the annual MIP acts like an add-on to the note rate, so sponsors should include it in effective debt service assumptions.

How long does a HUD 223(f) refinance usually take to close?

A realistic HUD 223(f) refinance timeline is often 7 to 9 months from lender engagement to closing when the sponsor is responsive and third-party reports stay clean. Typical components include 2 to 3 weeks to launch reports, 6 to 10 weeks for drafts, and 8 to 12 weeks for HUD review.

Why can a $60M HUD 223(f) loan still be cash-in at closing?

Net proceeds can fall short because HUD closings often include repair escrows funded at 100% at closing, upfront MIP, and initial replacement reserve deposits. Older assets may see initial reserve expectations up to about $1,000 per unit, and those escrows reduce cash available to pay off the existing bridge balance.

About FOCAL

FOCAL is an independent, principal-led commercial real estate investments and advisory firm headquartered in Beverly Hills, California. Through FOCAL Investments and FOCAL Advisory, the firm arranges debt and equity financing starting at $1 million for sponsors nationwide, and provides development consulting spanning land use and entitlements, owner’s representation, development advisory, and asset management.

FOCAL works across all major commercial asset classes, with deep roots in Southern California multifamily and ground-up development.

Capital Markets · Development Consulting · Closed Transactions · Contact FOCAL