Buyers spend weeks arguing about cap rates and about fifteen minutes thinking about debt, which is backwards. On a leveraged apartment purchase the loan terms move your return more than a quarter point of cap rate does, and unlike the cap rate, the loan is something you can actually shop.
There are three lanes of multifamily debt in Los Angeles, and they are not interchangeable. They want different buildings, they size loans by different math, they price differently, and they take very different amounts of time to close. Choosing the wrong lane is the most common reason an LA deal that looked financeable at the LOI stage falls apart at week five.
This walks through all three: who each lender is, what they want, how they size the loan, and the specific LA questions that come up in every one of these conversations, starting with what rent control does to your underwriting.
The Three Lanes
Agency debt means Fannie Mae and Freddie Mac multifamily loans, originated through approved lenders. This is the cheapest long-term money available on stabilized apartment buildings, it is generally non-recourse, and it comes with the most structure: minimum unit counts, occupancy seasoning, prepayment penalties, and reserve and escrow requirements.
Bank and credit union debt means a loan held on the lender’s own balance sheet. More flexible on the building, faster to close, more negotiable on terms, but usually recourse, with a shorter fixed-rate period and a deposit relationship expected.
Bridge and debt fund debt means short-term, usually floating-rate, usually interest-only money for a building that cannot qualify for the other two yet. Most expensive by a wide margin, fastest to close, and designed to be refinanced out of rather than held.
The rule of thumb that holds up: stabilized and holding long term, go agency. Stabilized but the building or the borrower has a wrinkle, go bank. Not stabilized, go bridge and plan the exit before you sign.
How a Lender Actually Sizes Your Loan
Buyers assume the loan amount comes from the purchase price. It does not. It comes from the lower of three tests, and on an LA rent-controlled building the binding test is almost never the one the buyer expected.
The loan-to-value test. A percentage of the appraised value. This is the test buyers think about.
The debt service coverage test. Net operating income divided by annual debt service has to exceed a minimum. This is the test that actually binds on most LA deals, because LA buildings have low in-place income relative to value. A building with a 4.5% in-place yield cannot carry much debt at a 6% borrowing cost no matter what the appraisal says.
The debt yield test. Net operating income divided by the loan amount, a floor the lender applies as a sanity check independent of interest rates. This one catches aggressive deals in low-rate environments.
The practical consequence for Los Angeles: your rent roll sizes your loan, not your price. Two buyers bidding the same number on the same building will be approved for different loan amounts if one of them is underwriting the in-place rents and the other is underwriting a projection, because the lender will underwrite the in-place rents regardless.
Run it on a real shape. Take a $6,000,000 building with in-place net operating income of $270,000, which is a 4.5% yield, and an agency loan at a 1.25x minimum coverage ratio and an 80% maximum loan-to-value.
The loan-to-value test allows 80% of $6,000,000, which is $4,800,000.
The coverage test works backwards from income. Annual debt service cannot exceed $270,000 divided by 1.25, which is $216,000. At an indicative 6.4% rate on thirty-year amortisation, every dollar of loan costs about 7.5 cents a year in debt service, so $216,000 of allowable debt service supports roughly $2,880,000 of loan.
That is the whole lesson in two numbers. The appraisal would support $4.8 million. The rent roll supports $2.88 million, which is 48% of the purchase price. The buyer who budgeted for 25% down is short by well over a million dollars, and no amount of shopping lenders fixes it, because every agency lender applies the same coverage floor.
This is why low-yielding Los Angeles assets so often need more equity than buyers expect, and why the single most useful thing you can do before making an offer is run the coverage test yourself.
The number to carry into every conversation is in-place NOI, not pro forma NOI. Lenders underwrite trailing twelve-month operating statements with their own expense assumptions, and their expense assumptions are usually higher than yours.
Where Rates Actually Are
Nobody publishes borrower rates for agency debt, so this section separates what is published by the government from what is quoted by the market. Both are useful. They are not the same thing.
Benchmarks, published:
| Index | Rate | As of |
|---|---|---|
| 10-year Treasury | 4.91% | 2 November 2026 |
| 5-year Treasury | 4.70% | 2 November 2026 |
| 30-day average SOFR | 3.91% | 2 November 2026 |
| Bank prime rate | 7.00% | 30 October 2026 |
Indicative agency rates, quoted. From a lender rate and spread sheet timestamped 1 November 2026. These are indicative, subject to change, and not a commitment to lend by anyone, including me. Agency pricing is a spread over the matching Treasury, so the all-in rate moves daily with the benchmark.
| Product | LTV | DSCR | Indicative all-in |
|---|---|---|---|
| Freddie Mac 10-year fixed | 65% | 1.35x | 6.00% to 6.15% |
| Fannie Mae 10-year fixed | 65% | 1.35x | 6.10% to 6.40% |
| Freddie Mac 7-year fixed | 80% | 1.25x | 6.19% to 6.39% |
| Fannie Mae 7-year fixed | 80% | 1.25x | 6.34% to 6.64% |
| Fannie Mae 5-year fixed | 75% | 1.25x | 6.49% to 6.79% |
Notice the pattern in that table, because it is the most practical thing on this page. Lower leverage prices better, and the spread between 65% and 80% leverage is worth more than the spread between lenders. A borrower who can live at 65% leverage and a 1.35x coverage ratio is looking at roughly thirty to forty basis points less than the same borrower pushing to 80% and 1.25x. On a $4 million loan over ten years that is real money, and it is a choice rather than a market condition.
Bank and portfolio debt. There is no published index for bank multifamily rates, so anything you read is a broker quote. A California quote sheet updated 3 November 2026 showed portfolio and bank multifamily around 6.5% at up to 75% leverage, with a five to ten year fixed band in the mid-6s. Treat that as a starting point for a conversation, not a rate.
Bridge debt. An aggregator tracking priced bridge programmes put the multifamily bridge band at 7.90% to 11.00% across October 2026, with a median maximum loan-to-value of 75% and a median maximum loan-to-cost of 72.5%. Terms run twelve to thirty-six months, interest-only, with a balloon.
Every figure in the last two paragraphs is market-quoted rather than published, which is why I have dated each one. If you are reading this more than a quarter after those dates, the benchmarks have moved and so has everything priced off them.
Agency Debt: Fannie and Freddie
Agency is where most stabilized LA apartment buildings should end up, and the programme structure changed meaningfully in 2026, which is the part owners have not caught up with.
Fannie Mae Small Loans is the workhorse for buildings in the $1 million to $9 million range. Published terms: maximum loan $9,000,000 nationwide, minimum five units, maximum 80% loan-to-value, minimum 1.25x coverage, terms from five to thirty years, amortisation up to thirty years, non-recourse available, fixed or variable. Prepayment is yield maintenance or a declining premium. Eligible properties are existing and stabilized. That $9 million cap has been in place for loans committed since August 2023 and did not change for 2026.
Freddie Mac retired the Small Balance Loan programme in April 2026. This is the change to know. On 15 April 2026 Freddie launched an integrated Conventional Small product on its core conventional platform, for loans from $2,000,000 to $10,000,000 on properties of fifty units or fewer, and legacy Small Balance Loan applications were only accepted through a two-week transition that ended 30 April 2026.
The published band moved from $1 million to $7.5 million under the old programme up to $2 million to $10 million under the new one, and the consequence for Los Angeles owners sits at the bottom end. There is now a $2 million floor. If you are buying a small building under $2 million, Freddie’s small programme no longer reaches you, and your options are Fannie Small Loans, a bank, or a credit union.
Conventional Small terms as published by lenders: five, seven, ten, twelve and fifteen year fixed; amortisation up to thirty years; maximum leverage 75% on a five-year amortising loan, 80% at seven years and beyond, with a maximum maturity loan-to-value of 70%; minimum 1.25x amortising coverage, with no refinance test where coverage is at least 1.40x and leverage is 60% or less; non-recourse with standard carve-outs; yield maintenance until securitisation, then a two-year lockout, then defeasance, with no premium in the final ninety days.
Fannie Mae conventional, for loans above the small-loan band, publishes 80% maximum leverage, 1.25x minimum coverage, five to thirty year terms, amortisation up to thirty years, non-recourse for most loans above $750,000, and a five-unit minimum. It also publishes the single most useful agency requirement for anyone with a vacancy problem: stabilized occupancy, typically 90%, for ninety days prior to funding. Pre-stabilized properties are considered case by case. If your building is at 82% occupancy, that sentence is why you cannot get agency debt yet, and it is why bridge exists.
Freddie’s conventional fixed-rate product carries a $10,000,000 minimum, so between $10 million and the Conventional Small ceiling the two programmes meet.
For context on capacity, the Federal Housing Finance Agency set 2026 multifamily purchase caps of $88 billion for each enterprise, $176 billion combined, with at least half of each enterprise’s business required to be mission-driven affordable housing and workforce housing loans excluded from the caps entirely. Capacity is not the constraint in this market. Underwriting is.
One thing I will not tell you, because I cannot verify it: there has been press reporting that Fannie and Freddie have tightened underwriting on rent-regulated buildings. That reporting is framed entirely around New York City’s rent freeze, it is sourced to another outlet rather than to any agency publication, and I found no Fannie or Freddie lender letter, guide bulletin or term sheet confirming it or extending it to California. Treat it as unconfirmed, and do not let anyone quote it at you as a reason to accept worse terms on an LA building.
Bank and Credit Union Debt
Banks and credit unions hold the loan on their own books, which is what makes them flexible and also what makes them unpredictable. Nothing here is standardised the way agency is.
What banks actually publish. The largest multifamily portfolio lender in Los Angeles publishes a programme with loan amounts from $500,000 to $25 million or more, a five-unit minimum, terms and amortisation up to thirty years, and three structures: hybrid loans fixed for three, five, seven or ten years and then adjustable; adjustable-rate loans; and a fifteen-year fixed with fifteen-year amortisation. Prepayment is stepdown or yield maintenance depending on the programme, and stepdown premiums on the three, five and seven year products are waived when refinancing with the same bank after two years. An application needs two years of operating history and a recent rent roll, and notably does not require tax returns. The loan process cost is the greater of $2,000 or 12.5 basis points.
What that lender does not publish is any rate, leverage limit or coverage ratio. Neither does anyone else. That is not evasiveness, it is the nature of balance-sheet lending: the terms come out of a credit committee looking at your specific building and your specific balance sheet.
The deposit question. You will hear that banks require a deposit relationship. I could not find a single bank that publishes one as a term. Treat it as a negotiated expectation that affects pricing rather than a published requirement, and ask directly what moving your operating accounts is worth in basis points.
Credit unions are underused on LA apartment buildings and worth a call. Published programmes typically start at five units, run from $500,000 to $15 million or more, offer fixed and variable options up to fifteen years with amortisation up to thirty, and two features stand out in their published terms: no prepayment penalty and zero origination points. On a building you might sell or refinance in four years, the absence of a prepayment penalty can be worth more than fifty basis points of rate.
The catch is twofold. Membership is geographic, so the credit union has to serve the county your business sits in. And there is a structural ceiling that explains the behaviour owners find baffling: a federally insured credit union’s total member business lending is capped at the lesser of 1.75 times its net worth or 1.75 times its required minimum net worth, commonly expressed as about 12.25% of assets. Loans on one-to-four unit residential property are excluded from that definition, but five-plus unit apartment loans count against the cap. That is why a credit union can be enthusiastic in January and closed for the year in September, regardless of rate. If a credit union is in your deal, ask early where they are against their cap.
The environment, as of now. Two things are worth knowing and neither is what people assume. First, the October 2026 Senior Loan Officer Opinion Survey reported that banks had tightened standards on net across every commercial real estate category, multifamily included, and also saw weaker demand. That is direction, not magnitude, and the survey covers the prior three months. Second, the bank capital rules everyone has been talking about are still a proposal. Comments closed in June 2026 and no final rule has issued. If anyone tells you bank CRE lending appetite changed in 2026 because of new capital rules, the rules have not been finalised.
Bridge and Debt Fund Debt
Bridge is the most expensive money on this page and the most misunderstood. It is not a faster version of agency debt. It is a different instrument, priced for risk the permanent lenders will not take, and designed from the first page to be refinanced out of.
What bridge lenders publish. One major lender’s current term sheet: loan size $10 million to $100 million, floating over one-month term SOFR, term typically up to three years including extensions, generally interest-only with some amortisation possible after year two, sized to up to 80% of current value and 70% of stabilized value, a 1.0% commitment fee, an exit fee that applies and may be waived in whole or part if the same lender provides the permanent loan, typically non-recourse with standard carve-outs, a required interest rate cap purchased by the borrower, and minimum equity of typically 10% to 15% cash. Partial recourse or completion guaranties may be required for a property in significant transition.
Another publishes a floating programme at SOFR-based pricing with a SOFR plus 250 floor, $10 million to $275 million or more, one to three years plus extensions, leverage under 75%, non-recourse with carve-outs, 1.00% origination and typically 1.00% exit, with the exit fee potentially waived if they refinance, and flexible prepayment with no lockout.
Read the sizing line in that first paragraph again, because it is the structural difference from agency. Agency sizes off in-place income. Bridge sizes off both current value and stabilized value, and tests both. That dual test is what lets a bridge lender fund a building that is 70% occupied, and it is also what caps your proceeds: you cannot borrow against a stabilized number you have not reached.
An honest gap. Every bridge lender whose terms I could verify sets a $10 million minimum. For an LA owner buying a $3 million or $6 million building, the published bridge market does not exist. Below $10 million you are dealing with private and relationship debt funds that publish nothing at all, so there is no benchmark to check a quote against. If you are offered sub-$10 million bridge debt, get every fee in writing, model the all-in cost including origination and exit, and compare it against simply putting more equity in. I have seen owners pay 11% plus two points for money that a larger down payment would have made unnecessary.
When bridge is the right answer. Five situations, and they are specific:
- Occupancy below the agency test. Fannie wants roughly 90% occupancy for ninety days before funding. Below that, the permanent market is closed to you until you fix it.
- Mid-renovation or mid-repositioning, where the income that will support permanent debt does not exist yet.
- An outstanding soft-story retrofit. A building under an unsatisfied order to comply is a classic bridge candidate, because the permanent lender will reserve against the work or decline it.
- A loan under $2 million, where Freddie’s new small-loan floor now excludes you and bank or credit union appetite is the only alternative.
- A closing timeline no permanent lender can meet. Bridge can close in two to four weeks. Agency cannot.
In each of those, the bridge loan is buying time to reach the condition that qualifies you for cheaper money. If you cannot describe the takeout loan in specifics, at today’s terms and today’s rates rather than hoped-for ones, you do not have a bridge plan. You have an expensive problem on a clock.
What Rent Control Does to Your Underwriting
This is the LA-specific question, and it comes up on every loan.
Rent regulation does not disqualify a building. Lenders finance City of LA rent-stabilized buildings every day. What regulation does is cap the income growth the lender is willing to assume, which matters at two points: loan sizing today, and the refinance you are planning in five years.
The mechanics worth understanding:
- In-place rents are what get underwritten. A lender sizing a loan on an RSO building will not give you credit for the market rent you expect to achieve on turnover, however well documented the loss-to-lease is. Your upside is yours; it is not collateral.
- The allowable increase caps the growth case. For City of LA RSO buildings the allowable increase is 3% for 1 July 2026 through 30 June 2027, under a formula adopted effective 2 February 2026 of 90% of average CPI with a 1% floor and a 4% ceiling. For buildings covered by AB 1482 instead, the LA metro cap is 8.7% for 1 August 2026 through 31 July 2027. A lender looking at a five-year refinance horizon models very different income on those two buildings.
- Clean registration matters to the loan, not just the sale. An unregistered RSO unit cannot lawfully take a rent increase. A lender reviewing a rent roll with registry gaps is looking at income that may not be legally collectible, and they will treat it accordingly.
- Deferred compliance work gets reserved for. An unretrofitted soft-story building will typically draw a holdback or a required repair escrow, which reduces your proceeds at closing.
The practical takeaway is that the compliance file is part of the financing package. Owners who keep registration current, document every increase against the correct allowable rate for that year, and close out their retrofit get better loan terms than owners who do not, on the same building. Which rulebook governs yours is worth confirming before you apply; I have mapped it out by city in the LA rent control map.
Choosing a Lane
A short decision path that covers most LA situations.
- Stabilized, 90%+ occupied for 90 days, five or more units, holding seven years or more. Agency. Take the prepayment penalty and the escrows in exchange for the rate and the non-recourse.
- Stabilized, but something is unusual. Unit count below the agency minimum, an unusual property type, a borrower with a thin multifamily record, a partnership structure that does not fit an agency box. Bank or credit union.
- Need to close fast to win the deal. Bank, or bridge if the timeline is genuinely extreme. Agency cannot move at that speed.
- Significant vacancy, mid-renovation, mid-retrofit, or a repositioning plan that needs capital. Bridge, with the agency or bank takeout underwritten before you sign.
- Planning to sell within three years. Avoid long-term fixed agency debt unless it is assumable and you expect that to be valuable. Prepayment penalties on a five-year hold structure can erase the rate advantage.
That last point is worth dwelling on, because it is the mistake I see most often. A buyer takes a ten-year fixed agency loan at an attractive rate, repositions successfully, and wants to sell in year four. The prepayment penalty on the way out eats a meaningful share of the gain. If your business plan has a three to five year exit, say so to the lender at the start and price the exit, not just the entry.
The Refinance Question
Many owners reading this are not buying, they are deciding whether to refinance or sell. Rate environment aside, the mechanical question is whether the building supports enough proceeds to justify the exercise after costs.
Refinancing has a real advantage in Los Angeles that gets undervalued: it does not trigger Measure ULA. A sale above $5.4 million carries the transfer tax on the gross consideration, from dollar one, at 4%, rising to 5.5% at $10.9 million or more. A refinance carries none of it. On an $8 million building, that is roughly $320,000 of difference in favour of staying in.
Against that, a refinance leaves you with the building, the regulation, the capital needs and the management. The full comparison, including when each one wins, is in the piece on whether to sell or refinance, and the transfer tax mechanics are in the Measure ULA guide.
Timeline and Documents
Agency loans typically run sixty to ninety days from application to funding. Bank loans forty-five to sixty. Bridge can close in two to four weeks. Build the longer number into your purchase contract, because a financing contingency that expires before the lender can realistically fund is how deposits get lost.
What every lender will ask for, and what you should have assembled before you apply:
- Trailing twelve-month operating statement, and two to three years of operating history or Schedule E
- Current rent roll with tenancy start dates and lease expirations
- Copies of all leases and addenda
- Twelve or more months of utility bills
- Service contracts
- RSO registration confirmation and rent registry history, if applicable
- Capital improvement records and permits
- Soft-story retrofit documentation, or the Order to Comply if not yet complete
- Personal financial statement, schedule of real estate owned, and two to three years of tax returns
- Property insurance quotes, which in California now take longer than people expect
Having that package ready shortens every process and improves your terms, because a lender’s confidence in the sponsor is priced into the deal whether or not anyone says so.
Common Mistakes
- Shopping rate alone. A quarter point of rate is usually worth less than the difference in proceeds between a lender sizing on a 1.20 coverage test and one sizing on 1.25.
- Underwriting pro forma income. The lender will underwrite trailing actuals with their own expense assumptions. Model from the same place they do.
- Ignoring the prepayment structure on a short expected hold.
- Applying with registry gaps unresolved. Income that may not be legally collectible is income the lender discounts.
- Forgetting insurance in the expense model. California multifamily insurance has repriced severely, and a stale insurance assumption breaks your coverage math.
- Taking bridge debt without an underwritten takeout. Bridge is designed to be refinanced out of. If the exit loan does not pencil at today’s terms, the bridge is a trap.
- Treating the compliance file as a sale issue. It is a financing issue first.
Frequently Asked Questions
What is the minimum number of units for an agency multifamily loan? Agency programmes carry minimum unit counts, and five units is the common threshold for multifamily treatment. Below that you are generally in residential or small-balance commercial territory. Confirm the current minimum with the lender, because programme parameters change.
Is agency debt non-recourse? Generally yes, subject to standard carve-outs for fraud, misapplication of funds and similar. That non-recourse feature is one of the main reasons to accept the additional structure.
Can I get a fixed rate for ten years from a bank? Rarely. Banks and credit unions typically fix for five, seven or sometimes ten years with a reset or a balloon, because they are funding from deposits. A long fixed term is the agency product.
Do lenders finance rent-controlled buildings in Los Angeles? Yes, routinely. Regulation constrains the income growth they will assume, not their willingness to lend. What actually causes problems is incomplete registration or undocumented increases, because that puts the collectibility of the stated rents in question.
Will my loan be sized off the purchase price? No. It will be sized off the lower of a loan-to-value test, a debt service coverage test and a debt yield test. On most LA buildings the coverage test binds, which means your in-place rent roll determines your proceeds.
Why is my loan smaller than the same loan would be in Texas? Because LA buildings carry low in-place income relative to value. The coverage test is a function of income and interest rate, not of price, so a low-yielding asset supports less debt regardless of what it is worth.
Can I use a 1031 exchange and new financing together? Yes, and it is common. The sequencing matters, because exchange deadlines and loan timelines both run on calendars that do not care about each other. Start the lender conversation before you identify, not after. The exchange mechanics are in the California 1031 guide.
What does a bridge loan actually cost compared to agency? Substantially more, and the gap is wider than most buyers assume once origination and exit fees are included. Bridge is priced for speed and for risk the permanent lenders will not take. Use it when the building cannot qualify otherwise, not to save time on a stabilized purchase.
When does bridge debt make sense on an LA building? Significant vacancy, mid-renovation, an incomplete soft-story retrofit, a lease-up, or a closing timeline that no permanent lender can meet. In each case the bridge buys you the time to reach the condition that qualifies you for cheaper money.
Does Measure ULA affect my loan? Not your loan directly, but very much your exit math. The transfer tax applies on sale above the threshold, assessed on gross consideration, so it changes the net proceeds your loan has to be repaid from. Factor it into the exit, not the entry.
How long should I allow for financing in my purchase contract? Agency sixty to ninety days, bank forty-five to sixty, bridge two to four weeks. Negotiate the contingency around the longer end of your lane’s range rather than the optimistic end.
What single thing most improves my terms? A complete, organised loan package, delivered early. It costs nothing, it shortens the process, and it changes how the lender reads the sponsor, which is priced into the deal whether anyone says so or not.
Where to Start
Before you talk to a lender, get your trailing twelve-month statement and your current rent roll in front of you and calculate the in-place net operating income with conservative expense assumptions. That single number, more than the price, determines what you can borrow. If it does not support the debt your plan needs, you have either the wrong lane or the wrong price, and it is much cheaper to learn that now.
If you are weighing a purchase, a refinance or a sale and want to see what the debt actually supports on a specific building, I am happy to run it. A free valuation includes the financing picture alongside the pricing, because on a leveraged deal those two answers are the same answer.
Evelyn Baez Nguyen Multifamily Investment Specialist, Lyon Stahl Investment Real Estate
