Why are multifamily properties decreasing in 2026

By Ken Lundin, Author, Operator and Investor

Why are multifamily properties decreasing in value? The gap between seller expectations and actual cash flow became unbridgeable. I’ve watched this movie three times. Prices inflate on projected rents and exit cap-rate dreams. Then they snap back when buyers underwrite what’s actually in place. Assets that traded at 4.2% cap rates in 2021 now bid at 5.8% to 6.5%. The buildings didn’t get worse. The lazy math finally stopped working.

Sellers anchor to comps from a world where money was free. Every pro forma assumed 95% occupancy by month eighteen. Buyers who survived the last correction know better. They price trailing-12 NOI. They stress-test rent growth at zero. They walk when the maximum allowable offer lands too far below ask. The same rent difference is never counted twice as both loss-to-lease and renovation premium, according to Lundin Capital acquisition records and screening standards from our August–October 2026 pipeline of 57 leads. Multifamily completions fell to 484,000 units in 2025, according to the National Association of Home Builders (analysis of Census Bureau Survey of Construction). That followed the highest number in nearly 40 years in 2024. 57% of new units were in buildings of 50 or more units. The question isn’t whether values are falling. It’s whether you know how to price what’s actually there today.

Key Takeaway: Multifamily property values are declining because buyers stopped accepting broker projections. They now underwrite actual in-place cash flow. Sellers remain anchored to 2021 transaction comps. Those comps reflected sub-4% cap rates and aggressive rent-growth assumptions. Assets now trade 140 to 230 basis points wider than peak pricing. The correction isn’t driven by deteriorating fundamentals. It’s the market finally rejecting inflated pro formas. Those pro formas counted the same income twice and ignored operating reality. Disciplined buyers price trailing-12 NOI. They walk when the numbers don’t clear their return threshold.

TL;DR

Cap rates expanded 140-230 basis points since 2021. Buyers now underwrite in-place NOI instead of broker pro formas built on fantasy rent rolls.
The 78% maximum-offer rule forces you to walk. Your top bid is based on stabilized NOI and realistic execution risk. If it lands too far below the seller’s ask, you’re done.
Most deals fail at first screen. The property isn’t bad. The gap between what you can pay and what the seller believes it’s worth is unbridgeable.
Voucher and LIHTC rents are held flat through the hold unless the contract says otherwise, and exit value uses trailing-12 NOI, according to Lundin Capital acquisition records and screening standards from our August–October 2026 pipeline of 57 leads. This eliminates the subsidy-growth fantasy that inflated 2021 valuations.

Price Today’s Cash Flow, Not Tomorrow’s Story

I don’t price deals off broker comps or market cap rates. I price them off what the property makes today. I price them off what it could make if I didn’t screw it up. I decide whether to keep going based on one test. Can my maximum offer get within 78% of the ask? If it can’t, I’m out before I waste time on a full model.

The first screen produces a price range, not a single recommended offer: the bottom of the range prices today’s in-place NOI, the top prices stabilized NOI through the 78% maximum-offer math, according to Lundin Capital acquisition records and screening standards from our August–October 2026 pipeline of 57 leads. That range tells me if there’s a deal here. Or if the seller is still living in 2021.

Step 1: Pull In-Place and Stabilized NOI

Step one of the funnel may hold nothing more than NOI and an asking price or cap rate, according to the same Lundin Capital screening standards. Its only job is to test whether the maximum allowable offer lands within 78% of the seller’s ask. Many declines at this step are expected. I’m not building a 10-tab model. I’m grabbing trailing-12 NOI from the broker package. Then I estimate stabilized NOI by plugging obvious holes. Vacancy that shouldn’t be there. Rents $150 below market. Property tax that’s clearly wrong. Pre-LOI work exists to decline deals Lundin Capital would not do under almost any circumstance, according to the same screening standards. It bubbles up the rest with real pricing guidance. Educated assumptions are acceptable because it is not a final negotiation.

Step 2: Price Both Ends of the Range

Lundin Capital does not price off a market cap rate, according to the same screening standards. It prices on stabilized NOI and the 78% factor. Every rent increase or expense reduction inside that stabilized NOI must be vetted. Broker expense figures are used only as conservative guidance. I take in-place NOI. I divide by my required yield. That’s the floor. Then I take stabilized NOI. I run it through the maximum-offer formula. That’s the ceiling. If the ceiling is below 78% of the ask, I decline and move on.

Step 3: Compare to the Ask and Decide

Lundin Capital prices from value today and the maximum-offer formula, according to the same screening standards. Never from the number a broker is trying to trade at. Once a broker has put a price in writing, that price is treated as the real ask. Why are multifamily properties decreasing in value right now? Because buyers like me finally stopped pretending the top of the range is real. It requires heroic assumptions. We’re walking when our maximum lands at 65% of what the seller still thinks the building is worth.

Vet Every Line of Stabilized NOI Like It’s Your Money

Step 1: Start with rent growth you can actually defend

I use 2% to 3% annual rent growth in stabilized NOI. Not because I’m conservative. Because I can cite local comps, trailing lease-ups, and renewal data that support it. Brokers love to show you 5% or 6%. That’s what happened in 2021. Or “the submarket is hot.” That’s not underwriting. That’s a story. Stabilized NOI carries 2% to 3% annual rent growth and is checked net of the capex the plan implies, according to Lundin Capital acquisition records and screening standards from our August–October 2026 pipeline of 57 leads. If you can’t defend the number with trailing data from comparable properties in the same zip code, you don’t use it.

Step 2: Use the expenses the property actually runs at, not the ones the broker invented

Broker pro formas love to show you a 42% expense ratio. The trailing twelve months say 51%. They’ll tell you the new owner will “operate more efficiently.” Maybe. But I underwrite what the T12 shows. Then I model specific line-item cuts I can name and execute. If you can’t identify the exact contract you’ll renegotiate, you don’t get credit for the savings. If you can’t name the exact service you’ll bring in-house, you don’t get credit for the savings.

Step 3: Never count the same rent increase twice

This is where most underwriting falls apart. The same rent difference is never counted twice as both loss-to-lease and renovation premium, according to Lundin Capital acquisition records and screening standards from the same pipeline. Market rent is $1,200. In-place is $1,050. You plan to renovate and push to $1,250. You don’t get to count the $150 loss-to-lease and the $200 renovation bump as separate sources of NOI growth. You get one: the total move from $1,050 to $1,250. I’ve seen deals modeled with 18% NOI growth. The underwriter counted turnover, loss-to-lease recovery, and unit upgrades as three separate lifts on the same 40 units.

Step 4: Hold subsidized rents flat unless the contract says otherwise

Voucher and LIHTC rents are held flat through the hold unless the contract says otherwise, and exit value uses trailing-12 NOI, according to the same Lundin Capital screening standards. I don’t model HUD fair market rent increases unless I have the contract in hand. The adjustment language must be clear. I never exit on a forward cap rate. I exit on trailing-12 NOI at the time of sale. That’s what the next buyer will underwrite.

Ready to Take the Next Step?

Book a Strategy Call

Frequently Asked Questions

Why are multifamily properties decreasing in value when demand is still strong?

Because buyers finally stopped paying for what could happen. They started pricing what’s actually happening today. Demand for units doesn’t set asset prices. Cash flow does. When you’re underwriting in-place NOI against 7% cap rates instead of the 4.5% comps from 2021, values drop 30% overnight. The Seven-Factor Market Score helps you separate real demand signals from narrative. But it doesn’t override the math. If the property can’t cover debt service at today’s rates with today’s rents, the price has to fall. Multifamily completions fell to 484,000 units in 2025, according to the National Association of Home Builders (analysis of Census Bureau Survey of Construction). That followed the highest number in nearly 40 years in 2024. 57% of new units were in buildings of 50 or more units. Supply contraction doesn’t rescue deals that were overpriced on bad assumptions.

Should I use market cap rates to price multifamily deals?

No. Market cap rates are averages of deals that already closed. Most were priced on someone else’s bad assumptions. I price the bottom of my range using in-place NOI. I price with the discipline to walk if that number doesn’t work. Not by backing into a valuation because “the market” says I should. If you’re using a market cap rate to justify an offer, you’re not underwriting. You’re inheriting a playbook written by people who don’t write the check. 82% of CEOs rely on an informal ‘kitchen cabinet’ of friends, acquaintances and former colleagues for advice on sensitive workplace issues, according to Stanford GSB / Hoover Institution / The Miles Group (2025 CEO Coaching and Kitchen Cabinet Survey). CEOs rely on four informal advisors on average. Many investors rely on market consensus instead of disciplined analysis. Both approaches substitute groupthink for independent judgment.

What is the 78% maximum-offer rule and why does it matter?

It’s the threshold that tells you when to stop talking to a seller. You stop before you waste time on full underwriting. The first screen produces a price range. The bottom prices today’s in-place NOI. The top prices stabilized NOI through the 78% maximum-offer math. If your maximum offer can’t reach 78% of the ask, you’re already done. I’ve seen buyers spend weeks modeling deals that were dead in the first fifteen minutes. They didn’t have the discipline to walk early.

How do I calculate stabilized NOI without inflating the numbers?

You defend every line like it’s your money. Because it is. Rent growth has to match what comparable units actually get. Not what the broker’s pro forma imagines. Expenses reflect the property’s deferred reality. Not some fictional 40% expense ratio. Capex matches the renovation plan you’d actually execute. The same rent difference is never counted twice as both loss-to-lease and renovation premium. If you can’t defend a line item to your LP without flinching, it doesn’t belong in the model.

Can I count both loss-to-lease and renovation upside in the same unit?

Not unless you want to count the same dollar twice. Not unless you want to lie to yourself about returns. A unit is either $150 below market because it hasn’t been turned. Or it needs $8,000 in capex to justify a $150 rent bump. It’s never both. I’ve reviewed dozens of broker packages that layer loss-to-lease on top of renovation premiums. Suddenly the pro forma shows 22% rent growth that doesn’t exist in reality.

What happens if my maximum offer is below the seller’s asking price?

You walk. Or you wait for the seller to get realistic. Step one of the funnel may hold nothing more than NOI and an asking price or cap rate. Its only job is to test whether the maximum allowable offer lands within 78% of the seller’s ask. Many declines at this step are expected. I don’t negotiate against myself. I don’t waste time building a full model when the range test already told me we’re $4 million apart. The seller still thinks it’s 2021.

How much rent growth should I assume in stabilized NOI?

Whatever you can defend with trailing-12 comps from comparable units in the same submarket. Nothing more. If renovated units three blocks away are getting $1,450 and in-place rents are $1,250, your stabilized rent is $1,450. Not $1,550 because you want the deal to work. Voucher and LIHTC rents are held flat through the hold unless the contract says otherwise, and exit value uses trailing-12 NOI. Rent growth is evidence. Not aspiration.

Do I need to adjust my underwriting for different property classes?

Yes. But the discipline stays the same. Class A properties in strong markets may support 3% rent growth if you can cite trailing comps. Class C in secondary markets may only support 1% to 2%. The framework doesn’t change. You’re still pricing in-place NOI at the bottom of your range. You’re still pricing stabilized NOI through the 78% maximum-offer math at the top. What changes is the evidence you use to defend each line item. A Class B value-add deal in a growing submarket gets different rent-growth assumptions. A Class C cash-flow play in a declining market gets different assumptions. Both get underwritten with the same rigor.

What if the seller won’t budge on price?

Then you walk. I’ve never seen a deal get better by convincing myself the seller’s number makes sense. Not when my underwriting says it doesn’t. Markets turn. Sellers get realistic. Or someone else overpays and you move on. The discipline is knowing your maximum offer before you start negotiating. Having the guts to walk when the gap is too wide. Most buyers waste months trying to make a bad deal work. They fell in love with the property or the story. I fall in love with cash flow. If the cash flow isn’t there at a price that works, neither am I.

How do I know if my stabilized NOI assumptions are realistic?

You stress-test them against three scenarios. What happens if rent growth is zero. What happens if vacancy runs 5% higher than projected. What happens if expenses come in 8% above your model. If the deal still clears your return threshold in all three scenarios, your assumptions are defensible. If it only works when everything goes right, you’re not underwriting. You’re hoping. Pre-LOI work exists to decline deals Lundin Capital would not do under almost any circumstance and to bubble up the rest with real pricing guidance. Educated assumptions are acceptable because it is not a final negotiation. But those assumptions have to survive contact with reality.

Should I adjust my cap rate assumptions based on current market conditions?

No. Because I don’t use market cap rates to price deals in the first place. Lundin Capital does not price off a market cap rate. It prices on stabilized NOI and the 78% factor. Every rent increase or expense reduction inside that stabilized NOI must be vetted. Broker expense figures are used only as conservative guidance. Market cap rates are backward-looking averages of deals that already closed. Most were mispriced. I price from cash flow and required returns. Not from what someone else was willing to pay six months ago.

Bottom Line

Values are falling because the underwriting that mattered in 2021 doesn’t matter anymore. What the broker said it could be worth doesn’t matter when your lender prices the loan on trailing-12 NOI. Your LP wants to see cash flow in year one. The operators still buying are the ones who figured out that the first screen produces a price range, not a single recommended offer. The bottom prices today’s in-place NOI. The top prices stabilized NOI through the 78% maximum-offer math. If your top can’t reach the ask, you walk before you waste a weekend on a model that was never going to work. The correction isn’t a crisis. It’s the market finally pricing what’s real instead of what a deck promised.