The Simple Pattern

The mistake is treating commercial real estate like one giant market. It is not. Office is not retail. Retail is not industrial. Industrial is not multifamily. Multifamily is not hospitality.

The real pattern is a split market. The broad CRE index can stop falling because some property types are holding up. At the same time, the office tail can still be in a real valuation crash.

The Broad Market Can Stabilize While Office Still Breaks

The Federal Reserve’s May 2026 Financial Stability Report says real CRE prices continued to stabilize after the steep decline from mid-2022 to early 2024. Green Street’s all-property index was also modestly positive year over year through April 2026.

That matters, but it does not mean the pain is gone. It means the broad average has stopped falling like it was during the first stage of the rate shock. The broad average can improve because retail, industrial, multifamily, and hotels are not all acting like distressed downtown office.

This is why the headlines feel contradictory. One side says CRE is stabilizing. The other side says office towers are selling for 80% or 90% off. Both can be true because they are not talking about the same part of the market.

The Sector Split

Office is still the center of the problem. CBRE reported U.S. office vacancy at 18.6% in Q1 2026. That is not normal. That is a damaged operating environment.

But retail is not acting like office. Multifamily is pressured in oversupplied markets, especially where too much new product hit at once, but it is not broadly an office-style collapse. Industrial is digesting supply. Hotels are cyclical, but their operating data are not the center of this credit problem.

The Debt Wall Is the Real Trigger

The real trigger is not just vacancy. It is debt. A building can limp along for a while if the loan does not have to be refinanced. But once that debt matures, the old valuation has to meet the current market.

MBA estimated total commercial and multifamily mortgage debt near $5.0 trillion by Q4 2025. Around $957 billion matured in 2025, another $875 billion matures in 2026, and another $652 billion follows in 2027. That is the pressure point.

This is where the old world meets the new world. The old world had cheaper debt and more forgiving assumptions. The new world has higher rates, tighter lending, more cautious buyers, and a very different office-demand reality.

The $10M to $2M–$3M Question

This is the part that sounds exaggerated until you look at the actual office transaction tape.

An 80% markdown takes a $10 million property to $2 million. A 70% markdown takes it to $3 million. Public office sales have already shown discounts in that range or worse.

A St. Louis office tower went from around $205 million to $3.6 million, which is about a 98% decline. A Chicago CBD office tower went from $306 million to $41 million, down about 87%. Another Chicago Loop tower went from $302 million to $45 million, down about 85%. San Francisco and Manhattan have also produced major markdowns.

That does not mean every CRE asset is down that much. It means the office tail is real.

Why Banks Make This Hard to See

CMBS is visible. Bank books are slower. That is the reason the public story always feels late.

CMBS data show delinquency, special servicing, maturity problems, and reappraisals more clearly. Private bank portfolios do not reveal the same kind of loan-level marks in real time. So the stress can be real before it is fully visible.

That does not mean every bank is hiding the same problem. Some banks have more CRE exposure. Some have less. Some have more office exposure. Some have more diversified books. That is why the banking risk is a concentration problem, not a universal statement.

What This Means

The CRE downturn is real, but the right answer is not panic and it is not denial.

The right answer is segmentation. Office is the epicenter. CMBS is the visible stress signal. Banks are the slower mark channel. The maturity wall is the forcing mechanism. Public transactions show the tail risk. The broad index shows why the whole market is not one giant collapse.

That is the actual Pattern Nexus read: the system is not dead. It is repricing through the weakest nodes first.

The CRE System Map

The commercial real estate downturn has to be mapped in layers: property type, geography, debt maturity, lender type, operating income, refinancing rate, and forced-sale price. If you collapse all of that into one headline, the article becomes useless.

The real CRE map looks like this:

Why the Index and the Fire Sales Can Both Be True

This is the part that needs to be explained better because it is where readers get lost. They see one headline saying CRE is stabilizing, then another headline saying an office tower sold for 90% off. They assume one of those stories has to be fake.

It does not. They are measuring different things.

A broad index is an average across sectors and markets. It can stabilize when better assets offset weaker assets. A forced-sale office tower is not the average. It is the tail. And in a credit cycle, the tail is where the truth shows up first because those are the assets that cannot hide behind time anymore.

Sector Divergence

Office is still the broken part because the operating model changed. Hybrid work, tenant downsizing, higher capex requirements, older building obsolescence, weaker downtown foot traffic, and higher rates all collide in the same place.

Retail is not in the same position. A lot of retail already went through its death narrative years ago, and the surviving formats are often supply-constrained. Grocery-anchored and necessity retail are very different from a half-empty downtown office tower.

Multifamily is more mixed. There is real pressure in markets where too much supply hit at once. But housing demand did not disappear. That makes multifamily a refinancing and supply-cycle issue, not a broad demand collapse.

Industrial is digesting the supply wave after the pandemic logistics boom. That is different from structural abandonment. Hotels are cyclical and can get hit by travel, labor, rates, and local demand, but they are not the center of this specific CRE credit story.

CMBS as the Visible Stress Layer

CMBS matters because it gives the public a cleaner window into distress. The April 2026 Trepp overall CMBS delinquency rate was 7.54%. KBRA reported office CMBS delinquency at 13.9% in January 2026. Fitch reported office special servicing at 14.5% in April 2026.

That is not background noise. That is a visible distress signal.

Special servicing also matters because it means the loan is not simply late in a clean way. It usually means the loan needs active workout attention: maturity default, payment stress, transfer, modification, extension, foreclosure path, or some kind of negotiated resolution. It is where the system admits the loan no longer fits the original box.

The Bank Mark Blind Spot

The bank side is harder because private bank books do not show the same real-time transparency. A bank can extend. A borrower can inject equity. A loan can be modified. A property can avoid a forced sale for a while. That does not mean the value is fine. It means the loss has not been forced into the open yet.

This is where the “hidden mark” idea becomes important. A building may still carry a valuation in a loan file that no real buyer would pay today. That gap can persist until a maturity, refinance, default, appraisal event, sale, foreclosure, or regulatory pressure forces recognition.

That does not mean every bank is lying. It means private credit systems reveal stress slower than public transaction systems. That is just how the structure works.

Transaction Evidence

The transaction tape is the hardest evidence because it shows what a real buyer actually paid. That is why the office fire-sale examples matter.

The St. Louis tower example around a 98% markdown is not “the CRE market.” It is a tail asset. But the tail asset proves the tail exists. The Chicago and San Francisco examples show the same thing. Manhattan repricing shows even stronger markets are not immune.

The PN Model

The PN model is simple: broad stabilization does not erase tail impairment. The system is not repricing evenly. It is repricing through weak collateral, debt maturity, and lender opacity.

That means the CRE downturn should be read as a delayed-discovery system. The first wave was rates. The second wave was refinancing. The third wave is recognition. And recognition does not happen evenly. It happens where the debt has to roll, where income cannot support the old loan, and where buyers refuse to pay the old mark.

Research Architecture

This article is built from three layers: evidence, credit mechanism, and Pattern Nexus synthesis.

The evidence layer includes Federal Reserve financial stability data, Green Street property price data, CBRE and JLL sector fundamentals, MBA mortgage maturity data, Trepp/KBRA/Fitch CMBS distress data, FDIC/OCC banking data, public SEC filings, academic bank-risk papers, appraisal research, and public distressed-office transaction examples.

The credit mechanism layer asks what forces hidden values into the open. The answer is debt maturity. A property can carry an old mark for a while, but when the loan matures, the owner has to refinance, sell, inject equity, extend, modify, or default. That is where the old valuation meets the current buyer and lender environment.

The PN synthesis layer asks what this means structurally: the system is not collapsing evenly. It is revealing stress through the nodes where income, debt, collateral value, and market liquidity no longer line up.

Executive Summary

The current U.S. CRE downturn is serious, but highly uneven. The strongest evidence does not support the idea that “all CRE” is down 70% to 80%. It does support the idea that older, commodity office buildings in troubled downtowns can be down that much or even more when forced into a real market-clearing sale.

By contrast, retail, multifamily, industrial, and hospitality are generally experiencing slower rent growth, refinancing pressure, or localized oversupply, not collapse. At the national level, the Federal Reserve says aggregate inflation-adjusted CRE prices have largely stabilized after the mid-2022 to early-2024 decline, and Green Street’s all-property index was modestly positive year over year in April 2026.

The specific $10 million to $2 million–$3 million claim is plausible for a subset of office assets. Publicly reported office sales show markdowns of approximately 85% to 98% in places such as St. Louis, Chicago, and San Francisco, with large discounts also showing up in Manhattan. Those examples are real, but they are not representative of average industrial warehouses, grocery-anchored retail, typical multifamily assets, or stabilized hotels.

Market-Wide Severity

The best way to state the severity is this: the downturn is not a uniform CRE crash; it is a severe repricing of specific office cohorts inside a broader, refinancing-driven credit cycle.

The Fed’s May 2026 Financial Stability Report says real CRE prices continued to stabilize after the steep decline from mid-2022 to early 2024, while Green Street reported the all-property index up 3.1% year over year through April 2026. That does not describe a market in free fall overall. It describes a market where the broad index has found footing while ugly pockets remain under heavy stress.

The biggest divide is by property type. Office remains the problem child. CBRE reported 18.6% U.S. office vacancy in Q1 2026, while JLL reported positive net absorption for a third straight quarter and modest same-asset rent growth. That suggests office may be stabilizing operationally, but from a weak base. By contrast, retail vacancy was 4.4% with rent growth still positive; multifamily vacancy was 4.8% with rents roughly flat to slightly up; industrial vacancy was 7.5% and expected to trend lower as supply is absorbed; and hotels posted modest year-over-year gains in occupancy, ADR, and RevPAR, even though urban hotels remained below 2019 occupancy levels.

Sector Table

Debt Transmission

The market is severe enough that it is now best understood through the credit channel, not just asset headlines. MBA estimated total commercial and multifamily mortgage debt outstanding at nearly $5.0 trillion by Q4 2025, and commercial banks held the largest share — about $1.8 trillion, or 37% — in Q3 2025. FDIC’s Q1 2026 Quarterly Banking Profile also said non-owner-occupied CRE loans had risen to $2.1 trillion, with multifamily portfolios at $625.1 billion.

The refinancing wall remains enormous. MBA says $875 billion — about 17% of outstanding commercial mortgages — matures in 2026, followed by $652 billion in 2027. Trepp adds that the CMBS market faces $76.6 billion of hard maturities in 2026 and that about $115 billion of CRE loans maturing before the end of 2026 had in-place DSCR below 1.20x.

Even if headline pricing is no longer cascading lower nationally, those maturities still force real mark-to-market decisions.

Bank Risk

The bank channel is more opaque and more contested. The OCC’s Spring 2026 risk report says overall bank credit risk remains manageable, but it specifically flags CRE refinancing risk and notes continuing legacy problem loans in office.

At the same time, recent academic work paints a more fragile picture for smaller and regional banks. Jiang, Matvos, Piskorski, and Seru estimate that a 10% CRE default rate could create roughly $80 billion in bank losses and a 20% default rate about $160 billion; by 2024 valuation conditions, that could put dozens to more than 300 mostly smaller regional banks at risk of solvency-run dynamics.

Hinzen, Severino, and Van Nieuwerburgh argue that reported delinquencies understate undercollateralized loan risk by a factor of four and that roughly one-third of U.S. commercial mortgage dollars sit on regional bank balance sheets.

But there is an important counterweight. Glancy’s 2026 Fed paper, using supervisory data, does not find that large banks broadly loosened extensions after the 2023 banking stress episode. Instead, it finds that after 2022, borrowers were generally required to make more principal paydown, provide more guarantees, or accept higher spreads to obtain extensions, with especially tight treatment for riskier and office-collateralized loans. In other words, extend-and-pretend may exist in parts of the system, but it should not be assumed as the whole story.

Appraisals and Transactions

The private bank mark question is directionally well grounded, but the evidence is indirect because bank-level marked valuations are mostly not public. What is public are distressed transaction prices, CMBS servicing and appraisal actions, bank disclosures, and academic evidence showing appraisal smoothing and valuation bias.

Recent research using NCREIF data finds meaningful deviations between appraised values and later transaction prices, reinforcing the long-known point that private-market appraisals can lag turning points.

That is why transaction evidence matters more than narrative in this cycle. The transaction evidence is strong enough to validate the possibility of a $10 million asset clearing at $2 million to $3 million if it is the wrong office building in the wrong submarket with the wrong lease roll and the wrong debt stack.

Case Study Matrix

These examples matter, but they are not the full market. They validate the tail, not the mean. That distinction is the difference between analysis and panic.

Where the Narrative Is Right and Wrong

The syndicator and broker narrative is right about three things. First, price discovery in older downtown office is often far worse than broad national indexes suggest. Second, private bank and debt-fund books are much less transparent than CMBS, so the public usually sees pain later than market participants do. Third, refinancing pressure is the mechanism forcing latent losses into view.

That narrative overstates the case when it implies that all CRE sectors are in comparable trouble, or that the average property is seeing 70% to 80% valuation destruction. Retail still has low vacancy and positive rent growth; multifamily demand has improved as completions slow; industrial is digesting supply rather than collapsing; and hotels are showing modest operating improvement.

Confidence Map

The most important limitation is that bank-held appraisal values, internal criticized-asset marks, and debt-fund workout books are not comprehensively public. If those were public, this report could say much more precisely how often a $10M to $2M–$3M reset is occurring outside CMBS and outside headline office trades.

Final Technical Model

The CRE cycle is best understood as a delayed price-discovery system.

Phase one was the rate shock. Higher rates broke the old valuation math. Phase two was the liquidity slowdown. Buyers, lenders, and borrowers stopped agreeing on price. Phase three is the maturity wall. Debt starts forcing decisions. Phase four is recognition. Some assets extend, some recapitalize, some sell, some default, and some finally clear at prices that expose how wrong the old mark was.

That is why this is not a clean crash and not a clean recovery. It is a staggered recognition cycle.

Pattern Nexus read: the broad index tells you whether the system is still falling as an average. The distressed office sale tells you what happens when the weakest node is forced to clear. The real signal is the gap between the two.