The $1.5 Trillion Commercial Real Estate Problem Sitting on Regional Bank Balance Sheets

The Office Market Has Not Bounced Back, It Has Restructured

Three years after the lifting of pandemic restrictions office vacancy rates in major US markets remain at record highs. The national average office vacancy rate has stabilized around 18-20% a level previously associated only with severe recessions. This vacancy rate is not a temporary dislocation awaiting resolution. It reflects a permanent structural shift in the way companies use office space driven by the hybrid work norm.which has taken root in all industries of the knowledge economy. The bifurcation within the office market is the central analytical fact: Class A office space in prime inner-city locations is actually doing reasonably well with vacancy rates in the range of 10-12% in major markets. Class B and C office space which comprises the bulk of office square footage in most cities is experiencing a structural collapse in demand with rates of25-35% vacancies common in suburban and secondary markets. The office market has not declined uniformly but rather has restructured concentrating viable demand at the upper end of the quality spectrum and leaving large amounts of secondary space economically stranded

The Banking System's Exposure

Regional and community banks hold approximately $1.5 trillion in commercial real estate loans about 28% of their total loan portfolios. The specific concern is the maturity wall: approximately $400 billion in CRE loans originated at interest rates of 3-4% in 2019-2021 were scheduled to mature in 2024-2026. Those loans now need to be refinanced at rates between 200 and 300 basis pointshigher at a time when property values are down 20-40% from their peaks and occupancy rates are structurally lower than when the loans were originated

Banks have largely been able to "extend and fake it" granting 12- to 24-month extensions to borrowers who can't refinance in the hope that values ​​will recover or the rate environment will improve. The Federal Reserve's maintenance of rates through 2026 has made this strategy more difficult: Each extension poses a problem down the road rather than solving it

A Worked Example: Why a Loan That Was Fine Cannot Refinance

The phrase maturity wall is used without anyone showing what happens when a specific loan hits it. Here's one with round numbers chosen for clarity

The loan drafted in 2019. A suburban office building generates net operating income of $10 million a year. At the current interest rate of about 5.5 percent the building was worth 10 divided by 0.055 or about $182 million. The bank lends 65 percent of that amount or about $118 million at only 3.5 percent interest. Annual debt service is about $4.1millions

Divide the income by the debt service and the coverage ratio is 10 over 4.1 or about 2.4 times. That is an extremely safe loan. The building could lose more than half of its income and still pay its interest. No one who approved this was being reckless

The same building in 2026. Tenants downsized after the renovation so net operating income fell 30 percent to $7 million. And with buyers now demanding a much higher return on office risk the top rate has been expanded from 5.5 percent to 8 percent. The building is now worth 7 divided by 0.08 which is about $88 million

Notice what just happened. Revenue fell 30 percent. Value fell 52 percent from $182 million to $88 million because the cap rate moved at the same time. That double whammy is the most important mechanism in commercial real estate and is the reason value drops of 20 to 40 percent are seen in buildings whose rent rolls fell much less

Now try to refinance it. A new lender looking at this building today will lend perhaps 60 percent of its current value which is about $53 million. The existing loan is $118 million. Therefore the borrower must write a check for about $65 million to close the gap

They won't. Their original 64 million of equity has been depleted and the position is below 30 million so the rational course of action is to hand over the keys to the bank and walk away. Nothing about that decision is emotional or strategic. There is simply no future rental amount that will recover 30 million of negative equity for someone who can walk away

Which puts the problem in the bank. If the mortgage is foreclosed today the bank maintains a loan valued at 118 million against collateral of 88 million a loss severity of about 26 percent. If the eventual sale approaches the 30 cents on the dollar that the struggling office operations have generated the loss is much worse

So look at the bank's actual choice. Extend the loan another two years and report it as current or recognize a loss of 26 percent or more today against the principal. Extending and faking it is not primarily dishonesty. It is a bank's arithmetic that compares a certain loss now with a possible recovery later and the incentive is in one direction

These figures are illustrative and each building is different. The structure is not. A secured loan with 2.4 times coverage became unfinanceable without a single late payment simply because the value of the collateral moved

The CMBS Market

Delinquency rates for commercial mortgage-backed securities in office CMBS pools have increased from less than 2% in 2022 to more than 8% in early 2026. The structure of CMBS creates specific risks: When the severity of the losses is high enough when a foreclosed office building sells for 30 cents on the dollar the losses penetrate through the capital structure to previously safe and investment grade tranches. VariousHigh-profile office foreclosures have already produced serious losses affecting investment grade tranches

The mechanism deserves to be explained clearly because it is what differentiates securitization from a bank loan. A CMBS arrangement stacks investors in order of who absorbs losses first. The equity and junior tranches are expected to take damage in a downturn which is what they are paid for. The upper investment-grade tranches are valued under the assumption that losses will never reach them because the lower layers are thick enough to absorb any plausible outcome. That assumption was calibrated againstthe historical severity of office losses and the previous worked example shows why they are no longer the correct reference. When the guarantee is worth half of what it was worth the cushion that seemed generous is no longer so

The path to overcoming the CRE office problem will take years: bank renegotiations on individual loans special CMBS administration of defaulted securitized loans and property conversion or demolition for the most economically unviable buildings. For regional banks with heavy concentrations of CRE in the affected markets the next two to three years will be characterized by high provisioning lower loan growth and pressure on net interest margins

Case Study: The Morning New York Community Bancorp Repriced the Sector

If you want to see how quickly it goes from slow adjustment to a crisis of confidence the clearest episode is New York Community Bancorp in early 2024

NYCB was not an obscure institution. It had grown substantially including by acquiring a large portion of the failed Signature Bank in 2023 which propelled it over a regulatory asset threshold that required higher capital and liquidity standards. It was considered one of the most conservative lenders in its niche with a long track record in rent-regulated multifamily lending in New York

On Jan. 31 2024 it reported a surprising quarterly loss cut its dividend by about 70 percent and disclosed a provision for credit losses of about $550 million driven by exposure to offices and rent-regulated multifamily buildings. The stock fell about 38 percent in a single session and about 60 percent in weeks. Moody's downgraded its credit rating to non-investment grade. In March the bank had raisedmore than $1 billion in emergency capital from a group of investors led by former Treasury Secretary Steven Mnuchin at a price that greatly diluted existing shareholders and had replaced their leadership

Three things about that sequence are important to anyone reading bank disclosures today

The losses did not come gradually. They arrived in a single provision one morning on loans that had been declared productive. That is exactly what the previous example predicts. A loan remains current until it matures or the bank decides to mark it so the deterioration is invisible and then total

Concentration was the whole story. The system-wide figure of 28 percent of loan books in commercial real estate is an average. Individual banks rose much more and those that did suffered violent revaluations while their diversified peers were barely affected. The sector average says almost nothing about a specific institution

Rent-regulated multifamily housing surprised people more than offices. Everyone was looking at the office. The provision that broke NYCB leaned heavily on apartment buildings whose revenues had been capped by a 2019 change to New York's rental law while their costs continued to rise. The lesson is widespread: The problem loan is usually one in which an assumption from the original underwriting quietly stopped being true and that's not always the asset class that shows up on the front page

Where the Doom Case Is Overstated

The $1.5 trillion headline has been used to imply a systemic banking crisis for three years running and it hasn't happened. Some of the reasons are worth taking seriously

The office is a minority of the exhibition. The 1.5 trillion figure covers all commercial real estate and offices are a modest portion of it. Most of it is multifamily industrial and retail and multifamily and industrial have held up much better than offices. Citing the total when describing the office problem overstates the exposure by a wide margin

A large part is occupied by its owners. Much of what banks classify as commercial real estate is a loan made to a business secured by the building in which it operates: a dentist a car dealership or a manufacturer. Those loans are repaid from businesses' cash flow not the rental market and behave like small business loans rather than office tower loans

Extend and pretend has a history of work. It is used as a term of abuse and after 2009 the same approach allowed a lot of deteriorated commercial real estate to recover as values ​​returned. It fails when the deterioration is structural and not cyclical which is precisely the open question about hybrid work. The strategy is a gamble and it is not stupid

Three years of supply have already passed. This has been the most anticipated credit problem in modern banking. Banks have been building reserves selling loans and reducing concentrations since 2023. Losses hitting prepared balance sheets are a very different event from losses hitting unprepared ones which is the real difference between 2024 and 2008

Much of the worst paper is not in the hands of the banks. The most aggressively underwritten office loans were securitized and those losses fall on CMBS investors insurers and debt funds. This is a genuine loss and it is dispersed among institutions that are funded with long-term capital rather than deposits that can disappear in an afternoon

My own view is that this is a long-standing earnings problem for a specific set of concentrated regional lenders rather than a systemic event and that the institutions at risk are identifiable in advance from their own disclosures

How I Would Analyse a Regional Bank's CRE Book

The industry average is useless in judging any individual bank so here is the order in which I would work with a specific one

I would start with the concentration ratio that is commercial real estate loans as a multiple of total capital and not as a proportion of loans. Regulators have long singled out banks with more than about three times capital for greater scrutiny and that ratio better identifies vulnerable institutions than any absolute dollar figure

Then you would specifically find shared offices and within them the division between the central business district and the suburbs and between Class A and everything else. The article above shows that those categories behave completely differently and a bank that reports only one office number is telling you something by default

Next you would look at the maturity schedule. The relevant question is how much of the book is due in the next twenty-four months because that is when the previous worked example stops being hypothetical. A loan that does not mature until 2029 has time for the maximum rate to rise again.Not one that expires next quarter

Then you would compare the reserve to the criticized loans rather than the overall reserve. A bank that reserves one percent of a portfolio in which five percent of the loans are already classified is telling you that it expects a recovery it has not yet seen

Finally I would focus on the deposit base because the NYCB episode was ultimately about trust more than credit. Losses can be survived in a bank with insured retail deposits much less one funded by large unsecured balances moving around on a holder

None of this is investment advice and I would like to emphasize that every single one of those figures is published making this an analyzable issue rather than a hidden one

The Bottom Line

The office market did not decline but rather restructured dividing into Class A spaces with 10 to 12 percent vacancy and Class B and C spaces with 25 to 35 percent that are economically stranded. Regional and community banks hold approximately $1.5 trillion in commercial real estate loans about 28 percent of their books with about $400 billion originated with maturities between $3 and4 percent in a market where rates are 200 to 300 basis points higher and values are down 20 to 40 percent

Arithmetic explains everything else. A building with 10 million in income at a cap rate of 5.5 percent was worth 182 million and had a loan of 118 million at 2.4 times coverage which is a conservative loan by any standard. If you reduce the income by 30 percent and expand the cap rate to 8 percent the building is worth 88 million a decrease of 52 percent. The borrower has 30 million underwater and walk and the bank chooses between a 26 percent loss today and another extension. That choice is why extend and fake exists

New York Community Bancorp showed how quickly this goes from footnote to crisis: a roughly $550 million provision on Jan. 31 2024 a 38 percent drop in one day a junk downgrade and a $1 billion bailout in March. Note the focus on capital the office division the maturity schedule and the deposit base. This is a multi-year identifiable solvable solution rather than a bomb.hidden systemic and the banks that are harmed will be those whose own disclosures say so first