Multifamily loan balances on U.S. bank books jumped 53 percent since the first quarter of 2019, reaching $665 billion by early 2026. That is the fastest growth of any real estate lending category, and it is the kind of figure that gets cited in investor letters as evidence of a sector's momentum.
The more revealing number is in construction and development lending. Indexed to 2019, construction balances rose to a peak near 142 in 2024 before declining to 128 by the first quarter of 2026. That rise-and-partial-reversal pattern is not a gentle plateau. It is a recognition that the projects underwritten in 2021 and 2022 are now delivering into a higher-rate, higher-cost environment, and the banks that funded them are no longer eager to double down.
According to CRED iQ data, total bank real estate loan balances stood at $6.14 trillion in the first quarter of 2026. The composition has shifted measurably since 2019. Multifamily grew fastest in percentage terms, but in absolute dollars, core commercial real estate added $467 billion and residential added $445 billion, each roughly double multifamily's $229 billion addition. The percentage-versus-dollar divergence is not a statistical curiosity. It reveals where banks concentrated their incremental exposure relative to the starting base, and where the maturity wall is thickest relative to the underlying asset's ability to refinance.
Multifamily's 53 percent growth was a bet on rent growth, low rates, and endless demand for housing. For a period, that bet paid. But the bet was also a bet on time: time for rents to catch up to basis, time for interest rates to stay low, time for exits to materialize. Time is the most expensive ingredient in any capital decision, and the market is now being charged for it.
The construction reversal matters because it is the leading indicator for the broader multifamily book. If development lending is pulling back, the refinancing of those completed projects will fall onto the existing multifamily balance sheet, which is already 53 percent larger than it was seven years ago. That is not a crisis in itself. It is a concentration of exposure that will test the relationship between basis and liquidity.
Residential lending grew only 17 percent over the same period, the slowest of any category. That is partly a function of securitization: a substantial share of newly originated one- to four-family mortgages is sold into agency or private-label MBS rather than retained on bank balance sheets. But it also reflects a market where rising rates suppressed origination volume and where banks chose to hold fewer of the loans they did make. The residential book is $3.1 trillion, roughly half of all bank real estate lending, but its growth has been the most restrained. That restraint is a form of discipline, and it stands in contrast to the multifamily build.
The question for multifamily owners and their lenders is not whether the 53 percent growth was justified. It is whether the assets behind those loans can generate enough cash flow to service the debt at today's rates, and whether the banks that hold the paper have the patience to wait for the answer.
Patience is not a market condition. It is a decision. And every decision to extend a loan is a bet that the borrower's discipline will outlast the lender's discomfort.
For sponsors with maturities in the next 18 months, the implication is straightforward: the banks that lent into the multifamily boom are now managing concentration, not growth. They will extend for credible sponsors with stabilized assets and defensible basis. They will not extend for stories. The difference between those two outcomes is the difference between a loan that gets modified and a loan that gets transferred to special servicing.
For lenders, the data is a reminder that balance sheet growth is not the same as balance sheet strength. The 53 percent increase in multifamily balances is a measure of how much capital was deployed. It is not a measure of how much of that capital will return on schedule. The construction lending reversal suggests that banks are already adjusting their risk appetite. The question is whether the adjustment is fast enough to matter.
The market is not rewarding the lenders who grew the fastest. It is rewarding the lenders who grew with the most structure. And structure, in this context, means underwriting that priced for the possibility that time would not cooperate.
Time did not cooperate. The 53 percent growth in multifamily balances is now a book of loans that will need to be refinanced, restructured, or resolved. The next phase of the market will not be defined by who owns the best story. It will be defined by who controls the cheapest capital and the longest clock.