AB CarVal Buys $340 Million Construction-Loan Portfolio as Private Credit Moves Deeper Into Housing

Share this article!

AB CarVal is putting $340 million of multifamily and build-to-rent construction debt onto its books, a transaction that offers another window into how private credit is moving into parts of real estate lending where traditional capital has become more selective.

Funds managed by AB CarVal, the alternative investment manager owned by AllianceBernstein, acquired a portfolio of 10 performing construction loans across seven U.S. metropolitan areas, according to a Sept. 14 announcement from the firm. The collateral includes both multifamily projects and single-family build-to-rent developments in various stages of construction.

AB CarVal did not identify the seller, individual borrowers, properties, loan balances by project, coupons, maturities, leverage ratios or the seven metropolitan areas. Those omissions limit how precisely outsiders can judge the credit risk or the price paid for it. The company characterized all 10 loans as performing.

North River Partners, which has previously worked with AB CarVal on construction lending, will manage the portfolio. The two firms funded $98 million of construction loans together in 2025, according to the announcement. The new acquisition takes that relationship from individual originations into a portfolio-scale investment.

“This transaction expands our footprint in multifamily construction credit and demonstrates our ability to uncover attractive relative value opportunities across commercial real estate credit,” AB CarVal principal Scott Greenfield said.

Greenfield said the firm believes the loans can generate attractive risk-adjusted returns. That is an investment manager’s assessment, not an independent valuation of the portfolio. What can be established from the transaction is that a large alternative-credit investor sees enough opportunity in residential construction debt to deploy capital across 10 projects at once.

The important part of the deal is the debt, not the buildings

Multifamily transaction headlines often focus on apartment values, cap rates and unit counts. This deal sits one layer earlier in the development cycle. AB CarVal is buying the loans financing projects that are still being built.

Construction lending carries a different risk profile from acquiring stabilized apartment debt. Future funding obligations can remain, projects can run over budget, lease-up assumptions can change and a property that is performing as a loan today still has to reach completion and stabilization. Build-to-rent projects add their own variables because the collateral consists of scattered or community-based single-family rental housing rather than a conventional apartment building.

That is also why the transaction is useful as a capital-markets signal. Private credit firms are not merely buying distressed office notes or rescue financing. They are competing for performing real estate credit where they believe traditional lenders have left enough spread or structural protection to make the risk worthwhile.

AB CarVal describes its real estate debt strategy as spanning bridge and construction lending, structured credit and special situations. The firm says it has invested $162 billion across 5,905 transactions in 82 countries since 1987 and currently manages approximately $26 billion.

Construction capital is becoming a strategic part of the housing supply story

The timing is notable because residential construction is slowing nationally. Zillow reported Monday that residential building permits over the 12 months ending in July fell 1.7% from the prior year and have now declined year over year for 44 consecutive months. A separate WRE News analysis examines how that pullback is concentrated in several pandemic-era Sun Belt building markets.

Those two developments are not contradictory. A slowing pipeline can make lenders more cautious while simultaneously creating opportunity for investors willing to underwrite individual projects. Capital does not disappear uniformly; it gets repriced and redistributed.

For developers, the identity of the lender can materially change how a project is financed. Banks operate under regulatory capital requirements, concentration limits and supervisory scrutiny that private funds do not face in the same form. Private lenders, in exchange, typically seek economics and protections that compensate them for illiquidity and construction risk.

The result is a financing market in which a viable project can still find capital even when conventional lending standards tighten — but potentially at a different cost and with a different set of covenants.

Build-to-rent remains part of institutional housing finance

The inclusion of single-family build-to-rent loans is particularly relevant to the broader housing industry. Build-to-rent sits between traditional multifamily and for-sale homebuilding: communities are constructed as single-family homes but operated as rental portfolios.

That structure attracts renters who want features associated with detached housing — additional space, yards, garages or suburban locations — without purchasing a home. It also gives institutional investors another way to gain exposure to single-family housing without assembling scattered homes one acquisition at a time.

Higher mortgage rates can strengthen part of that demand proposition by keeping some would-be buyers in rental housing longer. But the same rates also raise development financing costs and can reduce the value investors are willing to place on completed assets. Build-to-rent therefore benefits from one side of the affordability squeeze while remaining exposed to the other.

AB CarVal did not disclose how much of the $340 million portfolio is multifamily versus build-to-rent, so the transaction should not be read as a $340 million directional bet on single-family rentals alone.

The missing terms matter

Large portfolio announcements often sound more transparent than they are. In this case, the headline number is clear, but many of the details needed to evaluate the economics are private.

Without the acquisition price, it is not possible to know whether AB CarVal bought the loans at par, a discount or another negotiated value. Without loan-to-cost or loan-to-value ratios, there is no public measure of the equity cushion beneath the debt. Without maturities and remaining construction obligations, the duration and completion exposure cannot be independently calculated.

Those gaps do not make the transaction less real. They do change what can responsibly be concluded from it.

The strongest conclusion is narrower: a major alternative-credit manager has acquired a sizable portfolio of performing residential construction loans and is expanding an existing partnership specifically around multifamily and build-to-rent credit.

Why the transaction matters beyond $340 million

Housing supply is ultimately a financing story as much as a zoning, land and construction-cost story. A permitted project still needs equity and debt. When one source of capital retreats, another has to step in or the development can stall.

Private credit has increasingly marketed itself as that flexible source of capital. The attraction for borrowers is execution and the ability to finance transactions that may not fit a bank’s current appetite. The attraction for investors is the possibility of earning higher yields with negotiated collateral and covenant protections.

The tradeoff is cost. More expensive construction debt eventually has to be absorbed somewhere in a project’s economics — through land basis, developer returns, rents, sale proceeds or some combination of them.

That is what makes the AB CarVal acquisition worth watching. The portfolio itself is only 10 loans. The larger question is whether transactions like it become the normal route for financing a greater share of the next generation of apartments and build-to-rent communities.

If they do, private credit will not simply be an investor story. It will become part of the mechanism determining which housing projects get built.