The city wants 140 homes on a vacant lot in East Harlem. It wants them affordable under 485-x. It wants no public subsidy. And it wants a developer to make the math work.
That is the tension inside the RFP the New York City Economic Development Corporation issued Monday for B-East, a city-owned site on Second Avenue between East 125th and East 126th Streets. The lot has sat empty for nearly two decades, carrying approvals from a 2008 plan that envisioned roughly 1,000 homes across a three-parcel, six-acre assemblage. To date, more than 450 units have been delivered nearby, along with the 121,000-square-foot Proton Center. B-East is the piece that has not moved.
The RFP is not asking developers to build luxury condos and write a check to the city. It is asking them to underwrite a 140-unit mixed-use building with affordable housing under 485-x, neighborhood retail, community space, a local hiring program, a 30 percent M/WBE participation goal, and carbon-neutrality targets. No subsidy. The zoning is C6-3, which permits residential and commercial density. The question is whether the density, the rent restrictions, and the construction costs produce a return that a credible sponsor can defend to its capital partners.
This is where the counterparty map matters. The city needs housing production, political cover from the 2008 plan's unfinished promises, and a demonstration that 485-x can work without direct subsidy. The developer needs a basis low enough to absorb construction risk, rent growth that materializes on schedule, and an exit that does not depend on a cap rate compression that may not arrive. Those two sets of needs are not aligned. They are adjacent, and the gap between them is the negotiation.
The city controls the basis. It owns the land. It can set the ground lease or sale price at a level that makes the pro forma work, or it can hold out for maximum value and watch the RFP draw thin responses. The 2006 RFP that launched the East 125th Street Development was issued in a different rate environment, with different construction costs and different political expectations. The 2026 RFP lands in a market where hard costs have not come down, where 485-x is still being tested in the underwriting community, and where lenders are cautious about construction loans that depend on future affordable rent streams without a subsidy backstop.
The developer's constraint is time and certainty. A 140-unit building with 485-x affordability means rent growth is capped for the compliance period. The developer cannot count on market-rate upside to rescue a thin underwriting. The exit will likely be a sale to a permanent affordable housing operator or a REIT that understands the 485-x income stream. That buyer pool is real but narrow. The developer needs to know, before bidding, whether that exit will clear at a basis that covers the construction loan and the land cost.
The lender's constraint is similar. Construction financing for affordable housing without subsidy is not impossible, but it requires a sponsor with a strong balance sheet, a general contractor with a track record, and a rent roll that the lender can model with confidence. The lender will want to see that the 485-x rent schedule, the operating expenses, and the debt service produce a DSCR above 1.20 even in a stress case. If the land basis is too high, the debt yield will not support the loan.
The RFP also requires respondents to maximize the site's development potential within its C6-3 zoning. That is a signal. The city wants density. It wants the developer to push the zoning envelope to generate enough units to make the affordable requirement economically viable. But density also means more construction cost, more risk, and a longer lease-up. The developer must decide whether the marginal unit adds more to revenue than it adds to cost and timeline.
Manhattan Borough President Brad Hoylman-Sigal noted in a statement that the B-East lot has sat vacant for nearly two decades despite carrying its approvals. That is a political fact, not an economic one. The lot has been vacant because the math did not work. The 2008 plan was approved before the Global Financial Crisis, before the pandemic, before interest rates rose, before construction costs surged. The approvals survived. The economics did not.
The RFP lands alongside a larger reimagining of the district's remaining parcels. The A-Block sites, A-West and A-East, are expected to yield more than 1,000 homes, 40 percent of them affordable, plus two new public schools. Those sites are larger and may benefit from different subsidy structures or school construction funding. B-East is the test case for the no-subsidy 485-x model on a smaller, more constrained lot.
What should a developer test before responding? First, the land basis. The city has not disclosed a minimum price. The developer should model the maximum land cost that still produces a 12 to 15 percent levered IRR on a 10-year hold, assuming 485-x rent growth at the statutory cap, operating expense inflation at 3 percent, and a terminal cap rate of 5.75 to 6.25 percent. If that maximum land cost is below what the city expects, the bid will be a pass or a lowball.
Second, the construction cost. Hard costs in Manhattan for mid-rise residential with affordable requirements are running $500 to $650 per square foot, depending on foundation conditions, facade, and MEP systems. The developer should get a preliminary GC bid before committing to a land price. If the GC cannot guarantee a price within 5 percent for 18 months, the risk is too high.
Third, the exit. The developer should identify three to five potential buyers for a stabilized 485-x asset and get preliminary indications of the cap rate they would underwrite. If the buyer pool is thinner than expected, the exit risk is higher, and the land bid must be lower.
The RFP responses are due October 20, 2026. That gives developers roughly three months to run the numbers, talk to lenders, and decide whether the city's terms and the market's reality can meet in the middle. The lot has waited 18 years. It can wait a little longer for a deal that works.