DiamondRock Sold a Manhattan Courtyard for $175K Per Key. The Cap Rate Tells the Real Story.
A 13.3% trailing cap rate on a Manhattan hotel sale doesn't signal distress. It signals a REIT that ran the numbers on $12 million in deferred capex, a ground lease escalation, and July union negotiations, and decided someone else could hold that bag.
DiamondRock just sold its leasehold interest in a 189-room Courtyard by Marriott on Fifth Avenue for $33 million. That's $174,600 per key on a trailing 13.3% cap rate. Those two numbers don't belong in the same sentence for a Manhattan hotel... unless you decompose what's underneath them.
The 13.3% cap on trailing NOI looks like a fire sale. It's not. DiamondRock disclosed $12 million in required capital expenditure over the next 12 months, a contractual ground lease escalation, and anticipated labor cost increases (New York's hotel union contracts are up for renegotiation in July 2026). Adjust for all three and the company estimates a stabilized cap rate of 7.8%, or 6.5% on a fee simple basis. The gap between 13.3% and 6.5% is the cost of everything the next owner just inherited. That's not a discount. That's a price tag on deferred problems.
The per-key number needs the same treatment. $174,600 per key for a Manhattan select-service leasehold sounds cheap. Add $63,500 per key in near-term capex and you're at $238,000 per key before the ground lease reset and before union negotiations that haven't started yet. New York is projecting 4,852 new rooms in 2026. The buyer isn't getting a bargain. The buyer is making a bet that post-renovation, post-lease-reset, post-union economics still pencil at a select-service ADR in a market adding supply. I've seen that bet work. I've also audited portfolios where it didn't.
DiamondRock's own guidance adjustment tells you something about how they valued this asset internally. They cut $5.9 million from full-year Adjusted EBITDA and $5.1 million from Adjusted FFO. Those are company-level guidance reductions reflecting the sale. Separately, DiamondRock disclosed a 6.3x EBITDA multiple on the transaction, which implies the property was contributing roughly $5.2 million annually at the midpoint. For context, lodging REITs trading at 10-12x EBITDA are considered fairly valued. DiamondRock sold this asset at roughly half that multiple. CEO Jeff Donnelly said the expected returns didn't meet investment thresholds. Translation: we ran every scenario and none of them justified the capital.
The strategic read is straightforward. DiamondRock has been rotating out of urban select-service and into leisure and lifestyle for three years. This sale is consistent. But the financial read is more interesting. A REIT with a freshly authorized $300 million buyback program chose to sell an asset at 6.3x EBITDA rather than deploy $12 million in capex to stabilize it. That tells you what their internal model showed about post-renovation returns... and it wasn't enough. When a publicly-traded REIT would rather buy its own stock than renovate a Manhattan hotel, that's a statement about where they think value is. And where they think it isn't.
Here's what I want you to see if you're an asset manager or owner sitting on an urban select-service hotel with a ground lease. DiamondRock looked at $12 million in capex, a lease escalation, and a union negotiation cycle... and walked. That's a disciplined capital allocation decision, not a distress signal. But it should make you run the same math on your own portfolio. Take your trailing NOI, layer in your actual next-12-month capex obligations, any lease resets, and realistic labor cost increases. If the stabilized return doesn't clear your hurdle rate, you don't have a hold... you have a hope. And hope is not a capital strategy. The bid environment for Manhattan leaseholds is thin right now. If your math says sell, the window to find a buyer willing to inherit those problems is narrower than you think.