Bally's Bronx Casino Math: $4B Bet at a 7.1% Implied Yield on $1.5B Revenue

$157 million for 16 acres of parkland. That's $9.8 million per acre in the Bronx, before a single shovel hits dirt. Add the $500 million license fee to the MTA, the reported $115 million payout to the previous golf course operator, and $765 million in community benefit commitments, and Bally's is $1.5 billion deep before construction begins on a $4 billion project. The real number here is total capital deployed relative to projected revenue: $4 billion against a forecast of $1.5 billion in annual total revenue. That's a 2.67x revenue multiple, which implies Bally's needs roughly a 37.5% EBITDA margin to generate a 14% return on invested capital. For a casino resort that hasn't broken ground yet, in a market with two competing licenses coming online in the same window, that margin assumption deserves scrutiny.

Let's decompose the hotel component. 500 rooms in a 23-story tower attached to a 3-million-square-foot gaming complex. At $4 billion total project cost, the hotel is maybe 12-15% of that (call it $500-600M based on comparable integrated resort allocations). That's $1M-$1.2M per key. New York construction costs justify some of that premium, but the room block exists to feed the casino floor, not to compete on ADR with midtown Manhattan. The question asset managers should ask: what RevPAR does a Bronx casino hotel need to achieve for the room division to cover its allocated capital cost, or is the hotel permanently subsidized by gaming revenue? I've analyzed enough integrated resort models to know the answer is almost always the latter. Which is fine, until gaming revenue underperforms projections.

The competitive picture is the variable I can't model cleanly. Hard Rock near Citi Field and Resorts World's expansion in South Ozone Park are both targeting the same downstate New York gaming dollar. Three licenses collectively projected to generate $7 billion in state gaming tax revenue over a decade. That $7 billion number comes from somewhere, and the somewhere is GGR projections that assume each property captures its modeled share without significant cannibalization. I've audited casino revenue projections before. The base case always assumes rational market distribution. Reality distributes irrationally. One property wins the location battle, one wins the entertainment programming battle, and the third discovers its projections were the most optimistic of the three.

Bally's balance sheet adds a layer. Analysts carry a "Reduce" consensus on the stock. The company is simultaneously building a $1.7 billion casino in Chicago (opening late 2026), planning a Las Vegas project, and now committing $4 billion to the Bronx. Total development pipeline across three major markets while carrying significant existing debt. Gaming and Leisure Properties has provided $2.07 billion in financing, and the Chicago project alone required a $940 million construction facility. The math works if every project hits its revenue target on schedule. If one project delays or underperforms, the capital allocation pressure cascades across the portfolio.

The 15-year license term is the number that matters most and gets discussed least. Bally's needs to build by roughly mid-2027 (18 months from the February 2026 land closing), open by 2030, ramp to stabilized operations by 2032-2033, and then generate enough cash flow across the remaining 11-12 license years to justify $4 billion in capital. Back-of-envelope: $4 billion at a 10% target return requires $400 million annually in free cash flow from this single property. Against $1.5 billion projected revenue, that's a 26.7% FCF margin... achievable for a top-performing casino, aggressive for a new entrant in a three-way competitive market. The math works. The question is what "works" means for the equity holders if Year 1 GGR comes in at 75% of projection.