Edgewater — Phase One

Underwritten honestly, the base case does not pencil — a conventional build at market cost returns -17.1% levered IRR, below any institutional bar. The value here is not a more optimistic spreadsheet; it is a composed path of sourced moves that turns the deal into a financeable one.

Three moves do the work. The Capital Stack assembles public subsidy, tax-credit equity and HUD financing; the Cost agent value-engineers the residential basis with industrialised methods; and the Program agent adopts an efficient single-stair building type that adds roughly 30% rentable area on the same envelope. Composed in that order — cheapest, non-dilutive sources first — they lift the deal from -17.1% to 23.9% levered IRR, a 2.20× equity multiple, with a stabilised DSCR of 1.37 clearing the 1.20 floor.

The ask is $28.0M of LP cash equity into a ~$131.4M total-cost project. Every assumption behind the path carries its source; the moves not needed to pencil (revenue placemaking premium, on-site energy) are held in reserve as upside, not baked into the headline.

The Lead Agent composed the cheapest, non-dilutive path that clears every guard: Capital Stack → Cost → Program, reaching 23.9% levered IRR with a 1.37 DSCR — penciled, every validator satisfied. Revenue, Operations and Risk’s upside were held in reserve, not needed to clear the 15% line. When decision D lights up the reasoning agents, the composers will compose within the policies’ constraints and subject to the gate’s verdict.