An 11 room boarding house case: 38k NOI against a 52k underwrite, and the gap wasn't rents

The shape of this variance is worth studying because it shows up in more than one operator's books. Acquisition: 610k for an eleven room licensed house in a mid-size market with a hospital and a community college inside two miles. Rooms at 725, all furnished, utilities included, weekly common area cleaning. Potential gross 95,700. The underwrite: 8 percent vacancy, 5 percent credit loss, opex at 42 percent of effective gross. That gives 52k of NOI against 44k of debt service. Thin, but built to grow. What actually happens: economic occupancy lands at 74 percent. Rents hold fine. Every room that comes available takes between 19 and 34 days to fill, across 14 move-outs in 11 rooms. An 8 percent vacancy assumption borrowed from portfolio-level apartment data gets applied to a property where the unit of turnover is a room, and rooms turn three to four times faster than apartments. Fourteen turns times 26 average vacant days is 364 room-days, which is 9 percent right there before a single collection problem. Then collections. Three residents leave owing between 400 and 1,900. Credit loss comes in at 4,100, close to 4.3 percent, so that assumption holds. Then utilities. Budgeted at 700 a month. Actual runs 1,180. Eleven adults, individual space heaters, laundry running constantly. That's 5,760 of pure miss. Then the fire marshal flags the second floor rear egress, and two rooms sit empty five months during permitting and repair. 18,300 in work, on top of the lost rent already counted above. NOI lands at 38.4k against 44k of debt, a shortfall that has to be covered from the operating account. The lesson: underwrite turnover in room-days, expected turns times expected vacant days, rather than a borrowed percentage. Hold a licensing and code reserve separate from capex, since the inspection cycle doesn't respect a normal capex schedule. And model utilities per occupied bed with a winter number and a summer number instead of an annual average. Run all three at underwrite and this deal shows about 39k of NOI on paper, which doesn't support a 610k purchase price. That's the whole lesson.