Weighing a warehouse line and table funding against staying a pure broker in mortgage brokerage

Take a shop closing 61 files a year, 44 of them investor deals on the private and bridge side. On a brokered bridge file, compensation typically runs 100 to 150 basis points depending on the lender and how much of the fee the borrower bears. Some funding sources offer a correspondent arrangement instead, where the shop takes a small warehouse line, closes in its own name, and sells the paper within a week or two. The pitch is that the same file pays closer to 200 to 250 basis points because the shop absorbs the closing and early servicing work, and controls the timeline instead of waiting on someone else's draw desk. The other side of that trade is real. A line means covenants, minimum tangible net worth requirements, quarterly audits, and someone on staff who understands repurchase language on the sale side. Licensing for closing in a shop's own name is a different animal than brokering and varies state by state, so that is a legal question before it is a math question. And the moment a shop is the named lender on the note, its incentives stop being obviously aligned with the borrower, which cuts against the broker-neutrality pitch that built the relationship in the first place. The pure broker path stays light. Nobody can margin call a broker. But if private credit growth continues, the spread will sit with whoever holds paper for those two weeks, and a shop weighing this should model both the covenant risk and the opportunity cost of staying at 110 to 150 basis points before deciding.