Sub-to is largely a bet on the servicer staying asleep, and the exit should be priced that way

That month fourteen number tracks with what gets reported elsewhere, though some operators see closer to eighteen to twenty-four months before anything flags, which probably depends on the servicer's portfolio size and how manual their audit process still is. The point your mentor made about the rate being irrelevant if you cannot cover the exit is the only thing that actually matters in this whole structure. A private lender writing paper against one of these should price every loan assuming the borrower's exit gets bricked at some point, and that math has to close before a dollar is committed.

Month fourteen is actually generous, I have seen servicers flag at nine.

Your mentor is right that the exit liquidity is the real underwrite, but I would push back on the $180k to $240k range being some kind of safe band. I had a servicer sleep through month twenty-two on a $310k note in Myrtle Beach and wake up at month seven on a $195k one in Asheville, so loan size is not the variable that matters. The audit trigger I keep seeing is anything that touches the servicing file, an insurance claim, a tax reassessment, even a borrower name correction, because that is what pulls eyes onto it, not a calendar.

The forced refi scenario is the one most people model as a possibility instead of a certainty, and that shift in how you frame it changes everything about how much cash you need to hold against each deal. I ran four townhome acquisitions in 2022 and 2023 where I stress-tested every one of them against a day-one due-on-sale call, not a month-fourteen call, and the number that came back was uncomfortable enough that I restructured two of the four before I closed. The rate on the underlying was 3.1 percent on one of them and it genuinely did not matter once I priced in a forced conventional refi at whatever Fannie was pricing that month. The spread that looked like $800 a door became a break-even or worse on an exit refi, and that was before vacancy or carry costs.

What I have not seen you or your mentor mention yet is what the serviced loan balance looks like relative to current appraised value at the point the call is most likely to come. Because that ratio is doing a lot of work that the rate and the spread are not. If you are sitting at 55 LTV when the servicer wakes up you have refi options, ugly ones, but options. If the seller bought in 2021 at peak and you are at 88 LTV because the market in that submarket softened, you are not just scrambling for cash, you are scrambling for a lender who will even touch it. So what does the current value look like on the properties you are underwriting, relative to the balance you are assuming?

The part people skip is that servicers are not monolithic. Same loan, sold twice, and you can go from a servicer who has not looked at anything since origination to one who treats every portfolio acquisition as an excuse to run compliance sweeps. I bought into a deal in Mecklenburg last spring where the underlying was through a regional that got absorbed mid-2023 and the new servicer triggered a review inside eight months, not fourteen, because their onboarding process flagged any account where the mailing address on file did not match the property address. That was the tell. Nothing exotic, just an address mismatch from when the seller had updated their billing years before we closed.

Your mentor's $180k to $240k range is also worth interrogating on its own. That spread sounds clean until you try to refi out of a forced position right now in Charlotte or Raleigh where lenders on non-warrantable situations are pricing the risk premium hard. I was looking at numbers last quarter and the realistic forced exit on a sub-to in that bracket was not whatever the going rate headline said, it was plus 180 to 220 basis points over that depending on how the title chain read and how long you had been seasoning payments. So when he says the rate on the underlying is almost irrelevant if you cannot cover the exit, I would go one further and say the rate you are modeling for the forced refi is also probably too low by a material margin unless you are running it at current SOFR plus a full stress cushion.