The seller's lender called the due-on-sale after a routine tax record audit, not a missed payment.

Most lenders do not audit tax records routinely, but servicers sold into portfolios with due-on-sale enforcement mandates do.

The indexing exposure is real but the mitigation is at the entity layer, not the deed layer. If title transfers into a land trust with the seller named as beneficiary at closing, the deed that records shows a transfer into trust, and the beneficiary interest assignment to the buyer happens by separate document that does not record. That separation is the mechanical reason land trusts appear in subject-to structures: the recorded instrument looks like an estate-planning move rather than a sale. Whether a specific servicer treats that as a triggering transfer depends on how aggressively they read the Garn-St. Germain exceptions, so confirm the trust structure with a real estate attorney before closing, not after.

The second layer is servicer selection risk, which the original post illustrates exactly. A loan originated at a community bank and still serviced there behaves differently from one sold to a servicer whose entire business model is default and enforcement. Pulling the MERS record before closing tells you who is actually holding servicing rights, and that name matters more than the original lender's name on the note.

The third layer, which almost nobody prices, is a rate-bridge reserve. On a $187,000 balance, the $760 spread you calculated is $18,240 over 24 months. A deal that cannot survive that reserve being set aside at close was not priced for subject-to risk, it was priced for the assumption that the rate holds without incident.

What was the entity structure at the time of the deed transfer?

The assumption doing the most work in that deal was that a performing loan with no red flags would stay invisible. The county indexing removed that assumption at month 14, and it removed it permanently once the deed recorded.

The indexing exposure in a subject-to comes from the moment of recording, not from anything the buyer does afterward. The deed is public the day it hits the recorder's office. Some operators try to delay recording by using a land trust, where the deed transfers into the trust before or at closing and the trust instrument is what records, not a direct conveyance showing a change of beneficial ownership. The lender still holds a due-on-sale right under the Garn-St Germain Act if they pierce the trust structure, but the surface-level audit trigger, a grantor-grantee index showing a name change on a property with a live mortgage, is harder to automate against a trust transfer. Confirm with a real estate attorney in the relevant state whether that structure holds in your jurisdiction and what disclosure obligations attach.

A second layer is the contingency plan, and your deal illustrates exactly why "we'll refinance if called" is a plan in name only when the hold period has not seasoned enough equity or the rate spread makes a cash-out punishing. Operators who are serious about 24-plus month holds are sizing the equity cushion against the worst-case refinance, meaning they are underwriting the deal assuming the called-loan scenario and checking whether it still closes positive. If it does not survive that scenario with margin, the deal is mispriced at acquisition.

The third point you did not raise: the seller's credit was exposed for those 14 months. If that acceleration had not resolved cleanly, the seller's name was on the default. That is a documentation and relationship obligation that runs the entire hold, and it is the piece most sellers do not fully absorb at signing.

What was your contingency underwriting at acquisition? Specifically, did the deal close positive at a 7.4% refi on day one, or was the equity margin always dependent on the hold?