The exit you modeled when you bought is not the exit the borrower hands you

The bankruptcy mill thing is real and I think you are underselling how localized it is. There are specific zip codes where you can almost predict it from the address alone, because the same petition preparer is filing for half the street.

The most unexpected path I have watched unfold was not even on a note I held, it was on one my note fund reported out in a quarterly update in enough detail that I could follow the whole thing. Borrower on a Mississippi property, BPO somewhere around $80k, note bought somewhere in the mid-forties. Borrower quit-claimed the property to a relative two weeks after the note was purchased, relative filed a separate bankruptcy claiming homestead exemption as though they had been living there for years, and the servicer spent eight months establishing that the transfer was fraudulent before they could even get back to the original foreclosure track. Nobody priced a collusive quit-claim into that probability tree. The fund eventually resolved it but they were at something like 26 months before they got a clear path and the all-in crept up close to their recovery number.

What I keep thinking about reading your post is whether the $58k bid price in your example already has a meaningful cushion built in for timeline bleed, or whether the return math is doing the thing where it treats the base case timeline as the expected case timeline. Those are not the same number and a lot of NPN models I have seen conflate them.

This thread is doing something useful for anyone learning NPN underwriting, so I want to pull out the core lesson in plain terms before adding one thing.

A non-performing note is a mortgage where the borrower stopped paying. You buy it at a steep discount to what they owe, then try to resolve it, usually through a payment modification, a deed in lieu (the borrower signs the property over to avoid foreclosure), or foreclosure. Your profit depends on buying cheap enough that any of those paths still works.

The example here shows exactly why timeline is the variable most beginners underprice. The numbers in the bid ($58,000 purchase, $90,000 property value) look like they have room. And they do, until a borrower files bankruptcy and proposes a repayment plan a court confirms over your objection. At that point your legal costs keep running, your carrying costs (property taxes, insurance, loan servicing fees) keep adding up, and you cannot foreclose. The deal that penciled at 18 months and $58,000 all-in becomes 30 months and $71,000 in, and your return shrinks or disappears.

The thing a beginner often does not know to ask: in judicial foreclosure states (states where a court has to approve the foreclosure), the base timeline estimate already assumes no borrower resistance. Every escalation, a bankruptcy filing, a disputed payoff figure, a lost-note claim, adds months on top of that baseline. When you underwrite, carrying cost should be modeled on the pessimistic timeline, not the median one.

A licensed real estate attorney who handles distressed debt in the specific state where the property sits is the right person to estimate realistic foreclosure timelines, not a general estimate from a blog or a podcast.

What state is the property in, and have you looked at that state's average judicial foreclosure timeline from filing to sale?

The $190 plan confirmation over objection is the one that still bothers me.

Judge saw equity cushion, saw a breathing borrower, confirmed it anyway, and my partner's basis was already at $64k before that happened.

What I did not price was the borrower's cousin being a paralegal who knew exactly which chapter to file and when.

The path I find most underpriced in that UPB band is the borrower who locates a title defect, real or manufactured, and files a quiet title action shortly after the note transfers. It does not need to succeed to cost you. If the assignment chain has any gap, even a MERS transfer that looks clean on its face, a competent pro se borrower who found a legal aid attorney can tie the property in litigation long enough to blow past your foreclosure timeline entirely, and in some states the quiet title suit creates a lis pendens that makes a deed in lieu functionally impossible because no title company will insure it. On the numbers in your example, a 14-month judicial foreclosure timeline priced at $58,000 becomes a very different animal at 26 months with $8,000 in contested legal fees sitting on top of basis. The assumption doing the most work is not just behavioral, it is documentary: the model assumes the chain of title the servicer handed you will hold up under adversarial scrutiny, and that assumption is worth testing with a title attorney before bid, not after the lis pendens lands. I would confirm current state-specific lis pendens rules and quiet title timelines with a real property litigation attorney in the subject property's jurisdiction before pricing any bid on a note with more than two prior servicer transfers. What does your assignment chain look like on the deal you are modeling, and did the seller provide an endorsement in blank or a specific allonge to your entity?

The one that compounds the original scenario most aggressively is when the Chapter 13 plan gets confirmed, the borrower makes exactly enough payments to avoid immediate dismissal, then files a motion to value under section 506 arguing the property declined since petition date and asking the court to strip the lien down to current BPO. If the property dropped from $90,000 to $78,000 in the interim, and your claim was already underwater at $110,000 UPB, you are suddenly fighting a cramdown on collateral value while your basis is climbing past $65,000 in carrying costs. The confirmed plan at $190 a month was already a bad outcome; the 506 motion converts it into a potential permanent reduction in what you can ever recover, not just a timing problem.

The clause that matters in that scenario is whether your purchase agreement allocated any representation about the borrower's bankruptcy history in the prior 8 years, because a borrower who filed and received a discharge in year 6 cannot get another Chapter 7 discharge, which is why they run to 13, and a note seller who knew that and did not disclose it changed your probability tree before you ever opened it.

Did the seller's tape on this note include any bankruptcy history at all, or was the file silent on prior filings?

The successor-in-interest claim after a quiet title action mid-foreclosure is the path that breaks the model hardest, because it resets standing rather than just extending timeline.

A borrower's adult child, never on the note, files a claim asserting ownership through an unrecorded transfer executed before the default, sometimes a quitclaim deed the original borrower signed and tucked away precisely for this purpose. The note buyer must now litigate standing in a collateral proceeding while the foreclosure is stayed, and depending on the state's treatment of bona fide purchaser status, the buyer may face a real cloud on the security interest, not just a delay. At $58,000 basis on a $90,000 BPO property there is enough equity to make that fight worth someone's legal fees, and a contingency-fee title attorney will take it. The carrying cost math from the original post gets worse because this is not a bankruptcy timeline, which at least has structured milestones; quiet title litigation in a backlogged county court has no mandatory resolution date at all. The answer is to pull the chain of title back three to five years before bidding and treat any quitclaim or interspousal transfer in the 24 months before default as a flag requiring a title endorsement that specifically covers post-transfer claims, making that endorsement a condition of closing rather than a nice-to-have.

What did your pre-bid title work look like on the $58k scenario, and did you get gap coverage?