The calculation was correct. The input was wrong.

A mortgage-note calculation can be mathematically perfect and still lead to a bad investment decision. The reason is simple: formulas do not verify the facts you feed them.

Suppose a listing says the property is worth $250,000 and the unpaid balance is $150,000. The LTV calculation is straightforward: 60%. If the formula is implemented correctly, the answer is correct for those inputs. But if the $250,000 figure is stale, seller-stated, based on a different property condition, or contradicted by stronger evidence, the 60% result may describe a deal that does not actually exist.

That distinction matters because mortgage-note investors often spend their time checking formulas when the larger risk is upstream. Purchase price, UPB, payment status, property value, lien position, maturity, balloon structure and payment history can all look like ordinary input fields. They are not all equally supported facts.

A number can be precise without being proven

Precision is not provenance. A listing can show a property value down to the dollar. A spreadsheet can turn it into LTV, equity and recovery figures with several decimal places. None of that tells you where the original value came from, when it was observed, whether it reflects current condition or whether an independent source agrees.

The same problem applies to payment history. A ledger covering January through June can accurately prove payments during that period. It does not automatically prove that the note is current in September. A seller's label of “performing” is also not the same thing as dated payment evidence.

Four different kinds of inputs

A useful diligence process separates at least four states. Seller-stated means the seller or listing asserted the fact. Sourced means a document, public record, servicer record or other identifiable source supports it. Calculated means the number is derived from other inputs. Unresolved means the available evidence is missing, conflicting or insufficient. Seller-stated vs. verified data works through each class with a real deal.

This matters because a calculated result inherits the quality of the facts beneath it. If a yield calculation depends on an unverified payment amount and maturity date, the math can be correct while the investment conclusion remains provisional. If recovery depends on a disputed property value or lien priority, the recovery estimate should carry that uncertainty instead of hiding it.

The investor's real question is not “is the formula right?”

The better question is: which inputs are safe to underwrite?

That requires a chain of reasoning: identify the claim, locate the source, check its date and scope, compare it with other evidence, decide whether the sources are actually comparable, preserve conflicts that cannot be resolved, and only then allow supported inputs into the economic model.

This is where ordinary calculators and spreadsheets reach their natural limit. They can be excellent at arithmetic. They generally cannot know that a county record and an appraisal are describing different dates or conditions unless someone supplies that context. They cannot know that a payment ledger is stale simply because the last payment row is internally consistent. The software-by-job guide sets out which tool answers which question.

A better underwriting workflow

For every material input, record what was claimed, what evidence was found, its source and date, whether that source is independent, whether another source conflicts, and which downstream calculations use the fact. How to analyze a mortgage note puts those records in working order, stage by stage.

Then make the economics traceable. If the supported property value changes, show the effect on LTV, ITV, equity and recovery (ITV vs. LTV explains why both ratios move). If payment status becomes unresolved, show which yield or cash-flow assumptions are now conditional. If lien position is not established, do not let a confident-looking recovery figure imply that legal priority is settled.

What NoteGage is built to do

NoteGage exists for exactly this failure mode. Its report shows, beside every computed figure, which inputs are seller-stated, which are supported, and which are still open, so a correct calculation can never quietly borrow confidence it has not earned. How it works shows the stages in order; the methodology states the rules; the property value guide covers the input that most often moves the deal.

The point is not to distrust every seller or replace professional title, servicing or legal work. The point is to make sure a correct formula is not being used to give false confidence to an unsupported fact.

Before asking whether your mortgage-note math is correct, ask the question that comes first: are the numbers you are calculating with actually supported?