What is a loan tape? Structured finance data explained
A loan tape is a structured data file, typically a spreadsheet or CSV, that holds one row per loan and one column per loan attribute. It is the single source of truth for a pool of loans: when a lender securitizes a portfolio, draws on a warehouse line, or sells a book, the buyers, rating agencies, and auditors all evaluate the deal through the tape. Individual loan documents sit underneath it, but the tape is what the money looks at.
What actually lives in a loan tape
A tape is wide. A single consumer or SME pool can run to dozens of columns, grouped roughly like this:
| Column group | Example fields | Why it matters |
|---|---|---|
| Loan identity | Loan ID, origination date, original balance, current balance | Ties every row back to a real, traceable loan |
| Borrower and credit | Bureau score, income, debt-to-income, employment | Drives stratification and expected-loss models |
| Terms | Interest rate, term, amortization type, maturity | Feeds cash-flow and waterfall calculations |
| Collateral | Asset type, valuation, lien position (secured pools) | Sets recovery assumptions |
| Performance | Payment status, days past due, restructure or charge-off flags | Separates a clean pool from a distressed one |
The schema is not universal. A mortgage tape, an auto tape, and an unsecured consumer tape each carry their own standard columns, and every investor tends to ask for a few of their own. Part of the work in any deal is mapping your internal fields to the layout the counterparty expects.
What fields belong in a loan tape? The standard set
A loan tape's standard fields fall into four groups: loan identifiers and terms, borrower and collateral data, payment history and performance, and status and servicing fields. Buyers and securitization desks expect every group to be present and populated, because each one feeds a different part of how a pool gets priced and stress-tested.
| Field group | Fields lenders and investors expect | Why buyers and securitization desks need it |
|---|---|---|
| Loan identifiers and terms | Loan ID, origination date, original and current balance, interest rate, term, amortization type, maturity | Ties each row to a traceable loan and drives the cash-flow and waterfall math |
| Borrower and collateral data | Credit score for the asset class, income or DSCR, LTV and lien position, collateral type and valuation where secured | Sets expected-loss, recovery, and stratification assumptions |
| Payment history and performance | Payment status, days past due, delinquency bucket, modification and charge-off flags, prepayment history | Separates a clean pool from a distressed one and shapes the loss curves |
| Status and servicing | Current status, servicer, next payment date, escrow or draw balances, servicing flags | Tells a buyer what is happening on each loan now and who is handling it |
The right fields are asset-class-aware, not universal. A consumer tape leans on bureau score, debt-to-income, and standardized delinquency buckets; a residential transition loan (RTL) or private credit tape leans instead on collateral value, loan-to-value or after-repair value, debt-service coverage, and draw schedules, because the risk lives in the asset and the project rather than a consumer credit file. Copying a consumer template onto an asset-based book is a common way to ship a tape that looks complete and answers none of the questions its buyer will ask, a gap we see most in loan origination for private credit.
A loan tape is "clean" when its fields are consistent across every row, its balances reconcile to the general ledger, and its assumptions are documented, so a buyer can load it and trust it without a call to explain what a column means. That discipline is also what makes a tape survive a system change: when you migrate a loan origination system without stopping lending, the tape has to be reconciled in the new system rather than assumed to carry across.
Why the tape decides the deal
Everything downstream is computed from the tape. Stratification tables, weighted-average coupon, weighted-average life, expected loss, and the securitization waterfall are all just functions of those columns. That is the leverage and the risk: a single mislabeled status column or a stale balance can move a headline number enough to change pricing or trigger a diligence question that stalls the whole timeline.
This is why "clean tape" is a phrase you hear constantly in structured finance. A clean tape reconciles to the general ledger, has no missing values in required fields, uses consistent codes for things like payment status, and matches the loan origination and servicing systems it was pulled from. A messy tape, full of manual fixes and one-off exports, is the hidden cost that shows up as delayed closings and nervous investors.
Loan tape vs servicing report vs data room
Three terms get used interchangeably in early conversations, and they are not the same thing.
The loan tape is loan-level and machine-readable: one row per loan, point-in-time, built to be loaded into someone else's model. The servicing report is aggregated: pool-level performance for a period, collections, prepayments, and delinquency buckets, delivered to investors and trustees on a schedule. You can derive a servicing report from a good tape, but you cannot reconstruct a tape from a servicing report, because the loan-level detail is gone.
The data room is the container for the whole diligence exercise: credit policies, servicing agreements, sample loan files, legal documents, and the tape itself as one exhibit among many. When a counterparty asks for "the tape," they mean the loan-level file, not the summary report and not the folder. Getting the vocabulary right early keeps requests specific and turnarounds short, especially when several parties are asking for slightly different cuts of the same data.
Where tapes go wrong, and how to fix it
Most tape problems are not analytical, they are operational. The data is spread across an origination system, a servicing platform, and a few spreadsheets, and someone assembles the monthly tape by hand. Every manual join is a chance to drop rows, double-count balances, or carry forward last month's status.
The durable fix is to generate the tape directly from source systems on a schedule, with validation rules that catch breaks before the tape leaves the building: balances that do not reconcile, statuses that are not in the allowed set, dates that fall outside the loan's life. That is exactly what loan tape automation is for, and it is a recurring theme across the lending software work we do: the teams that win diligence are the ones whose tape is boring, because boring means trustworthy.
If you are standing up a securitization or warehouse program and the monthly tape is still a manual spreadsheet exercise, that is usually the first thing worth automating. It removes the error class that costs the most: the one that surfaces in front of an investor.
Get the due-diligence checklist
The concrete questions to ask any loan origination system vendor before you sign, across viability, data model fit, audit and reporting, integrations, migration, pricing, and reference calls.
Get the checklist→
