Securitization Data Tape: Clean Pools for DA and PTC Deals
Securitization Data Tape: How Lenders Prepare Clean Pools for Direct Assignment and PTC Deals
Last reviewed: November 2026
A securitization data tape is the loan-level file a lender hands to investors, rating agencies and trustees so they can test, price and monitor a pool offered in a direct assignment (DA) or pass-through certificate (PTC) deal. It records every loan's terms, balances, repayment history and collateral at a fixed cut-off date.
This guide is for CFOs, Treasurers and Heads of Capital Markets at NBFCs and housing finance companies. It covers the RBI eligibility rules your tape must prove, the fields investors expect, and how to produce the tape from systems rather than spreadsheets.
Key Takeaways
- Since 28 November 2025, NBFC securitisation and loan transfer rules sit in two consolidated RBI directions, both updated as on 1 July 2026.
- The tape must prove the minimum holding period of 3 or 6 months, usually counted from registration of the security interest.
- PTC originators retain 5% or 10% of book value, depending on original maturity, while DA sellers retain 10% if the buyer relies on portfolio-level due diligence.
- Most pool delays are data problems, not credit problems, such as missing registration dates or balances that do not match the ledger.
Where Do the RBI Rules for DA and PTC Deals Sit Now?
They sit in two directions issued on 28 November 2025. One covers securitisation through a special purpose entity, and the other covers direct transfers of loans.
| Instrument | What It Governs | What It Means for the Tape |
|---|---|---|
| NBFC Securitisation Transactions Directions, 2025 (RBI/DOR/2025-26/353) | PTC deals, retention, eligible assets, investor disclosures | Retention, asset eligibility and disclosure fields |
| NBFC Transfer and Distribution of Credit Risk Directions, 2025 (RBI/DOR/2025-26/352) | Direct assignment, holding periods, buyer due diligence | Holding period dates and sample-testing data |
Banks and small finance banks have parallel directions with their own paragraph numbers. Update the references in your board policy and transaction documents, and confirm current clause numbers before each deal.
Which Eligibility Rules Must the Tape Prove?
The tape must let any investor check each loan against the rules without asking you for extra files. If a rule cannot be tested from a field, expect a query or a haircut.
| Rule | Direct Assignment | PTC Securitisation | Field That Proves It |
|---|---|---|---|
| Minimum holding period | 3 months (tenor up to 2 years), 6 months (longer) | Same rule, applied to each loan | Security registration date, first repayment date, tenor |
| Retention | 10% economic interest if portfolio-level due diligence is used | 5% (maturity up to 24 months), 10% (longer or bullet), 5% for home loan pools | Original maturity, repayment type, product |
| Asset status | Loans in default go through the stressed loan route | Standard assets only | Asset classification, days past due |
| Excluded loans | Acquired loans held 6 months before resale | No revolving credit, residual maturity under 365 days, or recently restructured loans | Maturity date, restructuring flag, acquisition date |
| Structure | No credit enhancement or liquidity support | At least two tranches, minimum ticket of ₹1 crore per investor | Deal-level, not loan-level |
The holding period clock deserves attention. Under the transfer directions, it runs from registration of the underlying security interest, or from the first repayment where the security cannot be registered.
What Fields Belong in a Loan-Level Tape?
Investors need enough data to rebuild your cash flows and test your credit. The groups below cover what most investor and rating agency requests ask for.
| Field Group | Example Fields | Why Investors Need It |
|---|---|---|
| Identity and product | Masked loan ID, product, state, branch, sourcing channel | Concentration and channel analysis |
| Contract terms | Sanctioned amount, disbursement dates, rate type, rate, tenor, EMI | Cash flow modelling |
| Balances at cut-off | Principal outstanding, overdue principal and interest, charges | Pool size and pricing |
| Performance | Current days past due, 12-month payment history, bounces, prepayments | Delinquency and prepayment assumptions |
| Origination credit | Bureau score band, income, existing obligations | Credit quality and pool selection checks |
| Collateral | Type, value, valuation date, loan-to-value, registration date | Loss severity and holding period proof |
| Collections | Mandate type and status, repayment mode | Servicing risk and bounce sensitivity |
Mask personal identifiers and share only what the deal needs. Data shared with investors is still personal data under India's data protection law.
How Should Lenders Build the Tape From Their Systems?
Build it from the loan management system, reconcile it, then apply eligibility rules in code. A repeatable pipeline turns a two-week scramble into a routine run.
- Freeze the cut-off date. Extract balances and histories as at one date, and record the extraction time.
- Reconcile to the ledger. Total principal outstanding should match the general ledger, with every difference explained.
- Apply eligibility rules in code. Holding period, residual maturity, classification and restructuring checks should be scripted, versioned and repeatable.
- Produce stratification tables. Break the pool down by state, ticket size, rate, seasoning and days past due.
- Attach static pool history. Investors compare the pool with how your past disbursement cohorts performed.
- Lock and version the tape. Every later change should create a new version with a reason.
What Happens to the Tape After Closing?
The tape becomes the baseline for servicing reports. Collections on sold loans must be tracked separately and remitted to the special purpose entity without co-mingling with the originator's own cash.
The directions also ask originators to confirm holding period and retention compliance at least half-yearly, in September and March, using a prescribed disclosure format. Where a third-party servicer supplies the data, the originator should have it certified by auditors, preferably at least once a year.
For deals seeking simple, transparent and comparable treatment, loan-level or stratified data must stay available to investors at least quarterly. Your loan management system should therefore tag sold loans at the account level and produce pool reports without manual matching.
Collateral and registration fields often live in scanned documents rather than databases. An intelligent document extraction AI agent can pull valuation and registration details into structured fields, and a customer data quality AI agent can flag gaps before the tape leaves your building.
Could your team produce a reconciled, rule-tested data tape within a day of choosing a cut-off date?
Talk to Our Lending Technology Specialists
Visit Digiqt to see how we build lending platforms that keep loan, collateral and collection data deal-ready.
Illustrative Example: Cutting a ₹500 Crore Loan Against Property Pool
This illustrative scenario shows how data quality changes the size of a deal. An NBFC selects ₹500 crore of loan against property as a candidate PTC pool, with original tenors above five years.
| Step | Amount (₹ crore) |
|---|---|
| Candidate pool | 500 |
| Less: holding period not met, or registration date missing | (35) |
| Less: residual maturity under 365 days | (40) |
| Less: restructured within the specified period | (5) |
| Less: not classified as standard | (20) |
| Eligible pool | 400 |
| Retention at 10% of book value | (40) |
| Available to investors | 360 |
Because original maturity exceeds 24 months, retention is 10% of ₹400 crore, or ₹40 crore. That leaves ₹360 crore for at least two tranches.
Now look at the first exclusion. If ₹20 crore of the ₹35 crore failed only because registration dates were missing from the system, fixing that data would lift the eligible pool to ₹420 crore and investor capacity to ₹378 crore.
If the NBFC sold the same ₹400 crore by direct assignment instead, the buyer would test at least one-third of the pool loan by loan, or about ₹133 crore by value and a third of the loan count. The seller then keeps 10% if the buyer relies on portfolio-level checks for the rest.
Which Data Tape Mistakes Delay Deals?
Most of these errors surface during investor queries, when fixing them costs the most time.
- Tape totals that do not match the ledger. Investors read unexplained gaps as weak controls.
- Missing holding period evidence. Without registration or first repayment dates, eligible loans get excluded.
- Manual eligibility filters. Spreadsheet filters are hard to reproduce when a rating agency asks how the pool was cut.
- Inconsistent days past due logic. If the tape and your regulatory returns calculate overdue status differently, every number gets questioned.
What Is Changing Next?
Reporting is becoming more granular. The transfer directions require each loan transfer to be reported to a trade reporting platform once the RBI notifies it, and NBFCs must keep a transaction database with MIS until then.
The securitisation directions also call for quarterly reporting of deals to the RBI, in a format to be communicated separately. A tape built from governed systems makes both obligations easier.
Treasury teams should also link pool sales to capital planning. A capital adequacy forecasting AI agent can show how a planned sale changes capital ratios, and a prepayment risk forecasting AI agent can test the prepayment assumptions investors will challenge.
Want every pool you sell to pass investor queries the first time?
Talk to Our Lending Technology Specialists
Visit Digiqt to see how we build loan data pipelines, reporting and AI agents for lenders.
