How to Build a Bank-Grade Title & Collateral Evidence Pack for a Credit Committee

TL;DR

• Lenders and investors don't need a "title check" — they need a standardised, source-backed collateral evidence pack: a 30-year chain of title, a current encumbrance certificate (EC), a CERSAI search, a mutation/Khata-to-deed match, statutory approvals, a litigation sweep (including NCLT/IBBI for corporate borrowers), and spatial/zoning confirmation — assembled so a credit or investment committee can compare any deal on one risk framework.
• The pack must answer three questions in order: Who owns it? (chain + mutation), Is it already pledged? (EC + CERSAI), and Can it be enforced and is it litigation-free? (approvals, zoning, court/insolvency search).
• Every assertion needs its primary source attached — the RTC/Pahani, the registered deed copy from Kaveri 2.0, the EC, the CERSAI result, the K-GIS map — not a summary memo. A committee can only price risk it can see.
• The differentiator is structure and traceability, not volume: a 200-page deed dump is not diligence. A one-page risk grid backed by linked sources is.
• AI can gather records, translate them, reconcile names and extents, and draft the findings — but a lawyer reviews the chain and signs the title opinion. The pack supports that opinion; it does not replace it.

---

What collateral and title diligence do lenders and investors need to fund a real-estate deal?

A lender or investor needs enough evidence to conclude three things with confidence: the borrower (or its predecessor) holds a clean, marketable title; the asset is not already encumbered to someone else; and the security, once created, can actually be enforced without a litigation or insolvency surprise. Everything in a bank-grade pack exists to prove one of those three.

Where a developer's property due diligence checklist is about deciding whether to buy, a collateral pack is about deciding whether to lend against or invest in the same asset — and crucially, whether the security interest will survive the borrower's default. That shifts the emphasis. You care less about the perfect plot and more about defensible recovery: Is the charge registrable? Is the title good enough that a court would uphold the mortgage? Will an insolvency proceeding pull the asset into a moratorium?

The output is not a narrative. It is a structured, repeatable pack so that the deal in front of the committee can be set against last quarter's deals on the same axes. That standardisation — same checks, same sources, same scoring — is the entire point.

The four pillars, in lender terms

A complete pack maps to the four diligence pillars, each reframed around recoverability:

What goes into a bank-grade title and collateral evidence pack?

A bank-grade pack has eight components, each with the primary source attached and a plain-language finding. Below is the standard set, why a committee needs it, and what "good" looks like.

1. 30-year chain of title

The spine of the pack. You trace ownership back roughly 30 years — the long-standing market convention for high-value transactions, which sits well beyond the 12-year limitation period for most private adverse-possession claims under the Limitation Act, 1963 — through a continuous sequence of registered deeds: sale, gift, partition, succession. Each transfer should connect cleanly to the next, with no gap, no unexplained party, and no break in the recital. A single missing link is a defect, not a footnote. For the mechanics of how this chain is assembled and read, see what a Title Search Report actually is.

2. Current encumbrance certificate (EC)

The EC is the official record of registered transactions against the property over a stated period — sales, mortgages, releases. For a lender it answers "is there a registered charge ahead of mine?" In Karnataka, ECs and certified deed copies come from Kaveri Online 2.0. Pull the EC for the full search period, not just the last year, and confirm any prior mortgage shows a corresponding release deed. An EC with a live mortgage and no release is a red flag the committee must see, not the analyst alone.

3. CERSAI search

The EC captures registered charges; CERSAI captures the ones that may never reach the sub-registrar — equitable mortgages created by deposit of title deeds, the most common bank security in India. The Central Registry of Securitisation Asset Reconstruction and Security Interest (CERSAI) runs a search where anyone can check whether a security interest is already filed against an asset. Skipping CERSAI is one of the most common ways a lender ends up second in line behind another bank's equitable mortgage. Run it; attach the result; reconcile any hit.

4. Mutation and Khata-to-deed match

This is the reconciliation step that catches paper titles. The deed says X owns it; the revenue and municipal records must agree. That means matching the registered owner against the mutation register and the Khata / property-tax record. In Karnataka, this now includes the e-Khata / e-Aasthi record — a valid e-Khata has been made mandatory for property transactions within BBMP limits, so a mismatch or a missing e-Khata is both a title and a transactability problem. For revenue land, the Bhoomi RTC/Pahani shows the cultivator, extent, and — in Column 11 — registered encumbrances. Owner on deed, owner on RTC, and owner on Khata should be the same person. When they aren't, you have an unrecorded transfer, an inheritance gap, or worse.

5. Statutory approvals and land-use status

Title can be clean and the asset still un-financeable if the land can't legally be used as represented. The pack confirms the conversion order (for agricultural-to-non-agricultural land), the layout/building plan sanction, RERA registration where applicable, and any clearances the use requires. A common trap is agricultural land offered as collateral for a commercial project without a valid conversion — the security is over land that cannot lawfully host the stated development.

6. Litigation and dispute sweep

A search across eCourts (district courts), the relevant State High Court, and revenue tribunals for cases naming the borrower, the property, or prior owners in the chain. A pending partition suit, an injunction, or an attachment can freeze the asset regardless of how clean the deed looks. The sweep should cover the parties in the 30-year chain, not just the current owner.

7. NCLT / IBBI insolvency search (for corporate borrowers)

When the borrower or guarantor is a company or LLP, add a search of the National Company Law Tribunal (NCLT) and IBBI records for any ongoing or admitted insolvency. This is non-negotiable for corporate collateral: once the Insolvency and Bankruptcy Code moratorium kicks in, enforcement of your security can be stayed and the asset folded into the resolution process. A pristine title is cold comfort if the company is already in insolvency resolution.

8. Spatial and zoning confirmation

Finally, prove the parcel on paper is the parcel on the ground. A spatial overlay (in Karnataka, K-GIS) checks the survey number against the cadastral map for the correct location, extent, and any overlap with government land, tank beds, forest, or road-widening lines. Two adjacent survey numbers, a hissa split recorded wrong, or a parcel sitting partly in a buffer zone are all caught here — not at enforcement.

How should the evidence pack be structured for a credit or investment committee?

Lead with the verdict, support it with a risk grid, and put the raw sources behind every line. A committee should be able to read the first page and know the recommendation, then drill into any finding to its source document without leaving the pack.

A workable structure:

The discipline that makes this "bank-grade" is traceability: every assertion on the risk grid links to the primary record behind it. "EC clear for 30 years" links to the EC. "No CERSAI charge" links to the search result. "Survey extent matches deed" links to the K-GIS overlay. A committee can only price risk it can independently verify; a finding without a source is an opinion, and opinions don't get audited.

This is also why standardisation matters more than length. If every pack scores the same eight components the same way, the committee builds a portfolio view — which deals carry conversion risk, which carry chain gaps, which carry litigation — instead of re-learning each file from scratch. For why a bank needs a full report and not just a lawyer's letter, see the difference between a TSR, a legal opinion, and an LSR, and how banks verify title before sanctioning a mortgage.

What can these records NOT tell you?

The pack is only as good as the records behind it, and every source has a blind spot. An honest committee memo names them rather than implying certainty.

• An EC is not exhaustive. It records what was registered in that sub-registrar's index for the period and property described. Unregistered agreements, equitable mortgages, and transactions filed under a slightly different name or survey description can be missed — which is exactly why CERSAI and the mutation cross-check exist alongside it.
• CERSAI relies on correct filing. It only shows charges that institutions actually registered, indexed against the details they entered. A charge filed against a mis-typed address or an old survey number may not surface on your search.
• The RTC/Pahani is a revenue record, not a title deed. It establishes possession and cultivation, not ownership. Indian courts have repeatedly held that revenue entries do not by themselves confer title — they corroborate it.
• Court searches depend on indexing and naming. A case filed against a prior owner, a company's former name, or with a misspelt party can evade a name search. Coverage also varies by court and by how current the online cause-lists are.
• Spatial overlays carry survey error. Cadastral maps are digitised from older survey records; a K-GIS boundary is an indication, not a substitute for a physical demarcation by a licensed surveyor.

None of this makes the records useless — it makes the cross-checking the value. A single source can be wrong; eight sources that agree are hard to fake, and the points where they disagree are precisely the red flags worth a lawyer's attention.

Where does AI fit, and where does the lawyer?

AI does the gathering and the first draft; the lawyer reviews the chain and signs. That division is the whole model — and for collateral, the signature is what the committee and the auditor actually rely on.

An AI pipeline can pull the RTC, sweep the deeds and EC from Kaveri 2.0, run the CERSAI and court searches, translate Kannada records into English, reconcile names and extents across sources, and assemble the eight components into a standardised draft pack with every source linked. That turns a multi-week manual exercise — and an inconsistent one, since two analysts rarely produce identical packs — into a consistent, same-shape output the committee can compare deal to deal.

What AI does not do is opine on marketability or sign the title report. A qualified lawyer reads the chain, weighs the gaps, decides whether a defect is fatal or curable, and puts their name to the verdict. The pack exists to make that review fast and well-evidenced — not to replace it.

Frequently asked questions

What is the difference between a Title Search Report and a collateral evidence pack? A Title Search Report establishes whether title is clear and marketable. A collateral evidence pack does that and reframes everything around enforceability of the lender's security — adding a CERSAI search for equitable mortgages, an NCLT/IBBI insolvency check for corporate borrowers, and a standardised risk grid so a credit or investment committee can compare the deal against others. The TSR is a component of the pack, not the whole of it.

Why isn't an encumbrance certificate enough on its own? An EC only records transactions that were registered with the sub-registrar for the stated period and property description. The most common bank security in India — an equitable mortgage by deposit of title deeds — is often not registered there; it shows up on CERSAI instead. An EC can read "clear" while another lender already holds a charge. You need the EC and the CERSAI search and the mutation cross-check together.

Do I always need an NCLT or IBBI search? For corporate or LLP borrowers and guarantors, yes — treat it as mandatory. If insolvency proceedings are admitted, the IBC moratorium can stay enforcement of your security and pull the asset into the resolution process, so a clean title alone won't protect recovery. For an individual borrower with no corporate vehicle in the chain, the eCourts and High Court litigation sweep covers the equivalent ground.

How far back should the chain of title go? The market standard is around 30 years of continuous, registered transfers — a convention for high-value and dispute-prone deals that reaches well past the 12-year limitation period for most private adverse-possession claims. The chain must be unbroken — each deed connecting cleanly to the next — and for inherited or partitioned property you should see the succession or partition documents that explain every change of hands. A gap in the chain is a defect that needs explaining, not ignoring.

Can AI produce a bank-grade pack without a lawyer? AI can gather every record, translate Kannada documents, reconcile names and extents, run the searches, and draft the full pack with sources attached — fast and in a consistent format. But the marketability verdict and the signed title opinion must come from a qualified lawyer who reviews the chain and decides whether each gap is fatal or curable. AI gathers and drafts; the lawyer reviews and signs. That signature is what the committee and the auditor rely on.

What makes a pack "bank-grade" rather than just a pile of documents? Three things: standardisation (every deal scored on the same eight components), traceability (every finding linked to the primary source behind it), and an answer-first structure (verdict and risk grid on top, raw records in an annexe). A 200-page deed dump is not diligence; a one-page risk grid backed by linked, verifiable sources is. The goal is to let a committee compare collateral on one risk framework and audit any line back to its record.