Bank Reconciliation Software Built for Indian Banks
Indian bank statements arrive in formats that generic reconciliation tools do not read cleanly — MT940 and CAMT.053 from corporate accounts, CIB and YONO Business Excel exports, PDFs from co-operative banks, and narration fields carrying UPI VPAs, NEFT fragments, and cheque clearing codes. TransactIG is reconciliation software for India that parses every major Indian bank format natively, matches on narration patterns where references are missing, and produces the month-wise audit evidence that CARO 2020 statutory auditors examine.
Why Bank Reconciliation Is Harder in India
The narration field carries the match key
In most Indian bank statements, the structured reference field is either empty or carries a truncated fragment. The narration field — a free-text string up to 150 characters — is where the actual match key lives. A UPI credit arrives with narration UPI/VPA customer@bank/567890123456. A NEFT from a vendor arrives with narration NEFT N012345678 INV-2024-00456 ABC ENT. A salary payment batch arrives with narration SAL APR24 BATCH-1023. Generic reconciliation tools that look only at the reference field cannot match any of these — they route everything to an exception queue, where a finance analyst manually reads the narration and keys in the match.
The narration conventions are also bank-specific. HDFC's cheque clearing narration is not the same as ICICI's or SBI's. UPI VPA formatting differs across banks. A reconciliation system built for the Indian market has to treat narration parsing as a first-class capability, not an afterthought — and the rule set has to be maintained as banks change their narration templates.
Format fragmentation across banks
A typical Indian mid-market corporate holds accounts at 3 to 5 banks: an operating account, a collection account, a payroll account, a statutory payments account, and perhaps a foreign-remittance account. Each bank delivers its statement in a different format. HDFC and ICICI offer MT940 through their corporate cash-management platforms. SBI delivers YONO Business Excel — a different structure from private bank Excel exports. Axis and Kotak support MT940 but the field conventions vary. Co-operative banks typically deliver PDF, requiring OCR or pattern parsing.
Consolidating statements from five banks into a single reconciliation view — with intercompany transfers between accounts auto-matched — is an engineering problem, not a spreadsheet problem. The National Payments Corporation of India (NPCI) standardises the rails (NEFT, RTGS, UPI, NACH), but bank-side statement representation remains fragmented.
What TransactIG Bank Reconciliation Does
TransactIG ingests bank statements from every major Indian bank in their native formats, parses narration fields using India-specific pattern rules, and matches each bank line to the corresponding ERP voucher — taking match rates from the 51% typical of manual and narration-blind tools to 88% through automation.
Native Indian bank format support
MT940, MT942, CAMT.053, HDFC NetBanking Excel, ICICI CIB, SBI CMS, YONO Business, Axis Corporate Online, Kotak ActivMoney, and PDF statements from co-operative banks — all ingested and normalised into a single bank register without manual format conversion.
Narration pattern matching
Bank-specific narration rules extract cheque numbers, UTRs, UPI VPAs, invoice fragments, and employee ID prefixes from the free-text narration field. The match engine uses the extracted value as the match key even when the structured reference field is empty or corrupted.
Multi-account consolidation
Statements from every operating, collection, payroll, and statutory account roll into a single consolidated reconciliation view. Intercompany transfers between accounts auto-match by amount-and-date pair — a single NEFT from HDFC to ICICI is one matched pair, not two exceptions.
Bank charges and GST classification
Processing charges, NEFT/RTGS fees, cheque return charges, and CMS fees are classified into the correct expense category. 18% GST on bank charges is surfaced for input tax credit claim, cross-referenced against the bank's GSTIN in GSTR-2B.
Salary and payroll bulk matching
Bulk salary transfers appear in the bank statement as one aggregated debit covering hundreds of employees. TransactIG links the bulk debit to the underlying payroll batch and reconciles each salary credit on the employee account side against the payroll register and the TDS deduction schedule.
CARO 2020 audit evidence
Month-wise reconciliation register, ageing report for unmatched items (30/60/90/120+ days), maker-checker sign-off timestamps, and variance taxonomy — the evidence file statutory auditors examine under CARO 2020, produced automatically with every close cycle.
Indian Bank Format Coverage
Every bank has its own statement delivery format and narration convention. TransactIG handles each natively — the list below covers the banks that account for the large majority of Indian corporate banking volume. For detail on a specific bank's format, see the bank-specific guides linked under each row.
| Bank | Statement Formats Supported | Narration Characteristics | Volume Role |
|---|---|---|---|
| HDFC Bank | MT940 (corporate), CAMT.053, NetBanking Excel, CMS host-to-host SFTP | Cheque-clear: CHQ DEP 123456 / NEFT: NEFT N012345678 / UPI: UPI/VPA@bank | Highest-volume corporate bank in India — typical large enterprise has 2–4 HDFC accounts |
| ICICI Bank | MT940, CIB (Corporate Internet Banking) Excel, CAMT.053 on request | NEFT narrations carry partial reference; UPI narrations include VPA and RRN | Second-highest corporate volume — CIB statement pull is the standard integration path |
| State Bank of India | YONO Business structured Excel, SBI CMS batch files, PDF (branch-generated) | Distinct narration format vs private banks; government account narrations include scheme codes | Dominant for statutory payments, PSU-linked accounts, and government-vendor collections |
| Axis Bank | MT940 (corporate), Corporate Online Excel, MT942 intra-day reporting | Standardised narration format; cheque and NEFT references usually clean | Common for mid-market corporates and NBFC collection accounts |
| Kotak Mahindra Bank | MT940 (ActivMoney for corporates), Structured Excel, API (select customers) | Cleaner narration conventions than most PSU banks; UPI references consistent | Growing share among D2C and new-economy corporates |
| Co-operative and smaller PSUs | PDF statements (most common), occasional CSV; no MT940 availability | Inconsistent narration across branches; requires pattern-based parsing | Matters for NBFCs, manufacturing, and businesses with branch-level banking |
Bank-specific guides: HDFC · ICICI · SBI · MT940 format · narration patterns
How TransactIG Bank Reconciliation Works
Ingest statements from every bank
Host-to-host SFTP for banks that support it, scheduled pulls from NetBanking portals where they do not, direct MT940 delivery for corporate accounts. Statements are ingested on the cadence you define — daily for active operating accounts, weekly for low-volume statutory accounts.
Match against ERP voucher data
Vouchers from SAP FI, Oracle Fusion, Tally Prime, Zoho Books, Dynamics 365, or any mid-market ERP are mapped to the matching schema during implementation. Every bank line is matched against the voucher register on reference where clean, and on narration pattern extraction where reference is missing — the same approach finance teams use manually, automated and at scale.
Exception queue with variance classification
Unmatched items route to a structured exception queue with variance codes: TIMING_DIFF, MISSING_VOUCHER, BANK_ERROR, PARTIAL_MATCH, DUPLICATE_ENTRY, OPENING_BAL. Each exception carries its ageing, its suggested resolution action, and its maker-checker status — the audit trail CARO 2020 evidence requires, produced without finance team intervention.
What is Bank Reconciliation
Bank reconciliation is the process of matching entries in a company's cash book or general ledger against entries in its bank statement to verify that every cash movement recorded in the books has a corresponding entry in the bank, and vice versa. Any difference between the two — a deposit in the books that has not yet appeared on the statement, a bank charge debited by the bank but not yet recorded in the books, a cheque issued but not presented — is identified, classified, and resolved through a reconciliation register that becomes the auditor's evidence of control.
For Indian companies, bank reconciliation is not optional. The Companies Act 2013 requires a "true and fair" view of financial position, and the statutory auditor's CARO 2020 reporting includes a direct comment on bank reconciliation controls. Long-outstanding unreconciled items — a cheque issued 180 days ago and still unpresented, a deposit in transit ageing beyond the bank's clearing window, a bank charge repeatedly debited without corresponding ledger entry — are flagged in the audit report, and systemic failures can trigger an adverse or qualified opinion.
The practical challenge in India is that bank reconciliation has to work across formats (MT940, CAMT.053, CIB Excel, YONO Business, PDF), across narration conventions (every bank formats its narration differently), and across payment rails (NEFT, RTGS, UPI, NACH, IMPS, cheques). The system that handles this natively — rather than treating Indian bank statements as a special case of a generic global format — is what makes automated reconciliation viable at scale.
Related Guides
Bank Reconciliation Process
What changes at enterprise scale — the complete process guide.
Salary and Payroll Bank Reconciliation
Bulk transfer matching and TDS alignment for Indian payroll cycles.
CARO 2020 Bank Reconciliation Audit
What the statutory auditor examines and the evidence to produce.
Frequently Asked Questions
What is bank reconciliation software? +
Bank reconciliation software automates the matching of entries in a company's books — the cash book or general ledger — against entries in its bank statement. For Indian finance teams, that means ingesting bank statements from HDFC, ICICI, SBI, Axis, Kotak, and other banks in their native formats (MT940, CAMT.053, Excel, PDF), parsing narration text to identify the underlying transaction type (UPI, NEFT, RTGS, cheque, auto-debit), and matching each bank line to the corresponding voucher or journal in the ERP. The output is a reconciled bank register with an exception queue for unmatched items — the evidence file that statutory auditors examine under CARO 2020.
Which Indian bank statement formats does TransactIG support? +
TransactIG ingests statements from every major Indian bank in their native export formats. MT940 (SWIFT format) for corporate accounts at HDFC, ICICI, Axis, Yes Bank, and Kotak. CAMT.053 for banks that have migrated to the ISO 20022 standard. Structured Excel exports from SBI CMS, HDFC NetBanking, ICICI CIB, and YONO Business. PDF statements for smaller co-operative banks and PSU bank branches that do not yet offer structured exports. The parsing layer handles bank-specific narration conventions — for example HDFC's cheque-clearing narration format differs from SBI's, and both differ from ICICI's UPI narration pattern.
How is narration-based matching different from reference-based matching? +
Reference-based matching works when the bank statement carries a clean UTR, cheque number, or transaction ID that exactly matches a reference in the ERP voucher. Narration-based matching is what finance teams fall back on when the reference is missing or corrupted — which is most of the time in Indian bank statements, where UPI transactions carry VPA strings, salary payments carry employee ID fragments, and vendor payments sometimes carry only a truncated invoice number in the narration text. TransactIG parses the narration field using pattern rules trained on Indian bank conventions, extracting the cheque number, UTR, VPA, or invoice fragment, and using it as the match key even when the bank has not populated a structured reference field.
Does TransactIG handle multi-account and multi-bank consolidation? +
Yes. Large Indian corporates typically operate current accounts across three to eight banks — an HDFC operating account, an ICICI collection account, an SBI statutory payments account, and so on. TransactIG ingests statements from every account into a single consolidated bank register, reconciles each account to its own ledger in the ERP, and then presents a consolidated exception view across all banks. Intercompany transfers that move cash between accounts are recognised and auto-matched by the amount-and-date pair, so an NEFT from the HDFC account to the ICICI account is not treated as two separate unmatched items.
What audit evidence does TransactIG produce for CARO 2020 bank reconciliation? +
CARO 2020 requires statutory auditors to comment on the existence of bank reconciliation controls and on any long-outstanding unreconciled items. TransactIG produces the evidence file that covers both: a month-wise reconciliation register showing every bank line matched against its corresponding ledger entry, an ageing report for unmatched items (30/60/90/120+ days), maker-checker sign-off timestamps for each reconciliation cycle, and a variance taxonomy that classifies every unresolved difference by type (timing, data-entry error, missing voucher, bank error). The output is auditor-ready without further manipulation — the same register the auditor would have asked the finance team to build by hand.
How long does bank reconciliation implementation take? +
Implementation takes 2 to 4 weeks for most Indian corporates. Week one covers bank statement ingestion — either via host-to-host SFTP for banks that support it (HDFC, ICICI, Axis CMS integration), direct MT940 delivery for corporate accounts, or scheduled download from NetBanking portals where host-to-host is not available. Week two maps ERP voucher data — SAP FI, Oracle Fusion, Tally Prime, Zoho Books, Dynamics 365 — to TransactIG's matching schema. Weeks three and four apply matching rule configuration for narration patterns, tolerance windows, and exception routing. No code development is required. TransactIG is cloud-only and runs from AWS Mumbai.
Stop routing every narration-only bank line to a manual exception queue
TransactIG connects to your banks and your ERP in 2 to 4 weeks. MT940, CAMT.053, CIB, YONO Business, PDF — all ingested natively. ISO 27001:2022 certified. Runs from AWS Mumbai.