DQIApril 2026 · 9 min read

RBI's Data Quality Index for Credit Information — What Every Credit Institution Must Know

The Data Quality Index is one of the most underappreciated compliance metrics in credit information reporting. From July 1, 2026, the DQI framework becomes significantly more demanding — file-level DQI delivered within 3 days, monthly DQI brought forward, and a mandatory weighted average formula. Poor DQI now has direct regulatory consequences.

What Is the Data Quality Index

The Data Quality Index (DQI) is a composite score assigned by CICs to each Credit Institution's credit information submissions. It measures the quality of the data submitted — not just whether the file was received on time, but whether the content is accurate, complete, timely, and internally consistent.

The DQI is not merely an internal quality metric. It is reported to RBI through the DAKSH portal on a half-yearly basis, and CICs are required to place DQI performance before their Board sub-committees semi-annually. Credit Institutions with persistently poor DQI scores face escalating regulatory consequences — including being reported to RBI's Department of Supervision.

The Four DQI Parameters

ParameterWeightWhat It Measures
Accuracy30%Correctness of data fields reported — name, address, PAN, account details, outstanding balance
Completeness30%Percentage of mandatory fields populated without blanks or nulls. Missing fields are the most common DQI failure mode
Timeliness25%Adherence to reporting reference date deadlines. Late submissions directly reduce DQI score
Consistency15%Alignment between related fields — e.g., outstanding balance vs. overdue amount, account status vs. DPD

The weight distribution reflects the RBI's priorities. Accuracy and Completeness together account for 60% of the DQI score — because an incomplete or inaccurate credit record is worse than no record at all from a credit assessment perspective.

The New Two-Tier DQI Delivery Framework — Effective July 1, 2026

The Amendment Directions of December 2025 introduce a fundamentally different DQI delivery structure. Under the current framework, CICs provide one monthly DQI report per Credit Institution. From July 1, 2026, the framework becomes two-tier:

File-Level DQI
Coverage: Each individual file (all 4 reference dates) across all segments
Delivered by: CIC → CI
Within 3 calendar days of file receipt
CI-Level Monthly DQI
Coverage: Aggregate of all files submitted by CI in the month, all segments
Delivered by: CIC → CI
By 10th of the following month (brought forward from 15th)

The Weighted Average DQI Formula

Para 20(3) of the Amendment Directions mandates that the CI-level monthly DQI must be computed as the weighted average of file-level DQIs, weighted by number of records per file:

Monthly DQI = Σ(Records in File × File DQI Score) ÷ Σ(Total Records in all files)
Illustrative example: A CI submits 4 files in a month — 200, 250, 220, and 200 records — with file-level DQI scores of 95, 98, 100, and 98 respectively. Total records = 870. Weighted DQI = (200×95 + 250×98 + 220×100 + 200×98) ÷ 870 = 85,260 ÷ 870 = 98.00

DAKSH Portal Reporting — The Regulatory Escalation Path

CICs must report non-compliant Credit Institutions — those that miss submission deadlines — to RBI's Department of Supervision through the DAKSH portal on a half-yearly basis (March 31 and September 30 each year). This creates a direct regulatory escalation path from poor DQI performance to RBI supervisory action.

Additionally, the CIC Board sub-committee must review DQI performance data semi-annually. CICs that do not have adequate DQI governance frameworks — or that have Board sub-committees that are not genuinely engaging with DQI data — face potential findings in RBI inspections.

The Most Common DQI Failure Modes

From an advisory perspective, the most common DQI failures we encounter are not random — they follow predictable patterns rooted in LMS system limitations and operational process gaps:

Null mandatory fields

LMS does not populate all TUDF/Metro 2 mandatory fields. Data extraction leaves blanks rather than defaulting to required placeholder values.

Incorrect DPD calculation

LMS NPA logic does not accurately compute Days Past Due — leading to mismatches between the DPD field and the account status field, triggering consistency failures.

Late file submission

Manual processes for generating and transmitting the credit information file miss the reference date deadline — especially on months where the reference date falls around weekends or holidays.

Demographic data mismatch

Customer name, address, and PAN discrepancies between the LMS and the CIC-returned data — often stemming from data entry inconsistencies at origination.

Is your DQI score where it needs to be for July 2026?

A DQI audit will identify your specific failure modes and build an LMS-level remediation plan before the new framework comes into force.

Book a DQI Advisory Call