The True Cost of Fraud for NBFCs —
and Why Intelligence Pays for Itself
A data-driven case for why Collections and Fraud leaders must treat fraud intelligence as a P&L imperative, not a compliance checkbox.
KYCKART Research
Risk Intelligence Team
Executive Summary
India's NBFCs are growing at ~19% annually and look healthy on headline metrics. But RBI's own stress tests warn that NBFC gross non-performing assets (GNPA) could nearly double — from 2.9% to 5.8% — under a severe stress scenario by March 2026.
That stress doesn't materialise out of thin air. It is born the day a synthetic identity clears your KYC, the hour an overleveraged borrower adds your NBFC as their fourth lender, or the quarter a trusted insider manipulates disbursement records. This article maps the six cost layers that make fraud so expensive — and shows, with data, why fraud intelligence is the most underspent line item in your risk budget.
Direct Loss
The Write-Off Nobody Wants to Own
When a fraudulent borrower defaults, the NBFC absorbs a direct write-off. The loan was real; the disbursement was real; the loss is real. But the system-level scale of this problem is rarely appreciated in its totality. Over the last decade, Indian banks alone wrote off approximately ₹16.35 lakh crore in bad loans — with retail and consumer loans now emerging as the fastest-growing contributor.
Bad Loans Written Off
₹0L Cr
By Indian banks over the last decade
Parliament data / AngelOne
Written Off FY2023-24
₹0L Cr
Lowest in 5 years — still enormous
The Wire / Lok Sabha
Retail Share of Write-Offs
0%
In FY25 — up from 6.8% in FY15
Moneylife analysis
Annual Bank Write-Offs (FY2020–FY2025)
Scheduled Commercial Banks, India · ₹ Lakh Crore
Source: Lok Sabha data (The Wire); AngelOne analysis
Retail vs. Non-Retail Share of Write-Offs
% of total write-offs · FY2014-15 vs FY2024-25
Retail Loans
Non-Retail
Source: Moneylife analysis of MoF/RBI data — retail write-offs reached ₹45,404 Cr in FY25
"The cheapest fraud case to manage is the one you never onboard. Every rupee that tips into a write-off started as a 'normal' account at origination."
Collections Cost
Burning Ops Budget on Unrecoverable Accounts
Here is the number that should stop every Collections Head cold: Parliament data tabled in December 2024 shows that Indian banks recovered only ~20.5% of loans they had written off over the preceding five financial years. That means for every ₹100 of bad debt that reaches write-off, roughly ₹79–80 is gone permanently — after years of legal pursuit, collection calls, and operational expenditure.
Avg Recovery Rate
0%
On written-off loans — 5-year aggregate
Parliament reply
Best-Year Recovery
0%
FY2023-24 — still 73% unrecovered
Parliament reply
Lost Per ₹100 Written Off
~₹0
After years of collections effort — permanently
Derived from 20.5% recovery rate
Recovery Rate on Written-Off Loans by Cohort
% recovered from each annual write-off cohort · Indian Scheduled Commercial Banks
Source: Parliament reply, December 2024 (Sansad.in)
The Cost Model: A ₹500 Crore NBFC (Illustrative)
"RBI data shows banks recover barely one-fifth of what they write off. For a mid-size NBFC, that's not just a number — it's a permanent P&L drain that compounds every year fraud goes undetected at the gate."
Regulatory Cost
RBI Enforcement Is No Longer a Tail Risk
Fraud and collections misconduct used to be reputational risks. They are now enforcement risks. FACE's compilation of RBI penal and enforcement actions in FY 2024–25 counted 79 enforcement actions — 48 of them against NBFCs. In FY 2025–26, the count was 70 actions, with NBFCs again accounting for roughly 46% of the total.
RBI Actions FY2024-25
0
Total enforcement actions issued by RBI
FACE compilation
Actions vs. NBFCs
0
61% of all RBI enforcement actions in FY24-25 targeted NBFCs
FACE compilation
RBI Complaints Filed
0
About digital lending apps and recovery harassment (Apr 2021–Nov 2022)
Business Standard
RBI Enforcement Actions: NBFCs vs. Banks
Number of penal/enforcement actions · Compiled by FACE
Source: FACE FY2024-25 and FY2025-26 enforcement compilations
Beyond direct penalties, regulatory non-compliance affects provisioning requirements, audit scope, and — for listed or credit-rated NBFCs — ratings and investor confidence. The indirect cost of a regulatory action almost always exceeds the stated penalty amount.
"RBI enforcement did not peak and go away — NBFCs have been hit 48 times one year, 32 the next. Enforcement is now an annual pattern, not an exception."
Operational Drag
The Hidden Cost Inside Your Teams
Every fraud alert that has to be manually reviewed is a cost. Every good customer slowed by a false positive is a cost. Every collections agent working an account that was fraudulent at origination is a cost. Fraud does not only produce write-offs — it produces an invisible tax on the capacity of your operations function.
Micro-LAP 90+ DPD
0%
Dec 2025 — up 35bps YoY despite improving headline metrics
TransUnion CIBIL CMI, Mar 2026
Overlap Borrowers at Risk
0%
MFI + retail overlap borrowers already 30+ DPD on at least one obligation
CRIF High Mark MicroLend
MFI-Retail Overlap Pool
0%
Live MFI borrowers who also hold an active retail loan
CRIF High Mark MicroLend
The Overlap Borrower Problem
Bureau data reveals that borrowers in the MFI + retail overlap are 30+ DPD at rates exceeding one-third of the segment. Your collections team is likely working these accounts — without intelligence to know they were structurally distressed at origination.
~₹15,800 crore of microfinance loans (5% of sector portfolio) sit with overleveraged borrowers from more than 3 lenders, with ~10% already 30–180 DPD.
Reputational Cost
When Fraud Becomes a Brand Event
Fraud doesn't stay inside the balance sheet. When borrowers are harassed by collection agents operating on fraudulent-account instructions, when digital lending apps linked to an NBFC engage in abusive recovery, or when internal control failures make headlines — the damage extends to credit ratings, investor relations, partner trust, and customer acquisition costs.
India Fraud Exposure
0%
Indian organisations experiencing financial fraud in last 24 months — vs. 41% globally
PwC Economic Crime Survey 2024
Integrity Incidents
0%
Indian organisations reporting a significant integrity incident in the last 2 years
EY Global Integrity Report 2024
Internal Fraud Origin
>50%
Major Indian corporate frauds originate in procurement or operations, not cyber attacks
KPMG Fraudster Profile India
India Fraud Exposure vs. Global Benchmark
% of organisations reporting fraud in last 24 months · PwC 2024
Source: PwC Global Economic Crime Survey 2024 India Outlook
"Your biggest fraud risk isn't a hacker — it's the long-tenured insider sitting between origination and collections, operating in a system where 59% of Indian firms have already seen a major fraud in just two years."
The ROI Flip
What Fraud Intelligence Costs vs. What It Prevents
The five cost layers above — direct write-offs, collections ops waste, regulatory penalties, operational drag, and reputational damage — are not independent. They cascade. A bad account onboarded today generates write-off cost at month 6, collections cost at month 12, potential regulatory exposure at month 18, and reputational fallout if it becomes part of a pattern. Fraud intelligence interrupts the cascade at the source.
GNPA Trend: Banking System vs. NBFCs
% Gross NPA · 2018–2026 · *Severe Stress Projection (RBI FSR)
Source: RBI FSR June 2025; RBI Trend & Progress 2024-25; PIB; FIDC India
| Cost Layer | Without Intelligence | With Intelligence |
|---|---|---|
| Direct Write-Off | Full loss on fraudulent disbursements; 79.5% irrecoverable | Fraud blocked at origination; write-off risk eliminated for flagged accounts |
| Collections Cost | Ops team working unrecoverable accounts with no intelligence signals | Early-warning triggers; collections focused on viable recovery targets |
| Regulatory Cost | Exposure to RBI enforcement (₹21–33 Cr in annual penalties industry-wide) | Audit-ready controls; DL Directions and SRO-FT alignment documented |
| Operational Drag | High manual review burden; false positives slow good customers | Automated signals reduce review queue; straight-through processing improves |
| Reputational Cost | DLA harassment risk; court liability; rating and investor exposure | Documented controls reduce liability; proactive risk posture visible to auditors |
| Net Position | Compounding cost cascade per fraud event | One investment interrupts all five cost layers |
The Intelligence Equation
Prevent ₹1 Cr of Fraud
Save ₹3–5 Cr in Total Costs
When you factor in write-off loss, collections ops cost, regulatory exposure, and reputational drag — preventing a fraudulent account at origination is worth 3–5× the face value of the loan.
"RBI's own stress tests say NBFC GNPA can almost double under severe conditions. The question isn't whether your collections team can work miracles — it's whether your fraud stack can stop tomorrow's 90+ DPD pool from being born today."
Ready to Verify Smarter?
Stop Fraud Before It Becomes a Write-Off
KYCKART's fraud intelligence stack gives Collections and Risk teams the bureau-linked, real-time signals they need to act at origination — not at month 12.