MODULE-II

Banking Operations & Digital Banking

1. PAYMENT METHODS (NEFT, RTGS, IMPS)

1.1:NEFT (National Electronic Funds Transfer)
• Settlement: Processed in batches, taking approximately 2-3 hours for the recipient
to receive the funds.
• Availability: Operating window is limited to Weekdays & Saturdays only (closed on
Sundays and bank holidays).
• Charges: Economical transaction fee ranging from Rs 2.50 to Rs 10 depending on
the amount.
• Best Suited For: Routine, non-urgent, and regular fund transfers.

[Link] (Real Time Gross Settlement)

• Settlement: High-speed processing, typically settling within 30 minutes to 2 hours
on a continuous basis.
• Availability: Restricted to Weekdays & Saturdays only.
• Minimum Transaction Limit: Designed for high-value transfers; requires a minimum
amount of Rs 2 lakh.
• Charges: Premium service fee ranging from Rs 25 to Rs 100.
• Best Suited For: Large, high-value, and critical business/personal transactions.
[Link] (Immediate Payment Service)
• Settlement: Instantaneous processing; funds are transferred within 15-30 seconds.
• Availability: Accessible 24/7/365, including weekends, late nights, and public
holidays.
• Transaction Limits: Typically capped between Rs 1 lakh to Rs 2 lakh per day.
• Charges: Moderate fee structure ranging from Rs 5 to Rs 15.
• Best Suited For: Urgent, real-time, and immediate fund requirements.

Real Time scenario

• The Situation: Mr. Employee needs to transfer Rs 1,50,000 on a Sunday evening.
• The Recommendation: IMPS is the only viable and optimal method.
• Justification:
1. Availability: NEFT and RTGS are unavailable on Sundays, whereas IMPS works
24/7.
2. Speed: It ensures instant settlement within seconds.
3. Feasibility: The requested transfer amount of Rs 1,50,000 fits perfectly within the
IMPS daily limit.
4. Convenience: The transaction can be executed seamlessly via mobile banking
apps.
2. KNOW YOUR CUSTOMER (KYC)
Definition
A mandatory and legally binding process for financial institutions to verify the
authentic identity of their customers and evaluate potential legal or financial risks
before or during a relationship.

Importance of KYC

• Fraud Prevention: Actively identifies and blocks fake, stolen, or fabricated identities.
• Anti-Money Laundering (AML): Tracks and detects suspicious financial activities,
preventing illicit money movement.
• Risk Assessment: Allows banks to evaluate a customer's risk profile and transaction
patterns objectively.
• Regulatory Compliance: Strictly mandated and enforced by the Reserve Bank of
India (RBI).
Officially Valid Documents (OVDs)
• Proof of Identity (PoI): Passport, PAN Card, Voter ID, Aadhaar.
• Proof of Address (PoA): Utility bills (electricity/water), Registered Rental
Agreements, Official Bank Statements.

3. MOBILE BANKING
Definition
The provision and execution of financial and banking services directly through cellular
devices or smartphones.
Four Key Functional Features

1. Account Management: Allows users to check live account balances, view

transaction history, and download statements.
2. Fund Transfers: Enables seamless money movement via integrations with NEFT,
RTGS, and IMPS.
3. Bill Payments: Facilitates quick payments for utilities, insurance premiums, mobile
recharges, and loan EMIs.
4. Loan Management: Provides platforms to apply for new loans, upload documents,
and track approval status.
Core Security Measures

1. End-to-End Encryption:
• Mechanism: Encrypts all data transmitted between the mobile application and the
bank's core servers. Only the authorized sender and receiver possess the
cryptographic keys to decode it.
• Example: If you transfer money, the details are scrambled into unreadable
gibberish code; even if intercepted by hackers, it remains useless.
2. Two-Factor Authentication (2FA):
• Mechanism: Layered security that demands two distinct forms of identification
before granting access or approving a transaction.
 • Example: A user must enter their secret account password/PIN followed by a
dynamic, time-sensitive One-Time Password (OTP) sent to their registered mobile
number.
3. Biometric Authentication:
• Mechanism: Utilizes unique biological characteristics (fingerprints or facial
recognition) rather than traditional text passwords to verify identity.
• Example: Unlocking the banking application using a fingerprint sensor, where
biometric mathematical data is securely isolated and stored locally on the physical
phone hardware.
AI IN BANKING
Artificial Intelligence (AI) has transformed the banking sector from a traditional,
reactive system into a proactive, data-driven industry. By analyzing massive datasets
in real time, AI enhances security, automates customer engagement, and
personalizes the financial experience.

1. AI-Powered Fraud Detection & Mitigation

Traditional fraud detection relied on static, rule-based systems (e.g., "flag any
transaction over $5,000"). AI replaces this with dynamic, self-learning models that
analyze thousands of data points simultaneously to identify sophisticated financial
crimes.

A. Pattern Recognition
Detailed Explanation: AI models are trained on historical transaction data to
recognize complex signatures of legitimate vs. fraudulent behavior. Instead of looking
at a single transaction in isolation, it evaluates the broader context (merchant type,
geographic location, velocity of transactions, and device signatures) to spot
anomalies that human auditors or basic software would miss.

Real-World Example: If a credit card is used to buy coffee in New York, and 10
minutes later the same card is used to buy a luxury watch in Mumbai, the AI
recognizes this physically impossible geographical velocity pattern and instantly flags
the transaction.

B. Real-Time Monitoring
Detailed Explanation: Modern banking infrastructure processes millions of
transactions every second. AI infrastructure operates at millisecond scale, evaluating
risk scores for transfers, card swipes, and online payments instantaneously before
the transaction is finalized.

Real-World Example: During high-volume shopping events like Black Friday, a bank's
AI system evaluates millions of swipes per minute, approving legitimate high-velocity
holiday shopping while blocking compromised card details across global servers
concurrently.

C. Behavioral Analytics
Detailed Explanation: This creates a unique "behavioral biometric" profile for every
single customer. The AI learns your normal habits: what time you usually check your
account, how fast you type, your typical spending ranges, and your frequent device
IDs. If an action falls outside this individualized baseline, it triggers an alert.

Real-World Example: If a user typically logs in from an iPhone in London at 8:00 PM

and transfers small amounts, but suddenly logs in via an unknown Android device
from an IP address in another country at 3:00 AM and attempts to drain the account,
the behavioral analytics engine will halt the transfer and demand secondary multi-
factor authentication.

D. Predictive Analysis
Detailed Explanation: Instead of just reacting to fraud that has already occurred,
predictive AI uses historical trends, emerging global cyberthreat data, and machine
learning to forecast where vulnerabilities might appear next. It looks for systemic
shifts or testing patterns (like small micro-transactions used by hackers to verify if a
stolen card works) to stop large-scale breaches before they happen.

Real-World Example: An AI system notices a sudden, minor uptick in low-value ($1-

$2) charges from an obscure online merchant across thousands of seemingly
unrelated bank accounts. Predicting that this is a "card-testing" phase of a
coordinated cyber-attack, the bank proactively blocks that specific merchant
gateway, preventing millions of dollars in subsequent major thefts.

2. AI in Customer Service & Engagement

AI-driven customer service shifts banking from a rigid, branch-dependent model to an
instantaneous, hyper-personalized virtual experience.

A. Intelligent Chatbots & Virtual Assistants

Detailed Explanation: Moving far beyond simple automated phone trees or scripted
FAQ responses, Natural Language Processing (NLP) enables conversational AI bots to
understand human intent, context, and phrasing. These bots connect directly to
banking core systems to execute account actions securely without human
intervention.

Real-World Example: A customer types, "Hey, I lost my wallet at the airport, please
stop my card!" into their banking app at 2:00 AM on a Sunday. The AI chatbot instantly
understands the urgency, locks the specific debit card, confirms the user’s mailing
address, and orders a replacement card within seconds—completely bypassing the
need to wait on hold for a live call center agent.

B. Hyper-Personalized Recommendations
Detailed Explanation: AI aggregates a customer's income patterns, savings rates,
transaction histories, and life stages to build a custom financial profile. It uses this
data to recommend highly relevant financial products exactly when the customer
needs them, rather than blasting generic advertisements to all users.
Real-World Example: The AI observes that a customer has had a steady 15% increase
in savings over the last year, alongside frequent searches or payments related to
home improvement. Instead of showing them a generic credit card ad, the app
proactively presents a personalized home equity loan offer with an interest rate
tailored specifically to their excellent credit profile.

C. Sentiment Analysis
Detailed Explanation: This technology analyzes text strings, email inquiries, social
media mentions, or voice inflections during customer calls to detect the emotional
state of the customer (e.g., frustration, anger, satisfaction). This allows banks to
prioritize escalating issues and adapt communication styles dynamically.

Real-World Example: If a customer is typing an email about a reversed fee and uses
aggressive language or capitalization, the AI text analysis tool flags the message as
"High Frustration." It instantly bypasses the standard queue and routes the ticket to a
senior customer retention specialist, prepended with a summary note explaining the
customer's grievance to ensure a compassionate resolution.
CORE BANKING SYSTEMS (CBS)
Def: Core Banking System (CBS) is the back-end infrastructure that connects
multiple branches of a bank to a centralized data server. "CORE" stands for
Centralized Online Real-time Exchange. This network allows transactions to be
recorded and updated simultaneously across the bank's entire ecosystem, meaning
the customer is no longer tied to their specific "home branch" where they opened the
account; instead, they are a customer of the entire bank.

Four Function:
Anywhere Banking: Because account details reside on a central cloud or data server
rather than a physical ledger inside a specific branch, a customer can walk into any
branch nationwide to deposit cash, withdraw money, or update documents.

Example: A student who opened an account in Mumbai can seamlessly withdraw

cash or get a passbook updated at a branch in Delhi without needing to request data
transfers between physical locations.

Real-Time Information Integration:

Every modification to an account balance—whether via a cash withdrawal at a branch
teller, an online transfer, or an ATM transaction—is processed immediately. The
central database updates instantaneously, mitigating the risk of double-spending or
outdated balance readings.

Example: If you transfer money via a mobile app, a teller looking at your profile at a
physical branch counter a second later will see the updated balance immediately.

Integrated Financial Services: CBS unifies disparate banking systems (retail

banking, corporate lending, wealth management, and investments) into a single
customer profile. This gives the bank a 360-degree view of the customer’s financial
health and streamlines internal workflows.

Example: When applying for a home loan, the loan department can instantly see your
savings history, active fixed deposits, and credit card payments through the unified
interface without asking you to fetch statements from separate departments.

Structural Core Functions

Account Management: Maintaining master data profiles for clients, managing KYC
records, calculating interest on savings accounts automatically at the end of cycles,
and enforcing account status holds (e.g., freezing an account on regulatory orders).

Transaction Processing: Authenticating, clearing, and settling cash deposits,

physical check clearances, wire transfers, and automated clearing house (ACH)
salary credits.

Loan Management: Tracking the full lifecycle of a loan from its initial setup,
calculating interest accruals, scheduling automated Equated Monthly Installment
(EMI) deductions, and tracking overdue penalties.

Reporting & Analytics: Generating regulatory compliance reports for central banks
(like the Reserve Bank of India), analyzing liquidity positions, and measuring daily
balance sheets across the entire corporate entity.

SECURED VS. UNSECURED LOANS:

When a young professional decides to purchase an asset like a car, they are faced with two
primary financing architectures.

Types of Loans :Secured vs Unsecured Loans

Secured Loans: - Definition: Loans backed by collateral (an asset) –

Examples: * Home Loan (house is collateral) * Car Loan (vehicle is collateral) * Gold
Loan (gold jewelry is collateral) - Lower interest rates (8-12% typically) - If you don’t pay,
bank can seize the asset

Unsecured Loans: - Definition: No collateral required

Examples: * Personal Loan * Credit Card * Education Loan (sometimes) - Higher interest
rates (12-24%) - Higher risk for banks - Depends heavily on credit score

Real-world example: “Personal loans are popular for weddings, medical emergencies. But
if you default, the bank can’t take anything—they can only damage your credit score and
take legal action.”.
Interest Calculation Methods
Simple vs Compound Interest
Simple Interest: SI = (P × R × T) / 100 - P = Principal (₹1,00,000) - R = Rate (10% per
annum) - T = Time (3 years) - SI = ₹30,000 - Total = ₹1,30,000
Compound Interest: CI = P(1 + R/100)^T - P - Same values - CI = ₹33,100 - More
expensive for borrower!
Explain: “This is why credit cards are expensive—they compound interest daily!”
EMI Calculation (Conceptual)
Formula: EMI = [P × R × (1+R)^N] / [(1+R)^N - 1]
Where: - P = Principal loan amount - R = Monthly interest rate (Annual rate ÷ 12 ÷ 100)
- N = Loan tenure in months
Example on screen: ₹10 lakh home loan at 8.5% for 20 years - Monthly EMI ≈ ₹8,678 -
Total payment = ₹20,82,720 - Interest paid = ₹10,82,720 (more than the principal!)
Interactive moment: “Notice something shocking? You pay almost double the loan
amount over 20 years. This is why loan tenure matters!”
Show EMI breakdown chart: - First few years: Mostly interest - Last few years: Mostly
principal - This is called amortization

Part 3: Credit Appraisal Process

The 5 C’s of Credit
1. Character - CIBIL score (we’ll discuss in detail next) - Loan repayment history -
How responsible is the borrower?
2. Capacity - Income level - Debt-to-Income ratio - Can they afford the EMI?
Example calculation: “If Priya earns ₹60,000/month and already has EMIs of
₹20,000, banks typically won’t give more than ₹20,000 EMI (keeping total below 50-
60% of income)”
3. Capital - Down payment ability - Savings and investments - Shows financial
discipline
4. Collateral - For secured loans - Property valuation - Legal verification
5. Conditions - Purpose of loan - Economic conditions - Industry risk (for business
loans)
CIBIL Score Deep Dive
What affects CIBIL score?
1. Payment history (35% weightage) - Late payments hurt badly - Defaults are
disastrous
2. Credit utilization (30%) - Keep credit card usage below 30%
3. Credit history length (15%) - Older accounts are better
4. Credit mix (10%) - Mix of secured/unsecured is good
5. New credit inquiries (10%) - Too many applications lower score
Real scenario: “Amit has CIBIL 780. He gets home loan at 8.5%. His friend with CIBIL
650 gets the same loan at 9.5%. Over 20 years, this 1% difference means ₹2-3 lakhs
extra!”
Income Verification
Documents required: - Salaried: Last 6 months salary slips, bank statements - Self-
employed: ITR for last 2-3 years, business financials - Bank statement analysis: *
Checking for regular income * Bounce checks? * Gambling transactions? *
Suspicious patterns?
Collateral Evaluation
For property:
1. Legal verification - Clear title? - No pending cases?
2. Technical verification - Actual valuation by surveyor - Construction quality
3. Loan-to-Value (LTV) ratio - Banks typically give 75-90% of property value - Rest is
down payment
Example: “House worth ₹1 crore. Bank sanctions ₹80 lakhs (80% LTV). You need ₹20
lakhs as down payment.”

To prevent loans from deteriorating into NPAs in the first place, institutional banking desks
deploy five primary mitigation frameworks:

NPA:

1. Portfolio Diversification

• Banks must avoid concentrating too much loan capital inside a single domain.
Credit portfolios must be distributed across different industrial sectors, geographic
regions, and client types.
• Example: A bank establishes clear internal exposure caps: limiting commercial real
estate to 10% of total loans, retail auto loans to 15%, agriculture to 20%, and
manufacturing to 25%. If the real estate market crashes, the bank remains protected
because 90% of its loan book is tied to other, stable sectors.

2. Credit Insurance & Third-Party Guarantees

 • For large corporate loans or high-risk sectors, banks shift the default risk onto a
third party. This involves purchasing credit insurance policies or requiring
creditworthy corporate parents or government entities to act as co-signers who
assume liability if the primary borrower defaults.
• Example: An export business applies for a ₹50 crore loan to ship textiles overseas.
The bank secures a guarantee from the Export Credit Guarantee Corporation
(ECGC). If the international buyer goes bankrupt and defaults, the insurance policy
covers up to 90% of the bank's losses.

3. Continuous Monitoring & Early Warning Signals (EWS)

• Instead of waiting for an account to pass the 90-day default mark, banks track real-
time behavioral data to spot financial distress early. Accounts showing signs of
stress are flagged as Special Mention Accounts (SMA) before they officially
become NPAs.
• Example: A bank's automated monitoring system notices a corporate borrower has
started delaying utility bill payments, their factory's electricity consumption has
dropped by 40%, and the promoters are frequently selling off company stock. The
system flags these Early Warning Signals (EWS), prompting credit officers to visit
the factory and review their accounts before a formal default happens.

4. Dynamic Loan Provisioning

• Banking regulations require institutions to proactively set aside a percentage of

their operating profits to form a cash buffer against potential loan defaults. The
required provisioning percentage scales up aggressively based on how long the loan
has been overdue.
• Example: For a standard, healthy loan, a bank might only need to set aside a 0.40%
cash reserve. However, if that loan slips into the "Substandard NPA" category, the
regulator mandates that the bank instantly lock away 15% of the loan value into a
non-useable provisioning account, reducing the bank's reportable profit for that
quarter.

5. Advanced Technological Applications

• Utilizing artificial intelligence, machine learning algorithms, and big data to

analyze a borrower's creditworthiness more accurately than traditional paperwork
checks allow. This includes evaluating digital footprints, tax filings, cash flow
velocity, and transaction histories in real time.
• Example: An AI credit assessment engine processes a small business loan request
by analyzing their real-time point-of-sale digital transactions and GST tax filings.
The algorithm flags a hidden cash flow mismatch that a human loan officer might
have missed, prompting the bank to lower the approved loan limit to prevent a
future default.

A Secured Loan is backed by a tangible asset owned by the borrower (e.g., a house for a
home loan, or a car for an auto loan). If the borrower defaults, the bank holds the legal
right to seize and auction that asset to recover the unpaid loan balance. Because the lender's
risk is low, secured loans feature lower interest rates (typically 7%–10%) and longer
repayment terms.

An Unsecured Loan requires no collateral asset backing (e.g., personal loans or credit
cards). Approval is granted solely based on the borrower's steady income and credit
history. Because the bank faces a high risk of loss if the borrower defaults, unsecured loans
feature much higher interest rates (typically 12%–18%) and shorter repayment terms.

Why is a CIBIL score critical in the loan appraisal process?

A Credit Information Bureau (India) Limited (CIBIL) score is a 3-digit numeric summary
(ranging from 300 to 900) that rates an individual's financial credit history and repayment
reliability. It is critical because it serves as a bank's primary screening filter:

o Risk Evaluation: A high score (above 750) proves the applicant has a
spotless record of paying credit cards and EMIs on time, signaling a low
risk of default. A low score (below 650) flags past payment delays, defaults,
or excessive debt.
o Underwriting Speed: It allows banks to instantly filter out high-risk
applicants during the early stages, preventing future NPAs.
o Pricing Leverage: Borrowers with excellent CIBIL scores can negotiate for
lower interest rates and faster processing turnarounds, while low scorers are
either rejected outright or charged higher interest risk premiums.

What is a Non-Performing Asset (NPA), and when is a loan classified as such?

A Non-Performing Asset (NPA) is a classification given to a bank loan or advance that has
stopped generating regular income for the financial institution because the borrower has
stopped making payments.

Classification Timeline: A standard loan account is officially classified as an NPA if the

principal or interest payment remains overdue for a continuous period exceeding 90
days. Once it passes this 90-day threshold, the bank must stop recognizing accrued interest
as profit and move the asset into its bad loan ledger across specific regulatory categories
(Substandard, Doubtful, or Loss).