Core and Volatile Deposits: ALM Bucketing Rules for CAIIB BFM

Every deposit carries two maturities: the contractual one on the receipt, and the behavioural one the customer follows. Splitting the liability book into core and volatile deposits is how a bank turns that behaviour into a number it can slot, fund and hedge. For CAIIB BFM this single idea links the structural liquidity statement, the LCR, the NSFR and interest rate risk in the banking book.

🏦 What Core and Volatile Really Mean

Core deposits are the portion of the deposit base that has historically stayed with the bank irrespective of rate movements, seasonal pressure or competitive offers. They are typically retail, granular and relationship-driven — salary accounts, small savings balances, the sticky floor of an operating current account.

Volatile deposits are balances that leave quickly when price or sentiment changes: bulk term deposits, interbank and financial-institution money, broker-sourced funds and large single-name corporate balances.

The distinction is behavioural, not contractual. A savings deposit is legally repayable on demand, yet a large slice of it behaves like multi-year funding. A one-year bulk deposit may roll off entirely at the first repricing date if a competitor bids two basis points higher.

Regulation reinforces the split by size and price. Under RBI rules a bulk deposit for a scheduled commercial bank (other than an RRB) and a small finance bank means a single-rupee term deposit of ₹3 crore and above; for RRBs and local area banks the threshold is ₹1 crore. Because such deposits are priced aggressively, they behave volatilely. The chapter on Funding and regulatory aspects ties the pricing and regulatory threads together.

💡 Exam Tip: If a question describes a deposit by its contractual tenor and asks how ALM should treat it, the answer almost always turns on behaviour, not on the contract.

📊 How Banks Separate the Two

There is no single regulatory formula for carving out the core portion, so banks use a documented internal methodology approved by ALCO. Four approaches dominate in practice, and CAIIB questions can be built on any of them.

• Minimum balance method: take the lowest balance the pool has touched over a rolling 24 or 36 months and treat that floor as core. Simple, conservative and auditable.
• Statistical volatility method: compute the mean balance and subtract a multiple of the standard deviation of daily balances — commonly 1.65 or 2.33 deviations, a 95% or 99% floor. The excess is volatile.
• Decay or run-off modelling: fit the observed attrition of a deposit vintage over time, derive a half-life, and slot balances along that decay curve rather than at a single point.
• Regression modelling: relate balances to policy and market rates to estimate how much money walks out for a given rate shock, which also yields the deposit beta.

Segmentation matters as much as the maths. Balances are split by product, customer type, channel and ticket size, because a ₹15,000 salary-account balance and a ₹40 crore corporate current account do not behave alike.

The same behavioural curves drive the deposit beta and repricing lag used in margin planning, which is why this work sits beside cost of deposits and deposit pricing rather than apart from it.

🧾 Slotting in the Structural Liquidity Statement

The Structural Liquidity Statement is where classification becomes a supervisory number. RBI's ALM framework requires banks to distribute all inflows and outflows across defined time buckets — next day, 2–7 days, 8–14 days, 15–28 days, 29 days to 3 months, over 3 to 6 months, over 6 months to 1 year, over 1 to 3 years, over 3 to 5 years, and over 5 years.

Non-maturity deposits have no contractual date, so they are slotted on behavioural logic. The volatile portion goes into near-term buckets, the core portion into the over-1-year and longer buckets. Term deposits are slotted at residual maturity, adjusted for observed premature-withdrawal rates and for the share that historically renews.

The resulting mismatch is policed by tolerance limits: the cumulative negative gap must not exceed 5% of cumulative cash outflows in the next-day bucket, 10% in the 2–7 day bucket, 15% in the 8–14 day bucket and 20% in the 15–28 day bucket.

Because the core share directly relieves the short buckets, it is the most powerful assumption in the whole statement. The mechanics sit alongside Introduction to Treasury Management & Treasury Products, and current instructions are published on the RBI notifications portal.

⚠️ Common Mistake: Slotting the entire savings-bank balance into the 1–14 day buckets. That is not conservative, it is wrong: it overstates short-term outflow and distorts the gap ratios.

💧 How the LCR and NSFR Read the Same Deposits

Basel III replaces bank-specific judgement with prescribed factors. In the LCR, deposits attract stressed 30-day run-off rates; in the NSFR the same deposits earn Available Stable Funding factors. Retail and small-business money is treated as stickiest, financial-institution money as least sticky.

RBI's revised LCR instructions, effective 1 April 2026, add a 2.5 percentage point run-off add-on to retail and small-business deposits enabled for internet and mobile banking, on the reasoning that digitally reachable money can move in minutes. A retail deposit counts as stable only if it is fully covered by deposit insurance and sits in a transactional or established relationship account; everything else is less stable by definition.

Note the asymmetry candidates miss: a deposit can be 95% stable for the NSFR yet still bleed a run-off in the LCR, because the ratios answer different questions — a 30-day stress versus a one-year funding profile. System liquidity shifts behaviour too, hence the Standing Deposit Facility and liquidity corridor.

⚖️ Interest Rate Risk, Governance and Validation

The core split is not only a liquidity input. In interest rate risk in the banking book, non-maturity deposits are the largest source of modelling judgement, because assuming a long core life makes the balance sheet look naturally hedged. Basel's standardised approach therefore caps both the share treatable as core and its average repricing life — broadly 90% and five years for retail transactional accounts, 70% and 4.5 years for retail non-transactional, and 50% and four years for wholesale.

Governance follows the model-risk playbook: annual review, back-testing against realised balances, independent validation outside the business line, ALCO sign-off, and sensitivity analysis showing what breaks if the core share falls ten percentage points.

Two practical guardrails matter in the exam. First, deposit concentration — the share held by the top twenty depositors — because a granular ₹500 crore behaves nothing like a single ₹500 crore. Second, portal-sourced digital deposits, which show materially faster attrition than branch-sourced money of the same size.

Behavioural liability life plugs straight into asset-side measures such as duration and convexity, while valuation intent is covered in trading book vs banking book. For the supervisory frame, work through Risk Regulations in Banking industry.

📌 Remember: Core is an assumption, not a fact. Every basis point of comfort it buys must be earned with data and validated independently.

🧠 Practice MCQs: Core and Volatile Deposits

Want chapter-wise mock tests with 100+ MCQs? Start practising free →

❓ Frequently Asked Questions

Bringing it together for the exam

Classify a deposit, justify the split, slot it, then map it to an LCR run-off and an NSFR factor — that is most of what BFM asks on liabilities. Practise on the CAIIB course material and revise on the Bank Financial Management hub.