Deposit Modelling
Executive Summary
Key Takeaways
- ✓ Deposits should be segmented by product type — transactional (non-interest or low-interest, no contractual maturity), savings, and term — each with its own volume, cost, and behavioural assumptions.
- ✓ Behavioural stickiness, not just contractual maturity, drives how a deposit segment should actually be treated for funding and liquidity purposes, since transactional deposits have no contractual maturity at all but are often the most behaviourally stable funding source a bank has.
- ✓ Deposit pricing should be modelled with a deposit beta assumption — the proportion of a change in market interest rates that is passed through to deposit rates — rather than assuming deposit costs move one-for-one with market rates.
- ✓ Segment-level deposit balances feed directly into the interest expense build and into the loan-to-deposit ratio and liquidity metrics that constrain how much loan growth the funding base can support.
- ✓ A model that treats all deposits as a single undifferentiated funding pool cannot support meaningful liquidity or funding-concentration analysis, since the behavioural and cost differences between segments are exactly what liquidity risk management depends on.
Objective¶
This guide covers how to segment and model a bank's deposit base, within the Banking Financial Modelling pillar, as the liability-side counterpart to Loan Portfolio Modelling.
Segmentation¶
Deposits should be segmented by product type, since each carries a fundamentally different cost and behavioural profile:
| Segment | Contractual Maturity | Typical Cost | Typical Behavioural Stickiness |
|---|---|---|---|
| Transactional (current/checking) | None | Zero or near-zero interest | Often high — a core, stable funding source despite no contractual lock-in |
| Savings | None | Low, variable | Moderate — sensitive to rate competition over time |
| Term/time deposits | Fixed | Higher, often fixed for the term | Contractually locked during the term, but renewal at maturity is not guaranteed |
Behavioural Stickiness vs. Contractual Maturity¶
The central modelling insight for deposits is that contractual maturity and behavioural stickiness are not the same thing. Transactional deposits have no contractual maturity at all, yet are frequently the most behaviourally stable funding source a bank has. Term deposits have a fixed contractual maturity, but depositors may choose not to renew at maturity if competing rates elsewhere are more attractive. A deposit model should carry a behavioural stickiness or run-off assumption for each segment, distinct from its contractual terms, since this is what actually drives funding stability and feeds regulatory liquidity metrics.
Deposit Pricing: The Deposit Beta¶
Deposit costs should be modelled using a deposit beta assumption — the proportion of a change in market interest rates that the bank passes through to its own deposit rates. A deposit beta below 1.0, common for transactional and savings deposits, means deposit costs move by less than market rates, materially affecting the margin outcome in a rising- or falling-rate environment. Modelling deposit costs as moving one-for-one with market rates, without a segment-specific beta, overstates how quickly and how much deposit costs actually respond.
Segment Deposit Cost (Period t) = Segment Deposit Cost (Period t-1)
+ Deposit Beta × Change in Reference Market Rate
Feeding the Rest of the Model¶
Segment-level deposit balances and costs feed the liability side of the interest income build, and total deposit volume forms the denominator of the loan-to-deposit ratio, one of the primary constraints on loan book growth. The segment-level behavioural stickiness assumptions built here also feed directly into regulatory liquidity metrics — see Liquidity Coverage Ratio and Net Stable Funding Ratio.
Common Construction Pitfalls¶
- Modelling all deposits as a single blended balance and cost, losing the segment-level behavioural detail liquidity and funding-concentration analysis depends on.
- Assuming deposit costs move one-for-one with market rates rather than applying a segment-specific deposit beta.
- Treating contractual maturity as equivalent to behavioural stickiness, rather than modelling them as distinct properties of each deposit segment.
- Failing to connect segment-level deposit assumptions to the loan-to-deposit ratio and regulatory liquidity metrics that depend on them.
Continue Reading¶
Prerequisites¶
- Banking Financial Modelling — the parent pillar
Related Technical Guides¶
Related Glossary¶
Related Technical Guides (Liquidity)¶
How OXXON tests thisRun a free structural check with FMAE
Frequently Asked Questions
How should a bank's deposit base be modelled?
Segmented by product type — transactional, savings, and term deposits — each with its own volume, cost, and behavioural stickiness assumption, rather than as a single undifferentiated funding pool.
What is deposit stickiness, and why does it matter?
The degree to which a deposit balance remains stable over time regardless of its contractual terms. Transactional deposits have no contractual maturity but are often highly sticky in practice, while some term deposits, despite a contractual maturity, may not renew at maturity if rates move against the depositor — behavioural stickiness, not contractual maturity alone, is what actually matters for funding and liquidity planning.
What is a deposit beta?
The proportion of a change in market interest rates that a bank passes through to its deposit rates. A deposit beta below 1.0 means deposit costs rise (or fall) by less than market rates move, which is common for transactional and savings deposits and central to modelling net interest margin accurately in a changing rate environment.
How does deposit modelling connect to the loan portfolio module?
Segment-level deposit balances and costs feed the liability side of the interest expense build, and total deposit volume is the primary constraint the loan-to-deposit ratio measures loan growth against — see Loan-to-Deposit Ratio.
Why can't deposits be modelled as a single blended balance?
Because the behavioural and cost differences between transactional, savings, and term deposits are exactly what liquidity risk management and funding-concentration analysis depend on — collapsing them into one figure removes the information those disciplines need.
How does deposit modelling relate to liquidity metrics like LCR and NSFR?
The segment-level behavioural stickiness assumptions built here are a direct input to regulatory liquidity metrics, which apply different assumed run-off or stability factors to different deposit types — see Liquidity Coverage Ratio and Net Stable Funding Ratio.
Related Articles
Banking Financial Modelling
Banking financial modelling is structurally distinct from a standard corporate model: it is built balance-sheet-first, with earnings derived from asset and liability volumes and spreads rather than a top-line revenue forecast, and it must represent loan portfolio and deposit dynamics, credit loss provisioning, and a set of bank-specific KPIs that a generic corporate model has no equivalent for. This page is the hub for the Knowledge Centre's banking modelling content: how the bank business model translates into a model's architecture, how the three financial statements are structured for a bank, how interest income and the net interest margin bridge are built, and how loan portfolios, deposits, and credit loss provisions should be modelled.
Interest Income Modelling
Interest income modelling is the core mechanic of a bank financial model: interest income and expense are derived from forecast asset and liability volumes and their associated yields and costs, not from a standalone revenue assumption. This guide covers how to structure that build at a segment-by-segment level, how net interest income and net interest margin are calculated from it, and how to construct the net interest margin bridge that separates a period's margin change into volume, rate, and mix effects — the single most useful diagnostic output in a bank model.
Loan Portfolio Modelling
Loan portfolio modelling is the asset-side counterpart to deposit modelling: the loan book should be segmented by product type, risk grade, or business line, each carrying its own origination, repayment, yield, and expected loss assumptions. This guide covers how to structure that segmentation, how to roll forward segment-level balances period over period, and how the segmented output feeds both the interest income build and credit loss provisioning.
Loan-to-Deposit Ratio
The loan-to-deposit ratio compares total loans to total deposits, giving a core indicator of how much of a bank's lending is funded from its deposit base versus wholesale or other funding sources. A ratio above 100% means the bank is lending more than it holds in deposits, funding the difference through wholesale markets — a funding structure that carries more refinancing and liquidity risk than deposit-funded lending.
Liquidity Coverage Ratio
The liquidity coverage ratio (LCR) tests whether a bank holds enough high-quality liquid assets to survive a defined 30-day acute stress scenario. This guide covers how to model the LCR's two components — the stock of high-quality liquid assets and net cash outflows under the stress scenario — and how the deposit and funding behavioural assumptions built elsewhere in the model feed directly into the outflow calculation.
Net Stable Funding Ratio
The net stable funding ratio (NSFR) tests whether a bank's longer-term assets are backed by a stable enough funding profile over a one-year horizon, complementing the short-term liquidity coverage ratio. This guide covers how to model the NSFR's two components — available stable funding, weighted by the behavioural stability of each funding source, and required stable funding, weighted by the tenor and liquidity of each asset — and how it connects to the balance sheet forecast and deposit modelling already built elsewhere in the model.