Bank Treasury & Liquidity

Treasury keeps obligations funded at the time they fall due. Read the daily cash lifecycle before the ratios, then use the scenarios to connect funding, collateral, operations and governance.

Scope: banking students and practitioners. Regulatory explanations distinguish Basel standards from domestic implementation. All numerical cases and system designs are illustrative, with currencies, units and assumptions stated. Local rules, facility terms and accounting policy govern actual transactions.

1. Why bank treasury liquidity is the survival function

Credit risk can take years to show up in full. Market risk can hurt you in a week. Liquidity risk can kill a bank in a few days: sometimes in a few hours: even when the bank still looks solvent on paper.

That single fact is why Bank Treasury & Liquidity Management sits at the centre of modern banking. A bank can report healthy capital ratios, solid loan books and even decent profits, and still fail if it cannot meet cash obligations when they fall due. Solvency is about whether the value of assets exceeds the value of liabilities over time. Liquidity is about whether the bank can pay what it owes today, tomorrow morning, and under stress without destroying franchise value.

Treasury’s liquidity mandate is therefore not a side activity. It is the bank’s survival function. Everything else: lending, payments, client markets business, growth plans: depends on the bank remaining able to fund itself and meet outflows continuously.

In practical terms, liquidity management answers questions like these every single day:

• Do we have enough cash and near-cash to meet all known and reasonably expected outflows today?
• If a large corporate withdraws, or several facilities are drawn at once, what happens to our position?
• Are our liquid assets truly liquid, or only liquid in calm markets?
• Can we roll wholesale funding at a sensible cost, or are we becoming dependent on short-term money?
• If confidence weakens, what is our playbook, who acts, and in what order?
• Are our regulatory ratios (LCR, NSFR and internal stress metrics) telling the truth about resilience, or only about a formula?

This chapter is built to make those questions practical. You will learn how a treasury desk actually runs a day, how buffers are built, how LCR and NSFR work in real language, how intraday pressure arises, how funding strategy is designed, how contingency plans are tested, and how governance keeps the whole system honest.

If Foundations gave you the map of Markets & Treasury, this chapter walks the most important defensive territory on that map: keeping the bank liquid under normal conditions and under stress.

Learning objectives

By the end of this chapter you should be able to:

• Explain liquidity risk in plain banking language and distinguish it from solvency and from market risk.
• Describe how a treasury desk builds and manages the daily cash position.
• Read a sources-and-uses view of the balance sheet and connect it to funding needs.
• Explain what high-quality liquid assets are, why composition matters, and what “liquidity of liquid assets” means under stress.
• Walk through LCR and NSFR in practical terms: what goes into the numbers, what the ratios are trying to force, and where banks still get surprised.
• Understand intraday liquidity pressure and why payment timing can matter as much as end-of-day balances.
• Describe a sensible wholesale funding strategy and a funding ladder.
• Explain what a real Contingency Funding Plan contains and why testing matters more than the document.
• Connect FTP and ALM decisions to daily liquidity reality.
• Recognise the classic mistakes that still cause liquidity crises.
• Approach a liquidity dashboard or ALCO liquidity pack with practitioner questions, not passive reading.

2. What liquidity really means in a bank

Liquidity in banking is not “having a lot of assets.” It is the ability to meet cash obligations as they fall due, without unacceptable losses and without damaging the franchise.

Three ideas sit underneath that definition.

2.1 Timing is the core of liquidity

A loan that will be repaid in five years may be a fine credit asset. It does not pay the deposit that is withdrawn tomorrow. A bond that will mature in ten years may be credit-safe. If you must sell it tomorrow in a weak market, the cash you raise may be far less than the accounting value. Liquidity is about the schedule of cash in and cash out, not only the stock of assets.

This is why two banks with similar-looking balance sheets can have completely different liquidity profiles. One may fund long and hold short-duration liquid assets. Another may fund short and hold long-duration securities that look “safe” until rates move or markets freeze. The second bank is living closer to the edge, even if its capital ratio looks acceptable.

2.2 Market liquidity and funding liquidity are related but not the same

Market liquidity is the ability to sell an asset quickly without moving the price a lot. Funding liquidity is the ability to raise cash to meet obligations, whether by drawing on cash balances, selling assets, borrowing, or using central bank facilities.

In calm markets the two support each other. In stress they can fail together. Assets that were liquid yesterday become hard to sell. Counterparties who lent to you yesterday demand more collateral or refuse to roll. That twin failure is why liquidity crises feel sudden even when the underlying weaknesses built slowly.

2.3 Accounting solvency does not guarantee payment ability

A bank can be solvent: assets exceed liabilities in value terms: and still be unable to pay. Assets may be hard to sell, slow to repay, or only valuable if held to maturity. Liabilities may be demandable or short-dated. This mismatch between value and timing is the heart of liquidity risk.

That is also why liquidity management is a distinct discipline from credit analysis and from pure capital management. Capital absorbs losses over time. Liquidity ensures the bank can still operate while those longer processes play out: or while stress is still unfolding.

2.4 The practical question treasury asks every day

Strip away the theory and treasury’s daily question is simple:

Given everything we know about inflows, outflows, maturities, client behaviour and market conditions, can we meet our obligations today and over the near horizon without breaching limits, damaging relationships, or relying on emergency measures?

Everything else in this chapter: buffers, ratios, funding plans, stress tests, contingency playbooks: exists to answer that question under both normal and stressed conditions.

3. The daily cash position: how a treasury day actually runs

If you only remember one operational picture from this chapter, remember the daily cash position.

3.1 What the cash position is

The cash position is the bank’s running view of liquidity for the day and the near term: opening cash and central-bank balances, expected inflows, expected outflows, large known items, and the residual surplus or deficit that must be placed or covered.

It is not a single static number. It is a living picture that updates as payments settle, clients act, markets move, and forecasts are revised.

3.2 Opening the day

A typical treasury morning starts with:

• Overnight and opening balances at the central bank and in key nostro accounts
• Maturing money-market deals, repos, and wholesale funding
• Known large payments (tax, clearing, client settlements, internal transfers)
• Expected loan drawdowns and deposit movements based on pipeline and history
• Any market or operational alerts that could change behaviour

The desk is not trying to forecast the entire economy. It is trying to avoid being surprised by the bank’s own flows.

3.3 Building the intraday and end-of-day view

Through the day, treasury updates:

• Actual versus expected flows
• Large client movements that were not in the forecast
• Settlement outcomes
• Market conditions for placing surplus or covering deficit
• Regulatory and internal limit headroom

A surplus must be placed safely: usually in short-term instruments, interbank deposits, reverse repo, or central-bank facilities depending on the framework. A deficit must be covered: through wholesale borrowing, repo of liquid assets, or other approved channels: without creating a worse problem tomorrow.

3.4 The discipline of “ending the day square”

Most treasuries aim to end the day with a controlled residual: enough buffer, no accidental large uncovered deficit, and no lazy surplus left in the wrong place. “Square” does not mean zero. It means the residual is intentional, sized to policy, and understood.

Leaving a large unexplained deficit overnight is a control failure. Leaving a large surplus uninvested in a way that violates policy is also a failure. The standard is intentional management, not accidental balances.

3.5 Why this work is mostly invisible

Branch staff and relationship managers rarely see this process. Clients see payments that work. Boards see summary ratios. The desk sees the continuous adjustment that makes those outcomes possible. When the work is done well, nobody notices. When it fails, everybody notices at once.

That invisibility is why liquidity culture matters. A bank that only celebrates visible revenue and treats cash positioning as clerical work is training itself to be late to the real risk.

3.6 A funded payment from forecast to reconciled cash

The front-office cash dealer sees a projected USD deficit in the bank's own legal entity. Operations validates opening nostro cash from the correspondent's account record, rather than assuming yesterday's general ledger balance is current. A corporate drawdown, USD securities purchase and maturing wholesale deposit are booked with separate deal identifiers and expected settlement times. A loan drawdown credited to a customer's account initially increases a loan asset and deposit liability; external cash leaves when the customer sends the money to another bank. Forecasts must reflect that distinction and avoid recording both the drawdown and the resulting payment as two independent external outflows.

The dealer arranges approved overnight funding. Middle office checks product, tenor, counterparty limit, price and authorised booking entity. Back office independently matches the economics and authenticated settlement instructions, submits the cash instruction through the relevant payment route and monitors acceptance, pending status and final settlement. A confirmation or network acknowledgement is not proof that cash arrived. If borrowing fails to settle, funding remains expected, the original obligation remains due, and Treasury selects a tested alternative before its cut-off. Infrastructures, formats and field requirements depend on the currency, scheme and agreement.

The cash engine replaces the expected flow with the settled event using a stable event/deal reference and amount, currency, account, value date, event time and source status. Replays must not add the same receipt twice. Finance books the borrowing liability and cash asset under approved policy, while Operations reconciles the deal cash schedule to settlement evidence and the nostro statement. Liquidity Risk receives the changed funding maturity and remaining headroom. Treasury closes the day with aged breaks, the responsible owner and tomorrow's repayment included in the ladder.

3.7 Balanced journals and why a deposit transfer can consume cash

These simplified journals use USD millions, no fees or FX, and a transaction that qualifies for secured borrowing accounting. They show the economic entries, not a mandatory bank chart of accounts. For an overnight repo, the borrower retains the security on its balance sheet when the applicable derecognition analysis requires it, and separately tracks collateral encumbrance. A legal title transfer and accounting derecognition are different tests. See IFRS 9 classification and recognition for the IFRS framework; US GAAP and other frameworks require their own analysis.

Interest is 100 × 0.036 × 1/360 = 0.010 million, or USD 10,000. If an overnight deal spans three actual days over a weekend, the same simplified convention produces USD 30,000, not USD 10,000. Local holidays determine value dates; the contracted day-count determines accrual. Payment-clearing suspense accounts may intervene between customer posting and external settlement and must reconcile, rather than being confused with settled cash.

4. Sources and uses of funds: reading the balance-sheet flow

Liquidity management becomes clearer when you stop looking only at stocks and start looking at flows.

4.1 Sources of funds

Typical sources include customer deposits (retail, SME, corporate, institutional), wholesale funding (interbank, certificates of deposit, commercial paper, bonds), loan repayments and maturities, asset sales or repo of securities, capital over longer horizons, and central-bank facilities when used.

Not all sources are equal. Sticky retail deposits behave differently from flighty wholesale money. Long-term bonds behave differently from overnight interbank funds. A healthy funding mix is diversified by product, counterparty, tenor and behaviour under stress.

4.2 Uses of funds

Typical uses include customer lending and facility drawdowns, security purchases for liquidity and investment portfolios, payment and settlement obligations, maturing wholesale liabilities, collateral and margin postings, and operating cash needs.

The daily and structural job of treasury is to keep sources and uses aligned in timing and resilience, not only in average volume.

4.3 The practical reading skill

When you look at a balance sheet or a funding report, ask: What portion of funding can leave quickly? What portion of assets can turn into cash quickly without large loss? Where are the maturity walls? Which client segments are concentrated enough that a few decisions move the whole position? Is growth in lending being funded by equally stable growth in deposits, or by shorter wholesale money?

These questions turn a static statement into a liquidity story.

4.4 Structural versus tactical

Tactical liquidity is the day-to-day matching of cash. Structural liquidity is the longer design of the balance sheet: how much stable funding the bank holds against less liquid assets, how long the wholesale book runs, and how dependent the franchise is on confidence-sensitive money.

A bank can run a clean daily cash desk and still have a weak structural position if it funds long illiquid assets with short confidence-sensitive liabilities. Both layers must be managed.

5. High-quality liquid assets (HQLA) and the liquidity buffer

The liquidity buffer is the stock of assets treasury holds specifically so the bank can survive outflows without immediate collapse. Under modern regulation, much of that buffer is defined as high-quality liquid assets.

5.1 What “high quality” and “liquid” are trying to mean

An asset is useful as a liquidity buffer if, under stress, it can be sold or repo’d quickly, at a reasonably predictable value, with limited forced-sale discount. In practice that points toward central-bank reserves where applicable, high-quality government securities, and certain other assets that regulations and market practice treat as reliable under stress.

The label matters less than the behaviour. An asset that is liquid in textbooks but illiquid in a panic is not a true buffer.

5.2 Composition risk inside the buffer

Not all HQLA behaves the same way. Very short government bills may offer excellent liquidity and low market-risk noise, but lower yield. Longer government bonds may count as HQLA and yield more, but carry duration risk. If rates rise sharply and the bank must sell, mark-to-market losses become cash losses at the worst time.

A common institutional failure is to fill the buffer with assets that maximise yield or regulatory convenience while underestimating sale risk under stress. The buffer’s job is survival, not portfolio optimisation theatre.

5.3 Operational readiness of the buffer

A buffer is only real if legal and operational capacity to sell or repo exists, custodian and settlement arrangements work, encumbrance status is clear, people know which assets are designated for liquidity use, and market access is maintained in peacetime, not invented in crisis.

Assets that are theoretically liquid but operationally stuck are decoration.

5.4 Buffer size is policy, not only ratio compliance

Meeting LCR is necessary where required. It is not automatically sufficient for the bank’s own risk appetite. Internal policy may demand more buffer, different composition, or extra headroom for franchise-specific risks such as concentrated corporate deposits or rapid facility drawdowns. Strong treasuries treat regulatory minimums as floors, not targets.

5.5 Eligible HQLA, market collateral and accounting value

HQLA eligibility is a regulatory test. Central-bank collateral eligibility is a facility test. Repo eligibility is a counterparty/agreement test. An accounting carrying amount is a measurement under the bank's accounting framework. The same bond can pass one test and fail another. Required reserves count only to the extent the relevant authorities allow drawdown in stress; an unrestricted cash balance and restricted cash are not interchangeable.

Under the Basel design, Level 1 assets have no LCR haircut or composition cap; eligible Level 2A assets receive a 15% haircut. Level 2B eligibility is subject to supervisory permission and larger asset-specific haircuts. The total Level 2 stock is capped at 40% and Level 2B at 15%, after haircuts and the specified unwind of short-term secured transactions. These are not percentages of the original gross securities portfolio. The Basel LCR text, paragraphs 45-54 and Annex 1 sets the classification and cap mechanics; local implementations may differ.

For a teaching case with no secured transaction unwind and no Level 2B, Level 1 = 60 and Level 2A market value = 60, all USD millions. Level 2A after haircut is 51. Eligible Level 2 cannot exceed (2/3) × 60 = 40, so HQLA is 60 + 40 = 100. Simply adding 60 + 51 would report 111 and breach the composition cap. A separate repo advance on those assets uses its own negotiated haircut, not the LCR haircut automatically.

Pre-positioning collateral at a central bank does not always mean it is excluded from HQLA. Basel permits qualifying assets pre-positioned or pledged there but unused to generate liquidity, subject to availability and operational conditions. Once used to obtain borrowing, the associated encumbrance, cash, repayment and cap-unwind treatment must be evaluated. Do not count a bond and its repo cash as two independent resources. These distinctions follow Basel LCR paragraph 31.

6. Liquidity Coverage Ratio (LCR) in practical banking terms

LCR is designed to ensure a bank holds enough HQLA to survive a significant 30-day stress scenario of outflows.

6.1 The basic idea in one sentence

Can the bank cover stressed net outflows over the next 30 days using high-quality liquid assets?

6.2 The practical structure

LCR = High-Quality Liquid Assets ÷ Total Net Cash Outflows over 30 days (under stress assumptions)

The numerator is the buffer. The denominator is a regulatory stress estimate of outflows minus limited inflows.

6.3 Why outflow assumptions matter so much

Different liabilities receive different run-off factors. Stable retail deposits are assumed to leave more slowly than certain wholesale or less stable deposits. Undrawn credit facilities receive drawdown assumptions. Contingent exposures are included in stylised ways.

This is why two banks with similar total deposits can have different LCR profiles. The quality and behaviour of funding matter as much as the headline volume.

6.4 What LCR is good at: and what it does not solve

LCR forces banks to hold a real short-term liquidity buffer, confront sticky versus flighty funding, reduce casual reliance on short-term wholesale money without matching liquid assets, and bring liquidity into board conversation in a comparable way.

It does not by itself guarantee intraday payment survival, structural funding strength over a year, perfect HQLA behaviour under every market path, management competence during a confidence event, or full capture of internal concentration risks.

6.5 Practitioner reading tips

When you see an LCR report, ask: What is driving the ratio this month? Are we near the internal trigger or only above the regulatory floor? Is HQLA composition becoming longer-duration or more concentrated? Which outflow categories dominate the denominator? If our largest corporate depositors left faster than the assumed factors, how painful would that be?

6.6 The inflow cap and a complete ratio calculation

For the standard Basel calculation, net cash outflows = stressed outflows − min(eligible inflows, 75% of stressed outflows). The inflows themselves must first satisfy the applicable recognition rules. Receipts from assets counted in HQLA cannot be counted again as inflows. Internal forecasts can show contractual amounts while the LCR applies prescribed recognition. National implementation, including any specific exemptions or alternative treatment, must be documented. See Basel LCR paragraphs 69-72.

Illustrative USD millions: eligible HQLA after haircuts/caps = 100, stressed outflows = 160, eligible inflows = 140. Recognised inflows = min(140, 120) = 120, net outflows = 40 and LCR = 100/40 = 250%. Without the cap, the incorrect result would be 100/(160 − 140) = 500%. If inflows fall to 60, net outflows rise to 100 and LCR falls to 100%. This sensitivity shows why a high ratio can depend on receivables that do not solve an early-morning cash gap.

Basel's minimum is 100% in normal conditions. The buffer is designed to be used during stress, potentially taking LCR below 100%, with the supervisory response reflecting the cause and circumstances. Preserving a ratio by failing to meet due obligations defeats the purpose of the buffer. Internal appetite headroom and escalation still apply. See the Basel LCR overview.

Run-off categories require evidence, not a marketing label. An insured retail deposit is not automatically a stable deposit: qualifying relationship/transactional conditions also matter. SME treatment depends on the rule's small-business eligibility. Corporate operational deposit treatment applies to qualifying required operating balances, with excess balances treated separately. Credit facilities and liquidity facilities have different purposes and factors. An analyst must classify the customer, facility purpose, commitment terms and amount before applying the local factor.

7. Net Stable Funding Ratio (NSFR) in practical banking terms

NSFR is designed to push banks toward more stable longer-term funding of less liquid assets.

7.1 The basic idea in one sentence

Is the bank funding its less liquid assets with enough stable funding over a one-year horizon?

7.2 The practical structure

NSFR = Available Stable Funding ÷ Required Stable Funding

Available Stable Funding assigns higher weights to more stable liabilities. Required Stable Funding assigns higher weights to assets that are harder to monetise quickly.

7.3 What NSFR is trying to prevent

NSFR targets the business model that looks fine in calm markets while funding long illiquid assets with short confidence-sensitive liabilities. It is a structural discipline tool, not a daily cash tool.

7.4 Practitioner reading tips

Ask: Is the ratio healthy because funding truly lengthened, or because asset mix temporarily helped? Where is RSF concentrated? Are we dependent on a few wholesale instruments for ASF? What happens to NSFR if loan growth accelerates and deposit growth does not?

7.5 Weighted funding, with a reconciled example

Available stable funding (ASF) is the weighted eligible funding and capital; required stable funding (RSF) is the weighted requirement for assets and relevant off-balance-sheet exposures. NSFR is not a promise that every asset matures inside a year or that every liability is locked for a year. Residual maturity, counterparty, asset liquidity, encumbrance and specified derivatives treatment matter. Basel sets NSFR at at least 100%; domestic scope and rules govern a bank's reporting. See Basel NSFR.

Illustrative USD millions, deliberately simplified to selected categories with no derivatives or encumbrance:

The balance sheet balances at 200; the undrawn 20 is an additional off-balance-sheet amount. NSFR = 155/123.5 = 125.51%. If 20 of the qualifying stable deposits is replaced with financial-institution funding <6 months, ASF falls by 19 to 136, so NSFR becomes 110.12%, even though total funding stays at 200. This is a funding-quality change, not a cash loss at that instant. Production engines must also handle call options, encumbrance duration, derivatives, initial margin and interdependent assets/liabilities under the governing rules. The example factors follow the NSFR category tables; they are not bank-specific classification advice.

8. Intraday liquidity: the hidden pressure of the payment day

End-of-day liquidity is necessary. It is not sufficient.

Payments, clearing, settlement and margin flows do not wait politely for the end-of-day report. A bank can be fine at close of business and still face intense intraday pressure if large payments leave before large receipts arrive, or if market infrastructures demand timing precision.

Intraday liquidity risk is the risk of being unable to meet payment obligations at the moment they are due during the day, even if the end-of-day position would have been acceptable.

Typical sources of pressure include large client payments early in the day, clearing and settlement cycles, collateral and margin calls, time-zone effects, operational delays, and market infrastructure rules.

Strong banks monitor intraday peaks and troughs, largest payment concentrations, critical payment windows, and operational playbooks for delayed receipts. Intraday management is part liquidity, part operations, part market infrastructure literacy.

8.1 Measure time-of-day use separately from regulatory ratios

Basel's seven intraday tools are monitoring tools, not an additional LCR minimum: daily maximum intraday liquidity usage, opening available intraday liquidity, total payments and time-specific obligations for reporting banks; customer payment value and customer intraday credit lines for correspondent providers; and intraday throughput for direct participants. Supervisors determine the relevant application. The 2013 intraday framework explicitly states that intraday liquidity is outside LCR calibration.

Track cumulative settled receipts minus settled payments from the start of the day, along with opening usable balances and separately identified credit capacity. Do not add future receipts as current cash. At each critical time, compute the funding required to meet due obligations and the time needed to obtain it. Intraday credit is capacity, not cash already received; its collateral and repayment requirements remain visible. Borrowing that cannot be repaid by its required deadline can become an overnight funding problem.

9. Wholesale funding strategy and the funding ladder

Customer deposits are usually the foundation. Wholesale funding fills gaps, supports growth, and can diversify: or, if misused, can become the fuse.

A funding ladder is the schedule of wholesale maturities over time. A healthy ladder avoids large cliffs where too much must be refinanced at once.

Principles of a resilient wholesale strategy: diversify instruments, tenors, and investor types; avoid excessive reliance on short-term confidence-sensitive money; pre-fund when markets are open; maintain investor relationships in peacetime; align wholesale needs with asset growth and liquidity policy; know which funding sources disappear first in stress.

Short-term wholesale funding is often cheaper in calm markets and less reliable under stress. Long-term funding costs more and stabilises the structure. Treasury’s job is not to minimise cost in isolation. It is to buy an appropriate amount of resilience.

10. Contingency Funding Plan: from document to drill

A Contingency Funding Plan is the bank’s playbook for stressed liquidity conditions. A serious CFP contains clear stress stages and triggers, roles and decision rights, ordered contingent actions, communication protocols, operational checklists, and honest assumptions.

A plan that has never been tested is a story. Testing reveals whether data can be produced fast enough, whether legal and operational steps actually work, whether decision-makers understand their roles, and whether market-access assumptions are realistic.

In a healthy culture, calling the CFP early is professional. In a weak culture, people delay escalation to avoid causing alarm. Liquidity crises punish delay.

11. Liquidity stress testing that boards can trust

Regulatory ratios are standardised stress lenses. Internal stress testing asks: what would hurt us specifically?

Useful tests are specific to the business model, severe enough to matter, clear about assumptions, linked to management actions and CFP options, and reported with residual risk after actions.

Boards should see which assumptions drive survival, how fast buffers burn, which actions are realistic in the first 48 hours and first 30 days, and where the bank would still be fragile after management actions.

12. Funds transfer pricing and the internal price of liquidity

If business units can raise or use funds without facing a realistic internal price for liquidity and stability, the bank will grow the wrong balance sheet. FTP is the nervous system that connects product decisions to treasury reality.

A clean FTP framework credits deposit-gathering businesses for funding value, charges uses of funds for the funding they consume, differentiates stable from flighty funding where policy requires, concentrates risk in treasury, and makes product profitability more honest.

Weak FTP leads to products that look profitable while consuming scarce stable funding, business incentives that fight resilience, and growth that later must be funded in expensive or unstable ways.

13. Organisation of a bank treasury liquidity function

A treasury liquidity function can look different from bank to bank, but the work normally separates into a few stable responsibilities. The names may differ. One bank may call a team Liquidity Management. Another may call it Cash and Funding. Another may split treasury dealing, liquidity risk, ALM and regulatory reporting into separate reporting lines. The labels matter less than the ownership. The bank must know who owns daily cash, who owns structural funding, who owns buffer assets, who owns regulatory liquidity reporting, who owns assumptions, who challenges Treasury independently, and who escalates when the position moves outside appetite.

The daily cash or liquidity desk is closest to the operating day. It watches opening balances, expected inflows, expected outflows, money-market maturities, repo flows, large client payments, clearing obligations and central-bank balances. This team answers the immediate question: do we have enough cash in the right places today, and what do we do with the surplus or deficit? It is a practical desk. It needs numbers, cut-offs, market contacts, settlement awareness and calm communication. A good cash desk does not wait for perfect certainty. It works with the best available view, updates continuously and escalates early when assumptions start breaking.

The funding desk usually manages wholesale issuance and investor-facing funding activity. It may issue certificates of deposit, commercial paper, covered bonds, senior unsecured debt, subordinated debt or other funding instruments depending on the bank and jurisdiction. Its job is not only to raise money cheaply. It must raise the right type of funding for the balance sheet: right tenor, right currency, right investor base, right legal entity, right timing, right stability. Cheap overnight funding can be useful tactically but dangerous structurally if it becomes the default answer for long-term assets.

The liquidity buffer or investment portfolio team manages securities held for liquidity and, within policy, yield. It chooses assets that can support LCR, central-bank eligibility, repo capacity, collateral needs and internal stress survival. This team must understand credit quality, duration, market liquidity, custody, encumbrance, haircut behaviour and regulatory eligibility. The mistake is to treat the buffer like an ordinary investment portfolio. A liquidity buffer is insurance. It can earn income, but income is not its first job.

The ALM team looks at structural interest-rate risk, balance-sheet maturity mismatch, behavioural assumptions, FTP and ALCO decision support. ALM is not always inside the liquidity desk, but it is inseparable from liquidity thinking. Deposit behaviour, loan maturity, security duration, hedging and FTP all affect how stable or fragile the bank becomes. ALM provides the medium-term and long-term view that the cash desk cannot see if it focuses only on today.

Regulatory liquidity reporting teams translate source-system data into LCR, NSFR and other required reports. In some banks this sits under Finance. In others it sits under Risk or Treasury. The key is independence of evidence and clarity of ownership. Regulatory reports are not management decoration. They are formal statements to supervisors and senior governance. They need lineage, controls, reconciliations, documented assumptions and review.

Independent Liquidity Risk should challenge Treasury. This is a critical distinction. Treasury manages liquidity. Liquidity Risk measures, challenges and monitors whether management is inside appetite. It tests assumptions, designs stress scenarios, reviews contingency funding plans, escalates breaches and reports to risk committees. If Treasury both creates the position and marks its own homework without challenge, the bank has a weak control environment.

Good organisation has one more feature: named decision rights. Who can decide to use the liquidity buffer? Who can approve funding outside the plan? Who can trigger the Contingency Funding Plan? Who communicates with supervisors? Who approves changes to behavioural assumptions? Who can override a liquidity classification? In calm conditions, unclear decision rights look harmless. In stress, unclear decision rights waste hours the bank may not have.

14. Governance and risk appetite in liquidity management

Liquidity governance turns concern into control. Without governance, everyone agrees liquidity is important, but nobody knows exactly what the bank will or will not tolerate. A strong governance framework defines appetite, limits, triggers, escalation, reporting, independent challenge and decision authority. It gives Treasury room to manage the bank while making sure the defensive mandate is not diluted by short-term commercial pressure.

Risk appetite starts at the board and senior management level. The board does not need to approve every money-market placement. It does need to approve the broad level of liquidity risk the bank is willing to run. This includes minimum liquidity buffer expectations, minimum regulatory ratio headroom, concentration limits, reliance on short-term wholesale funding, intraday liquidity tolerance, survival horizon expectations, contingency planning standards and escalation triggers. Appetite should be written in a way that can be translated into measurable limits.

Limits are the daily language of appetite. A liquidity framework may include a minimum LCR, minimum NSFR, minimum internal stress survival horizon, maximum short-term wholesale funding reliance, maximum single depositor concentration, maximum unsecured funding by counterparty, minimum unencumbered HQLA by currency, maximum maturity cliff by time bucket, maximum intraday overdraft or peak usage, and required CFP test frequency. These limits protect the bank from becoming fragile slowly while reports still look acceptable.

Triggers are reached before hard limits as risk worsens. For a minimum liquidity ratio, an early-warning threshold normally sits above the minimum; for a maximum concentration or usage limit, the warning sits below the maximum. A trigger warns management before a formal breach. For example, the bank may have a regulatory minimum, an internal board limit and an early-warning trigger above both. If the trigger is crossed, Treasury and Risk may need to explain drivers, propose actions and increase monitoring. Triggers are valuable because liquidity crises rarely announce themselves politely. They start as signals: deposit concentration increases, wholesale spreads widen, rollover becomes harder, market prices gap, media sentiment changes, collateral haircuts rise, or large clients begin asking questions.

Governance must also separate normal management actions from contingency actions. Under business-as-usual conditions, Treasury may place cash, borrow overnight, issue funding, buy securities or execute hedges inside policy. Under stress, some actions require committee escalation or CFP activation. The framework should define when the operating mode changes. A bank should not argue in the middle of stress about whether stress exists.

ALCO is usually the core senior forum for structural liquidity governance. It reviews liquidity position, funding plan, LCR and NSFR drivers, buffer composition, concentration risk, wholesale market access, stress results, FTP changes, funding issuance plans and contingency readiness. A good ALCO does not only receive information. It makes decisions. It asks what changed, what is fragile, what management action is required and who owns it.

Risk committees add independent oversight. They should ask whether Treasury is operating inside appetite, whether assumptions remain valid, whether stress tests are severe enough, whether contingency options are real, and whether breaches are handled with urgency. Finance ensures numbers reconcile to official records. Internal Audit later tests whether the process works as documented. Regulators may review both the numbers and the governance behaviour behind them.

The human culture matters. In strong liquidity cultures, early escalation is respected. Nobody is punished for bringing uncomfortable liquidity news early. In weak cultures, people hide behind ratios, wait for perfect evidence, or delay action because they fear causing alarm. Liquidity punishes pride and delay. Governance exists so the bank can act before the problem becomes public.

15. Deposit behaviour and concentration risk

Deposits are the foundation of many banks, but they are not all equal. A balance is not stable simply because it is called a deposit. Treasury must understand who owns the deposit, why the money is there, how rate sensitive it is, whether it is operational, how quickly it can leave, whether it is concentrated, and how it behaved in past stress. The same headline deposit growth can be excellent or dangerous depending on composition.

Retail transactional deposits often behave differently from rate-chasing savings balances. A salary account used for daily banking may have a stable core because customers use it for payments, cards, salary, bills and convenience. A high-rate digital savings account may be more sensitive to competitor offers. SME operating deposits may be tied to business cycles, payroll, supplier payments and tax dates. Large corporate deposits can move in large blocks because corporate treasurers actively manage cash. Institutional and financial-sector deposits can be especially confidence sensitive.

Operational deposits deserve careful treatment. These are balances linked to real operational relationships such as cash management, payments, clearing, custody or transaction services. They may be more stable than pure investment deposits because the customer keeps them to run daily business. But even operational deposits are not permanent. If the customer loses confidence, changes provider or centralises cash elsewhere, balances can fall quickly. Treasury should never treat labels as guarantees.

Concentration risk is one of the most dangerous liquidity weaknesses. A bank may have a large total deposit base but depend heavily on a small number of corporate, public sector, institutional or financial customers. If the top ten depositors can materially change the liquidity position, Treasury needs dedicated monitoring. The question is not only whether they left before. The question is what would happen if they left under the same conditions when market funding is also difficult.

Deposit behaviour changes when rates change. In a low-rate environment, customers may not care much about small rate differences. In a high-rate environment, they may actively move money. Digital channels reduce friction. Social media and fast news can accelerate confidence movement. A historical stability model built in one rate regime may be weak in another. ALM and Treasury must therefore review assumptions, not merely inherit them.

Business teams often see deposit growth as success. Treasury sees the quality of that growth. A large short-term corporate deposit arriving just before reporting date may help a ratio temporarily but may not provide stable funding. A retail deposit campaign may bring balances quickly but at a high cost and uncertain persistence. A relationship deposit tied to operating services may be more valuable than a rate-sensitive balance of the same size. FTP should transmit those differences.

Good deposit analytics segment by customer type, product, rate sensitivity, operational relationship, balance size, tenure, channel, legal entity, currency, geography and historical behaviour. It should identify single-name concentrations, correlated segments and early-warning signals. A sudden increase in withdrawals from one sector may matter more than a small aggregate movement. A shift from stable operating balances to rate-sensitive term balances may change funding quality.

From a requirements perspective, deposit stability cannot be a vague field. Who defines stable? Which source system provides product classification? How are operational relationships identified? How often is behaviour reviewed? Where are overrides approved? How are top depositor movements monitored? Are deposit flows visible intraday or only at end of day? Which balances are insured and which are not? Which balances are from financial counterparties? These questions make liquidity reporting real.

The professional lesson is simple: deposits are not money in a jar. They are customer behaviour sitting on the liability side of the balance sheet. Treasury's job is to understand that behaviour before the customers change it.

16. Undrawn facilities and contingent liquidity risk

A committed credit facility can be quiet for years and then become very loud in stress. This is why undrawn facilities are part of liquidity management. A borrower may have the legal right to draw funds when certain conditions are met. From the borrower's perspective, the facility is insurance. From the bank's perspective, it is a contingent cash outflow. The bank does not hold the cash as an asset today, but it may need to provide cash tomorrow.

Corporate revolving credit facilities are the classic example. In normal times, a corporate may leave the facility mostly undrawn. In stress, companies may draw facilities to secure liquidity while they still can. If many customers do this together, the bank experiences correlated outflows. The legal commitment becomes a cash event. During market-wide stress, this can happen exactly when wholesale funding is harder and deposit behaviour is more nervous.

Retail and SME facilities can also matter. Overdrafts, credit cards, committed working-capital lines and trade finance commitments may be used differently under stress. The drawdown pattern may depend on customer confidence, sector stress, seasonality, product design and macro conditions. A bank with large committed but undrawn exposure must understand behavioural drawdown risk.

Liquidity reporting frameworks apply assumptions to undrawn commitments, but internal management should not rely blindly on generic factors. A facility to an investment-grade corporate with strong liquidity may behave differently from a facility to a stressed sector. A committed facility supporting essential operations may draw faster than a relationship line with low usage history. A facility with material adverse change clauses may still create franchise pressure if the bank refuses funding at the wrong time.

Pricing should reflect contingent liquidity. A facility commitment fee that does not cover the cost of liquidity capacity can encourage business growth that looks profitable locally while consuming Treasury resilience centrally. FTP and product pricing should recognise that unused commitments are not free. They require funding planning, liquidity buffer support and stress capacity.

Systems often struggle with contingent exposures. Loan outstanding balances are easy to see. Undrawn commitments may sit in different systems, with different product codes, limits, expiry dates, conditions and currencies. Treasury needs the committed amount, drawn amount, undrawn amount, maturity, borrower, sector, legal entity, currency, product type, collateral status, cancellation rights and historical utilization behaviour. Without this data, stress outflow estimates become weak.

Testing should include a facility drawdown event. When a corporate draws a large committed line, does Treasury forecast update? Does liquidity reporting capture cash outflow? Does credit exposure update? Does payment processing reflect value date? Does FTP adjust? Does ALM cash-flow view change? Does concentration monitoring identify the customer? A drawdown is not only a lending event. It is a liquidity event.

The key mindset is this: undrawn does not mean irrelevant. In liquidity stress, some of the most important cash outflows come from promises made earlier when everyone was calm.

17. Repo, collateral and liquidity transformation

Repo is one of the most important practical tools in treasury liquidity management. In simple terms, a repo allows a bank to raise cash against securities by selling securities and agreeing to repurchase them later. Economically, it behaves like secured borrowing. The cash provider receives collateral. The cash borrower receives funding. The collateral, haircut, tenor, counterparty and legal agreement define the practical risk.

Repo turns securities into cash, but not magically. The bank must own or control eligible securities, the counterparty must be willing to transact, the haircut must be acceptable, settlement must work, legal agreements must be in place, collateral must be transferable, and the repo market must remain open. In stress, any of these conditions can weaken. A security that financed easily yesterday may require a bigger haircut tomorrow. A counterparty may reduce lines. A settlement failure may block cash. A legal or operational issue may stop collateral movement.

Reverse repo is the opposite side: the bank places cash against collateral. It can be a way to invest surplus cash securely, manage collateral, or support market activity. From a liquidity perspective, reverse repo may be safer than unsecured placement because collateral protects exposure. But the bank must still understand collateral quality, liquidity, legal enforceability and operational settlement.

Haircuts are central. If a bank repos a bond with a 2 percent haircut, it receives cash slightly below the market value of the collateral. If haircuts widen under stress, the same securities generate less cash. This is why Treasury should not assume nominal securities value equals repo capacity. The practical question is: how much cash can be raised after realistic stress haircuts, in the right currency, within the required time?

Collateral encumbrance matters. A security used in repo is encumbered. It cannot simultaneously support another liquidity action unless released. A bank may own a large securities portfolio but have a smaller unencumbered amount. Liquidity reports must distinguish owned assets, pledged assets, available collateral, operationally movable collateral and assets trapped by legal entity or custodian constraints. A top-level holding number can mislead management badly.

Central clearing, bilateral repo and central-bank repo-like facilities may have different operational and legal frameworks. A Treasury team must know which channel it can use under which condition. Peacetime relationships matter. A bank that never uses a repo counterparty and then asks for large stress funding may find the market less open than its plan assumed.

Repo also connects Treasury to Markets. Trading desks use repo to finance inventories and cover shorts. Treasury uses repo for liquidity. Collateral teams manage eligibility and substitution. Risk monitors counterparty exposure. Operations settles securities and cash. Finance accounts for transactions. A repo break can therefore affect several teams at once.

A strong repo liquidity playbook includes collateral inventory, eligibility, haircuts, counterparties, legal agreements, settlement cut-offs, operational contacts, test transactions, central-bank facility mapping and fallback actions. It should be tested. If the bank discovers during stress that collateral cannot be mobilized because of an account setup issue, the liquidity buffer was overstated.

The practical memory line: repo is powerful because it converts securities into cash, but it is only as strong as collateral quality, market access, legal agreement, haircut behaviour and operational readiness.

18. Central-bank facilities and operational readiness

Central banks provide monetary-policy operations, standing facilities, intraday credit and, in some frameworks, exceptional liquidity assistance. Their purposes and access conditions differ. Routine eligible participation can be intended and appropriate; exceptional support has separate legal, supervisory and operational conditions. Central-bank access can be essential in system stress, idiosyncratic stress or market disruption. It can provide secured liquidity against eligible collateral when private funding is difficult. But it comes with eligibility rules, operational requirements, potential stigma, supervisory visibility and governance expectations.

The first practical requirement is knowing what facilities exist and which legal entities can access them. A multinational bank may not have equal access across all entities. Collateral eligibility may differ by jurisdiction and central bank. A security eligible in one facility may not be eligible in another. Currency matters. Legal entity matters. Operational accounts matter. A group-level liquidity view can hide entity-level constraints.

The second requirement is collateral readiness. Eligible collateral must be identified, valued, unencumbered where required, documented and operationally movable. The bank must know how to pledge or transfer collateral, how long it takes, which systems are used, who approves the action and what cut-offs apply. A facility that exists legally but has never been operationally tested may not be reliable in a fast event.

The third requirement is governance. When can Treasury use central-bank facilities? Is use part of normal monetary-policy operations in that jurisdiction, or does it signal stress? Does ALCO need approval? Does senior management need to be informed? Does Risk need to challenge the action? Are supervisors notified? How is use disclosed or reported? These questions should be answered before stress, not during stress.

Stigma is a real behavioural factor in some markets. Even when facilities are designed to provide liquidity support, banks may worry that use will be interpreted as weakness. In a system-wide crisis, stigma may be lower because many banks use facilities. In an idiosyncratic event, the signal can be more sensitive. Treasury must understand the local context and communication approach.

Central-bank capacity should be included in contingency planning realistically. It should not be overstated. If collateral mobilisation takes a day, it cannot solve a payment due in one hour. If collateral is already encumbered, it cannot be counted twice. If the facility supports one currency, it may not solve another currency need. If access requires operational approval not yet tested, capacity should be discounted in internal stress thinking.

For learners, the balanced view is important. Central-bank facilities are not shameful by nature; they are part of the financial system's liquidity architecture. But they are not a substitute for strong liquidity management. A well-run bank knows how to use them, tests access, maintains eligible collateral and distinguishes routine operations, intraday capacity and contingent support within its funding and liquidity plan. The Federal Reserve standing repurchase agreement operations, for example, are intended for eligible counterparties when economically sensible, supporting overnight rate control and market functioning. That example does not give every bank access or set another central bank's policy.

19. Legal entity, currency and trapped liquidity

One of the biggest beginner mistakes is thinking cash is always freely movable inside a banking group. It is not. Legal entities, regulatory requirements, local liquidity rules, tax, capital restrictions, resolution planning, ring-fencing and operational constraints can all limit how liquidity moves. A group may look liquid in aggregate while one entity or currency is tight.

Legal entity liquidity matters because obligations are legal-entity specific. If Bank A's subsidiary in one country owes payments today, cash held in another subsidiary may not be immediately usable. Even if the group owns both entities, moving cash may require legal, regulatory, tax or internal approvals. In stress, local regulators may expect liquidity to remain in the local entity. Resolution frameworks can strengthen the need for entity-level self-sufficiency.

Currency matters because obligations settle in specific currencies. A bank may have a surplus in EUR and a deficit in USD. It can convert or swap currencies, but that is a market transaction with price, settlement, counterparty and cut-off risk. FX markets may be liquid in normal times and stressed at exactly the wrong moment. A currency-level liquidity report is therefore essential for material currencies.

Trapped liquidity appears when cash or assets exist but cannot be used where needed. Causes include legal entity restrictions, local regulatory buffers, collateral already pledged, securities held in an account that cannot be repoed quickly, operational cut-offs, currency controls, tax constraints or internal policy. Trapped liquidity is frustrating because it looks like liquidity until the moment the bank tries to use it.

Resolution and recovery planning make this more important. Supervisors may expect banks to show how liquidity would be available to material entities in stress and how support arrangements work. Internal service agreements, parental guarantees, committed lines and liquidity transfer mechanisms must be legally and operationally credible. A simple group treasury assumption that cash will move where needed may not be enough.

Treasury reporting should therefore show liquidity by group, legal entity, branch where relevant, currency and material product. It should identify trapped balances, transfer restrictions and operational constraints. ALCO and risk committees should not rely only on group averages. A group average can hide a local fire.

Requirements should ask whether the system can report by legal entity and currency. Can it distinguish freely transferable cash from restricted balances? Can it apply local regulatory minima? Can it identify encumbered assets? Can it model FX swap availability? Can it show where support lines exist and where they are only assumptions? These are not exotic questions. They are practical survival questions for multi-entity banks.

20. Survival horizon as a management tool

Survival horizon is an internal management metric that asks: under a defined stress scenario, how many days can the bank continue meeting obligations before available counterbalancing capacity is exhausted or management actions fail? It is not a replacement for LCR or NSFR. It is a practical way to express time under stress in language senior management can understand.

A survival horizon framework begins with a stress scenario. The scenario may include deposit outflows, wholesale funding non-renewal, committed facility drawdowns, collateral calls, market-value haircuts, operational constraints and limited inflows. The bank then projects cash inflows and outflows over days or weeks and applies available liquidity sources. The result is not only a number of days. It is a burn-down story.

The value of survival horizon is that it exposes timing. A bank may have enough liquidity over 30 days in aggregate but face a severe gap in the first three days. Another bank may survive 30 days only if management actions are assumed unrealistically fast. A survival horizon forces the question: when does pain arrive, and what can we actually do before then?

Management actions should be separated into certain, likely and uncertain. Cash already at the central bank is more certain than a planned bond issuance. Unencumbered government securities with tested repo access are more certain than assets that may need new legal setup. Central-bank access that has been operationally tested is more certain than theoretical eligibility. A good survival horizon discounts action quality.

Survival horizon also helps communication. Saying the bank has a 132 percent LCR may satisfy a ratio discussion. Saying the bank can survive this internal severe scenario for 42 days before exhausting defined counterbalancing capacity gives management a more intuitive view. Both can be useful. The second often supports better action planning.

But survival horizon can be misused. If assumptions are optimistic, the number becomes false comfort. If management actions are counted without capacity or timing evidence, the horizon is inflated. If legal entity and currency constraints are ignored, group-level survival may hide local weakness. Independent Liquidity Risk should challenge the assumptions and report sensitivity.

A strong ALCO does not ask only what the survival horizon is. It asks what drives it, what shortens it, what would extend it, which assumptions are least certain, what actions are needed now, and whether the CFP remains credible.

21. Assumption risk in liquidity management

Liquidity management depends on assumptions because the future is not fully known. Deposit run-off, facility drawdown, asset haircuts, market access, collateral eligibility, settlement timing, central-bank capacity and management action speed all require assumptions. Assumption risk is the risk that those assumptions are wrong.

Deposit assumptions are especially sensitive. A model may say a certain retail deposit segment is stable because historical balances stayed within a narrow band. But the history may not include a modern digital-rate war, a social-media confidence event, a sharp rate hiking cycle or a bank-specific reputational shock. A corporate deposit model may understate how quickly treasury departments move cash when they perceive risk. Assumptions must be reviewed against current conditions, not only history.

Facility drawdown assumptions can also fail. Corporate customers may draw committed lines together when market liquidity is tight. This creates wrong-way liquidity pressure: the bank must provide cash just when it is harder to raise cash. A low historical utilization rate does not guarantee low stress utilization. Treasury should run severe but plausible drawdown scenarios.

Haircut assumptions matter for asset monetization. A bond that can be repoed with a small haircut in calm markets may need a larger haircut in stress. Less liquid securities may become difficult to finance. If a stress test assumes sale or repo proceeds based on normal market conditions, it overstates liquidity. HQLA should be tested for market-value sensitivity, haircut widening and settlement capacity.

Operational timing assumptions are often too optimistic. A plan may assume securities can be sold same day, central-bank collateral can be pledged quickly, or funding can be raised in a certain window. But approvals, cut-offs, documentation, system access, custodian processes and people availability can slow action. A tabletop exercise should test timing, not only strategy.

Assumption governance should include ownership, evidence, review frequency, sensitivity analysis, approval and documentation. If Treasury proposes a behavioural maturity for deposits, who approves it? If Liquidity Risk challenges it, how is disagreement resolved? If market conditions change, what triggers review? If an assumption is overridden manually, where is the evidence?

The professional approach is humility. Liquidity assumptions are necessary, but they are not truth. A strong bank knows which assumptions are fragile and watches them closely. A weak bank treats assumptions as facts because the spreadsheet looks neat.

22. Liquidity stress testing in full practical language

Liquidity stress testing asks how the bank behaves when normal assumptions break. It is one of the most important tools in Treasury because liquidity crises are nonlinear. Outflows can accelerate. Funding sources can close. Asset haircuts can widen. Confidence can change. Stress testing forces the bank to rehearse those combined effects before they arrive.

A useful stress test starts with a scenario narrative. For example: a bank-specific rating downgrade combined with negative media, partial deposit outflows, wholesale market caution, increased collateral calls and slower inflows. Or: a market-wide stress where repo haircuts rise, unsecured funding becomes expensive, clients draw facilities and securities values fall. Or: a currency-specific stress where USD funding tightens while the bank has large USD payment obligations.

The scenario must then translate into cash-flow assumptions. Which deposits leave? How fast? Which wholesale funding rolls and which does not? Which facilities are drawn? Which collateral calls occur? Which assets can be sold or repoed? What haircuts apply? Which inflows are capped or delayed? Which currencies are affected? Which legal entities can transfer liquidity? Without this translation, the scenario is only a story.

Counterbalancing capacity is then applied. This includes central-bank reserves, unencumbered HQLA, repo capacity, maturing assets, committed liquidity lines where credible, central-bank facilities, asset sales and other management actions. Each action should have capacity, timing, haircut, operational confidence and governance status. A plan that says raise funding from market should specify instrument, investor appetite, timing and stress realism.

The output should show time. Day 1, day 2, day 7, day 30 and beyond can look very different. The first few days often matter most because they reveal operational liquidity pressure. Later days reveal structural funding endurance. A single end-state number hides the burn path.

Stress testing should produce decisions. If a scenario shows survival is too short, management may increase HQLA, reduce short wholesale reliance, adjust deposit strategy, change FTP, pre-fund, reduce asset growth, improve collateral readiness, or revise the CFP. A stress test that produces no action despite weakness is a performance, not risk management.

Back-testing and experience review are important. When real stress events occur, even mild ones, compare actual behaviour with assumptions. Did deposits behave as expected? Did wholesale funding roll? Were haircuts stable? Did payment flows change? Were dashboards timely? This feedback improves future tests.

For BA and testing work, a stress engine should support scenario versioning, assumption ownership, source data traceability, currency and legal entity views, management action modelling, output comparison, audit trail and approval workflow. Stress testing is not just analytics. It is governance technology.

23. Contingency Funding Plan design in real detail

A Contingency Funding Plan is not a binder. It is a usable operating manual for liquidity stress. It should be short enough to act on, detailed enough to guide decisions, and rehearsed enough that people do not read it for the first time during a crisis. The best CFPs are living playbooks, not annual compliance artifacts.

A CFP begins with stress levels or stages. Stage 0 may be normal monitoring. Stage 1 may be heightened awareness: early warning indicators triggered, increased reporting, management notified. Stage 2 may be moderate stress: funding spreads widen, deposit outflows exceed threshold, wholesale rollover weakens, liquidity buffer falls near internal trigger. Stage 3 may be severe stress: significant outflows, market access restricted, emergency actions considered. Each stage should have criteria, owners and actions.

Early warning indicators are crucial. They may include deposit outflows, concentration movements, funding spread widening, failed funding rollovers, rating outlook changes, credit default swap spread movement, equity price movement, negative media, social sentiment, collateral haircut changes, payment delays, increased facility drawdowns and internal ratio deterioration. No single indicator tells the whole story. The pattern matters.

Action menus should be ordered and realistic. Some actions are fast and reliable: use existing central-bank balances, stop discretionary placements, allow maturing assets to roll off into cash, repo ready collateral with existing counterparties. Some actions take longer: issue wholesale funding, sell less liquid assets, restructure balance sheet, change product pricing. Some actions have signalling or franchise cost: use emergency facilities, restrict business growth, sell securities at loss. A CFP should not pretend all actions are equal.

Communication protocols matter. Who speaks to supervisors? Who speaks to rating agencies? Who speaks to investors? Who speaks to major clients? Who speaks internally? What messages are allowed? How are facts verified before communication? In a liquidity event, inconsistent communication can worsen confidence. Silence can also worsen confidence. The CFP should define roles and approval.

Payment prioritisation may be needed under severe stress or operational liquidity pressure. The bank should understand which payments are critical, which are legally required, which are customer-impacting, which are market-infrastructure obligations and which can be delayed within policy. This is sensitive and must align with legal and regulatory expectations. It cannot be invented casually during stress.

Testing the CFP should include tabletop exercises and operational drills. A tabletop tests decision-making and communication. An operational drill tests whether collateral can be mobilized, reports produced, approvals obtained and contacts reached. The test should record what failed. A CFP test where everything always works perfectly is suspicious. Real testing finds gaps.

The final CFP question is: if stress begins at 8:30 tomorrow morning, do people know what to do by 8:45? If the answer is no, the plan is not ready.

24. Intraday liquidity management and payment timing

Intraday liquidity is where Treasury meets payment operations most directly. A bank may have enough end-of-day liquidity but still struggle during the day if payment obligations fall due before receipts arrive. High-value payment systems, securities settlement, CLS, clearing systems, margin calls and instant payments can all create timing pressure.

The first intraday question is opening capacity. How much cash or central-bank balance is available at the start of the day in each critical currency and system? Are there expected early outgoing payments? Are there expected incoming receipts? Are any large settlements dependent on counterparties? Are there cut-offs that cannot be missed? The desk needs a live view, not a static yesterday report.

Payment sequencing matters. Some payment systems allow queue management or prioritisation. Others settle instantly. A large outgoing payment early in the day can consume liquidity that was expected to be replenished later. If the incoming payment is delayed, the bank may need to borrow intraday, use collateral, delay lower-priority flows where allowed, or escalate. Poor sequencing can create unnecessary stress.

Securities settlement adds complexity. Delivery versus Payment links cash and securities. A securities purchase may require cash at settlement time. A securities sale may generate cash only when settlement completes. A failed sale means expected cash does not arrive. A failed purchase means expected securities or collateral do not arrive. Treasury should receive timely information on material settlement fails because they affect cash and buffer availability.

FX settlement can create currency-specific intraday needs. A bank may have to pay one currency before receiving another, depending on settlement mechanism. Payment-versus-payment arrangements reduce risk where available, but not all currencies or trades are covered. Time zones matter. Cut-offs matter. Nostro funding matters.

Instant payment systems change the rhythm. Retail payments that once settled in batches may now move continuously. Corporate APIs may trigger faster flows. Fraud holds, sanctions holds and operational repair can delay outgoing or incoming flows. Treasury may not need to watch every small payment individually, but it needs aggregated, timely visibility and exception alerts for large or unusual patterns.

Intraday liquidity metrics may include peak usage, largest net debit position, time spent below threshold, delayed payments, gridlock events, incoming/outgoing timing mismatch, payment value by hour, and usage of intraday credit. These metrics help the bank understand whether stress appears inside the day before it appears in end-of-day reports.

For requirements, intraday dashboards should answer: what is the position now, what is expected next, what changed unexpectedly, which critical payments are pending, which receipts are delayed, what are the cut-offs, what buffers are available and who needs to act? A dashboard that only shows historical totals is not an intraday control.

25. Funding strategy, investor confidence and market access

Funding strategy is not only a spreadsheet of maturities. It is the bank's plan for maintaining market confidence and stable access to money. Investors lend to banks because they believe the bank is creditworthy, liquid, well governed and transparent enough. That belief can change. Treasury must manage funding before it is needed.

A resilient strategy diversifies by instrument. Deposits, interbank borrowing, certificates of deposit, commercial paper, repo, covered bonds, senior unsecured debt, subordinated debt and capital instruments each serve different purposes. Repo can provide secured short-term funding. Covered bonds can provide relatively stable funding backed by eligible assets. Senior unsecured debt supports broader term funding. Capital instruments support regulatory capital, not ordinary liquidity. The mix should match the bank's balance sheet and risk appetite.

Tenor diversification prevents maturity cliffs. A bank should avoid too much funding maturing in one narrow window. Maturity walls create refinancing risk. Even if market access is good today, the bank cannot know future conditions. Smoothing maturities gives management options. Pre-funding can be sensible when markets are open and spreads are acceptable.

Currency diversification must be real, not cosmetic. If the bank lends in USD, funds in EUR and relies on FX swaps to transform currency, it has basis and rollover risk. FX swap markets are deep but can stress. A currency funding strategy should consider natural currency deposits, wholesale issuance, swaps, central-bank facilities and liquidity buffers by currency. Material currencies need dedicated management.

Investor diversification matters. A funding base concentrated in one investor type can weaken quickly if that segment changes appetite. Money-market funds, asset managers, banks, insurers, pension funds, central banks, retail depositors and corporate treasurers behave differently. Treasury and Investor Relations should maintain relationships in peacetime. Stress is a poor time to introduce yourself.

Market access indicators should be monitored. Funding spreads, failed issuance, investor feedback, secondary-market bond spreads, CDS levels, rating outlooks, deposit pricing pressure, repo haircuts and peer movements all provide signals. Treasury should not wait until a funding transaction fails. It should watch the market's willingness to fund the bank.

Funding decisions affect product strategy. If stable funding becomes expensive, long-term lending may need repricing. If short-term wholesale reliance is near limit, balance-sheet growth may need moderation. If covered bond capacity is constrained, mortgage growth strategy may change. Treasury should be integrated into business planning, not consulted after growth has already happened.

A strong funding plan includes base-case issuance, contingency windows, currencies, instruments, investor targets, maturity profile, internal triggers, alternative actions and communication plan. It should be reviewed at ALCO. It should be linked to asset growth, deposit strategy, liquidity buffer and regulatory ratios. Funding is not the last step of banking. It is part of product economics from the beginning.

26. Liquidity buffer duration and the yield temptation

A liquidity buffer must be safe, usable and available. But because banks are commercial institutions, there is always temptation to earn more yield from the buffer. Longer-duration securities may yield more than very short cash-like instruments. Lower-quality or less liquid assets may offer spread. In calm markets, stretching for yield can look sensible. In stress, it can become dangerous.

Duration is the key. A long fixed-rate government bond may have low default risk but high interest-rate sensitivity. If rates rise, its market value falls. If the bank can hold the bond and does not need cash, the loss may remain unrealised depending on accounting classification. But a liquidity buffer exists because the bank may need cash under stress. If the bank must sell the bond after rates rise, the loss becomes realised. The asset was credit-safe but not liquidity-safe in the forced-sale scenario.

This is why buffer design should consider both regulatory eligibility and market-value sensitivity. A security may qualify as HQLA but still create mark-to-market volatility. Treasury should monitor weighted-average duration, price sensitivity, unrealised loss, repo haircut, sale liquidity, currency, issuer concentration and maturity ladder of the buffer. A big HQLA number is not enough if the buffer is fragile under rate shock.

The yield temptation is subtle because it can be rewarded for years. The bank earns extra carry. Management sees income. No stress occurs. The strategy looks smart. Then rates rise or deposits leave, and the bank discovers the buffer cannot be monetized without painful loss. This is exactly why risk governance must challenge buffer strategy during calm periods, not only after losses appear.

A strong buffer has layers. Some cash or central-bank reserves for immediate needs. Very liquid short securities for near stress. High-quality securities that can be repoed or sold with predictable haircuts. Possibly other eligible assets with clear limitations. The composition should reflect the bank's business model, currencies, legal entities, stress horizon and appetite. It should not be built only by chasing yield inside regulatory eligibility.

Product Control and Finance should understand buffer valuation. ALM should understand duration. Liquidity Risk should understand monetization. Treasury should understand market executability. ALCO should understand the trade-off between income and survival. This multi-function review prevents a liquidity buffer from quietly becoming a hidden trading position.

The clean phrase is: a liquidity buffer is allowed to earn yield, but it is not allowed to forget why it exists.

27. HQLA operations: encumbrance, custody and mobilization

High-quality liquid assets are only useful if they can be mobilized. Mobilization means turning the asset into cash through sale, repo, central-bank pledge or other approved mechanism. A bank may own eligible securities, but if they are encumbered, in the wrong account, missing from collateral systems or operationally blocked, they may not help when needed.

Encumbrance means the asset is already pledged or otherwise unavailable. A security used to secure derivatives, repo or central-bank borrowing is not freely available to support a second action. A release path must be timed and evidenced. Pre-positioned central-bank collateral unused to generate liquidity can have different HQLA treatment under the applicable rules, as explained in section 5.5. Liquidity reporting must distinguish total holdings from unencumbered holdings. Double counting collateral is a serious error.

Custody location matters. Securities are held through custodians, central securities depositories or internal structures. To repo or sell quickly, settlement accounts, counterparty links, legal agreements and operational processes must be ready. If the bank's HQLA sits in a custodian account not linked to a repo counterparty, the timeline to raise cash may be longer than assumed.

Collateral eligibility data must be accurate. Eligibility depends on issuer, instrument type, rating, currency, maturity, jurisdiction, central-bank rules, haircut category and sometimes internal policy. These attributes can change. A rating downgrade, corporate action, maturity change or regulatory update can alter eligibility. Static data governance is therefore part of liquidity management.

Mobilization playbooks should list which assets can be used first, through which channel, with what haircut, in which currency, and by what time. The playbook should be tested. A small test repo or central-bank collateral drill can reveal operational gaps early. Waiting until stress to discover that a signatory list is outdated is poor discipline.

HQLA reporting should show operationally available value after haircuts by legal entity and currency. It should also show assets unavailable because of encumbrance, settlement pending, legal restriction, operational issue or data uncertainty. Management needs the real usable number, not a comforting gross number.

A BA working on HQLA systems should ask: where is the security master? Where is encumbrance stored? How are haircuts applied? How are central-bank eligibility rules maintained? How is custody location captured? How are collateral movements updated? How is double counting prevented? How are manual overrides approved? These questions turn a regulatory concept into a functioning control.

28. Liquidity reporting data model

A strong liquidity reporting data model connects products, customers, legal entities, currencies, maturities, behavioural assumptions, encumbrance, market data and regulatory classification. This is why liquidity reporting projects are often harder than they look. The calculation may be documented, but source data is spread across the bank.

Core entities include customer, account, product, balance, transaction, facility, security, derivative, funding instrument, legal entity, currency, counterparty, collateral, market data curve, stress scenario, assumption and report output. Each entity has attributes that matter. For a deposit: customer type, insured status, operational relationship, rate sensitivity, currency, legal entity and product type. For a facility: committed amount, drawn amount, cancellation rights, maturity, borrower, currency and product. For a security: issuer, maturity, rating, HQLA category, encumbrance and haircut. For wholesale funding: instrument, investor type, maturity, currency and legal entity.

Time dimensions are critical. Liquidity is about timing. The model must handle trade date, value date, maturity date, repricing date, settlement date, call date, report date and effective date of assumptions. A balance without the right date context can be misleading. A transaction booked today but settling tomorrow affects liquidity differently from cash already settled.

Behavioural assumptions need versioning. If a deposit product moves from one runoff assumption to another, the bank should know when, why and who approved it. If stress haircuts change, historical reports should remain reproducible. If management overlays are applied, they should be stored with reason and approver. Versioning protects auditability.

Legal entity and currency dimensions must not be optional. Group aggregation is useful, but liquidity stress often bites at entity or currency level. The data model should support both consolidated and standalone views. It should show restrictions and transfer assumptions clearly.

Reconciliation should be designed into the model. Source balances should reconcile to GL where appropriate. Securities should reconcile to custody. Funding instruments should reconcile to debt systems. Facilities should reconcile to credit systems. Reports should show control totals and breaks. A liquidity data model that cannot reconcile will lose trust.

The output layer should support management, regulatory and operational views without mixing definitions. LCR is not the same as internal stress. Daily cash forecast is not the same as regulatory outflow. ALM behavioural maturity is not necessarily the same as NSFR treatment. The data model should allow shared source data with different approved lenses.

29. Requirements and acceptance criteria for liquidity platforms

Liquidity platform requirements should begin with decisions, not screens. What decision is the user trying to make? Place or cover cash today? Issue funding? Raise contingency liquidity? Monitor LCR? Identify top outflows? Report to ALCO? Each decision needs different data freshness, granularity and controls.

For a daily cash dashboard, requirements should specify currencies, legal entities, accounts, opening balances, known inflows, known outflows, forecasts, actuals, variance, cut-offs, payment status, market transaction maturities, large-item alerts and user actions. Acceptance criteria should verify that a large payment changes the forecast, a delayed receipt creates variance, and a maturing repo updates the position on value date.

For LCR reporting, requirements should specify HQLA classification, haircut, encumbrance, deposit runoff category, wholesale funding category, facility drawdown treatment, inflow cap handling, legal entity scope, currency treatment, data lineage, override workflow and report approval. Acceptance criteria should test each material category and reconciliation to source balances.

For NSFR, requirements should cover available stable funding, required stable funding, product mapping, maturity buckets, off-balance-sheet exposures, derivatives treatment, encumbrance, legal entity, currency and effective dating. Test cases should show that changes in loan maturity, deposit stability or wholesale tenor affect the ratio correctly.

For stress testing, requirements should support scenario definition, assumption sets, management actions, time buckets, currency/legal entity views, output comparison, sensitivity, approval and audit trail. Acceptance criteria should verify that scenario changes alter survival horizon and that management actions are applied only when permitted.

For CFP workflows, requirements should include trigger monitoring, stage activation, owner assignment, action checklist, communication log, evidence capture, decision record and post-event review. A tabletop test should prove the system supports the first hour of stress.

Non-functional requirements matter deeply. Liquidity systems need availability during stress, controlled access, audit trail, data lineage, reconciliation, performance, export controls, backup procedures and clear error handling. A liquidity dashboard that crashes during market stress is not only an IT issue. It is a governance failure.

30. Testing scenarios for topic 2

Testing Bank Treasury & Liquidity requires scenario thinking. A tester should not only verify that fields display. The tester should prove that the system behaves like a bank.

Start with a normal surplus day. Opening cash is positive. Known inflows exceed outflows. A maturing security adds cash. Treasury places surplus overnight inside policy. The system should show forecast, actual, placement, residual and audit. Then introduce an unexpected large outflow. The forecast should update, variance should appear, and the residual should change.

Test a deficit day. Opening cash is low. A corporate facility drawdown and outgoing payment create deficit. Treasury covers through repo of eligible securities. The system should reduce the freely available securities inventory, show settled cash raised after the agreed haircut, update collateral encumbrance and reflect funding maturity. Recalculate the full LCR treatment: eligible cash can replace securities in the numerator, while funding outflows and composition-cap adjustments may also change. A repo does not mechanically reduce total HQLA by its gross collateral amount.

Test a deposit concentration movement. One of the top depositors withdraws a large balance. The system should identify concentration impact, update cash forecast, update LCR category if applicable, and trigger alert if thresholds are crossed. It should not hide the movement inside aggregate deposits.

Test committed facility drawdown. A previously undrawn corporate revolver is drawn. Credit exposure, cash outflow, liquidity forecast, facility utilization and reporting should all update. If multiple facilities draw under stress scenario, the stress output should burn liquidity faster.

Test HQLA downgrade or encumbrance. A security moves to a lower eligibility category or is pledged in repo. The available buffer should reduce or haircut should change. Reports should show why. Manual overrides should require approval.

Test intraday delayed receipt. An expected incoming high-value payment is delayed past a cut-off while outgoing payments continue. The system should show intraday position pressure and alert the user. End-of-day position may still be positive, but intraday breach should not be ignored.

Test CFP trigger. Deposit outflows, funding spread widening and ratio deterioration cross predefined triggers. The workflow should activate heightened monitoring or CFP stage, assign actions, record decisions and notify owners.

Test reporting reconciliation. LCR source balances should reconcile to underlying systems. A difference should create a break with owner. Reports should not silently publish with unexplained control breaks.

These scenarios make the topic real. They teach that liquidity is timing, classification, behaviour, collateral, controls and governance working together.

31. Reading a liquidity dashboard like a practitioner

A liquidity dashboard should not be read like a decorative scorecard. It should be read like an operating cockpit. The practitioner asks what changed, why it changed, whether the change is temporary or structural, whether the bank is inside appetite, what action is required, and what could make the position worse. A dashboard that simply shows green or red indicators is not enough. Liquidity management needs drivers.

Start with the current position. What is the bank's cash position by material currency and legal entity? Is the view actual, forecast, or a mix? When was it refreshed? Which large flows are still expected? Are there unsettled securities trades, FX settlements, margin calls, clearing obligations or payment batches not yet reflected? A number without refresh time and scope can mislead.

Then look at buffer quality. How much HQLA is available? How much is unencumbered? Which portion is central-bank reserves, government securities, other eligible assets or lower-quality liquid assets? What is the duration of the securities buffer? What haircut has been applied? Is the buffer located in the legal entity and currency where the stress could occur? Is any asset counted as available while actually pledged, pending settlement or operationally restricted?

Next read LCR and NSFR as driver reports, not medals. If LCR improved, why? Did HQLA increase, outflows reduce, inflows change, or classification move? If NSFR weakened, was it because asset growth consumed required stable funding, or because available stable funding fell? A ratio trend without driver decomposition can hide the real business story.

Look at concentrations. Top depositor movements, wholesale counterparty concentration, investor concentration, secured funding dependence, currency concentration and maturity cliffs all matter. A bank can have a healthy aggregate ratio while depending on a few fragile relationships. Concentration reports should show names, sectors, product types, currencies, legal entities and behavioural assumptions where appropriate.

Read the funding ladder. Which maturities occur in the next day, week, month, quarter and year? Are there cliffs? Are planned issuances already executed or still dependent on market conditions? Has pre-funding been done? Are maturing instruments secured or unsecured? Are they in currencies where markets are currently open? Does the ladder align with asset growth plans?

Read intraday measures. Peak usage, delayed receipts, pending critical payments, payment queue size, settlement failures and cut-off risks tell a different story from end-of-day liquidity. If the dashboard has no intraday view for a bank that participates in high-value or real-time payment systems, the control is incomplete.

Read early-warning indicators. Funding spreads, deposit outflows, media sentiment, rating outlooks, investor feedback, failed rolls, repo haircuts, collateral calls, unusual client queries and market indicators should be visible in a structured way. Liquidity stress is often seen first through signals rather than ratios.

Finally, ask what action is required. A dashboard should support decisions. Place surplus, cover deficit, pre-fund, adjust buffer, trigger heightened monitoring, review concentration, prepare CFP, call ALCO, contact supervisors, or investigate data. If no decision follows from the dashboard, either the dashboard is not useful or the reader is not using it properly.

32. A normal surplus day from opening to close

A normal surplus day is not a day where Treasury relaxes. It is a day where discipline is easier to maintain, but still required. Suppose the bank opens with strong central-bank balances, several expected corporate inflows, maturing reverse repo cash, and moderate outgoing payments. Forecast suggests a surplus by midday and a larger surplus by close.

The cash desk first validates the opening position. Overnight balances are reconciled. Known maturities are confirmed. Large expected flows are checked with relationship managers or payment operations where needed. The desk compares today's flow pattern with normal day-of-week behaviour. It checks whether any clearing, securities settlement or tax event could change the picture.

As inflows arrive, the surplus grows. Treasury must decide how much to leave at the central bank, how much to place overnight, how much to place term, whether to reverse repo, whether to buy short securities, and whether any action affects internal or regulatory liquidity. The cheapest or highest-yielding action is not automatically best. The placement must respect counterparty limits, product limits, tenor limits, currency needs, HQLA policy, settlement cut-offs and tomorrow's forecast.

If surplus comes from a large one-off corporate deposit, Treasury should not automatically treat it as stable funding. It may need to be placed short because the customer may move it tomorrow. If surplus comes from a broad stable increase in retail operational balances, ALM and FTP may see structural value. The same cash amount has different meaning depending on behaviour.

Throughout the day, actual flows are compared with forecast. If expected inflows are late, the surplus may be smaller. If unexpected outflows appear, the desk may cancel or reduce placements. A strong desk keeps optionality until the position is clear. It does not lock too much cash away early simply to earn yield.

At close, the desk documents the residual. What was placed? What remains? What surprised us? Are there open exceptions? Are tomorrow's maturities clear? Did any large client behaviour need follow-up? The surplus day becomes a data point for behaviour analysis, not only an easy day.

The lesson is that surplus needs governance too. Poor banks create future problems during comfortable days. Strong banks use comfortable days to maintain relationships, test markets, improve forecasts and avoid lazy risk.

33. A normal deficit day and the discipline of covering

A normal deficit day occurs when expected outflows exceed available cash after normal inflows. It is not a crisis if planned and covered inside policy. It becomes a problem when the deficit is unexpected, persistent, concentrated or covered with fragile funding that creates tomorrow's problem.

Suppose the bank opens with moderate balances. A large loan drawdown is due, several securities purchases settle, and a wholesale funding maturity rolls off. Expected incoming deposits are smaller than usual. The cash forecast shows a deficit by midday. Treasury must decide how to cover.

The first step is confirmation. Are the outflows real and value today? Are any expected receipts missing? Are there maturing assets that can provide cash? Is the deficit in one currency or several? Is it in the correct legal entity? Can group liquidity help legally and operationally? Is this a one-day timing issue or a structural pattern?

If the deficit is short-term and expected, Treasury may borrow overnight or use repo. If it is part of a longer trend, the funding desk may need term issuance, deposit strategy adjustment or asset-growth moderation. Covering a structural issue with overnight borrowing every day is like using painkillers instead of treating the disease.

Repo can be a strong covering tool if collateral is available and market access is normal. The desk selects eligible securities, applies haircut, executes repo, settles cash and records encumbrance. The action solves today's cash deficit but changes tomorrow's collateral and funding maturity. The system should show both effects.

Unsecured borrowing may be available, but reliance should be monitored. If the bank repeatedly covers deficits with short unsecured wholesale funding, Risk should ask whether the funding profile is becoming fragile. Cheap short money can be addictive. Liquidity governance exists to stop that addiction.

If the deficit is due to a forecast miss, the desk should investigate. Was a large client outflow not communicated? Did payment operations classify a flow incorrectly? Did a system feed arrive late? Was a holiday calendar wrong? Did a facility draw unexpectedly? Forecast accuracy is a control, not a clerical metric.

The close-of-day note should explain how the deficit was covered, whether any limits were used, what residual remains, what caused variance and whether the issue is expected to recur. A normal deficit day handled clearly is good Treasury work. A deficit covered without explanation is a warning.

34. Mild liquidity stress: the moment discipline matters

Mild stress is the stage where a bank can still act calmly if it recognises the signals. The danger is denial. Early stress rarely looks dramatic. Deposit outflows are slightly higher. Funding spreads widen. Investors ask more questions. Repo counterparties shorten tenor. Large customers request reassurance. Market rumours appear. Ratios remain above formal limits, so weak cultures say everything is fine.

A strong Treasury response begins with heightened monitoring. Daily reports may become intraday reports. Top depositor movements are reviewed more frequently. Funding maturities are rechecked. Buffer composition is examined. Unencumbered HQLA is confirmed. Repo capacity is validated. Central-bank collateral readiness is reviewed. CFP trigger levels are watched.

The bank should identify drivers. Are outflows concentrated in one segment? Are they rate-driven, confidence-driven or seasonal? Are wholesale spreads moving for the whole sector or only the bank? Are clients drawing facilities because of their own liquidity needs? Are payment flows changing? Diagnosis matters because actions differ.

Management should prepare before formal crisis. ALCO or a smaller liquidity steering group may meet more often. Investor Relations may prepare messaging. Relationship managers may contact major clients with factual reassurance where appropriate. Treasury may pre-fund or shorten placements. Risk may challenge assumptions. Legal and communications may review CFP protocols.

Mild stress is also a test of culture. If Treasury raises concerns and management says ratios are still green, the bank may lose the advantage of early action. Ratios are lagging and stylised. Behavioural signals can move faster. Early action may feel unnecessary if stress fades, but that is not a failure. The point is to preserve options.

Actions in mild stress should be proportionate. The bank should not panic-sell assets or send alarming messages. It should improve visibility, preserve liquidity, reduce avoidable outflows, avoid new fragile positions, communicate internally and prepare contingency steps. The best stress response is often quiet preparedness.

35. Severe liquidity stress: first 48 hours

Severe liquidity stress is not the time for theory. The bank needs speed, truth, authority and calm. The first 48 hours can determine whether the situation stabilises or accelerates. A severe event may involve rapid deposit outflows, failed wholesale rollovers, market rumours, rating pressure, collateral calls, negative media, or operational payment pressure.

The first action is to establish a single command structure. The bank must know who is leading the liquidity response, which committee or crisis group is active, who reports to senior management, who contacts supervisors, who approves liquidity actions, who handles communications and who maintains the event log. Multiple uncoordinated groups create confusion.

The second action is to establish a trusted liquidity picture. What is cash now by currency and legal entity? What outflows are confirmed today? What outflows are expected tomorrow? What HQLA is unencumbered and mobilizable? What repo capacity exists? Which central-bank facilities are available? Which wholesale maturities are due? Which depositors are moving? Which facilities are being drawn? Which data is uncertain? Management must know both the numbers and the confidence level.

The third action is to activate the appropriate CFP stage. This is why the CFP must be practical. Actions should be sequenced: preserve cash, stop nonessential placements, allow maturities to roll into cash, repo ready collateral, prepare central-bank collateral, contact key funding counterparties, review asset sale options, manage payment priorities within legal and policy boundaries, and control internal/external messaging.

The fourth action is communication. Supervisors should receive coherent, factual updates as required. Senior management and board members need concise situation reports. Employees need instructions about communication discipline. Major clients and investors may need carefully approved messages. Poor communication can trigger more outflows. Overconfident communication that later proves wrong destroys credibility.

The fifth action is to document decisions. Stress creates legal, regulatory and audit scrutiny. Who approved using buffer assets? Who approved central-bank facility access? Who approved asset sales? What assumptions were used? What information was available? Why were certain payments prioritised? Documentation protects the bank and improves later learning.

The first 48 hours should not be spent inventing tools. If dashboards, collateral playbooks, contact lists and approval chains are not ready, stress becomes much harder. Severe stress does not create weaknesses. It reveals them.

36. Liquidity, confidence and communication risk

Liquidity is not only numbers. It is also confidence. Banks are confidence institutions. Depositors, investors, counterparties, rating agencies, regulators and clients all form views about whether the bank is safe enough to trust. Those views influence behaviour, and behaviour influences liquidity. This feedback loop is why communication risk matters.

A technically correct liquidity action can be misread. Using a central-bank facility may be normal in one jurisdiction and alarming in another. Selling securities may be seen as prudent liquidity management or as a sign of desperation depending on context. Raising expensive funding may reassure because liquidity improves, or worry investors because cost is high. Treasury actions need communication framing.

Internal communication is equally important. Relationship managers should not improvise explanations to major clients. Operations should know escalation rules. Front-office Markets teams should understand any funding or collateral constraints. Payment teams should know whether any prioritisation or heightened monitoring is active. Senior management should receive consistent updates. Inconsistent internal messaging creates external confusion.

External communication must be factual and controlled. Overly defensive language can sound weak. Vague reassurance without evidence can fail. Technical language can confuse non-specialists. A good message explains the bank's position, actions and resilience without disclosing sensitive tactical details or creating false certainty. Legal, Communications, Investor Relations, Treasury, Risk and senior management must coordinate.

Social media and fast news have changed liquidity behaviour. Confidence can move faster than traditional reporting cycles. A rumour can trigger client questions within minutes. Digital banking makes withdrawal easier. Treasury cannot control public sentiment, but the bank can monitor signals and respond with disciplined communication.

The CFP should include communication protocols because in stress, words are actions. A wrong word can create outflows. Silence can also create outflows. The goal is not marketing polish. The goal is accurate, calm, timely communication that supports trust and regulatory expectations.

37. Liquidity and payments: real-time banking pressure

Payment systems are the arteries through which liquidity moves. A bank can have a sound funding plan but still suffer operational stress if payment flows are not visible or timed correctly. This is especially true in real-time and instant payment environments, where customer expectations and settlement speed compress the management window.

High-value payment systems create large intraday movements. Corporate treasury payments, interbank payments, central-bank payments and financial market settlements can move material cash. If a large outgoing payment settles before expected incoming cash, the bank's intraday position tightens. If this happens across currencies or legal entities, the situation becomes more complex.

Instant retail payments create continuous liquidity motion. Individual transactions may be small, but aggregate flows can matter. Fraud holds, sanctions screening, system outages and repair queues can change timing. If a bank offers instant payments, Treasury needs enough visibility to understand whether payment behaviour is normal, spiking or concentrated.

Securities settlement links cash and asset movement. If a sale fails, expected cash does not arrive. If a purchase settles, cash leaves and securities arrive. If securities are intended for the liquidity buffer or repo collateral, settlement failure can affect HQLA and collateral capacity. Treasury should receive material settlement-fail information quickly, not days later.

FX settlement links currencies. A customer or bank trade may require one currency to be paid and another received. If one leg is delayed, the bank may face currency funding pressure or settlement risk. Time zones and cut-offs are important. A USD funding need late in the Asian day, for example, may be different from the same need during US market hours.

For business analysts, the question is: how does payment status feed liquidity status? Submitted, validated, released, settled, rejected, returned, repaired and failed states can all mean different liquidity timing. A payment that is initiated but not settled should not be treated the same as cash already gone. A delayed incoming payment should change forecast confidence.

Payment operations and Treasury should share a practical language. Payment teams know queues, cut-offs, repair and settlement status. Treasury knows cash, funding and buffer. Together they manage the timing of value. Separating these teams too strongly creates blind spots.

38. Liquidity and sanctions, fraud, AML and operational holds

Compliance and fraud controls can affect liquidity timing. This does not mean controls are a problem. It means Treasury should understand their liquidity impact. A payment held for sanctions review, fraud investigation or AML concern may delay outgoing cash or incoming credit. At individual level this may be small. At aggregate or high-value level it can matter.

Suppose a large outgoing payment is held after being included in expected cash outflows. Treasury may temporarily have more cash than forecast, but the cash is not freely available if the payment may release later. Treating it as usable surplus could create a later deficit when the payment clears. Conversely, an expected incoming payment held elsewhere may not arrive when forecast, creating a funding gap.

Fraud incidents can create unusual payment patterns. A spike in suspicious payments may lead to holds, reversals, customer complaints and operational workload. AML investigations may delay funds. Sanctions screening may block or reject cross-border flows. These statuses should be visible enough for liquidity forecasting where material.

Market-related payments can also be affected. FX settlements, securities cash legs, derivative payments and collateral movements may pass through screening or operational checks. A delayed collateral payment can create counterparty exposure. A blocked FX payment can create settlement risk. Treasury may need to know if the delay is temporary, final, or uncertain.

Requirements should define how held payments are treated in cash forecasts. Are they excluded, included with status, included with probability, or manually adjusted? Who owns the judgement? How are material holds escalated? Can Treasury see amount, currency, value date, expected release and reason category without breaching confidentiality? These are practical design questions.

The key point is not to weaken compliance. It is to integrate control reality into liquidity forecasting. A bank that manages liquidity as if all payments settle instantly after initiation is ignoring its own control process.

39. Treasury liquidity and customer product design

Product design shapes liquidity. A bank can create a fragile balance sheet through products that seem attractive individually but produce unstable funding or cash-flow optionality. Treasury should therefore be involved in product design before launch, not after volumes grow.

Deposit products are the obvious example. A high-rate savings product may bring large balances quickly, but if customers are rate-sensitive, those balances can leave quickly. A notice deposit may be more stable if notice periods are enforceable and customer behaviour supports it. A transactional account linked to payroll and bill payments may provide stable operating balances. Product features such as rate tiers, withdrawal limits, notice periods, digital switching ease and promotional rates affect liquidity.

Loan products also matter. Fixed-rate lending creates interest-rate and funding considerations. Revolving facilities create contingent liquidity outflows. Prepayment options change cash-flow behaviour. Balloon maturities create repayment concentration. Multi-currency loans create currency funding needs. Product teams may focus on customer demand, but Treasury asks how the cash behaves.

Payment products affect intraday liquidity. Instant payments, bulk corporate payments, request-to-pay products, card settlement and cross-border payment services all create timing patterns. A new corporate payments product may generate large outflows earlier in the day. A new instant payment channel may increase weekend liquidity needs. Treasury needs these implications before launch.

Investment and wealth products can also affect bank liquidity if they move deposits into off-balance-sheet instruments or create settlement obligations. A campaign that encourages customers to move from deposits to mutual funds may reduce stable funding. A securities platform may increase settlement and FX flows. Product success can create Treasury work.

FTP is the mechanism that translates liquidity impact into product economics. If a product consumes stable funding, it should be charged. If it provides stable funding, it may receive credit. If it creates contingent drawdown risk, pricing should reflect it. Product profitability that ignores liquidity cost is incomplete.

A new product approval process should ask: what funding does this product create or consume? How stable are expected balances? What are withdrawal or drawdown rights? What is the currency profile? What are value-date and settlement patterns? Does it affect LCR or NSFR? Does it need HQLA support? What stress behaviour is expected? What data must be captured for Treasury? Without these questions, the bank may launch a product that looks good in sales and bad in stress.

40. Treasury liquidity in multi-currency banks

Multi-currency banks face a harder liquidity problem because cash cannot be treated as one universal pool. A surplus in one currency may not solve a deficit in another unless the bank can convert or swap currencies in time and at acceptable cost. Currency liquidity therefore requires dedicated management.

USD is often especially important in global banking because many trade, investment and market flows settle in USD even outside the United States. A European or Asian bank may have large USD assets or payment obligations but a more limited natural USD deposit base. It may rely on FX swaps, wholesale USD funding, correspondent balances or securities that can generate USD cash. In stress, USD funding can tighten.

FX swaps are a common currency funding tool. A bank with surplus EUR and need for USD may swap EUR for USD and agree to reverse later. This helps liquidity, but it introduces rollover risk, market access risk, basis cost and settlement timing. FX swaps are deep markets, but not risk-free funding magic. Treasury should monitor maturity ladders and stress assumptions by currency.

Currency-level HQLA matters. A bank may hold liquid assets in one currency while outflows occur in another. Regulatory frameworks and internal stress tests often require attention to currency mismatch. If stressed outflows are in USD, EUR securities may not be immediately sufficient unless convertible through markets. During stress, conversion cost and market access can change.

Legal entity and currency interact. A branch may have local currency surplus but head office may need another currency. Local rules may restrict transfer. Tax or resolution constraints may apply. Nostro account locations matter. A group-level treasury dashboard should allow drill-down into currency by legal entity.

Payment cut-offs are currency-specific. A funding action possible in one time zone may be too late for another. A multi-currency liquidity desk needs awareness of global market hours, settlement systems, nostro banks and internal handover. Treasury is a time-zone business.

For requirements, currency cannot be a display filter added at the end. It must be a core dimension in forecasting, stress, LCR/NSFR analysis, HQLA, funding ladder, intraday monitoring and CFP actions. Currency mismatches should be visible, not buried in converted totals.

The professional memory line: group liquidity is useful; currency liquidity pays the actual obligation.

41. Treasury liquidity and technology resilience

Liquidity management depends on technology. If systems fail during stress, the bank loses visibility when it needs it most. Technology resilience is therefore part of liquidity risk management, not only IT operations.

The critical systems include cash-positioning platforms, payment systems, treasury management systems, trade capture, collateral management, security master, market data, ALM, LCR/NSFR reporting, stress testing, general ledger, data warehouses, dashboards and communication tools. Failure in any one may not stop the bank, but combined failures can create serious risk.

Data latency matters. A daily cash view that is several hours late may be acceptable for long-term analysis but not for intraday liquidity. A payment feed delay can make the bank think cash is available when it has already gone. A collateral feed delay can overstate available HQLA. System timestamping and freshness indicators are essential.

Resilience requirements should include availability, backup, manual fallback, recovery time, recovery point, access control, audit, monitoring and incident escalation. Treasury should know how it will operate if the main cash dashboard fails. Is there a fallback report? Can balances be obtained from payment systems? Can central-bank accounts be checked directly? Who has access? When was the fallback tested?

Cyber events can become liquidity events. If customers cannot access accounts, confidence may weaken. If payment systems are disrupted, settlement obligations may fail. If market rumours start during an outage, deposit behaviour may change. Treasury and operational resilience teams should understand this connection.

Change management also matters. A small mapping change in a liquidity feed can distort reports. A release that changes payment status codes can break forecasting. A reference-data migration can change HQLA classification. Treasury systems require controlled testing, regression evidence and business sign-off because reporting errors can create governance errors.

For developers, the domain lesson is that every timestamp, status, currency, legal entity, maturity date and source identifier matters. Liquidity systems are not simple dashboards. They are decision systems used when the bank may be under pressure.

42. Developer and data-engineering view of liquidity

A developer building liquidity functionality needs to understand the domain enough to avoid dangerous simplifications. Liquidity is not simply sum balances by currency. It is source data, transformation, classification, timing, assumptions, stress logic, legal entity, encumbrance, reconciliation and audit.

Event design matters. A payment initiated event is not the same as a payment settled event. A trade booked event is not the same as a trade settled event. A repo initiated event is not the same as cash received. A facility approved event is not the same as facility drawn. Systems should carry state clearly because liquidity timing depends on state.

Idempotency matters in event-driven systems. If the same payment event is processed twice, cash forecast may double count. If a cancellation arrives before original booking because of ordering issue, the system must handle it. If replay is used after outage, calculations must not create duplicates. Liquidity systems need robust event handling.

Schema governance matters. A change in product code, counterparty type, maturity field, legal entity identifier or payment status can break liquidity classification. Data contracts should be versioned. Downstream liquidity reporting should receive alerts when schemas change. Silent schema drift is dangerous.

Lineage and audit are functional requirements. Users should be able to trace a liquidity number to source records and assumptions. If HQLA changed, the system should show whether holdings changed, eligibility changed, haircuts changed, encumbrance changed or market value changed. If LCR moved, drivers should be visible. If stress survival changed, assumption version should be identifiable.

Reconciliation should be automated where possible. Cash balances reconcile to bank accounts or GL. Securities reconcile to custody. Funding instruments reconcile to debt records. Facilities reconcile to credit systems. Breaks should be surfaced with owner and age. A system that calculates quickly but cannot reconcile will not be trusted.

Performance design should match business need. Intraday dashboards require faster refresh than monthly NSFR. Stress testing may require scenario batch capacity. ALCO reporting may require version-controlled snapshots. Regulatory reporting may require locked reporting periods. Different use cases need different architecture.

The developer's best question is: what decision will this data support, and what could go wrong if it is late, duplicated, misclassified or unauditable? That question turns technical design into banking design.

43. BA, tester and operations handover notes

A BA working on Treasury Liquidity should produce requirements that operations, developers, testers, Risk and Treasury can all understand. The document should define business purpose, scope, data sources, calculations, assumptions, user actions, exceptions, controls, audit and reporting. It should avoid vague phrases like show liquidity position without defining the position.

A good BA workshop starts with flow. What happens from source event to liquidity view? For a deposit outflow, what system records it, when does Treasury see it, what status matters, how is it classified, how does it affect forecast, how does it affect LCR, and who investigates if it is unexpected? For a securities repo, what trade, collateral, cash and encumbrance events occur? Flow thinking exposes missing requirements.

Testers should build domain scenarios, not only UI scripts. A tester should know why a wrong value date matters, why an encumbered security should not count as available, why a committed facility drawdown affects liquidity, why legal entity aggregation can mislead, why a held payment is different from a settled payment, and why stale market data should be flagged.

Operations handover should be practical. What daily checks must be performed? Which exceptions appear? Which breaks are urgent? Which reports must be signed off? Who approves manual adjustments? How are late feeds handled? What is the fallback if a system is down? What evidence is required for audit? A beautiful implementation without operational handover will fail slowly.

Training material should use fictional, sanitised examples with stated assumptions. Show a surplus day, deficit day, top depositor outflow, repo funding, HQLA downgrade, facility drawdown, intraday delay and CFP trigger. Let users see how the system reacts. Treasury users learn by flow and consequence.

Sign-off should include Treasury, Liquidity Risk, Finance, Operations, Technology and sometimes Compliance or Regulatory Reporting. Each team sees a different risk. Treasury asks if the tool helps manage cash. Risk asks if assumptions and limits are controlled. Finance asks if numbers reconcile. Operations asks if breaks can be handled. Technology asks if the system can run. Good sign-off brings those views together.

44. Practical source-backed regulatory section

The official Basel LCR framework states the objective of promoting short-term resilience of banks' liquidity risk profile by ensuring adequate unencumbered HQLA that can be converted into cash to meet liquidity needs during a 30-day stress scenario. In practical terms, this supports the chapter's repeated point: liquidity buffer assets must be real, unencumbered and usable under stress, not merely impressive in accounting value.

The Basel NSFR framework addresses a longer-term structural funding problem by requiring a stable funding profile relative to assets and off-balance-sheet activities. This supports the chapter's structural message: daily cash management is not enough if long and illiquid assets are funded with unstable short money.

Basel's IRRBB standards matter for liquidity because interest-rate movement can change liquidity-buffer securities values and deposit behaviour. IRRBB measures economic value of equity (EVE) and earnings such as net interest income (NII), rather than cash survival. The July 2024 shock recalibration has a Basel implementation date of 1 January 2026; national transposition determines the applicable local requirements. Do not assume the original 2016 shock sizes remain current everywhere. IRRBB is not the same as liquidity risk, but in practice the two interact. A liquidity portfolio with high duration can become painful when rates rise and cash is needed. A deposit base can change behaviour when rates move.

Market-risk capital rules and trading-book standards matter where Treasury interacts with Markets through securities, repo, derivatives, FX and hedging. Some positions are banking-book liquidity assets. Others are trading-book inventory. The prudential boundary affects capital, risk measurement and governance; accounting classification and valuation must be assessed separately. Liquidity stress can expose weaknesses if classification is not honest.

The SEC's T+1 implementation in the US and ESMA's EU T+1 work matter because shorter securities settlement cycles compress funding, FX, allocation, confirmation and repair windows. Treasury liquidity cannot ignore settlement-cycle changes. A faster settlement market may reduce counterparty and time risk, but it increases operational and funding discipline requirements.

The key is to use official rules as anchors while teaching the human flow. A learner should know where LCR, NSFR, IRRBB, market risk and settlement-cycle changes come from. But the purpose of this chapter is to make those frameworks operationally understandable inside a real bank.

45. Source references for Topic 2

• BIS Basel Committee - Basel III Liquidity Coverage Ratio and liquidity risk monitoring tools - primary source for LCR, HQLA and the 30-day short-term liquidity stress concept.
• BIS Basel Committee - Basel III Net Stable Funding Ratio - primary source for stable funding over a one-year horizon.
• BIS Basel Committee - Interest rate risk in the banking book - official source for IRRBB governance and measurement standards.
• BIS Basel Committee - Minimum capital requirements for market risk - official source for trading-book market-risk capital standards.
• SEC - T+1 settlement cycle implementation statement - official US source confirming the T+1 settlement cycle implementation date of 28 May 2024.
• ESMA - Shortening the settlement cycle to T+1 in the EU - EU transition preparation. 11 October 2027 is a future transition date, not the current general settlement cycle; see the securities chapter for the applicable legislative scope.

46. Final long-form checklist for Bank Treasury & Liquidity

Before leaving this chapter, the learner should be able to explain liquidity without hiding behind acronyms. Liquidity means the bank can meet obligations when due, in the right currency and legal entity, without unacceptable loss or franchise damage. It is not the same as solvency. It is not the same as profit. It is not the same as owning many assets. It is a timing, confidence and operational-access discipline.

The learner should be able to describe a treasury day. Opening balances are checked. Forecasts are built. Known flows are confirmed. Actuals are compared. Surplus is placed safely. Deficit is covered responsibly. Large variances are investigated. Residual is intentional. Exceptions are logged. Tomorrow's position is prepared.

The learner should be able to explain HQLA. The asset must be high quality, liquid, unencumbered, operationally available and useful under stress. Gross holdings are not the same as usable buffer. Duration and forced-sale risk matter. Repo capacity, central-bank eligibility, custody and haircut behaviour matter.

The learner should be able to read LCR and NSFR. LCR is the 30-day stressed liquidity buffer view. NSFR is the longer-term stable funding structure view. Both are necessary. Neither replaces internal stress, intraday liquidity, legal-entity analysis, currency analysis or management judgement.

The learner should understand intraday liquidity. Payments, settlements, margin, FX and securities flows move during the day. End-of-day adequacy does not guarantee intraday comfort. Payment status, cut-offs, delayed receipts and critical obligations must be visible.

The learner should understand wholesale funding. Tenor, investor base, instrument mix, currency, maturity ladder and confidence all matter. Cheap short-term funding can be useful but dangerous if overused. Funding strategy is resilience design, not only cost minimisation.

The learner should understand CFP. A Contingency Funding Plan is a tested playbook with stages, triggers, actions, communications, authority and evidence. It must be rehearsed. It must identify real capacity, not theoretical comfort.

The learner should understand FTP. Liquidity has an internal price. Products that consume stable funding must feel that cost. Products that provide stable funding may receive credit. Without FTP, business units can grow a fragile balance sheet while believing they are profitable.

The learner should understand governance. ALCO, independent Liquidity Risk, Finance, Operations, Treasury, Risk Committee and Audit each play a role. Early escalation is a strength. Ratio-only thinking is weak. Assumptions must be owned and challenged.

The learner should be ready for topic 3, Money Markets, because the defensive liquidity logic is now clear. Money Markets will explain the instruments Treasury uses every day to place surplus, cover deficit, transact repo, manage short-term funding and interact with central-bank operations.

47. Final human summary

Bank Treasury & Liquidity Management is the bank's discipline of staying able to pay. Everything else depends on that. A bank can have strong products, good customers, clever technology and impressive strategy, but if liquidity confidence breaks, the conversation changes immediately. Treasury's job is to make sure the bank does not discover its fragility too late.

The work is quiet because success looks ordinary. Payments settle. Depositors stay calm. Funding rolls. Securities can be monetized. Ratios remain above appetite. ALCO actions close. Reports reconcile. Stress tests lead to decisions. CFP drills find and fix gaps. Nothing dramatic happens. That quietness is not lack of value. It is the value.

For a professional learner, this chapter is not only about becoming a treasurer. It is about understanding how customer behaviour, payment systems, balance-sheet growth, securities, repo, central-bank access, regulatory ratios, data, controls and governance connect. Once you understand that, you can work better as a BA, tester, developer, operations analyst, risk analyst, product owner or architect in any banking environment that touches liquidity.

The chapter's simplest truth is also its most important: liquidity is timing under trust. When trust is high, timing is easy. When trust weakens, timing becomes survival. Treasury exists so the bank is ready before that moment arrives.

48. Worked example: one week in liquidity management

A one-week example helps make the discipline real. Imagine a bank starts Monday with comfortable liquidity. The LCR is comfortably above internal appetite, NSFR is stable, wholesale maturities are manageable, and the HQLA buffer is mostly unencumbered government securities and central-bank reserves. Nothing looks dramatic. That is exactly why discipline matters: most liquidity trouble begins while everything still looks manageable.

On Monday morning, the cash desk sees a normal opening position. Several corporate inflows are expected. A small repo matures and returns securities. Retail payment flows are close to seasonal pattern. Treasury places a modest surplus overnight and leaves enough cash at the central bank for known payment obligations. The day closes cleanly. A weak desk would stop there. A strong desk notes that two large corporate balances increased materially and asks relationship managers whether these are temporary operating balances or short-term parking.

On Tuesday, one of those corporate balances leaves through a high-value payment. The outflow is not a crisis, but it creates a forecast variance. The desk updates the top-depositor concentration report and notices that the customer has moved balances in large blocks several times around quarter-end. ALM reviews whether this segment should receive a less stable behavioural treatment. FTP may not change immediately, but the deposit is not treated as a stable gift. The lesson is that liquidity analysis is not only about stress. It is about learning from normal behaviour.

On Wednesday, a wholesale certificate of deposit maturity is due. The funding desk had expected to roll it for one month, but market pricing is slightly wider because peer bank spreads have moved. Treasury can still roll, but at higher cost. The desk decides to roll only part and use a planned repo for the rest, preserving investor relationship without overpaying blindly. Risk asks whether repeated partial rolls would increase secured-funding dependence. The funding ladder is updated. The lesson is that market access is not binary. It weakens in shades before it closes.

On Thursday, a securities settlement fails. The bank expected cash from a security sale, but settlement does not complete because of a counterparty instruction mismatch. The expected inflow is delayed. Treasury does not panic because the amount is manageable, but the intraday dashboard shows lower cash than forecast. Operations owns the fail, Treasury adjusts the cash position, and the liquidity report marks the inflow as uncertain until settlement confirmation. The lesson is that operations events are liquidity events when values are material.

On Friday, a central-bank holiday in one currency and a normal business day in another create timing complexity. A multi-currency payment flow requires cash in the open market currency, while some expected receipts in the holiday currency are delayed. The desk uses an FX swap already inside policy to manage the currency need. The action works because currency-level forecasts existed. If the bank had only looked at consolidated cash converted into one reporting currency, it might have missed the issue. The lesson is that currency liquidity pays obligations, not group totals.

At the end of the week, ALCO receives a short summary. No ratio breach occurred. No crisis occurred. But the week produced four learning points: one corporate deposit segment is more volatile than assumed, wholesale roll pricing is widening, securities fails can affect intraday forecast, and currency calendars need better visibility in a specific product flow. A strong bank treats these as improvement items. A weak bank says the week was green and learns nothing.

That is the difference between passive liquidity reporting and active liquidity management. The best Treasury teams learn during normal weeks so they are not surprised during abnormal ones.

49. Worked example: product launch liquidity review

Now imagine the bank wants to launch a new digital business savings product for SMEs. The product team is excited. The rate is attractive, onboarding is simple, and marketing expects rapid balance growth. The commercial story looks good. Treasury should not block the idea automatically, but it must ask what kind of funding the product creates.

First, Treasury asks who the customers are. Are these operating balances used for payroll, supplier payments and tax? Or are they excess cash balances chasing yield? Operating balances may be more stable. Rate-sensitive surplus balances may move quickly. If the product is marketed heavily around headline rate, the bank should assume some rate sensitivity. If the product is integrated with payment services, it may have more operational stickiness.

Second, Treasury asks about withdrawal rights. Can customers withdraw instantly? Are there limits? Is there a notice period? Are promotional rates temporary? Can balances move through APIs or bulk payments? The easier it is to move money, the more behavioural modelling matters. Digital convenience is excellent for customers but can accelerate outflows under competition or confidence stress.

Third, Treasury asks about currency and legal entity. Is the product only domestic currency? Does it sit in one bank entity? Can balances be used to fund local assets? Are there local liquidity requirements? If balances grow in one currency while lending grows in another, Treasury must plan transformation or separate funding.

Fourth, FTP must be agreed. The business may want full credit for deposit balances. Treasury may argue that only the stable core deserves full funding credit, while rate-sensitive balances receive a lower credit or shorter behavioural life. This is not internal politics. It is economic truth. If FTP over-rewards unstable balances, the product appears more profitable than it is.

Fifth, liquidity reporting and data must be ready. The product code should map correctly to deposit category, customer segment, insured or uninsured status, operational relationship flag where relevant, runoff assumptions, concentration monitoring and ALM behaviour. If the product launches before data mapping is ready, Treasury may be blind to fast growth.

Sixth, stress behaviour should be discussed. What happens if the bank cuts the rate? What happens if a competitor offers higher rate? What happens if negative news appears? What happens if customers withdraw at weekend through instant channels? What alerts should fire? What threshold should trigger ALCO review?

The launch decision may still be yes. But it will be a better yes. Product gets growth. Treasury gets visibility. FTP reflects behaviour. Risk gets assumptions. Operations understands payment patterns. Technology captures the right data. That is how liquidity thinking supports business rather than merely saying no.

50. Worked example: liquidity issue after a payment system delay

A payment system delay can create liquidity confusion even when the bank's balance sheet is healthy. Suppose the bank expects a large incoming settlement payment at 10:00 and several outgoing customer payments between 10:15 and 11:00. Normally this timing works. Today the incoming settlement is delayed because the counterparty has a technical issue. Outgoing payment files are ready.

Payment operations sees a delayed receipt. Treasury sees forecast cash not arriving. Customer teams see outgoing payments waiting. Risk wants to know whether internal intraday limits are at risk. The first mistake would be for each team to work separately. The correct response is to connect status, cash and decision.

Treasury should update the intraday forecast immediately. Move the payment from on-time expected inflow to delayed expected inflow; actual cash stays unchanged until final receipt. Outgoing payments should be reviewed by criticality and legal obligation according to policy. If payments must settle, Treasury may use intraday credit, repo, central-bank balance or other approved liquidity source. If some payments can wait legally and operationally, sequencing may help. The decision must be controlled.

Operations should chase the counterparty or infrastructure status and provide realistic expected resolution. Not hopeful guesses. Realistic status. If the incoming payment is only delayed by minutes, the action differs from a delay with no expected time. Communication between Operations and Treasury must be fast.

Customer communication may be needed if outgoing payments are delayed. It should be accurate and not blame internal systems casually. If the bank can settle outgoing payments using liquidity, customers may never notice. If delays occur, relationship managers need approved language.

After resolution, the bank should review whether forecast logic was correct. Did the dashboard show the delay? Did alerts fire? Did Treasury know before outgoing cut-offs? Were payment statuses granular enough? Did teams know the playbook? Did intraday liquidity limits work? Did the customer impact process work?

This example shows that liquidity is not only funding markets and deposits. It is also operational timing. Payment systems, statuses, queues, settlement confirmations and cut-offs are part of the liquidity world.

51. Review questions for serious learners

Answer these in your own words before moving forward.

1. Why can a solvent bank fail from liquidity pressure?
2. What is the difference between cash, liquidity buffer and HQLA?
3. Why is unencumbered status more important than gross securities holdings?
4. What does LCR try to prove, and what does it not prove?
5. What does NSFR try to prevent?
6. Why are large corporate deposits treated differently from stable retail operating balances?
7. Why do undrawn committed facilities create liquidity risk?
8. How does repo convert securities into cash, and what can go wrong?
9. Why is central-bank facility access useful but not a substitute for normal liquidity management?
10. Why must liquidity be reported by currency and legal entity?
11. What is survival horizon and why can it be more intuitive than a ratio?
12. What are three examples of assumption risk in liquidity stress testing?
13. What should a real Contingency Funding Plan contain?
14. Why does intraday liquidity matter even if end-of-day balances are positive?
15. How can a new product create hidden liquidity risk?
16. Why does FTP matter for liquidity discipline?
17. What should a liquidity dashboard show beyond green ratios?
18. How do payment delays affect liquidity forecasting?
19. What data fields are critical for liquidity reporting?
20. What is the difference between early-warning trigger and hard limit?

If these questions feel easy, the chapter has landed. If they feel difficult, revisit the matching sections. Liquidity knowledge becomes powerful only when you can explain it without reading from a slide.

52. Final workplace checklist

When you join a Treasury, Liquidity Risk, ALM, payments, data or reporting project, use this checklist.

Ask for the business purpose first. Is the project about daily cash, intraday liquidity, LCR, NSFR, stress testing, HQLA, funding ladder, FTP, ALCO reporting, CFP workflow or regulatory reporting? Do not let a project simply be called liquidity dashboard. Name the decision.

Ask for scope. Which legal entities, currencies, products, customers, systems, time horizons and reports are included? What is explicitly out of scope? Liquidity projects fail when scope is assumed.

Ask for source systems. Where do deposits come from? Where do loans and facilities come from? Where do securities come from? Where is encumbrance stored? Where are payments and settlements sourced? Where is market data sourced? Where is GL reconciliation done?

Ask for assumptions. Which runoff factors, behavioural maturities, drawdown rates, haircuts, inflow caps, management actions and stress assumptions are used? Who owns them? How are they versioned? How are changes approved?

Ask for controls. How are data breaks handled? How are manual overrides approved? What reconciles to what? Who signs off the report? What audit trail exists? What happens if a feed is late?

Ask for users and actions. Who reads the output? What action can they take? What alert requires escalation? What is the workflow if a limit is breached? Who receives notification? What evidence closes the action?

Ask for non-functional needs. How fresh must data be? What is the recovery plan? Can the system work during stress? Who has access? Can the report be reproduced later? Can numbers be traced?

This checklist makes you useful quickly. It proves you understand liquidity as a bank operating discipline, not just as a set of definitions.

53. Topic 2 closing standard

Apply this chapter by explaining a full cash lifecycle: source obligation, correct entity/currency, value date and due time, approved funding, independent settlement controls, final cash evidence, accounting and reconciliation, exception ownership and the next repayment. Link that lifecycle to buffer eligibility, internal stress capacity and regulatory ratio mechanics without treating them as identical numbers.

For the practical test, combine a delayed receipt, a concentrated withdrawal, wider repo haircuts and an approaching cut-off. Identify usable cash now, executable actions by the due time and remaining risks after action. The advanced chapter and numerical workbooks develop these cases further.

Bridge to Money Markets

You now have the defensive liquidity core of bank treasury. The next major section: Money Markets: makes the tactical instruments concrete: interbank, repo, short-term paper, and central-bank operations where short-term cash is actually placed and raised every day.

Bank Treasury & Liquidity Management - Practical Addendum

A practical companion for implementing the core chapter's treasury operating flow, forecasting, product review, data controls and exception management. The architecture and tests are illustrative bank designs.

Why this addendum exists

The main Topic 2 chapter already explains liquidity, HQLA, LCR, NSFR, intraday liquidity, wholesale funding, contingency funding, stress testing, FTP, governance, dashboards, payments and workplace readiness in substantial detail. The useful gap is not another definition of liquidity. The useful gap is a more explicit operating view of how a bank turns liquidity theory into daily decisions, system logic and testable requirements.

In a real bank, Treasury and Liquidity Risk do not simply read a ratio and relax. They run a daily operating algorithm. They collect balances, validate source feeds, classify flows, forecast outflows, check liquidity in the correct currency and legal entity, compare positions with appetite, decide whether to place surplus or cover deficit, monitor intraday settlement, update stress capacity, escalate exceptions and record actions. That operating discipline is what keeps liquidity management alive.

This addendum is written for business analysts, developers, testers, data engineers, operations teams, risk teams and students who need to understand how liquidity management becomes a real system and workflow.

1. The practical liquidity operating algorithm

A bank liquidity process begins before the dashboard is opened. The first step is source collection. Cash balances come from central-bank accounts, nostro accounts and treasury systems. Deposit balances come from core banking and channel systems. Loan drawdowns and repayments come from lending platforms. Securities and repo data come from treasury or custody systems. Wholesale funding maturities come from treasury deal capture. Derivative margin calls come from collateral systems. Payment flows come from payment hubs, clearing systems and operations queues. If any source is missing, the liquidity view is not complete.

The second step is classification. A cash balance must be linked to currency, account, legal entity and availability. A deposit must be linked to product type, customer segment, stability assumption and concentration. A loan must show maturity, drawdown schedule and undrawn commitment. A security must show HQLA eligibility, encumbrance, custodian, haircut and monetisation route. A repo must show maturity, collateral, counterparty and settlement status. A payment flow must show expected time, status, currency, account and cut-off. Classification is where liquidity becomes usable.

The third step is projection. Treasury projects expected cash inflows and outflows across time buckets. The shortest buckets are operational: now, next hour, today, tomorrow and this week. Regulatory and structural buckets may cover thirty days, three months, six months, one year and beyond. A practical projection separates confirmed flows, forecast flows, behavioural flows and stress assumptions. A confirmed security maturity is not the same as an expected corporate deposit retention. A committed loan drawdown under stress is not the same as a scheduled repayment.

The fourth step is comparison against appetite. The bank compares cash, HQLA, intraday usage, LCR, NSFR, survival horizon, funding gaps, concentration, encumbrance and currency positions against limits and triggers. A trigger is early warning. A hard limit is a boundary. The process should show not only whether the metric is green, amber or red, but why it moved.

The fifth step is action. If there is surplus cash, Treasury decides where to place it safely. If there is deficit, Treasury decides whether to borrow, repo assets, use central-bank balances, delay non-critical internal flows where policy allows, or activate contingency actions. If there is structural weakness, ALCO may change funding plan, deposit pricing, loan growth, securities strategy, hedging or FTP. Liquidity management is incomplete unless calculation leads to decision.

The sixth step is evidence. The bank records key assumptions, source completeness, decisions, approvals, exceptions, actions and closure. In calm markets this may feel administrative. In stress it becomes vital. If a regulator, auditor, board member or crisis committee asks why Treasury acted, the evidence should already exist.

2. Liquidity forecasting logic

Liquidity forecasting is not one number. It is a layered estimate of what cash will be available and what cash will leave. A good forecast separates contractual flows, behavioural flows, operational flows and stress flows.

Contractual flows are flows the bank can derive from contracts or booked events: funding maturities, repo maturities, bond coupons, securities maturities, scheduled loan repayments, derivative coupons, known wholesale roll-offs and confirmed settlement obligations. These flows are usually easier to model but still require correct data. A wrong maturity date can create a false liquidity view.

Behavioural flows depend on customer or counterparty behaviour. Non-maturity deposits, early loan repayment, facility drawdowns, deposit renewals and customer payment behaviour fall here. Behavioural flows require assumptions. Those assumptions should be based on evidence and reviewed regularly. A deposit segment that behaved as stable in a low-rate world may behave differently when rates rise or confidence weakens.

Operational flows are flows created by banking operations: payment files, clearing obligations, card settlement, securities settlement, payroll, tax, internal transfers, nostro funding and margin calls. These flows may be predictable by pattern but still volatile in timing. A payment expected at 10:00 and arriving at 15:00 is not the same liquidity event.

Stress flows are assumptions applied under severe conditions. They may include faster deposit runoff, higher facility drawdowns, reduced wholesale market access, wider repo haircuts, lower monetisation value, higher collateral calls and delayed inflows. Stress flows should not be hidden inside a normal forecast. They should be visible as scenario layers.

For system design, each forecast item should carry source, confidence level, time bucket, currency, legal entity, account, status, owner and last update time. This allows users to distinguish confirmed cash from expected cash and expected cash from assumed cash.

3. Product liquidity review algorithm

Every new banking product can change liquidity. A product may bring deposits, consume funding, create undrawn commitments, increase payment volumes, change intraday timing, create FX needs or require collateral. Treasury should therefore review product liquidity before launch.

The first question is whether the product creates funds or uses funds. A deposit product provides funding, but not all deposits are equally valuable. A loan product uses funding, but not all loans consume liquidity in the same way. A credit line may show low drawn balance today but create stress liquidity need tomorrow. A payment product may not create large balance-sheet exposure but can create intraday liquidity and operational timing pressure.

The second question is behaviour. Can the customer withdraw instantly? Is there a notice period? Is the product rate sensitive? Is the customer relationship operational or investment-like? Are balances concentrated? Can balances move through API or bulk channels? Does the product attract stable operating balances or promotional money? Product design shapes liquidity behaviour.

The third question is currency and legal entity. If a product gathers deposits in one currency but lending grows in another, Treasury must manage transformation. If a product sits in one legal entity, liquidity may not be available to another. If the product depends on cross-border transfers or payment systems, cut-offs and settlement timing matter.

The fourth question is FTP. The product should receive or pay an internal price that reflects funding value, liquidity cost, optionality and behavioural maturity. If FTP over-rewards unstable deposits, the product may grow quickly while weakening resilience. If FTP undercharges long-term assets, lending may appear more profitable than it is.

The fifth question is reporting readiness. Product codes, customer segments, maturity, rate type, commitment amount, currency, legal entity, channel, withdrawal rules and behavioural flags must feed ALM and liquidity reports from day one. A product launched without data mapping creates blind spots.

4. Currency and legal-entity liquidity

A bank does not pay obligations with consolidated liquidity. It pays with cash in the required currency, account, entity and settlement system. This is why currency and legal-entity liquidity are central to real treasury management.

A banking group may have surplus liquidity in one subsidiary and deficit in another. Transfer may be restricted by regulation, local supervision, tax, capital, operational rules or internal appetite. A parent-level report can look comfortable while a local entity is under pressure. This is not a theoretical issue. In stress, liquidity trapped in the wrong entity may not save the entity that needs cash.

Currency liquidity is equally important. A bank may be comfortable in EUR but tight in USD, GBP, SEK, DKK or another currency. FX swaps can transform currency liquidity, but swap markets can become more expensive or less available in stress. The bank should not assume that a strong group cash number can instantly become the currency it needs.

Systems must therefore capture currency, legal entity, branch, account location, central-bank account, nostro, custodian, settlement system, encumbrance and transferability. Reports should show both consolidated and granular views. The consolidated view is management summary. The granular view is survival reality.

5. Collateral monetisation and repo readiness

A liquidity buffer is useful only if the bank can turn it into cash. Securities may be sold, repoed, pledged to central bank facilities or used as collateral. Each route has operational and market requirements. A bond sitting in a custodian account is not automatically usable cash.

Repo readiness means the bank knows which securities are eligible, which counterparties can accept them, what haircuts apply, what settlement timings apply, which legal agreements are active, which systems can process the transaction and which people can execute under stress. A security that is high quality but operationally unavailable is weak liquidity.

Collateral monetisation also requires encumbrance control. If an asset is already pledged, it cannot be counted as freely available. If it is pledged in one transaction and assumed available in another report, liquidity is overstated. Encumbrance should update from relevant reserved, instructed, settled, released and failed events, with periodic reconciliation. A once-daily feed alone can overstate intraday availability.

Stress testing should include haircut widening and market access reduction. In calm markets, repo may be easy. In stress, haircut increases or counterparty limits may reduce cash that can be raised. The bank should know stressed monetisation value, not only market value.

6. Central-bank facility readiness

Central-bank facilities can be an important liquidity backstop, but they are not a substitute for normal liquidity management. A bank should know which facilities exist, what collateral is eligible, what operational steps are required, which entity can access them, what pre-positioning is needed and what governance applies.

The key word is readiness. A facility that exists legally but has never been tested operationally may not be reliable in crisis. Collateral may need to be pre-positioned. Documentation may need to be in place. Systems may need to generate instructions. Treasury and operations staff must know the process. Senior management must know the reputational and signalling considerations.

Using central-bank liquidity may be normal in some frameworks and exceptional in others depending on the facility and market. The chapter should therefore avoid simplistic language such as “central bank funding is bad.” The accurate view is that central-bank access is a tool, but reliance, timing, disclosure, stigma and policy context matter.

7. Dashboard design for liquidity management

A liquidity dashboard should support action. It should not only show green ratios. A good dashboard shows current cash, projected cash, HQLA, encumbrance, LCR, NSFR, survival horizon, intraday position, top deposit movements, wholesale maturities, collateral calls, currency gaps, legal-entity gaps, limit utilisation, trigger status, source completeness and data freshness.

Each metric should have drill-down. If LCR falls, users should see whether the driver is HQLA movement, deposit outflow, committed facility drawdown, derivative collateral or data issue. If intraday cash is tight, users should see which payment flows or delayed receipts are driving the pressure. If HQLA changes, users should see whether assets became encumbered, sold, matured or reclassified.

Dashboard colours should be meaningful. Green should not mean “no one looked deeply.” Green should mean data is complete, assumptions are current and metrics are within appetite. Amber should indicate early warning. Red should indicate breach or urgent action. Grey should indicate missing or unreliable data. Missing data should never silently appear as green.

For BA work, dashboard requirements should include metric definition, source, refresh frequency, owner, threshold, escalation rule, drill-down path, export evidence, historical trend and reconciliation point.

8. Exception ownership and liquidity workflow

Liquidity breaks should have owners. A late source feed, stale balance, unexplained forecast variance, missing encumbrance flag, failed settlement, margin dispute, large unforecast payment, LCR trigger breach or currency gap should not sit in an inbox without accountability.

A practical workflow records issue type, severity, metric impacted, source system, owner, raised time, expected resolution, temporary workaround, approval, closure evidence and lessons learned. The workflow should distinguish between data issue, operational issue, market issue and true liquidity pressure. Each type needs different action.

Exception ageing matters. A one-day data issue may be tolerable with manual control. A repeated data issue is a system weakness. A recurring forecast variance indicates poor assumptions. A repeated large payment surprise indicates weak communication with business or payment operations. Treasury should learn from exceptions.

9. BA and testing scenarios to add to Topic 2

Test an unexpectedly early top depositor withdrawal. The deposit leaves earlier than forecast. The dashboard should update cash, concentration, LCR impact and forecast variance. The workflow should assign owner and capture explanation.

Test a delayed incoming payment. Expected cash at 10:00 arrives at 15:00. Intraday liquidity should show pressure. Payment operations status should feed Treasury. The end-of-day report may still be green, but the intraday stress must be visible.

Test collateral encumbrance. A security moves from unencumbered to pledged in repo. HQLA usability should change. Available buffer should update. Reports should not double count the asset.

Test a wrong legal entity mapping. Liquidity appears in the group view but belongs to a different entity. Entity-level report should show the constraint. Consolidated comfort should not hide local deficit.

Test currency mismatch. EUR surplus and USD deficit should not net away silently. The report should show FX swap dependency or funding action needed.

Test product launch mapping. A new deposit product should feed correct product code, customer segment, withdrawal rule, currency, legal entity and behavioural assumption. If fields are missing, the report should flag exception.

Test FTP change. A new FTP curve version should apply from the correct effective date. Existing balances and new balances should be treated according to policy. Profitability and Treasury reports should reconcile.

Test CFP trigger. A survival horizon falls below early-warning level. The system should notify the right owner, record escalation and show available actions. A trigger without workflow is weak control.

9.1 Cash-flow identity and acceptance evidence

Use one cash-flow identity per economic leg and preserve the identifiers from the booking, payment and statement systems. A trade amended from 50 to 60 must replace the expected 50 with 60, not create a new 60 alongside it. A cancelled instruction should remove the unexecuted expected payment only after validating cancellation state; it must not erase cash that has already settled. A returned payment creates a new incoming flow and does not undo history. A recalled payment remains uncertain until the scheme/counterparty outcome is known.

This is an acceptance pattern, not a prescribed architecture. A bank may implement batch or event ingestion, but it must explain how expected and actual cash avoid double counting and how a corrected report remains reproducible.

10. Final addendum takeaway

The missing practical layer in many liquidity chapters is the operating logic. Real liquidity management is not only LCR, NSFR and HQLA definitions. It is source collection, classification, projection, limit comparison, action, evidence and learning. It is the ability to see cash in the right currency, entity, account and time. It is the ability to separate confirmed cash from expected cash. It is the ability to know whether a buffer is usable, whether collateral is encumbered, whether a product creates hidden optionality and whether a dashboard is trustworthy under pressure.

For a banking BA, tester, developer or operations analyst, the standard is clear: do not only ask whether the liquidity report runs. Ask whether the bank can act from it. If the answer is yes, the system is useful. If the answer is no, the report is only decoration.

Source anchors

• BIS Basel Committee - Principles for Sound Liquidity Risk Management and Supervision
• BIS Basel Committee - Basel III Liquidity Coverage Ratio and liquidity risk monitoring tools
• BIS Basel Committee - Basel III Net Stable Funding Ratio
• BIS Basel Framework - LCR high-quality liquid assets
• BIS Basel Framework - SRP31 Interest rate risk in the banking book