Money Markets

How banks place and raise short-term cash, manage collateral, settle transactions and handle funding exceptions. All numerical cases are fictional training examples; contract terms and local rules govern real transactions.

1. Why money markets matter to a bank

Every day a bank ends with a surplus or a deficit of cash relative to what it needs to meet payments and maintain policy balances. That residual does not wait for the bond market’s long end or for a quarterly funding plan. It must be placed or covered in the short-term markets - often overnight, sometimes out to a few weeks or months.

Money markets are where that happens.

In plain language: money markets are the set of markets and instruments used to borrow and lend cash for short periods, secured or unsecured, between banks, other financial institutions, large corporates, and official sector counterparts. For bank treasury, they are the tactical engine of liquidity management. For markets desks, they are also a client and trading franchise. For the banking system as a whole, they are part of how monetary policy transmits into the cost of short-term money.

If Bank Treasury & Liquidity Management taught you why survival cash matters, Money Markets teaches you the instruments and market practices used to move that cash every day.

This chapter is written for people who will sit near a cash desk, a funding desk, a rates desk, or an independent risk function that oversees short-term markets risk. It is not a school summary of definitions. It is operating knowledge.

Learning objectives

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

• Explain what money markets are for in a bank’s daily life
• Distinguish unsecured interbank from secured (repo) funding and placement
• Describe how a treasury desk uses money markets to square the cash position
• Explain the main risks: credit, liquidity, operational, and market risk on short-term positions
• Read a simple short-term funding and placement stack
• Understand why money markets can freeze or become one-way in stress
• Connect money-market activity to LCR, buffers, and the funding ladder
• Recognise common mistakes in short-term funding dependence

2. What “money markets” means in banking practice

Money markets are short-term by convention. Tenors typically run from overnight to one year, with the heaviest activity at the very short end - overnight, tom-next, spot-next, one week, one month, three months.

The economic function is the redistribution of cash between parties who have temporary surplus and parties who have temporary need. A bank with excess reserves or surplus customer deposits may lend short-term. A bank with loan drawdowns or deposit outflows may borrow short-term. A money-market fund, a corporate treasurer, or an official institution may sit on either side depending on the day.

Instruments vary by market, but the practical families a bank treasury cares about are:

• Unsecured interbank deposits and loans
• Repurchase agreements (repo) and reverse repo
• Short-term bank paper (certificates of deposit and similar)
• Short-term corporate or public paper where the bank is an investor or conduit
• Central-bank facilities and open-market operations that set the floor and ceiling for short-term rates

The exact menu depends on jurisdiction. The operating logic does not: move cash safely, inside limits, at a sensible price, without creating a worse problem tomorrow.

3. Unsecured interbank markets

3.1 The basic transaction

Bank A lends cash to Bank B for a stated term at a stated rate, unsecured. Bank A takes credit risk on Bank B. Bank B gets cash and pays interest. At maturity, principal and interest are repaid.

This is the oldest and simplest form of interbank money. It is also the form that disappears first when confidence falls, because there is no collateral to reduce credit exposure.

3.2 Why unsecured capacity is fragile

Unsecured lines are based on credit appetite, relationship, and perceived soundness. In calm markets, large banks trade actively. In stress, credit lines are cut, shortened, or cancelled. A bank that funds a structural need with continuous unsecured overnight borrowing is one confidence event away from a funding crisis.

Practical rule: use unsecured interbank for tactical residual management inside limits, not as the permanent foundation of the balance sheet.

3.3 Limits and name risk

Every unsecured placement is a credit decision. Treasury and credit risk must agree limits by counterparty, tenor, and product. Name concentration matters: placing surplus with one attractive-looking counterparty is not diversification. On the start date, the lender pays cash and acquires an unsecured claim; repayment occurs later. This intentional term credit exposure is not eliminated by payment-versus-payment (PvP), which links two currency payments in FX settlement. Control unsecured deposits through authorised credit exposure, independently validated instructions, cash-status monitoring and maturity collection. A failed repayment is both a credit event and a funding problem for the lender.

3.4 Pricing

Unsecured rates embed credit risk, liquidity, and the policy rate environment. In stress, credit spreads in interbank unsecured markets can gap wider even when official policy rates are stable. That is a signal, not noise.

4. Repo and reverse repo - the secured backbone

4.1 Repo from the cash borrower’s view

In a classic repurchase agreement, the cash borrower sells securities and agrees to repurchase equivalent securities at an agreed price on the end date. At a positive repo rate the repurchase price is higher; at a negative rate it can be lower. Economically this is secured financing, but legal title, reuse rights, income payments and accounting follow the agreement and applicable law. The price difference reflects repo interest, not a mark-up on the security's market price.

Repo is the workhorse of secured short-term funding. It reduces credit risk relative to unsecured borrowing, but it introduces collateral, margin, and operational complexity.

4.2 Reverse repo from the cash lender’s view

In a reverse repo, the bank buys a security and agrees to sell it back later. Economically, the bank has lent cash against collateral. This is a primary way to place surplus cash with credit protection, subject to haircut and collateral eligibility rules.

4.3 Why haircuts matter

A haircut can reduce the cash advanced below the collateral's market value. Zero haircuts also exist in some arrangements; overcollateralisation is not guaranteed by the product name. A haircut creates an over-collateralisation buffer against price moves. In volatile markets, haircuts can rise, which means the same securities raise less cash. Capacity that looked large in calm markets can shrink exactly when needed.

4.4 General collateral versus specials

General collateral (GC) refers to a class of securities acceptable at a common rate. Special repo occurs when a specific security is in high demand to borrow, and the cash lender accepts a lower rate (sometimes much lower) for the privilege of receiving that security. Cash desks care about GC for funding and placement. Securities financing desks also care about specials for inventory and short covering.

4.5 Open repo, term repo, and roll risk

An overnight repo has a contractual end date on the next applicable settlement business day; a new trade is required to refinance it. An open repo has no fixed end date and continues until terminated under agreed notice terms, with rate changes subject to agreement. A term repo has a fixed end date, but early-termination rights, floating rates, margin obligations or break clauses can still matter. A book that depends on continuous overnight repo of the same collateral faces roll risk: the risk that capacity disappears or haircuts jump on a bad morning. Funding ladders should include secured roll risk, not only unsecured maturities.

4.6 Encumbrance

Securities used in repo are encumbered. They cannot be simultaneously treated as the bank's freely available buffer. Release at maturity depends on completed return settlement, not merely a scheduled end date. Received cash and collateral treatment must be assessed separately under the relevant liquidity rules. Liquidity managers who forget encumbrance double-count capacity. Unencumbered HQLA is the number that matters for survival actions.

5. Short-term bank paper and wholesale certificates

Banks issue short-term paper - certificates of deposit and similar instruments - to raise wholesale funds from institutional investors, money-market funds, and corporates. Tenors may run from weeks to months.

5.1 Role in the funding stack

Short-term paper can diversify funding away from pure interbank and can reach a different investor base. It is still wholesale and often confidence-sensitive. A large maturity wall of paper is a cliff like any other.

5.2 Investor behaviour in stress

Money-market investors can shorten, exit, or demand higher yields quickly when credit or liquidity concerns rise. Programme capacity that was open in peacetime can close. CFP assumptions about paper issuance in stress should be conservative.

5.3 Documentation and programme maintenance

Issuance programmes need current documentation, authorised signatories, and operational settlement readiness. Capacity on a slide is not capacity in a crisis if the programme is stale.

6. Central-bank operations and the short-term rate environment

Central banks influence money-market rates through policy rates, reserve regimes, open-market operations, and standing facilities. The exact design differs by jurisdiction, but the practical consequences for bank treasury are similar.

6.1 The corridor idea

In a corridor model, standing deposit and lending rates help bound overnight market rates when eligible banks can use both facilities. In an abundant-reserves or floor-style framework, rates can trade near the administered remuneration/deposit rate. These are operating models, not a guarantee that every participant can access a floor or ceiling. Credit, collateral, balance-sheet costs and segmented access can move individual rates away from policy benchmarks.

6.2 Why treasury watches official operations

Official operations change the amount of reserves in the system, the scarcity or abundance of collateral, and the incentive to place or borrow in private markets. A desk that ignores the official calendar will misread private market tightness.

6.3 Regular, standing and contingent central-bank access

Separate regular reserve-supply operations, on-demand operational facilities and emergency assistance. Facility use can be routine or contingent according to its published purpose. For example, the Bank of England identifies Short-Term Repo and Indexed Long-Term Repo as regular sources of reserves, while its Discount Window addresses unexpected needs. Eligibility, collateral and tested access matter in every case. An internal policy must not misclassify all official borrowing as emergency funding. Bank of England operations guide.

7. How the treasury cash desk uses money markets daily

7.1 Surplus day

The desk estimates residual surplus after contractual and behavioural flows. It places cash in reverse repo, unsecured placements inside limits, short paper, or central-bank facilities according to policy. It does not chase the last basis point into an unapproved name or an excessive tenor.

7.2 Deficit day

The desk covers via unsecured borrowing inside lines, repo of designated collateral, or other approved channels. It ranks options by cost, capacity, signalling, and the residual risk created for later days. Covering today by creating a large cliff next week is not success.

7.3 Multi-day view

Even tactical money-market activity should sit inside a multi-day residual view. A string of overnight borrowings that grows each day is a structural signal, not a series of independent one-day events.

8. Credit risk, settlement risk and operational risk in money markets

Credit risk on unsecured placements and on the residual exposure in secured trades after haircuts. Liquidity risk if markets close or capacity shrinks when you need to roll. Collateral risk if haircuts rise or collateral prices fall. Operational risk in confirmations, settlement instructions, and system failures. Legal risk if documentation is incomplete or unenforceable.

Controls: limits, eligible collateral lists, haircut schedules, settlement discipline, documentation standards, and independent monitoring. Money markets are simple economically and unforgiving operationally.

9. Money markets in stress - what changes

In stress, unsecured markets shorten or close first. Repo markets may continue but with higher haircuts, narrower eligible collateral, and reduced counterparties. Central-bank operations often expand. Investors in short-term paper flee to quality or to official facilities. Basis relationships that were stable can break.

A bank’s CFP must assume degraded private money-market capacity, not peacetime maximums. Testing should include the operational steps to move from private markets to contingent secured and official options.

10. Connection to LCR, NSFR and the liquidity chapter

Liquidity treatment depends on counterparty, residual maturity, collateral, availability and the applicable rules. Unsecured financial wholesale funding can attract high LCR outflows, but secured-funding runoff is collateral-dependent: the Basel treatment differs for Level 1, Level 2 and other collateral. A reverse-repo receivable is not automatically HQLA; assess whether received collateral can legally and operationally be reused, and prevent double counting with inflows. Central-bank reserve eligibility is also conditional. Internal stress can assume less rollover than the regulatory formula. Basel LCR, paragraphs 112-115 and HQLA operational conditions.

Money-market choices are therefore not only rate decisions. They are liquidity-regulation and survival decisions.

11. Desk organisation around money markets

Typical touchpoints:

• Cash and liquidity desk: daily residual, overnight and short placements/covers
• Funding desk: term paper, investor programmes, ladder management
• Securities financing / repo desk: deeper repo, specials, collateral optimisation
• Rates desk: client short-end flow and positions permitted by its mandate, applicable law and limits
• Credit risk: counterparty limits
• Market risk: where trading positions exist
• Operations and settlements: confirmation and settlement integrity

Reporting lines vary. What matters is that defensive cash management is not subordinated to trading P&L pressure on the same short-end positions without clear book separation and limits.

12. Common mistakes in money-market usage

1. Funding structural needs with continuous overnight unsecured money
2. Ignoring repo roll risk and haircut expansion in stress
3. Double-counting encumbered securities as free HQLA
4. Placing surplus for yield into concentrated or weak names
5. Assuming paper markets remain open in a name-specific stress
6. Stale issuance programmes and settlement instructions
7. No multi-day view - only today’s square
8. Treating central-bank borrowing as ordinary funding without policy clarity

These mistakes are ordinary. Their consequences are not.

13. Worked tactical examples (illustrative)

Example A - Surplus placement choice

Residual surplus 500 for one day. Options: unsecured to a limit-approved bank at a higher rate; reverse repo against government collateral at a lower rate; leave excess at the central bank at the floor rate. Policy that prioritises survival over yield chooses collateralised or official options when credit or concentration is a concern, and documents the carry cost as insurance.

Example B - Deficit cover with collateral

Residual deficit 400. Unsecured lines are tight. Designated HQLA bills are available. Desk repos bills, notes encumbrance, and schedules the roll or unwind against the multi-day forecast. It does not repo the last unencumbered tranche without escalation if that tranche is also the CFP’s first line of defence.

Example C - Building a cliff with paper

The bank issues a large amount of one-month paper every month on the same cycle. Average maturity looks acceptable. The maturity wall each month is a repeated cliff. Strong practice staggers issuance. Weak practice hopes investors always roll.

14. Short-term rates, benchmarks and the curve’s front end

Money-market rates form the front end of the yield curve. Benchmark reform in many markets replaced older interbank fixings with robust reference rates based on actual transactions (for example overnight risk-free rates compounded in arrears for many products). Practitioners must know which benchmarks their contracts, FTP curves, and funding products reference, and how fallback language works.

For the cash desk, the practical issue is the rate paid or earned on each placement and cover. For the funding desk and ALM, the issue is how short-end benchmarks feed product pricing and hedge design. For risk, the issue is basis risk between funding instruments and hedges when benchmarks differ.

15. Client money-market business on the markets side

Beyond pure treasury, markets desks may provide short-term investment outlets for corporate and institutional clients, make markets in short-term paper, run repo as a client and interdealer business, and structure short-term solutions linked to rates.

This activity sits in the commercial Markets mandate. It requires the same credit, operational and market-risk controls as other trading businesses, plus awareness that client flows can interact with the bank’s own funding and collateral position.

16. Governance and limits specific to money markets

Useful limit types: counterparty limits by product and tenor (unsecured vs secured); aggregate short-term wholesale reliance limits; collateral concentration and eligibility lists; maximum overnight funding share of total wholesale; settlement and operational risk controls; book-level market risk limits where trading books hold short-end positions.

Breach escalation should be same-day for material overnight capacity loss or limit breaks.

17. Full practitioner expansion: the short end as the bank's daily operating layer

The money market is sometimes introduced as if it were a list of instruments. That is useful for exams, but it is too thin for a person who has to work inside a bank. In a bank, the money market is the daily operating layer between the payment system, the deposit base, the securities portfolio, the central bank account, and the wholesale investor base. It is where yesterday's customer balances meet today's payment obligations. It is where the treasury dealer learns whether a surplus is genuinely available or whether it is only an accounting balance that will disappear when large payments are released. It is where policy rates become tradeable funding costs. It is also where quiet fragility first shows up, because a bank that cannot borrow overnight, renew repo, issue short paper, or place excess cash safely has a problem that cannot be hidden by a long strategy note.

A good money-market chapter must therefore be read in two ways. The first reading is product based: what is an unsecured deposit, what is repo, what is a certificate of deposit, what is commercial paper, what is a treasury bill, what is a central-bank facility, what is an overnight benchmark rate? The second reading is balance-sheet based: what does this trade do to cash, collateral, liquidity ratios, encumbrance, counterparty exposure, settlement workload, interest-rate risk, transfer pricing, and crisis survivability? The best treasury people read both layers at the same time. They do not say, "I borrowed cheaply overnight," and stop there. They ask what the funding is supporting, whether it can be rolled, what collateral has been tied up, whether the counterparty can pull the line tomorrow, how the position behaves if policy expectations move, and whether the trade creates an operational dependency on one system, custodian, settlement window, or dealer relationship.

This is why money markets are central to Markets & Treasury. They sit below the bond desk, but they influence the bond desk. They sit beside liquidity management, but they change liquidity management. They are often lower duration than the fixed-income book, but they can create very fast losses or funding gaps if they are mishandled. They look simple because many trades mature tomorrow, but the discipline is demanding because tomorrow arrives every day. A five-year bond has price risk that can be measured over a curve. Overnight funding has renewal risk that can become binary: either the cash arrives or it does not. A short paper program has investor diversification until the investor base decides it no longer wants the name. A repo book has collateral discipline until haircuts rise, settlement fails increase, or the needed securities are no longer available in the correct place.

For a learner, the practical question is not "Which instrument is best?" The practical question is "Best for what purpose, under what constraint, and under what stress?" If the bank has temporary surplus cash, unsecured placement may look attractive because it pays a spread over a risk-free account, but it creates name exposure. Reverse repo may reduce credit risk through collateral, but it needs collateral eligibility, valuation, margining, and settlement infrastructure. Central-bank placement may be operationally simple and credit clean, but may pay less or be available only to eligible firms in eligible currency accounts. If the bank needs funding, repo may be more reliable than unsecured borrowing, but it uses collateral and may be constrained by haircuts. Term deposits can reduce rollover risk, but may cost more and may be hard to unwind. Short-term paper can diversify funding, but investor confidence can disappear in stress. The answer is always conditional.

A mature money-market desk therefore behaves less like a price-taking cashier and more like an operating nerve center. It watches balances, flows, rates, collateral, investor appetite, payments, and market tone. It knows the opening cash position but respects that the opening position is only a forecast. It distinguishes money that can be placed for a week from money that should stay overnight because client behavior is uncertain. It understands that the cheapest funding is not necessarily the best funding if it concentrates maturity risk. It speaks to liquidity risk, ALM, settlements, collateral management, finance, and front-office sales. It is commercial, but it is also defensive. The bank survives because this short-end discipline works even when the market is boring. The crisis version is simply the same discipline with less room for error.

18. The boundary between money markets and capital markets

Money markets are usually defined by tenor: borrowing and lending of cash for up to one year. Capital markets usually refer to longer-term debt and equity financing. That boundary is helpful, but bank practice treats it as a working convention, not a wall. A three-month certificate of deposit, a six-month treasury bill, and an overnight repo clearly sit in the money-market world. A two-year note clearly sits in fixed income. But a bank's one-year senior unsecured note, a 364-day committed facility, and a rolling program of one-month paper can all affect the same liquidity ladder. The treasury officer cares less about the textbook boundary and more about whether the cash is available, stable, callable, encumbered, secured, unsecured, pledged, repayable, and eligible for regulatory treatment.

The money-market boundary also matters because short-end instruments are often priced from policy expectations rather than long-run credit stories. A three-month deposit rate reflects the central-bank policy path, the bank's credit spread, term liquidity, balance-sheet constraints, and the timing of quarter-end or year-end. A ten-year bond has all of that plus duration, inflation expectations, term premium, and long-horizon investor appetite. At the short end, tiny changes in operational detail can matter. A trade settling today versus tomorrow can have a different value if the bank has a cash need today. A secured trade using general collateral may behave differently from a trade using a scarce security. A deposit callable by the investor is not the same as a firm term deposit. A repo that settles through tri-party infrastructure has different operational dependencies from a bilateral delivery-versus-payment repo.

One of the first professional upgrades for a learner is to stop treating "cash" as one thing. Cash in a central-bank account, cash expected from maturing reverse repo, cash due from a counterparty in another time zone, cash raised through a repo against encumbered securities, and customer deposits available after payment cut-off are not operationally identical. The money market converts some forms of cash availability into other forms, but not without cost and not without friction. The chapter's core discipline is to notice those frictions before they become losses.

19. The central-bank anchor

Most modern money markets sit around a central-bank operating framework. The framework differs by jurisdiction, but the principle is similar: the central bank sets the policy stance and uses facilities, reserves, open-market operations, remuneration rates, and standing tools to influence short-term market rates. In the United States, the New York Fed describes repo and reverse repo operations as tools used to help keep the federal funds rate in the target range set by the Federal Open Market Committee (https://www.newyorkfed.org/markets/desk-operations/repo). The Federal Reserve's standing repo arrangements are designed to supply liquidity to eligible counterparties against securities and help limit upward pressure on overnight money-market rates (https://www.federalreserve.gov/monetarypolicy/standing-overnight-repurchase-agreements.htm). In the euro area, the ECB publishes the euro short-term rate, or €STR, as a reference for wholesale euro unsecured overnight borrowing costs (https://www.ecb.europa.eu/stats/financial_markets_and_interest_rates/euro_short-term_rate/html/index.en.html). In sterling markets, the Bank of England administers SONIA, which reflects overnight sterling wholesale borrowing rates based on eligible transactions (https://www.bankofengland.co.uk/markets/sonia-benchmark). These are not academic details. They are part of the operating map a bank uses every morning.

The central-bank anchor does not eliminate market risk. It frames it. A bank may know the policy rate, but it still has to manage where actual trades clear. Repo rates can move differently from unsecured rates. A bank's own funding spread can widen even when the policy rate is unchanged. Quarter-end balance-sheet constraints can change dealer appetite. A scarcity premium in specific collateral can pull repo levels away from general collateral. A central-bank facility may exist but carry eligibility, collateral, operational timing, or stigma considerations. A money-market practitioner should therefore understand policy operations without assuming that policy tools guarantee effortless private-market funding.

The central bank also shapes the floor and ceiling psychology of the market. If surplus cash can be placed safely at an official rate, private counterparties must pay enough to attract it. If eligible banks can borrow against high-quality collateral at an official standing facility, private repo rates are influenced by that alternative. The details matter: eligible counterparties, eligible collateral, transaction timing, caps, minimum bids, settlement conventions, and whether the facility is automatic or discretionary. A beginner often memorizes the policy rate. A practitioner studies the plumbing around the policy rate, because the plumbing determines where the marginal trade happens.

20. Money-market instruments as operating choices

The main money-market instruments can be understood as different ways to answer the same question: who gives cash to whom, for how long, against what protection, with what exit, and with what operational burden? Unsecured interbank deposits answer: cash is lent on the borrower's credit, for an agreed tenor, with no collateral. Repo answers: cash is lent against securities, with a repurchase promise and margin protections. Certificates of deposit answer: the bank borrows from investors by issuing short-dated negotiable paper or deposit instruments. Commercial paper answers: a company or financial issuer borrows short-term through paper sold to investors. Treasury bills answer: a sovereign raises short-term funding and investors hold a highly liquid discount instrument. Central-bank tools answer: eligible institutions transact with the official sector under defined rules. FX swaps, which are treated more deeply in the Foreign Exchange chapter, often answer a multi-currency version of the same problem: one currency's cash is exchanged for another currency's cash over a short tenor.

Instrument choice changes the bank's risk profile. A bank that borrows unsecured overnight has no collateral drag but heavy confidence reliance. A bank that borrows through repo has more collateral protection for the lender but consumes securities and may create encumbrance. A bank that issues three-month paper reduces tomorrow's rollover need but becomes dependent on investor appetite at maturity. A bank that parks money in treasury bills may improve liquidity quality but accepts market price movement if sold before maturity. A bank that relies on central-bank facilities as normal funding may face internal governance questions even when the facility is legitimate. The instrument is never just a rate. It is a package of cash timing, legal terms, regulatory effect, operational steps, market convention, and contingency behavior.

21. Unsecured interbank deposits in depth

Unsecured interbank trading looks clean on the ticket: counterparty, currency, amount, value date, maturity date, rate, day-count convention, settlement instructions. Under that clean ticket sits a serious credit decision. The lender is exposed to the borrower's ability and willingness to repay at maturity. Because there is no collateral, the lender's comfort comes from counterparty credit assessment, limits, relationship knowledge, tenor discipline, and market confidence. That is why unsecured markets can be liquid in normal conditions and suddenly thin in stress. The same transaction that feels routine at 9:00 a.m. in calm markets can feel unattractive at 9:00 a.m. after a credit headline.

The unsecured desk must therefore work inside a limit framework that is more detailed than a single total number. A proper framework distinguishes counterparty name, legal entity, currency, product, tenor, settlement exposure, and group aggregation. It may permit a large overnight exposure but only a smaller one-month exposure. It may allow placements with a parent bank but not a weaker branch or subsidiary. It may reduce limits around rating actions, CDS widening, deposit outflow stories, sanctions risk, or operational failures. The desk should not treat unused credit limit as an invitation to use it. The better question is whether the placement is necessary, fairly priced, diversified, and consistent with the bank's defensive posture.

The borrower's view is different. Unsecured interbank funding is attractive because it does not require collateral. It can be fast, relationship driven, and flexible. But that attractiveness is exactly why risk managers dislike structural dependence on it. If the bank needs the same unsecured overnight borrowing every day to fund stable assets, the bank has effectively built a maturity mismatch on confidence. In a benign market, it may roll for months. In a stress event, the line may be cut before management has time to execute a calm replacement. A bank can use unsecured markets tactically; it should not build its survival plan on the assumption that unsecured markets will remain open on the worst day.

22. Repo as secured money

Repo is often described as a sale and repurchase of securities, economically equivalent to secured borrowing. One party sells securities for cash and agrees to repurchase equivalent securities at an agreed later price, reflecting the agreed repo rate, which may be positive or negative. The price difference represents the repo interest. For the cash lender, the transaction is a reverse repo. The lender provides cash and receives securities as collateral. For the cash borrower, the transaction is a repo. The borrower receives cash and provides securities. Legal form, accounting treatment, and regulatory treatment can vary by jurisdiction and documentation, so practitioners must know the local rulebook rather than relying only on the economic shorthand.

Repo is attractive because collateral reduces counterparty credit exposure. It does not remove risk. Collateral must be eligible, valued, delivered, monitored, and returned. If collateral falls in value, the cash lender may demand margin. If collateral becomes scarce, the economics of the trade can change. If settlement fails, the cash and securities legs may not move as expected. If the counterparty defaults, the non-defaulting party must rely on close-out, netting, collateral sale, and legal enforceability. This is why repo documentation and market practice matter. ICMA's European repo best practice guide exists precisely because orderly trading and settlement conventions reduce disputes and delays (https://www.icmagroup.org/market-practice-and-regulatory-policy/repo-and-collateral-markets/icma-ercc-publications/icma-ercc-guide-to-best-practice-in-the-european-repo-market/). BIS work on repo market functioning also stresses that repo supports the flow of cash and securities across the financial system, while excessive dependence and procyclical margining can amplify stress (https://www.bis.org/publ/cgfs59.pdf).

The haircut is one of the most important repo concepts. If a borrower provides securities worth 100 and receives only 98 of cash, the 2 percent difference is a haircut. The haircut protects the cash lender against market value changes, liquidation costs, and counterparty default timing. Haircuts are low for very high-quality, liquid government collateral and higher for riskier or less liquid collateral. In stress, haircuts may rise because lenders become less confident that collateral can be sold quickly at fair value. A rising haircut can create a liquidity squeeze even when the borrower still owns securities. The borrower must either provide more collateral for the same cash or borrow less cash against the same portfolio.

23. General collateral and specials

Repo markets distinguish general collateral from special collateral. General collateral repo is mainly about financing cash against a broad class of acceptable securities, such as eligible government bonds. The specific security is less important as long as it meets the collateral schedule. Specials are different. A specific security may be in high demand because it is needed for settlement, short covering, delivery into a futures contract, relative-value trading, or collateral transformation. In a special, the security itself drives the economics. The cash rate may be unusually low because the cash lender is effectively paying to borrow the scarce security.

For a bank treasury desk, the general-collateral market is usually more relevant for daily funding. For a securities financing desk, specials can be a revenue source and a market-making need. But even a treasury learner should understand specials because they reveal that repo is not just cash funding. It is also securities funding. The money market moves both cash and collateral. A bank may have cash but need a security; another bank may have the security but need cash. Repo and securities lending connect those needs. In stress, this two-sided nature becomes critical because collateral location, eligibility, and scarcity can matter as much as the headline cash balance.

The distinction also matters for control. A treasury repo used for funding should not quietly become a specials trading strategy without the right risk ownership. A specials book may carry basis risk, settlement risk, short-covering risk, and client franchise obligations. It requires different expertise from a basic cash-funding operation. Strong organizations make the purpose clear in desk mandates, product approval, limits, and reporting.

24. Tri-party, bilateral, and cleared repo

Repo can be arranged in several operating models. In bilateral repo, the two counterparties agree the trade and manage collateral and settlement directly, often through custodians and securities settlement systems. In tri-party repo, a third-party agent supports collateral management, valuation, substitution, and settlement mechanics. In centrally cleared repo, a central counterparty becomes the buyer to every seller and the seller to every buyer, subject to membership, margin, default fund, and rulebook requirements. Each model changes operational burden, counterparty exposure, netting, settlement timing, and liquidity usage.

Tri-party infrastructure can reduce operational friction, especially for broad collateral pools. It can make collateral allocation and margining more standardized. But it also creates dependency on the agent's timelines, eligibility logic, and operational processes. Bilateral repo can be flexible and precise, especially for specific securities, but it requires stronger internal collateral and settlement capability. Cleared repo can reduce bilateral counterparty exposure and create netting benefits, but it introduces central counterparty rules, margin requirements, membership or sponsored access arrangements, and potential concentration in market infrastructure. BIS work on repo clearing and settlement notes that arrangements differ across markets and that the crisis experience showed the importance of robust infrastructure (https://www.bis.org/cpmi/publ/d91.htm).

A practitioner should not describe repo as safe simply because it is secured. Safety depends on collateral quality, legal enforceability, operational settlement, valuation, haircut policy, counterparty behavior, and market liquidity. The repo form is a risk mitigant, not a magic shield.

25. Short-term bank paper

Banks often fund themselves through short-term paper such as certificates of deposit, commercial paper, Euro commercial paper, or local equivalents. These instruments allow the bank to borrow from a broader investor base than the interbank market. Investors may include money-market funds, corporates, asset managers, insurers, official institutions, and other banks. The attraction for the issuer is diversification and tenor. Instead of relying only on overnight interbank borrowing, the bank can issue one-week, one-month, three-month, six-month, or similar maturities depending on market appetite and program terms.

The hidden risk is that short-term paper can feel stable until it is not. Investor demand may be strong when the issuer is perceived as safe and the market has ample liquidity. If the issuer's credit story weakens, if the sector comes under pressure, or if money-market funds become defensive, issuance windows can close quickly. Maturing paper still has to be repaid. If the bank has not built a maturity ladder, back-up liquidity, and diversified investors, a closed paper market can become a near-term cash problem. This is why treasury teams track not only the amount outstanding but also maturity concentration, investor type, dealer distribution, currency mix, and the proportion of paper that must be rolled in the next several days.

Short-term paper also connects to internal transfer pricing. If business lines benefit from funding that treasury raises through paper, the cost should be reflected in FTP so products are not priced as if overnight money is free and infinite. The rate paid on paper is not just a treasury desk outcome. It is an input into product economics, balance-sheet growth, and ALCO decisions.

26. Treasury bills and sovereign short instruments

Treasury bills are short-term sovereign instruments usually issued at a discount and redeemed at par. They are a core money-market asset for many investors because they combine short maturity, high credit quality in major sovereign markets, and strong secondary-market liquidity. For banks, bills can be part of the liquidity portfolio, collateral pool, trading inventory, or client market-making franchise. They can also be a safe parking place for surplus cash when yields and liquidity treatment are attractive.

A bill is not identical to cash. It has price, settlement, and liquidity characteristics. If sold before maturity, the price depends on the current short-end yield and market liquidity. In major markets the price risk may be modest, but it is not zero. The bank must also consider whether the bill is unencumbered and operationally available. A bill pledged in repo is not free HQLA for another use. A bill held in the wrong legal entity, custodian, or currency may not solve the immediate cash need. The treasury learner should think in layers: credit quality, market liquidity, settlement availability, regulatory eligibility, and internal transferability.

Bills also influence money-market rates because they compete for cash. When bill supply rises, investors may demand higher yields, which can affect repo and unsecured pricing. When bill supply is scarce, cash may flow into repo or central-bank facilities. This is one reason money-market desks follow government cash management and bill issuance calendars closely. Sovereign issuance is not background noise. It changes the short-end investment menu.

27. Benchmarks after LIBOR

Modern money markets are heavily shaped by overnight risk-free or near risk-free reference rates. SOFR in US dollars, €STR in euros, and SONIA in sterling are prominent examples. The New York Fed describes SOFR as based on transaction-level repo data and publishes it each business day (https://www.newyorkfed.org/markets/reference-rates/sofr). The ECB publishes €STR based on eligible euro unsecured overnight borrowing transactions (https://www.ecb.europa.eu/stats/financial_markets_and_interest_rates/euro_short-term_rate/html/index.en.html). The Bank of England administers SONIA and describes it as based on actual overnight sterling wholesale transactions (https://www.bankofengland.co.uk/markets/sonia-benchmark). The common lesson is that benchmark design has moved toward rates grounded in observable overnight transactions rather than expert judgment about unsecured term borrowing.

For a bank, this matters in three practical ways. First, money-market trades supply the data and economic reality behind benchmarks. Second, benchmarks influence derivatives, loans, floating-rate notes, FTP curves, and valuation systems. Third, the difference between secured and unsecured rates can reveal market conditions. SOFR is secured by Treasury collateral, while €STR and SONIA reflect unsecured overnight wholesale borrowing in their respective designs. The spread between secured and unsecured rates is not fixed by nature. It changes with collateral scarcity, balance-sheet constraints, credit concerns, central-bank operations, and investor behavior.

The post-LIBOR world also requires operational discipline. Systems must know whether a trade references an overnight rate compounded in arrears, a term rate, a simple average, or a fixed money-market rate. Accruals, reset dates, observation lags, lookbacks, payment dates, and fallback terms must be accurate. A money-market chapter cannot cover all benchmark mechanics, but it must make the learner alert: short-end rates are not merely quotes on a screen. They are legal, operational, valuation, and conduct objects.

28. The daily cash position

The money-market day begins before the market fully wakes up. Treasury needs an opening cash position by currency and legal entity. That position is built from yesterday's closing balances, known maturities, expected customer flows, loan drawdowns, securities settlements, derivative collateral calls, wholesale maturities, central-bank account movements, fees, tax payments, and operational adjustments. The first number is rarely perfect. It is a working estimate that improves as payments are confirmed and unexpected flows arrive.

A strong desk separates confirmed flows from forecast flows. A maturing repo is more reliable if settlement is confirmed and collateral is in place. A large corporate deposit outflow is less reliable if the client has only indicated intent. A securities sale due today may still fail. A derivative margin call may change after market moves. A payment file may be delayed. The cash position should therefore show confidence levels, not only totals. The question is not "What is the balance?" The question is "How much of this balance can be treated as available for placement, for how long, without creating a shortfall if the uncertain flows move against us?"

The desk then decides whether to borrow, lend, roll, repay, or hold. That decision is influenced by rate levels, risk appetite, liquidity buffer targets, intraday needs, expected next-day flows, reserve requirements where relevant, regulatory metrics, ALCO guidance, and market tone. A junior person may see a surplus and ask where to place it at the best rate. A practitioner asks how much of the surplus is sticky enough to place beyond overnight, which counterparties are available, whether collateralized placement is better than unsecured placement, and whether the bank should deliberately hold more cash because the next two days are uncertain.

29. Intraday liquidity and payment timing

Money markets are often discussed by maturity date, but banks also live inside the day. Intraday liquidity is the ability to make payments as they fall due during the business day, not only the ability to end the day with a positive balance. A bank can be solvent and still create a serious operational problem if it cannot release payments on time. Payment systems, securities settlement systems, foreign-exchange settlement, clearing houses, and correspondent banking arrangements all demand cash or collateral at specific moments.

The money-market desk must therefore coordinate with payments and operations. For a cash lender, a reverse repo maturing today is an expected cash inflow against returning equivalent securities. For the cash borrower, repo maturity is an outflow against receiving the securities back. Under delivery-versus-payment (DvP), linked final transfers reduce principal risk; they do not guarantee timely settlement or eliminate replacement-cost and liquidity risks. A large outgoing payment may leave early. A central-bank facility may settle at a defined time. A securities purchase may consume cash before a sale proceeds. A derivative margin call may be due before noon. End-of-day cash is not enough if the bank misses intraday obligations. The more complex the bank, the more important the timing grid becomes.

This is also where collateral and liquidity meet. Intraday credit may require collateral. Payment-system access may depend on eligible assets. Securities settlement may need pre-positioned securities. A treasury team that thinks only in rates can miss these dependencies. A treasury team that understands money-market plumbing knows that good liquidity management is partly choreography: get the right cash, to the right account, in the right currency, against the right collateral, before the relevant cut-off.

30. Practitioner scenarios and desk judgement

Use each scenario as an incident or decision exercise. Identify the currency and entity, cash deadline, confirmed versus forecast flows, counterparty or collateral constraint, accountable owner and evidence of closure. The individual cases below focus on the additional judgement required by that specific event.

30.1. Overnight surplus placement

The desk sees surplus cash after client inflows.

The common mistake is placing the full amount for one week because the quoted rate is slightly better. The better control is to split the balance between central-bank placement, reverse repo, and short unsecured placements based on forecast confidence.

Evidence: Treasury retains the client-flow forecast, placement maturity, usable intraday buffer and approved counterparty exposure. A forecast outflow at 09:00 tomorrow cannot be covered by placement proceeds expected at 15:00.

30.2. Unexpected corporate outflow

A large operating account leaves earlier than forecast.

The common mistake is funding the gap with whatever overnight line is cheapest without checking tomorrow's maturities. The better control is to treat the flow as a forecast error, revise the ladder, and check whether the event changes deposit behavioral assumptions.

Operations confirms whether the payment settled or remains queued. Treasury changes actual cash only for the debit that occurred, includes queued obligations in its forward ladder, and records the forecast variance against the deposit segment.

30.3. Repo haircut increase

A lender raises haircuts on a collateral class.

The common mistake is calling the position funded because nominal collateral value is still high. The better control is to measure the cash actually available after haircut, stress the next margin call, and review substitute collateral.

For new or rolled cash funding, collateral worth 100 million supports 98 million at a 2% haircut and 95 million at 5%. An existing contract is governed by its margin and amendment terms; a market haircut change is not automatically a unilateral change to that contract.

30.4. Specific collateral demand

A security becomes special because dealers need it for delivery.

The common mistake is booking it as routine cash investment income. The better control is to separate specials activity from cash management and report securities financing purpose, settlement risk, and recall risk.

Compare the special repo rate with GC in the same currency, tenor and settlement basis. Confirm whether the contract allows early termination or substitution; a right to recall securities must not be assumed for every fixed-term repo.

30.5. Quarter-end balance-sheet pressure

Dealers reduce repo balance-sheet capacity near reporting dates.

The common mistake is assuming yesterday's quotes are still executable size. The better control is to pre-fund earlier, diversify counterparties, and include quarter-end dates in the funding ladder.

30.6. Short paper maturity wall

A large amount of three-month paper matures in the same week.

The common mistake is celebrating low historical issuance cost while ignoring the roll concentration. The better control is to spread future maturities and maintain back-up liquidity sized to a failed issuance window.

30.7. Central-bank facility use

Private repo rates jump above the official standing facility rate.

The common mistake is avoiding the facility for image reasons without quantifying the cost. The better control is to escalate transparently, test operational eligibility, and compare facility use with market alternatives.

30.8. Benchmark mismatch

A system accrues a compounded overnight rate incorrectly.

The common mistake is treating benchmark selection as a legal-team problem only. The better control is to reconcile trade economics, confirmation language, benchmark calendar, and system accrual logic before go-live.

Finance reproduces the interest using the agreed observation dates, day weights and rounding. Operations agrees any economic correction with the counterparty; a technology repair must preserve the old calculation and authorised correcting event.

30.9. Failed securities settlement

The cash leg is ready but collateral delivery fails.

The common mistake is ignoring the fail because the economics are small. The better control is to escalate settlement fails, track repeated counterparties, and include fail costs in desk performance review.

In a linked DvP start-leg fail, neither final cash nor final securities transfer should be assumed. Treasury removes the expected funding receipt from available cash and arranges cover. Operations records the matching or inventory cause, any partial settlement and the next viable settlement window.

30.10. Wrong legal entity

A group entity has excess cash while another has shortage.

The common mistake is netting the group position mentally without checking transfer restrictions. The better control is to manage by legal entity first, then evaluate whether treasury can move funds under documented arrangements.

Capture both entities, accounts, transfer authority and intercompany terms. A branch and a subsidiary are not interchangeable legal concepts; local restrictions and settlement-account access must be assessed before instructing a transfer.

30.11. Currency mismatch

Dollar funding is comfortable but euro cash is short.

The common mistake is assuming total group liquidity solves a currency-specific need. The better control is to run currency ladders separately and account for FX swap capacity, settlement timing, and convertibility limits.

Show each FX-swap near and far leg separately. The near leg funds today; the far leg creates a later repayment obligation and requires a viable refinancing or natural cash source.

30.12. Unsecured line withdrawal

A counterparty reduces the bank's overnight line.

The common mistake is looking only at today's replacement rate. The better control is to review all name limits, market signals, investor feedback, and stress scenarios because the withdrawal may be an early warning.

30.13. Collateral concentration

Most repo funding uses the same sovereign collateral pool.

The common mistake is calling it diversified because there are many repo counterparties. The better control is to diversify both counterparties and collateral sources, and test what happens if one collateral class becomes less financeable.

30.14. Open repo complacency

An open repo has rolled for months without issue.

The common mistake is treating it as permanent funding. The better control is to report it as callable short funding and stress immediate termination.

30.15. Tri-party dependency

Collateral allocation depends on one agent's cut-off and data feed.

The common mistake is assuming operational convenience means operational certainty. The better control is to map cut-offs, substitution rules, data controls, and contingency procedures for agent disruption.

Reconcile the agent allocation file to custody inventory and the collateral subledger. A received allocation message does not prove settlement; use the agent's final status and account evidence.

30.16. Client franchise pressure

Sales wants better rates for a large client placement.

The common mistake is accepting uneconomic funding to protect the relationship. The better control is to price client balances through FTP and document exceptions with business-owner accountability.

30.17. Money-market fund behavior

Fund investors become defensive after sector headlines.

The common mistake is assuming fund capacity renews because documents allow investment. The better control is to track actual appetite by fund, tenor, and rating constraint rather than treating eligibility as demand.

30.18. Central-bank reserve abundance

System liquidity is high and rates are stable.

The common mistake is letting discipline fade because cash is easy. The better control is to maintain stress ladders, collateral testing, and roll analysis because abundant reserves can decline.

30.19. Reserve scarcity

System liquidity tightens and repo rates become jumpy.

The common mistake is assuming a higher paid rate always solves funding. The better control is to recognize balance-sheet and collateral constraints; rate is only one part of capacity.

30.20. Dealer balance-sheet constraint

A dealer quotes wide or refuses term repo.

The common mistake is blaming relationship coverage only. The better control is to consider regulatory balance-sheet costs, quarter-end, netting, client flow, and inventory pressure.

30.21. Collateral upgrade trade

A desk wants to exchange lower-quality collateral for HQLA temporarily.

The common mistake is treating it as simple liquidity improvement. The better control is to measure cost, encumbrance, maturity, legal ownership, and unwind risk.

30.22. Same-day maturity cliff

Several overnight trades mature before replacement cash settles.

The common mistake is looking at end-of-day totals only. The better control is to build an intraday ladder and pre-arrange replacement funding before the earliest maturity window.

30.23. Rate chasing

The desk maximizes yield on surplus cash.

The common mistake is ignoring credit, tenor, and operational friction. The better control is to set a hierarchy: safety, liquidity, regulatory effect, diversification, then yield.

30.24. Operational static data error

A settlement instruction is wrong for a counterparty.

The common mistake is calling it a back-office detail. The better control is to treat static data as front-to-back risk because a wrong SSI can turn a good trade into a failed payment.

Use independently verified standard settlement instructions (SSI), effective dates and dual approval for changes. Repair the affected instruction with a linked version and check whether it was accepted or settled before retransmitting.

30.25. Documentation gap

A repo counterparty is active but annex terms are unclear.

The common mistake is relying on relationship memory. The better control is to keep enforceable master agreements, collateral schedules, margin terms, and close-out rights current.

30.26. HQLA encumbrance

The bank funds cheaply by pledging liquid securities.

The common mistake is counting the same securities as freely available buffer. The better control is to report encumbered and unencumbered HQLA separately and trace release to actual settlement. See the Basel LCR source.

The collateral team owns reservation and release status. Regulatory Reporting reconciles the HQLA stock and related contractual flows without treating the same security as both freely available and already committed to funding.

30.27. Stable funding illusion

Short assets are funded by rolling overnight liabilities.

The common mistake is calling it matched because average maturity looks short. The better control is to assess contractual and behavioral rollover risk and connect the position to the NSFR stable-funding framework, using the applicable funding and asset categories.

Contractual overnight liabilities require renewal even when the financed asset matures soon. Report the actual maturity and counterparty; an average tenor cannot represent the largest single-day funding need.

30.28. Liquidity buffer monetization

Treasury considers using bills or bonds for cash.

The common mistake is waiting until the stress day to test execution. The better control is to pre-test sale and repo channels, collateral eligibility, settlement timing, and haircuts.

30.29. Wrong day-count convention

A deposit trade has a rate but incorrect interest calculation.

The common mistake is assuming the difference is immaterial. The better control is to validate currency convention, holidays, maturity date, and accrual before confirmation.

On 100 million at 5.25% for 3 calendar days, Actual/360 interest is 43,750.00; Actual/365 Fixed is 43,150.68. The 599.32 difference is a reproducible cash claim, not a display-only rounding issue.

30.30. Holiday calendar miss

A tom-next trade crosses a holiday in one market.

The common mistake is settling based on a generic date rule. The better control is to use currency and market calendars, especially around year-end and multi-currency funding.

Record the original contractual dates, adjusted settlement dates and resulting interest period. A next-business-day maturity may extend actual accrual; never change both date and interest using an undocumented generic holiday rule.

30.31. Margin call spiral

Market moves create collateral calls across derivatives and repo.

The common mistake is treating margin as separate from money-market funding. The better control is to include expected and stressed margin calls in the same daily cash and collateral dashboard.

30.32. Client deposit volatility

A large non-operational deposit enters at month-end.

The common mistake is treating it as stable franchise funding. The better control is to apply behavioral assumptions, concentration monitoring, and FTP that reflects likely outflow risk.

30.33. Secured funding overconfidence

Repo capacity exists in normal markets.

The common mistake is assuming collateralized funding cannot disappear. The better control is to stress counterparty appetite, haircut changes, settlement fails, and collateral eligibility changes.

30.34. Liquidity hoarding

The desk refuses all placements in mild stress.

The common mistake is freezing commercial activity without escalation. The better control is to define stress modes and approved placement hierarchy so defense is disciplined rather than panicked.

30.35. Cross-border restrictions

Cash is trapped in a jurisdiction during stress.

The common mistake is counting group surplus as universally available. The better control is to apply legal-entity and jurisdictional transfer limits before relying on group liquidity.

30.36. Market rumor

A peer bank headline moves short-term funding sentiment.

The common mistake is waiting for rating agencies to act. The better control is to escalate market intelligence early and test whether counterparties shorten tenors or ask for collateral.

30.37. System outage

Treasury management system is unavailable near cut-off.

The common mistake is improvising trades without audit trail. The better control is to maintain fallback tickets, maker-checker controls, and reconciliation procedures.

A fallback trade log needs unique IDs, approved economics, external confirmation and actual settlement evidence. When service returns, import once and reconcile ticket counts and amounts before releasing queued messages.

30.38. Wrong collateral valuation

A stale price overstates collateral coverage.

The common mistake is assuming high-quality collateral never gaps. The better control is to use current prices, independent valuation, concentration add-ons, and dispute processes.

30.39. Securities lending confusion

A securities loan is booked beside repo but reported differently.

The common mistake is aggregating products without understanding legal form. The better control is to map product taxonomy, accounting treatment, collateral type, and liquidity treatment.

30.40. Internal liquidity transfer

One desk wants to use another desk's cash assumption.

The common mistake is treating internal balances as free funding. The better control is to require clear ownership, transfer price, maturity, and cancellation rules.

30.41. Contingency funding trigger

Several early warnings flash but ratios remain compliant.

The common mistake is waiting until the ratio breaks. The better control is to use qualitative triggers, market access signals, and management judgment before formal metrics deteriorate.

30.42. Public disclosure timing

A results announcement approaches during funding renewal.

The common mistake is assuming investors ignore calendar risk. The better control is to plan maturity dates, communication, and liquidity buffer around disclosure windows.

30.43. Clearing house liquidity demand

A CCP margin call is due after market volatility.

The common mistake is funding it after ordinary money-market placements are done. The better control is to prioritize critical payment obligations and reserve cash before placing discretionary surplus.

30.44. Regulatory reporting date

The bank optimizes around a reporting snapshot.

The common mistake is creating window dressing that weakens actual resilience. The better control is to align reporting, management information, and genuine liquidity risk appetite.

30.45. FTP lag

Business product pricing uses stale short-end funding curves.

The common mistake is letting business growth consume treasury liquidity cheaply. The better control is to refresh FTP curves and charge optionality, concentration, tenor, and liquidity costs.

30.46. New product launch

A digital deposit product gathers fast balances.

The common mistake is assuming growth means stable funding. The better control is to study customer behavior, rate sensitivity, withdrawal channels, concentration, and stress outflow assumptions.

30.47. Collateral substitution

A borrower replaces collateral during the repo term.

The common mistake is approving substitution without checking liquidity quality. The better control is to control eligibility, concentration, valuation, settlement timing, and impact on buffer composition.

Reserve the incoming collateral and agree eligibility before release. Coordinate linked delivery or agent procedures so a substitution does not leave an unprotected interval; update custody, margin coverage and encumbrance after final status.

30.48. End-of-day cash excess

The bank ends with large idle balances repeatedly.

The common mistake is treating it as harmless conservatism. The better control is to investigate forecast quality, transfer pricing, placement limits, and whether surplus is structural.

30.49. Stress communication

Senior management asks why funding cost jumped.

The common mistake is answering only with market rates. The better control is to explain volume, tenor, counterparty appetite, collateral, benchmark moves, and defensive liquidity choices.

31. How to read the money-market ladder

A money-market ladder is a time map of inflows and outflows. It should show what matures today, tomorrow, this week, next week, and across the relevant short-term buckets. At minimum it should include unsecured deposits, repo and reverse repo maturities, short paper issuance and redemptions, central-bank balances or operations, securities settlements, derivative collateral movements, large known customer flows, and committed liquidity facility drawings where applicable. The ladder should be by currency and legal entity, not only by consolidated group. A consolidated surplus can hide a local shortfall. A dollar surplus cannot automatically pay a sterling obligation. A parent-company surplus may not be freely transferable to a regulated bank subsidiary under stress.

The best ladders distinguish contractual cash flows from behavioral assumptions. A maturing certificate of deposit is contractual. A forecast retail deposit outflow is behavioral. A planned new issuance is an assumption until the cash is settled. A repo trade agreed but not yet settled is not the same as cash received. This distinction matters because stress turns assumptions into questions. When markets are calm, a desk may roll short funding as a matter of routine. In stress, the ladder must be rebuilt using conservative rollover assumptions, reduced unsecured capacity, higher repo haircuts, delayed settlement proceeds, and larger client outflows. The LCR standard is built around the idea that banks should hold enough high-quality liquid assets to withstand a 30-day liquidity stress scenario (https://www.bis.org/publ/bcbs238.pdf); a desk ladder is the operating tool that helps treasury understand how that concept appears in the next hours and days.

A ladder should not be a beautiful spreadsheet that nobody challenges. It should be argued over. Are the biggest inflows truly reliable? Are the same counterparties providing several sources of liquidity? Are maturities clustered on dates when market liquidity is usually thin? Are central-bank-eligible assets already pre-positioned? Are payment cut-offs respected? Are unsecured assumptions too optimistic? Does a customer outflow assumption reflect the current rate environment? Does the ladder include collateral calls that could rise if rates or spreads move? The purpose of the ladder is not to prove that the bank is fine. The purpose is to reveal where action is needed while action is still voluntary.

32. Pricing: the rate is an outcome, not the whole decision

Money-market pricing starts with the risk-free or policy-rate environment, but the executed rate is a mixture of many forces. Tenor matters because a lender giving cash for three months loses optionality compared with overnight placement. Counterparty credit matters because unsecured lenders need compensation for default risk. Collateral matters because secured lenders care about the securities received, haircut, liquidity, and margin process. Balance-sheet cost matters because dealers and banks face internal capital, leverage, liquidity, and reporting constraints. Calendar matters because quarter-end, year-end, tax dates, government issuance, and large settlements can change cash demand. Relationship and franchise matter because banks may pay or accept different economics to serve clients or maintain market access.

A professional desk therefore explains pricing as a bridge from market reference to trade-specific spread. For example: overnight general-collateral repo may trade near the central-bank operating corridor, adjusted for collateral supply, dealer balance sheet, and settlement demand. A one-month unsecured deposit for a bank name will include policy expectations and a bank credit/liquidity spread. A three-month certificate of deposit includes investor appetite, issuer credit, dealer distribution, and maturity concentration. A treasury bill yield reflects sovereign credit, bill supply, safe-asset demand, and money-market fund allocation. These differences are not noise. They are the market's way of pricing the exact package of cash, tenor, credit, collateral, and optionality.

The rate must also be judged against internal cost. If treasury raises expensive term money to reduce rollover risk, that cost should be visible to businesses that consume liquidity. If a business line gathers rate-sensitive deposits only by paying above-market rates, the FTP framework should not pretend those balances are stable cheap funding. If a desk lends surplus cash for an extra few basis points but increases name concentration, the additional return may not be worth the risk. Money-market pricing is best understood as risk-adjusted funding economics, not a race to the highest placement yield or lowest borrowing cost.

33. Liquidity regulation and the money market

The Liquidity Coverage Ratio and the Net Stable Funding Ratio changed how banks think about short-term markets. The LCR focuses on short-term resilience by requiring banks to hold an adequate stock of unencumbered high-quality liquid assets that can be converted into cash to meet liquidity needs under a 30-day stress scenario (https://www.bis.org/publ/bcbs238.pdf). The NSFR promotes more stable funding over a one-year horizon by comparing available stable funding with required stable funding (https://www.bis.org/bcbs/publ/d295.pdf). These standards do not tell a cash dealer exactly which trade to do at 10:05 a.m., but they frame the environment in which the trade is judged.

A repo transaction can improve today's cash position and reduce tomorrow's liquidity freedom. If the bank pledges HQLA to raise cash, the cash arrives, but the securities become encumbered. The bank must ask whether the cash remains available, whether the collateral is still counted in the buffer, and how the trade affects reported liquidity. A reverse repo can be a secured placement of cash, but the received collateral may or may not be operationally useful in the way the bank needs. A short-term wholesale borrowing can solve an immediate funding need but increase contractual outflows in the LCR window. A term funding trade can improve rollover profile but may be expensive and may change NSFR treatment depending on tenor and funding type.

The deeper lesson is that regulatory liquidity and desk liquidity must talk to each other. A bank can satisfy a ratio and still have operational weaknesses, such as concentration in one funding channel or collateral trapped in the wrong entity. A bank can also make a trade that looks attractive on a daily cash report but weakens a regulatory or management stress view. Strong treasury functions connect the dealer blotter, liquidity reporting engine, ALCO dashboard, stress testing model, collateral system, and finance general ledger. The systems do not have to be one platform, but the numbers must reconcile and the explanations must be consistent.

34. Collateral management as a money-market skill

State the contractual calculation explicitly. A 2% haircut on collateral market value means cash of 98 against value 100. An initial margin ratio of 102% means collateral of 102 against cash 100, equivalent to a haircut of about 1.9608%, not 2%. Agreed haircut or margin terms ordinarily apply for the contract's term; a new market schedule primarily affects new or rolled funding unless the agreement permits a change. ICMA's March 2025 European repo guide explains these conventions; it is market guidance, not a universal legal rule.

Collateral management is often treated as a support function, but in money markets it is a core economic skill. The desk needs to know which securities are eligible for which counterparties, which are already pledged, which can be substituted, which are central-bank eligible, which are needed for client settlements, which carry concentration constraints, and which can be financed at attractive haircuts. The difference between owning a security and being able to use it is large. A security may be high quality but in the wrong custodian. It may be legally owned by a different entity. It may be needed to cover another trade. It may be subject to settlement delay. It may be eligible today but vulnerable to rating or policy change.

Haircut schedules deserve close attention. A haircut is not only a credit mitigant for the lender. It is a liquidity cost for the borrower. If a bank can finance 100 of government bonds at a 2 percent haircut, it receives 98 of cash. If the haircut becomes 5 percent, the same collateral produces only 95 of cash. That three-unit difference must be funded somewhere else. In a stress scenario, haircut increases can combine with falling collateral values and higher margin calls. The bank may need to deliver more securities just when securities are hardest to mobilize. This is why collateral stress testing is not optional.

Collateral also has opportunity cost. The same high-quality bond might support central-bank borrowing, private repo, clearing-house margin, payment-system collateral, or liquidity buffer monetization. If treasury uses it for a cheap repo today, the bank may lose flexibility tomorrow. A good collateral allocation process ranks uses by criticality, liquidity value, cost, and contingency importance. It does not simply allocate the cheapest security to the loudest desk.

35. Documentation, legal form, and close-out

Money-market trades depend on documentation more than beginners expect. Unsecured deposits require agreed terms, confirmations, settlement instructions, and legal entity clarity. Repo requires master agreements, collateral schedules, margining terms, events of default, close-out mechanics, substitution rights, interest calculation, and dispute processes. Commercial paper and certificate of deposit programs require issuance documentation, dealer agreements, investor disclosures, paying-agent arrangements, and jurisdiction-specific rules. Central-bank operations require eligibility documentation, account access, collateral pre-positioning, and operational testing.

The reason documentation matters is simple: the worst day is when the document is most needed. In normal markets, counterparties may resolve small disagreements through relationship goodwill. In stress or default, goodwill is not a control. The bank needs enforceable rights to net exposures, liquidate collateral, calculate close-out amounts, and resolve disputes. It also needs internal clarity about who can sign, trade, confirm, amend, and approve exceptions. A repo trade booked under unclear collateral terms may look fine until a margin dispute appears. A short paper program with stale documents may fail exactly when issuance capacity is needed. A central-bank facility that was never operationally tested may be theoretically available and practically useless.

For learning purposes, treat legal documentation as part of product knowledge. If you cannot explain the basic legal mechanics of the product, you cannot fully explain the risk of the product. The front office does not need to become legal counsel, but it must understand where legal enforceability, netting, collateral rights, and documentation status influence trade approval.

36. Operations: confirmation, settlement, reconciliation

The money market is operationally unforgiving because settlement dates are close. A one-day error is not a small delay when the whole trade is overnight. The operating chain begins with trade capture.

The ticket must contain the correct counterparty, legal entity, product, direction, amount, currency, value date, maturity date, rate, day-count basis, settlement instructions, collateral details if relevant, and trader or approver identifiers. The confirmation must match the counterparty's understanding. Any mismatch should be resolved quickly because there is little time before cash moves.

Settlement then has to move cash and securities through the correct systems. For unsecured deposits, the focus is cash settlement. For repo, both cash and securities legs matter. For central-bank operations, collateral eligibility and account setup may be decisive. For short paper, issuance and redemption involve dealers, paying agents, clearing systems, and investors. After settlement, reconciliation must confirm that cash moved, securities moved, interest accruals are correct, maturities are scheduled, and breaks are investigated. A break in a money-market book is not administrative dust. It can hide a funding gap, wrong exposure, wrong interest amount, or failed settlement.

Good operations teams use cut-off discipline, maker-checker controls, standard settlement instructions, exception queues, escalation rules, and daily reconciliations. Good front-office teams respect those controls because they know controls protect market access. A desk that treats operations as an obstacle eventually pays for it in fails, disputes, losses, or damaged counterparty trust.

37. Risk taxonomy for money markets

Money-market risk begins with liquidity risk: the risk that the bank cannot obtain cash when needed or cannot place cash safely without unacceptable loss. Rollover risk is a specific form: funding that matures soon may not renew. Funding concentration risk is another: too much dependence on one counterparty, investor type, product, currency, tenor, platform, or collateral class. Intraday liquidity risk is the timing version: payments may be due before cash arrives.

Credit risk is most obvious in unsecured lending, but secured trades also carry counterparty risk because collateral may be insufficient, disputed, illiquid, or delayed. Market risk appears through short-end rate movements, mark-to-market changes on bills or short securities, basis between secured and unsecured rates, and spread changes on short paper. Collateral risk includes haircut changes, wrong-way risk, concentration, valuation errors, eligibility loss, and settlement failure. Operational risk includes wrong bookings, static data errors, system outages, failed payments, bad confirmations, and manual workaround failures. Legal risk includes unenforceable close-out, unclear documentation, or jurisdictional transfer restrictions. Conduct risk includes unfair client pricing, misleading liquidity representations, benchmark misuse, and inappropriate exception handling.

The risk taxonomy matters because each risk has different controls. Credit limits do not solve settlement cut-offs. Haircuts do not solve legal uncertainty. A rate limit does not solve investor concentration. A liquidity buffer does not solve a broken payment instruction if the buffer cannot be moved. Strong risk management maps each product to each risk type and assigns an owner, metric, limit, escalation path, and stress treatment.

38. Stress: how money markets break

Money markets break by losing confidence, losing collateral mobility, losing settlement certainty, or losing balance-sheet capacity. Unsecured lending disappears when counterparties do not want name exposure. Repo becomes more expensive or smaller when haircuts rise, collateral is questioned, or dealers cannot intermediate. Short paper issuance closes when investors shorten tenors or refuse names. Treasury bill markets can become crowded when everyone wants the same safe asset. Central-bank facilities may become more important, but only eligible institutions with eligible collateral and tested access can rely on them.

The dangerous feature of money-market stress is speed. A bank can lose meaningful overnight capacity in a day. A short paper program can fail to roll during a week. A large deposit can leave in hours. Margin calls can arrive after market moves. A settlement fail can cascade into other obligations. Because tenors are short, stress management must begin before formal default or ratio breach. Early warnings include widening funding spreads, shorter offered tenors, reduced counterparty line availability, larger haircuts, failed or delayed settlements, unusual collateral requests, investor questions, rating-watch headlines, deposit concentration movements, and increasing use of internal exceptions.

Contingency funding plans should specify actions, not slogans. Possible actions include increasing central-bank balances, pre-funding maturities, lengthening tenor where available, reducing new asset growth, raising term funding, monetizing securities, drawing committed lines where appropriate, using central-bank facilities if eligible, restricting discretionary placements, and escalating client concentration issues. The CFP should also define communication: who informs ALCO, risk, finance, regulators where needed, board committees, business heads, and operations. In money-market stress, silence creates confusion and confusion consumes time.

39. Behavioural deposits and money markets

Customer deposits are not money-market instruments in the narrow trading sense, but they shape the money-market need. A bank with stable deposits has less need for wholesale short-term funding. A bank with volatile rate-sensitive deposits has more need to manage sudden inflows and outflows. Treasury must understand deposit behavior by client type, product type, currency, channel, insurance status, rate sensitivity, operational purpose, and concentration. An operating account used for payroll behaves differently from a large non-operational treasury deposit placed for yield. A retail savings balance with many small customers behaves differently from a few large corporate balances.

Money markets often absorb deposit volatility. When deposits surge, treasury may place surplus cash overnight, buy bills, conduct reverse repo, or repay wholesale funding. When deposits leave, treasury may borrow, roll repo, issue paper, sell or repo securities, or use buffers. But if deposit behavior is poorly understood, money-market activity becomes reactive. The desk may place cash too long before outflows occur or fund too short for a structural shift. The bank then pays through missed income, emergency borrowing, or liquidity stress.

This is why FTP and deposit analytics matter. Businesses should see the liquidity value or cost of the deposits they gather. Stable operational balances may deserve different internal pricing from hot money gathered through promotional rates. Large concentrations should be charged for liquidity risk. Optionality should be recognized: a customer who can withdraw tomorrow has sold the bank no term funding, even if the balance has historically stayed. Money markets are where these assumptions become cash outcomes.

40. Multi-currency money markets

Large banks do not manage one money market. They manage several currency money markets that overlap but do not merge perfectly. Dollar, euro, sterling, yen, and local currency cash each have their own central-bank framework, settlement systems, holidays, benchmarks, collateral practices, investor base, and stress behavior. A group may be liquid in one currency and short in another. Converting liquidity across currencies usually requires FX swaps or other foreign-exchange transactions, which introduce settlement timing, counterparty limits, basis spreads, and market access risk.

A multi-currency treasury desk must run separate ladders and then a consolidated view. The separate ladders answer survival questions by currency. The consolidated view answers group strategy questions. The danger is relying on consolidation too early. If euro cash is short today, a dollar surplus helps only if the bank can swap dollars into euros in time, at acceptable cost, within limits, and through settlement systems that are open. In stress, cross-currency basis can widen and swap capacity can shrink. A theoretical group surplus may not prevent a local currency liquidity squeeze.

Central-bank access is also currency-specific. Eligibility for one facility does not create eligibility for another. Collateral accepted by one central bank may not be accepted by another. Local liquidity rules, ring-fencing, or supervisory expectations may restrict transfer. A money-market practitioner should therefore treat currency as a structural feature, not a reporting column.

41. Governance: ALCO, limits, and escalation

Money-market activity should sit inside clear governance. ALCO sets the broad appetite for liquidity risk, funding mix, tenor profile, buffer size, and balance-sheet growth. Treasury executes within that appetite. Market risk, liquidity risk, credit risk, operations, finance, and compliance provide independent challenge and control. The board or board risk committee receives higher-level reporting on liquidity resilience, funding concentration, stress results, and material limit breaches.

Limits should be practical. Counterparty limits control name exposure. Tenor limits stop overextension. Product limits separate unsecured, secured, central-bank, short paper, and investment activity. Currency limits prevent hidden cross-currency dependence. Maturity concentration limits reduce cliff risk. Collateral limits manage eligibility, concentration, and encumbrance. Settlement limits and operational thresholds identify processing risk. Stress triggers force escalation before crisis. A limit that nobody reads is decoration. A useful limit changes behavior.

Escalation must be fast because money-market tenors are fast. A material loss of unsecured lines, sharp haircut increase, failed issuance, large unexpected deposit outflow, central-bank account problem, settlement system outage, or breach of liquidity buffer should not wait for a monthly meeting. The escalation note should state facts, impact, actions taken, decisions needed, and next update time. Senior management does not need a poetic market essay in the first call. It needs a clear operating picture.

42. Conduct and client fairness

Money-market businesses often serve institutional clients that rely on banks for deposits, investments, repo, liquidity products, and market access. Conduct risk appears when the bank prices unfairly, hides risks, misrepresents liquidity, mishandles confidential flow information, or sells a product whose behavior the client does not understand. A client placement that is described as cash-like may still carry issuer risk, market risk, early withdrawal restrictions, or settlement timing. A repo transaction may be secured but still subject to margin and collateral rules. A short paper investment may be high quality but not government-guaranteed.

Human language matters. Clients should understand what they are buying or entering. "Safe" should not be used casually when the product has conditions. "Liquid" should be explained by market conditions, maturity, transferability, and exit process. "Secured" should be explained by collateral type, haircut, margining, and legal terms. Internal teams should avoid using client franchise pressure to override risk discipline without proper approval. A bank can be commercial and fair at the same time. In fact, durable money-market franchises depend on that combination.

Benchmark conduct is also important. If a product references SOFR, €STR, SONIA, or another benchmark, the bank must handle benchmark use, fallback language, systems, and client explanation carefully. Official benchmark administrators publish methodology and governance information for a reason (https://www.newyorkfed.org/markets/reference-rates/sofr; https://www.ecb.europa.eu/stats/financial_markets_and_interest_rates/euro_short-term_rate/html/index.en.html; https://www.bankofengland.co.uk/markets/sonia-benchmark). The desk does not need to turn every client call into a benchmark seminar, but it must know when benchmark mechanics affect economic outcome.

43. Technology and data

A modern money-market operation depends on technology. Core systems include treasury management, front-office trading or order capture, market data, counterparty limits, collateral management, settlement, confirmation matching, general ledger, regulatory reporting, liquidity risk, FTP, and data warehouses. The architecture may be elegant or messy, but the data elements are unavoidable: counterparty, legal entity, product, amount, currency, value date, maturity date, rate, benchmark, collateral, haircut, margin, settlement status, accounting classification, liquidity treatment, and owner.

Data quality failures create real risk. A wrong maturity date distorts the ladder. A wrong counterparty identifier breaches limits. A wrong collateral tag overstates HQLA. A stale settlement instruction creates failed payments. A missing benchmark convention creates bad accruals. A wrong legal entity hides transfer constraints. Technology controls should therefore be designed around business consequences, not only around system completeness. The test case should ask: if this field is wrong, what decision becomes wrong?

For business analysts and technology teams, money markets are a demanding domain because the same trade has front-office economics, credit exposure, collateral movement, settlement events, accounting entries, risk metrics, and regulatory implications. Requirements should be written with end-to-end flows. A repo ticket is not done when the trade is captured. It is done when collateral is allocated, cash settles, margin can be called, maturity is scheduled, accounting posts, risk updates, liquidity reporting receives the right treatment, and exceptions can be escalated.

44. Product approval and change management

New money-market products or platform changes should go through disciplined approval. The approval should identify product purpose, target clients or internal users, legal form, documentation, accounting, regulatory treatment, risk taxonomy, valuation method, settlement process, systems impacted, controls, limits, stress behavior, and exit plan. A new repo collateral type, a new short paper program, a new benchmark convention, a new central-bank facility workflow, or a new electronic trading venue can all change risk. The fact that the tenor is short does not make the change small.

Change management is especially important when markets migrate from old benchmarks to new overnight rates, when settlement cycles shorten, when central banks adjust operating frameworks, or when clearing rules change. The desk may see these as market events, but technology, operations, legal, finance, and risk must implement them. A benchmark change affects confirmations, curves, accruals, valuation, risk, client statements, and fallback language. A settlement change affects cash timing, staffing, cut-offs, reconciliation, and fails management. A collateral eligibility change affects funding capacity and liquidity buffers.

Good product approval avoids two bad outcomes: blocking sensible business because nobody understands it, and approving risky business because everyone assumes someone else checked the details. It creates shared clarity.

45. Worked example: funding a same-week cash need

Assume a bank expects a 500 million cash shortfall in three days because a large corporate deposit is scheduled to leave and several wholesale maturities fall in the same week. The beginner response is to ask for the cheapest three-day borrowing rate. The practitioner response is broader. First, confirm the forecast: is the deposit outflow certain, what currency, which legal entity, what payment time, and is there any offsetting inflow? Second, check the ladder: what maturities occur before the outflow, what trades can be rolled, and what discretionary placements can be allowed to mature? Third, check available instruments: unsecured borrowing, repo, short paper issuance, securities sale, central-bank balances, or internal transfer. Fourth, check constraints: limits, collateral, settlement windows, investor appetite, regulatory liquidity, and stress signals.

If repo is available against high-quality collateral, the bank might raise part of the funding through overnight or one-week repo. If unsecured lines are stable, it might use a smaller unsecured amount to preserve collateral. If short paper investors are receptive, it might issue one-month paper to avoid rolling the whole need overnight. If the liquidity buffer is above target, it might let a reverse repo mature or sell bills. The final solution could be a mix: 200 million one-week repo, 150 million one-month paper, 100 million unsecured overnight rolled with daily monitoring, and 50 million from maturing placements. The mix may cost more than the cheapest single trade, but it reduces dependency on one channel.

The post-trade review matters. Did the outflow occur as forecast? Did funding execute at expected spreads? Were any limits stressed? Did collateral move smoothly? Did the ladder improve or merely postpone the problem? Should deposit behavior assumptions change? A money-market desk that learns from each event becomes steadily better. A desk that treats each event as a one-off stays fragile.

46. Worked example: placing temporary surplus

Now assume the bank receives 700 million of unexpected surplus cash after a client transaction. The balance may leave within a week, but the exact timing is uncertain. The highest quoted rate is a one-month unsecured placement with a weaker bank. A lower rate is available through overnight central-bank placement. A moderate rate is available through reverse repo against high-quality government collateral. The beginner chases the one-month unsecured quote. The practitioner ranks the choices by purpose. If the surplus is uncertain, liquidity and safety dominate yield. A one-month unsecured placement may create exactly the wrong profile: name exposure, tenor lock-up, and limited flexibility for a few extra basis points.

A better action may be to keep a portion in central-bank cash, place a portion through overnight or short reverse repo, and only place a smaller amount unsecured inside strong limits and short tenor. The desk should also update the forecast, ask the client coverage team about likely timing, and inform liquidity management if balances are changing materially. If the surplus proves sticky over several days, treasury can gradually extend tenor or adjust the investment mix. The key is not to turn a forecast surprise into a risk concentration.

This example shows why money-market income must be judged carefully. A few extra basis points on surplus cash can look good in daily P&L. But if the trade reduces flexibility, increases counterparty exposure, or complicates liquidity reporting, the true risk-adjusted return may be poor. Treasury is not an asset manager with unlimited risk appetite. It is the steward of the bank's cash survival and funding efficiency.

47. Learning path for bankers, analysts, and developers

For a banker, the money-market learning path starts with cash purpose. Understand why clients hold operating balances, how corporate treasurers choose deposits or short investments, why institutional investors buy bank paper, and what repo solves for securities holders. Then learn the products and risks. A banker who can explain secured versus unsecured funding in plain language earns client trust. A banker who oversells liquidity or ignores product conditions creates conduct risk.

For a treasury analyst, the learning path starts with the ladder. Build and read cash ladders by currency and legal entity. Trace a trade from execution to settlement to maturity. Learn how LCR, NSFR, internal stress metrics, FTP, collateral encumbrance, and ALCO reporting connect to the same trade. Sit with operations and ask what breaks. Sit with collateral management and ask what can actually be mobilized. Sit with risk and ask which limits bite first in stress.

For a developer or business analyst, the learning path starts with data lineage. A money-market trade is a small object with many downstream consequences. The system must carry product type, legal entity, counterparty, dates, rates, benchmark conventions, collateral, settlement status, and liquidity treatment accurately. Test cases should include holidays, failed settlement, early termination, benchmark fallback, collateral substitution, margin call, counterparty limit breach, and maturity roll. The best technology teams in treasury are not only coders. They understand why the field exists.

48. Source map for this chapter

This chapter uses several official and market-standard anchors. The Basel Committee's Liquidity Coverage Ratio document explains the short-term resilience logic behind HQLA and liquidity stress (https://www.bis.org/publ/bcbs238.pdf). The Basel Committee's Net Stable Funding Ratio document explains the stable funding lens that sits behind structural funding judgement (https://www.bis.org/bcbs/publ/d295.pdf). The New York Fed's repo operations materials describe repo and reverse repo operations used in US monetary policy implementation (https://www.newyorkfed.org/markets/desk-operations/repo) and the standing repo facility (https://www.newyorkfed.org/markets/repo-agreement-ops-faq). The Federal Reserve Board's standing repo page explains the liquidity-supplying role of standing overnight repo operations (https://www.federalreserve.gov/monetarypolicy/standing-overnight-repurchase-agreements.htm). The New York Fed's SOFR page documents the transaction-based secured overnight rate (https://www.newyorkfed.org/markets/reference-rates/sofr). The ECB's €STR materials document the euro short-term rate (https://www.ecb.europa.eu/stats/financial_markets_and_interest_rates/euro_short-term_rate/html/index.en.html). The Bank of England's SONIA materials document sterling overnight benchmark administration (https://www.bankofengland.co.uk/markets/sonia-benchmark). ICMA's European repo best practice guide provides market-practice guidance for orderly repo trading and settlement (https://www.icmagroup.org/market-practice-and-regulatory-policy/repo-and-collateral-markets/icma-ercc-publications/icma-ercc-guide-to-best-practice-in-the-european-repo-market/). BIS work on repo market functioning and repo clearing provides broader system context (https://www.bis.org/publ/cgfs59.pdf; https://www.bis.org/cpmi/publ/d91.htm).

Sources are anchors, not substitutes for judgement. Local regulation, local market convention, product documentation, tax treatment, accounting policy, and internal risk appetite always matter. When using this chapter in a real bank, reconcile these concepts with the bank's approved policies, jurisdictional rules, and systems.

49. Self-check questions

1. Why can a bank be cash-rich at group level and still short in one currency or legal entity?
2. What is the difference between unsecured interbank borrowing and repo borrowing?
3. Why does a haircut protect the cash lender but reduce cash available to the borrower?
4. Why can repo funding become difficult even when a bank owns high-quality securities?
5. What is the difference between general collateral and special collateral?
6. Why does central-bank facility eligibility not automatically mean practical liquidity access?
7. How do SOFR, €STR, and SONIA differ in broad design?
8. Why should a money-market ladder separate confirmed flows from forecast flows?
9. Why does intraday liquidity matter even if end-of-day liquidity is positive?
10. How can short-term paper diversify funding and still create rollover risk?
11. Why does HQLA encumbrance matter when using repo?
12. What early warning signs suggest private money-market capacity is weakening?
13. Why should FTP reflect liquidity value, tenor, optionality, and concentration?
14. What can go wrong if settlement instructions are wrong on an overnight trade?
15. Why is a high placement yield not automatically a good treasury decision?

50. Key takeaways

Money markets are the bank's daily cash operating layer. They connect customer behavior, payment timing, collateral, central-bank policy, wholesale funding, liquidity regulation, benchmarks, and risk appetite. The products look short and simple, but the decisions are rich. An unsecured overnight loan is a credit decision. A repo is a collateral, legal, and settlement decision. A short paper issue is an investor-confidence and maturity-ladder decision. A central-bank facility is an eligibility and operational readiness decision. A treasury bill is a liquid asset, but not identical to cash. A benchmark rate is not only a number; it is a methodology, system input, and legal reference.

The strongest practitioner habit is to translate every trade into balance-sheet consequences. What cash arrives? When? In which account? Against which collateral? For how long? What happens at maturity? Which limit is used? Which ratio changes? Which system must capture it? Which stress assumption is affected? Which person needs to know if it fails? When these questions become natural, money markets stop being a vocabulary list and become a working craft.

51. Bridge to the next chapter

The next topic, Bank Treasury & Liquidity Advanced, builds on this chapter's daily money-market mechanics and moves deeper into advanced liquidity design, stress governance, collateral strategy, recovery planning, and the harder decisions that appear when normal funding assumptions are no longer comfortable. Money markets give treasury the instruments. Advanced liquidity management teaches the judgement needed when those instruments become scarce, expensive, or politically sensitive.

End of Money Markets expanded practitioner chapter.

Topic 3 Supplement: Money Markets Missing Practical Areas

Operations workbook: interest and discount calculations, repo legs and journals, cash certainty, programme operations and exception controls.

1. Why this supplement is needed

Use this operations workbook to reproduce a short-term trade's interest and cash movements, then test its settlement, accounting and exception states. A simple rate ticket can still produce a wrong cash amount through an incorrect day count, calendar, benchmark rule, rounding convention or maturity amendment.

All numbers below are illustrative. Contract terms determine the economics; market conventions, central-bank rules, accounting policy and local law determine implementation. The workflow is an example of role separation, not a universal banking architecture.

2. Front-end benchmarks and rate governance

Money-market trades sit close to official policy rates and overnight benchmarks. A treasury desk must understand not only the rate on the ticket, but the benchmark environment around that rate. The New York Fed explains repo and reverse repo operations as tools that support monetary policy implementation and smooth market functioning by helping maintain the federal funds rate within the FOMC target range (https://www.newyorkfed.org/markets/domestic-market-operations/monetary-policy-implementation/repo-reverse-repo-agreements). The ECB publishes the euro short-term rate, or €STR, as a reference for wholesale euro unsecured overnight borrowing costs (https://www.ecb.europa.eu/stats/financial_markets_and_interest_rates/euro_short-term_rate/html/index.en.html). The Bank of England administers SONIA as the sterling overnight benchmark and describes it as based on actual transactions in overnight sterling wholesale borrowing (https://www.bankofengland.co.uk/markets/sonia-benchmark).

For a bank, benchmark governance means knowing which benchmark is used, who sources it, when it is published, what happens if it is missing, how fallback logic works, how historical corrections are handled, and which downstream systems consume the rate. A cash trade may be fixed-rate and simple. A floating money-market instrument, OIS-linked product, FTP curve or overnight indexed calculation may depend on official benchmark data. If a benchmark feed is stale, the wrong number can flow into interest accrual, P&L, customer statements, hedge valuation and treasury dashboards.

The practical control is a benchmark inventory. Each benchmark should have source, administrator, publication time, calendar, usage, fallback, owner, tolerance, override process and audit trail. A BA should not write a requirement saying only “use SONIA” or “use €STR.” The requirement must say where the rate comes from, which date applies, how weekends and holidays are handled, how corrections are treated, and which products are affected.

3. Interest calculation, day count and broken dates

For a fixed-rate simple-interest deposit, interest = principal × annual rate × actual days / basis, where rate is a decimal and basis is the contractual convention. Count the agreed start/end boundaries, apply the contractual business-day adjustment and use the resulting interest period. Actual/365 Fixed has a fixed denominator of 365, including a leap year; it must not be confused with Actual/Actual or Actual/365L.

Assume USD 100,000,000 lent at 5.25% annual simple interest from Friday to Monday, with no intervening holiday, under Actual/360. The agreed interest period is three calendar days:

100,000,000 × 0.0525 × 3 / 360 = USD 43,750.00

The lender pays USD 100m at the start and expects USD 100,043,750 at maturity. The borrower has opposite cash directions. Under a separately agreed Actual/365 Fixed convention, interest would be USD 43,150.68 after rounding to cents. The difference is USD 599.32. A trade called overnight can span multiple calendar days; use actual contractual dates. Tom-next starts on the next applicable business day and ends on the following one; spot-next starts on the applicable spot date. Currency calendars determine those dates.

Discount instruments require a different quote convention. Assume a fictional bill with USD 10m face value, 90 actual days to redemption and a bank-discount quote of 4.80% using a 360 denominator. Price is 10m × (1 - 0.048 × 90/360) = USD 9.88m. The discount is USD 120,000. A simple investment yield on price using Actual/360 is 120,000 / 9,880,000 × 360/90 = 4.8583%. The 4.80% quote is based on face value, not the purchase price. Do not use this formula for a market quoting a different bill-yield convention.

An overnight reference-rate contract can compound daily instead of using one fixed rate. A simplified contractual factor is product(1 + r_i × d_i / basis) - 1, where each observation's rate is applied for its assigned day weight. On USD 1m, a 5.00% rate weighted for one day followed by 5.10% weighted for three days on Actual/360 produces [(1 + 0.05/360) × (1 + 0.051 × 3/360) - 1] × 1m = USD 563.95, rounded only at the end. This is a synthetic arithmetic example, not a published SOFR/SONIA fixing. Actual contracts may use observation shift, lookback, lockout, spread treatment and different rounding rules.

Store the calculation inputs, convention/version and intermediate precision. Test weekends, currency holidays, leap years, broken dates, early termination, negative rates, rate amendments and maturity extensions. Compare confirmation cash, accrual, settlement amount and the maturity ladder. Do not infer contract mechanics from the benchmark name alone.

4. CP and CD programme operations

Commercial paper (CP) and certificates of deposit (CD) raise short-term funding through different legal instruments and issuance arrangements. Negotiability, discount or coupon pricing, investor eligibility, denomination, identifier and settlement agent depend on the programme and jurisdiction. A market label does not establish a universal legal maturity limit or investor-protection status.

The issuance chain starts with authorised funding need and available programme capacity. Treasury agrees currency, size, tenor and price; dealers or direct channels allocate investor orders. Operations validates instrument terms and identifiers with the issuing/paying agent, reconciles allocations and sends settlement instructions. The agent/depository records issuance; external cash receipt establishes the realised funding. Finance recognises the liability and the relevant discount, coupon and transaction-cost treatment. Regulatory Reporting classifies the instrument and cash flows under applicable rules.

Keep face amount, issue price/proceeds, discount or interest, fees, outstanding amount and maturity redemption separately. For the illustrative discounted 10m bill mechanics above, proceeds and face amount differ; a discounted funding liability likewise creates a redemption obligation larger than net proceeds. At maturity the paying agent requires funded cash and sends redemption; do not assume refinancing has occurred because investor orders were indicated.

A programme limit is a legal capacity, not demand. Reconcile issued versus settled amounts and investor allocations; watch failed issuance, unallocated pieces, dealer/investor concentration, tenor shortening and maturity walls. Keep a cancelled allocation linked to its original order so it cannot later produce duplicate issuance or cash.

5. Money-market funds as investors and liquidity transmitters

Money-market funds are important investors in short-term bank paper, repo and other short-end instruments in many markets. They can provide useful funding diversification, but they can also transmit stress quickly because fund managers are professionally focused on liquidity, credit quality, maturity, regulatory constraints and investor redemption behaviour. A bank should never assume that because a fund can buy its paper, the fund will buy it during stress.

The practical investor-appetite view should track fund name, fund type, currency, approved tenor, rating constraints, concentration limit, historical ticket size, current appetite, last trade date, spread paid, investor feedback and redemption pressure where observable. If funds shorten from three months to one week, the bank still has funding but the quality of funding has weakened. If funds stop taking new paper and only roll overnight, that is an early-warning indicator. If funds reduce exposure after a rating outlook change, the funding desk must show the effect on the maturity ladder and contingency plan.

For learners, the key point is that money-market investors are not passive cash pools. They manage their own liquidity and credit risk. When the bank is under pressure, those investors may become more defensive at the same time. That correlation is exactly why stress testing cannot assume normal-market issuance.

6. Collateral transformation and upgrade trades

Collateral transformation is the process of exchanging one collateral profile for another, often to obtain assets that are eligible for a specific purpose. A bank may need high-quality securities for central-bank operations, clearing margin, repo funding, payment-system collateral or regulatory buffer composition. It may hold lower-quality or less eligible assets and use repo, securities lending or swap-like structures to obtain better collateral temporarily. This can support liquidity, but it creates cost, maturity, unwind, encumbrance and counterparty risk.

The trade should never be described only as improving liquidity. It may improve one liquidity measure while creating another dependency. If the bank borrows HQLA against lower-quality collateral, it must know the maturity of the transaction, haircut, substitution rights, counterparty exposure, legal ownership, accounting treatment and what happens if the trade cannot be rolled. A temporary upgrade can become a cliff if the bank relies on the upgraded collateral in a stress plan.

A BA should capture original collateral, received collateral, eligibility purpose, haircut, maturity, substitution rules, encumbrance, legal entity, counterparty, settlement location and downstream liquidity treatment. A tester should prove that the same collateral is not double counted and that unwind reverses the eligibility and encumbrance correctly.

7. Quarter-end, year-end and balance-sheet constraints

Money-market prices and capacity often behave differently around quarter-end, year-end, regulatory reporting dates, tax dates and major settlement dates. Dealers may reduce balance-sheet usage. Repo rates may become jumpy. Investors may prefer shorter tenors. Some counterparties may avoid trades that increase reported leverage or liquidity usage. Collateral demand can shift. A desk that treats every business day as normal will misread these effects.

The BIS CGFS report on repo market functioning noted that repo market activity and pricing can be affected around balance-sheet reporting dates and that repo intermediation has changed across jurisdictions (https://www.bis.org/publ/cgfs59.htm). That matters in practical treasury. The bank should not assume that because a repo counterparty funded the bank last Tuesday, the same capacity and rate will be available on quarter-end. Forecasting should include calendar effects, dealer balance-sheet tone, known tax dates, securities settlement pressure, central-bank operation dates and public holidays.

A good money-market dashboard shows calendar flags. It marks quarter-end, month-end, holidays, coupon dates, tax dates, large maturities, central-bank operations, reserve maintenance periods where relevant and expected dealer constraints. This allows treasury to pre-fund, stagger maturities, avoid last-minute concentration and explain rate movements correctly.

8. Multi-currency money markets and FX swap dependency

Money-market funding is currency-specific. A bank may have surplus in one currency and deficit in another. FX swaps can transform liquidity across currencies, but they are market transactions with counterparty limits, settlement timing, collateral or credit requirements, basis risk and cut-offs. A group-level surplus does not automatically solve a dollar, euro, sterling, yen or local-currency shortage.

The practical question is cash by currency and value date. If euro cash is short today and dollar cash is surplus tomorrow, the desk cannot simply offset the two without a transaction that settles correctly. FX swap markets may be deep in major currencies but can become expensive or constrained in stress. Cross-currency basis can widen. Some currencies have local restrictions. Holidays can prevent same-day movement. Correspondent banks can reduce limits. The money-market desk must therefore coordinate with FX treasury, nostro teams and liquidity risk.

Testing should include currency-specific cash ladders, FX swap settlement mismatch, failed swap settlement, holiday mismatch, restricted currency, cut-off missed, correspondent limit reduction and basis-cost shock. The expected output is not only funding cost. It is usable liquidity by currency after realistic transformation.

9. Settlement, confirmations and cash certainty

A fictional bilateral classic repo shows the complete start, term and end sequence. Assume Bank A borrows USD 98m from Bank B for seven actual days at 4.50% simple interest, Actual/360, against securities initially worth USD 100m, using a 2% haircut on market value. No fees, coupon event, tax, substitution or interim margin occurs in the base case. Legal terms permit the illustrated DvP transfers.

1. Agree and book. Bank A's dealer records repo borrowing; Bank B records reverse repo lending. Store the legal entities, master agreement, security/quantity, cash amount, dates, rate, haircut and SSI. Credit and product checks apply to both sides; legal title and margin rights follow the agreement.
2. Validate and confirm. Independent trade support verifies economics, security eligibility and date conventions. Operations matches the confirmation and custodian instructions, reserves securities/cash and identifies cut-offs. If central clearing applies, CCP acceptance and the resulting obligations are separate states; a tri-party collateral agent is not automatically a CCP.
3. Start-leg settlement. Bank A delivers securities to Bank B's custody path against receiving USD 98m. DvP links final securities and cash transfers under the settlement system's rules. A full start-leg fail leaves the expected funding unreceived; a pending message is not usable cash. Record any actual partial settlement separately.
4. During the term. Collateral teams reconcile holdings and encumbrance, revalue collateral and apply agreed margin rules. Treasury includes margin calls and the end-date repayment in its forecast. Interest accrues on cash, not on the USD 100m collateral value.
5. End-leg settlement. Interest is 98m × 0.045 × 7/360 = USD 85,750; Bank A pays USD 98,085,750 against receiving equivalent securities back. Bank B receives that cash and returns securities. Release encumbrance only after the return is evidenced.
6. Reconcile and close. Operations matches cash/custody movements to each leg and maturity. Finance reconciles subledger/GL and interest. Risk and liquidity systems update remaining exposure and available collateral. A new roll is a new or properly linked contractual event, not proof the old repo repaid.

For a simplified secured-borrowing accounting assumption in which Bank A retains the security's risks and rewards, its security remains on balance sheet. The journal sketch is:

Every row balances in each bank's books. Custody movement of collateral is recorded separately; the lender does not recognise a purchased investment security merely because it holds legal collateral title under this assumption. Actual derecognition, netting, impairment and collateral accounting follow the applicable framework and contract. Interest receivable/payable may be embedded in carrying amounts in a real chart of accounts.

For a collateral sensitivity, cash capacity at the start is market value × (1 - haircut). If market value falls from 100m to 99m while the 2% haircut remains, capacity is 97.02m. Against unchanged 98m principal, the cash-equivalent coverage shortfall is 0.98m; restoring it with more same-type securities requires about 1m market value. A real call may include accrued exposure, netting, thresholds, minimum transfer amounts and rounding. A haircut is a buffer; variation margin responds to agreed exposure/valuation changes. Do not confuse the two or assume a new market haircut automatically amends a fixed contract.

If DvP fails, identify which instruction, holding or funding prevented completion, revise the cash ladder and keep the outstanding obligation. If a cash borrower defaults, follow Legal/Credit and contract/CCP default procedures for close-out and collateral realisation, not routine maturity cancellation. DvP reduces principal risk at transfer, while term credit, liquidation, replacement-cost and liquidity risks remain. ICMA European repo best-practice guide.

10. Treasury money-market system architecture

An illustrative platform connects deal capture, counterparty/reference data, independent validation, confirmation, payments or securities settlement, collateral, accounting, liquidity and reporting. It may use one integrated system or several reconciled applications. Each business object needs a named authoritative record and each interface needs a completeness control.

Keep confirmation status, instruction status, matching status, settlement status, accounting status and reconciliation status separately. For a repo, keep a state for each cash/securities leg and each partial settlement. The dealer's executed trade feeds expected liquidity; external final account evidence updates actual liquidity. A confirmation mismatch can coexist with a settled movement that still requires investigation.

The example in the diagram uses an opening usable balance of USD 60m. An 08:30 payment of 80m creates a 20m cash gap unless funded beforehand. A reverse-repo repayment of 50m expected at 11:00 cannot fund that earlier payment. Even if it arrives and the closing arithmetic looks adequate, intraday timing still requires cover. If the receipt fails and an additional 40m margin payment is due at 14:00, the desk needs 60m of total extra funding to meet both outgoing obligations while ending at zero, before any policy buffer. These amounts are decision inputs, not universal funding limits.

Reconcile event IDs and versions, source timestamps and ingestion timestamps, expected cash and actual cash, instrument/quantity and custody inventory, collateral reservations and available assets, calculated interest and settlement amounts, then subledger and GL. A replayed event must not create a second payment or journal. An amendment after instruction release needs linked cancellation/repair and external status checks; changing a dealer ticket alone is insufficient.

11. Limits, controls and exception ownership

Money-market limits must reflect the speed of the business. Useful limits include unsecured counterparty limits, secured counterparty limits, tenor limits, currency limits, product limits, settlement limits, maximum overnight reliance, maximum maturity concentration, maximum investor concentration, maximum collateral class concentration, haircut floors, minimum central-bank eligible collateral, maximum open repo dependency, maximum manual adjustment tolerance and benchmark override limits.

Each limit needs owner, measurement frequency, breach threshold, escalation route, permitted action and closure evidence. An overnight counterparty limit breach cannot wait for a monthly meeting. A failed settlement that affects same-day cash must be escalated immediately. A benchmark override must be approved and visible. A collateral eligibility override must be controlled because it can change funding capacity and regulatory liquidity treatment.

Exception ownership is the difference between a dashboard and a control process. A red item without an owner is not managed. The system should show who owns the exception, when it was opened, what decision is needed, and when it must be resolved.

Regulatory reporting is a distinct lifecycle control. For an entity and transaction within EU SFTR scope, map the applicable reporting obligation and delegation arrangement, transaction identifiers, counterparty LEIs, collateral composition, haircuts, reuse and substitutions to the trade-repository submission. Preserve schema/validation version, acknowledgement, rejection, correction and reconciliation evidence. A trade settled in custody can still have a reporting break. ESMA's SFTR reporting guidance specifies the regime's technical standards, ISO 20022 XML schemas and validation framework. Other jurisdictions and products have different requirements; this is not a global reporting prescription.

12. BA and testing scenarios for Topic 3

A strong BA/test pack for money markets should include practical end-to-end scenarios. Book an overnight unsecured placement and verify trade capture, counterparty limit, interest, settlement, maturity and accounting. Book a one-month unsecured borrowing and verify funding ladder, FTP input and maturity wall reporting. Book a repo and verify collateral eligibility, haircut, encumbrance, cash after haircut, settlement and unwind. Book a reverse repo and verify cash placement, collateral receipt, margining and maturity. Issue short-term paper and verify investor allocation, programme limit, maturity ladder and liability accounting.

Additional stress scenarios are essential. A counterparty cuts an unsecured line. A repo haircut rises intraday. A money-market fund refuses rollover. A settlement instruction fails. A benchmark rate is missing. A payment cut-off is missed. A quarter-end dealer refuses term repo. A central-bank operation is available but collateral is not pre-positioned. A CP programme document has stale signatories. An FX swap needed for currency funding fails to settle. A collateral upgrade trade cannot be rolled. A forecast inflow does not arrive. A maturity wall forms after repeated short issuance.

For each scenario, the expected result should include cash impact, liquidity impact, limit impact, accounting impact, operational status, escalation owner and audit evidence. This is how Topic 3 becomes usable for BAs, developers, testers, operations, treasury, risk and finance.

13. Source map for this supplement

The official source anchors for this supplement are conservative and practical. The New York Fed explains repo and reverse repo operations as part of US monetary policy implementation and smooth market functioning (https://www.newyorkfed.org/markets/domestic-market-operations/monetary-policy-implementation/repo-reverse-repo-agreements). The Federal Reserve's overnight reverse repo page explains ON RRP operations and their role in helping control the federal funds rate (https://www.federalreserve.gov/monetarypolicy/overnight-reverse-repurchase-agreements.htm). The ECB's €STR page is the official source for the euro short-term rate (https://www.ecb.europa.eu/stats/financial_markets_and_interest_rates/euro_short-term_rate/html/index.en.html). The Bank of England's SONIA page is the official source for the sterling overnight benchmark (https://www.bankofengland.co.uk/markets/sonia-benchmark). BIS CGFS material on repo market functioning provides official system context for repo markets and stress behaviour (https://www.bis.org/publ/cgfs59.htm). ICMA's European repo best-practice material is useful for market convention and settlement discipline in European repo markets (https://www.icmagroup.org/market-practice-and-regulatory-policy/repo-and-collateral-markets/icma-ercc-publications/icma-ercc-guide-to-best-practice-in-the-european-repo-market/). BIS CPMI material on repo clearing and settlement gives additional infrastructure context (https://www.bis.org/cpmi/publ/d91.htm).

These sources do not replace local law, central-bank rulebooks, market documentation, accounting policy, tax treatment, internal treasury policy or supervisory expectations. Where this supplement explains bank practice, it is practical industry explanation unless a cited source directly states the rule.

14. Final practical standard added to Topic 3

Check that you can reproduce the Friday-to-Monday interest, distinguish a discount quote from yield on price, explain both repo cash directions, balance the borrower and lender journals, and identify a DvP fail's effect on actual cash and usable collateral. Then follow the corresponding event, instruction and account references through the system map.

A short-term trade remains a lifecycle obligation until all agreed cash, collateral and accounting effects are resolved. A maturity label, confirmation match or transmitted message cannot substitute for final movement and reconciliation evidence.