This module teaches coverage from zero, but goes deep enough for a professional
It assumes no prior knowledge of corporate banking and builds every idea from its foundations. Yet it goes deep enough to be useful to someone who already sits on a coverage desk, runs a client-data programme, or sells transformation work to a bank.
Wherever a concept is technical, we stop and explain it in plain language before using it again. Wherever a term is jargon, we define it once, clearly, and then rely on it. French terms are given alongside the English ones where the French market has its own vocabulary, because much of this module is illustrated through the French universal banks.
Two threads run through everything: the business and the technology
The business thread covers what coverage is, why banks organise around it, how a relationship earns its return on capital, and where relationships go wrong.
The technology thread covers the systems that hold a relationship together, how a single client is represented across a dozen platforms, and why the quality of client data at the front door decides whether onboarding, pricing, risk aggregation and reporting work at all.
These two threads are not separate. In modern corporate banking, the relationship and the record of the relationship are the same thing seen from two angles. A banker who does not know the client loses the mandate; a bank whose systems do not know the client loses money, time and, occasionally, its licence.
This module opens the Corporate Banking track
Coverage is the common front to every business line in corporate banking. Financing, trade, cash management and the lifecycle modules that follow all assume that a client already exists, has been onboarded, and is owned by someone. This module explains how that happens. It is therefore the right place to start, and it introduces vocabulary — share of wallet, RAROC, obligor group, golden record, three lines of defence — that every later module reuses without redefining.
How to read this module
Read it as a narrative, not a reference card
Each part assumes you have read the ones before it. If you are short on time, read the section titles and the "In a nutshell" boxes first, then come back for the detail. The titles are written to carry the argument on their own.
Each coloured box plays a fixed role
Knowing what each box signals lets you skim or dive as you choose.
Key insight distils the one idea from a section worth remembering even if you forget everything around it. Going deeper adds nuance for readers who already have the fundamentals; in this module it is often where the French market specifics live. IT & data lens connects the business step to the systems and data beneath it. Worked example grounds an abstract mechanic in concrete numbers, usually through a running case. Watch out flags a common misunderstanding or a place where relationships fail. In a nutshell closes a section with a short recap. Common misconception names a plausible wrong answer and dismantles it. Practice sets out calculations and judgement calls to attempt yourself. Explain it back asks you to teach the idea to someone outside finance, which is the sharpest test of whether you have it.
One running case ties the module together
Most worked examples follow a single fictional client, Helion Industries. Helion is a family-controlled industrial group headquartered in Lyon that makes precision components for the aerospace and energy sectors. It has revenue of about 2.8 billion euros, operates 38 legal entities in 12 countries, carries roughly 900 million euros of net debt, and works with six banks. The bank whose coverage team we follow is currently the third most important of those six. Every number attached to Helion is invented to teach a mechanic; none describes a real company.
The module is built to be worked, not only read
Reading this document end to end will leave you able to follow a conversation about corporate coverage. That is not the same as being able to hold one. The difference is made by four kinds of box that ask something of you rather than telling you something, and skipping them will cost you most of the value here.
Practice boxes contain the calculations and judgement calls a coverage team actually makes: a share-of-wallet estimate, a weighted pipeline, a RAROC test on a facility, the identification of a beneficial owner through two layers of holding companies. Attempt them before looking anything up. Getting one wrong and then finding out why is worth more than reading the correct method three times.
Explain it back boxes ask you to teach the idea to a specific person who does not work in finance. This is the sharpest available test of whether you understand something or merely recognise it, because jargon becomes unavailable at exactly the moment you were about to reach for it.
Common misconception boxes name the plausible wrong answer directly. Most of these are beliefs held confidently by people who work in the industry, which is why they are worth stating rather than leaving to be inferred.
Check your understanding boxes close each part with recall questions. The answers are deliberately not printed beside them; they are collected in the Answer Key near the end of the document. Retrieving an answer from memory is what fixes it. Reading one you have just been shown mostly produces the feeling of knowing, which fades within a day.
A suggested path through the document
The module is not designed to be read in one sitting. Four sessions works well for most readers.
The first session covers Part 1, which sets out what coverage is and how a relationship makes money. The second covers Part 2, the coverage lifecycle from prospect to renewal, which is the heart of the module. The third covers Parts 3 and 4, the actors and the systems, which will read very differently depending on whether you arrived from a business or a technology background. The fourth covers Part 5, the summary and the cheat sheet, and is the right moment to return to the practice boxes from the earlier parts, because spacing the retrieval out is what makes it stick.
Readers who already work in corporate banking can treat Part 1 as revision and start at Part 2, with one exception: the worked example on relationship economics is frequently new even to experienced practitioners, because few people outside finance and pricing teams ever see the whole calculation laid out.
What you will master
By the end of this module you will be able to do six things. Each maps to one part of what follows.
You will understand what coverage actually does
You will know why a bank puts one banker in charge of a client who will buy products from a dozen other teams, what that banker owns and what they merely convene, and why the role is one of influence rather than authority.
You will be able to explain how a relationship makes money
You will understand why the loan that opens a relationship is often unprofitable on its own, how capital-light products sold around it rescue the return, and how share of wallet, return on equity and RAROC turn that logic into numbers a bank can steer.
You will be able to walk the coverage lifecycle end to end
You will follow a client from a name on a target list through onboarding, know-your-customer checks, account planning, pitching, pricing, mandate and hand-off, and into the ongoing service and annual review that restart the loop.
You will be able to map every actor
You will know who does what inside the bank and inside the client, how coverage and product specialists share the same revenue, and where the first line of defence sits when a relationship carries credit, financial-crime or conduct risk.
You will be able to trace client data front to back
You will follow one corporate group through the CRM, the onboarding platform, the master-data hub, the credit and limits system and the profitability engine. You will understand why a single company is represented in four different hierarchies that never quite agree, and why that mismatch is the hardest data problem in corporate banking.
You will recognise the risk and transformation landscape
You will know the regulations that shape every step, the failure modes that have cost real banks billions, and the places where better data and technology genuinely change the economics of coverage.
1Context and the Big Picture
Before any process makes sense, you need the map. This part defines what coverage is, places it inside the bank, and lays out the segments, the competitive landscape and the economics that everything else builds on.
A company has many banks but only a few relationships
A corporate client buys banking the way it buys anything important: from people it trusts
A company of any size needs banks for a surprisingly wide range of things. It needs somewhere to hold its cash and a way to pay its suppliers and staff. It needs credit lines to smooth out the months when customers pay late. It needs to borrow for a new factory or an acquisition. If it sells abroad, it needs to manage foreign-currency risk and to be sure its overseas buyers will pay. Occasionally, it needs advice on buying a competitor or selling a division, or access to the bond and equity markets.
No single team inside a bank can do all of this. Lending is a different craft from cash management, which is different again from hedging or mergers advice. Banks therefore organise their expertise into product lines. But the client does not want to deal with twelve product lines. The chief financial officer wants one person who knows the company, picks up the phone, and makes the rest of the bank behave as one institution.
Coverage is the function that gives the client that one person
Coverage, also called relationship management, is the part of the bank that owns the client relationship. The banker at its centre is the relationship manager, or RM; in larger banks the same role may be called a coverage banker, a senior banker or a client executive, and in French banks the chargé d'affaires entreprises for mid-sized clients or the senior banker for large groups.
The RM is accountable for knowing the client, anticipating what it will need, and mobilising every product the bank sells behind it. The RM sells nothing directly and everything indirectly.
The most useful image is an orchestra conductor. The conductor plays no instrument, yet no music happens without them. They choose which sections come in and when, keep everyone in time, and take responsibility for the performance as a whole. The RM chooses which product specialists to bring to the client and when, keeps the bank's many teams coordinated, and is accountable for the relationship as a whole. The RM's real product is trust.
Three ways of looking at the same role
The coverage role looks different depending on who in the bank is describing it, and all three descriptions are correct at once.
From the commercial angle, coverage is a sales role. The RM manages a portfolio of named clients, grows revenue across them, and is measured on origination and cross-selling.
From the advisory angle, coverage is a trusted-adviser role. The RM understands the client's strategy, financing needs and stage of life, and brings ideas before the client asks for them. A relationship that only ever responds to requests is a supplier relationship, not a coverage relationship.
From the risk angle, coverage is a gatekeeper role. The RM is the first line of defence on credit risk, on financial crime and on reputational risk. Every client the bank takes on is a risk decision, and the RM is the person who proposes it.
Coverage owns the front door of the corporate banking value chain
Every corporate banking deal passes through six stages
A corporate banking transaction, whether a new credit line, a cash-management mandate or a trade-finance facility, follows broadly the same chain. Coverage finds and wins the client and the opportunity. Risk assesses the borrower's creditworthiness, assigns a rating and decides how much the bank is willing to lend. A product team shapes the facility: its size, tenor, pricing, covenants and any hedging. Legal and operations document the deal, book it in the bank's systems and activate the limits. Operations then services it day to day, handling drawdowns, repayments and amendments. Finally, risk and coverage together monitor it, reviewing the client periodically, watching for early-warning signs and preparing the renewal that loops back to the start.
Coverage owns stage one and stays accountable for all six
Coverage owns origination outright. It does not underwrite, which is Risk's job; it does not structure, which is the product team's job; and it does not book, which belongs to operations and legal. But it convenes all of them, and it remains the client's point of contact throughout.
This matters because quality at the front door determines quality everywhere downstream. A client that was badly qualified, a need that was mis-scoped, or a promise made before Risk had been consulted creates rework, credit surprises and servicing pain later. The cheapest place to get a relationship right is at the beginning.
Each stage in the chain is a different system domain: the CRM at origination, the rating and limits engine at credit, structuring and pricing tools at product, booking and legal-documentation systems at execution, payment and loan-servicing platforms afterwards, and monitoring and early-warning tools at the end. The hand-offs between them are control boundaries and data boundaries at the same time. The same client is re-identified at every gate, often under a different internal identifier, which is exactly why straight-through processing rates in corporate banking remain low compared with retail banking or listed markets.
The bank is a matrix: coverage owns the client, product owns the capability
Two axes cross at every client
Every corporate and investment bank is organised as a matrix. One axis is coverage, aligned to clients: who owns the relationship with this aerospace group? The other axis is product, aligned to capabilities: who runs debt capital markets, cash management or trade finance?
A single deal touches both axes. The RM originates the opportunity and owns the client. The product banker structures, prices and executes the solution. Both are essential, and neither can deliver alone.
How the two axes divide the work
| Axis | What it owns | What it is measured on |
|---|---|---|
| Coverage | The client relationship, the account plan, the wallet target, first-line KYC and the right to convene the bank | Client revenue, share of wallet, return on the relationship, cross-sell |
| Product | Deep expertise, structuring, pricing, execution and the profit and loss of its product line | Product revenue, margins, volumes, product-level return |
| Shared | The revenue, the client experience and the risk that a deal goes wrong | Both scorecards, by design |
| Friction | Who leads the client, whose budget receives the revenue, whose priority wins when resources are scarce | Arbitrated by governance |
Why the matrix exists at all
No bank can put one person in charge of both the client and every product, because the expertise is too deep and the clients too many. A banker who understood cash pooling, export credit, interest-rate swaps and leveraged loans equally well would be rare; one who also knew forty clients intimately would not exist. The matrix trades clean accountability for scale and specialisation. Its cost is coordination overhead. Its benefit is that each client gets both a trusted owner and specialists who are genuinely expert.
Revenue sharing is a permanent, governed negotiation
Because both axes are measured on the same revenue, banks deliberately count it twice: once in the coverage scorecard, once in the product scorecard. This keeps both sides motivated to collaborate. In French banks this is known as the partage de PNB, the sharing of net banking income (produit net bancaire), and it is a formal process with written rules, steering committees and arbitration.
The double count is a management device, not an accounting one. The bank's published accounts record each euro of revenue once. Only the internal scorecards add up to more than the whole. Part 2 returns to the mechanics when the mandate is handed from coverage to product.
Segmentation is the operating model
Banks slice their client base by size, type and complexity
A bank does not serve a family business with 30 employees the way it serves a global aerospace group. It therefore divides its corporate clients into segments, and each segment receives a different coverage model, a different set of products, a different credit process and different systems.
| Segment | Typical client | Coverage model | Entry product and credit approach |
|---|---|---|---|
| Multinationals and large caps | Global, often listed groups with hundreds of entities | Global RM plus sector team, global account structure | Full product shelf; bespoke credit, senior committees |
| Large corporates | National champions and large private groups | Dedicated RM | Loans, cash, trade, hedging, selective capital markets; rating model plus committee |
| Mid-caps (ETI in France) | Mid-sized, often family-owned firms | Regional coverage close to the client | Lending-led; scorecard plus delegated limits |
| SMEs | Smaller firms in high volume | Portfolio or digital coverage | Standardised lending and payments; automated scorecard |
| Financial institutions (FIG) | Banks, insurers, asset managers, funds | Specialised FIG team | Correspondent banking, repo, custody, markets; FI-specific models |
| Public sector and sponsors | Sovereigns, agencies, private-equity funds | Public-sector desk or sponsors coverage | Loans, bonds, fund financing; sovereign or fund-specific logic |
The segmentation decides who owns the client and how much authority they have
Placing a client in a segment is not a filing exercise. It decides who owns the client, what level of service it receives, how much credit authority is delegated to local teams, and which systems the relationship runs through. Getting it wrong quietly destroys either service quality, when a complex client is pushed into a high-volume model, or profitability, when a simple client is given a bespoke team it cannot pay for.
France defines one of these segments in law. The entreprise de taille intermédiaire, or ETI, was created by the 2008 Law on the Modernisation of the Economy and its implementing decree: broadly, a company with 250 to 4,999 employees and either turnover not exceeding 1.5 billion euros or a balance sheet not exceeding 2 billion euros. French banks treat the ETI segment as a strategic battleground. They cover it through dense regional networks of business centres (centres d'affaires entreprises), while CAC 40 groups are covered by global sector teams in Paris and London. The result is a hybrid matrix of geography for the middle of the market and sector for the top.
Large caps and multinationals are the most complex clients
At the top of the market, one senior banker, the global relationship manager, coordinates coverage across every country in which the group operates. Underneath sits a global account structure: a parent-level umbrella with shared global limits, a common pricing framework and a single account plan for all subsidiaries. Local teams cover subsidiaries under global agreements, which means dozens of legal entities, currencies and regulators.
Everything about these clients is bespoke. Credit decisions on large, correlated exposures go to senior committees. Board-level access is earned with sector insight, so sector bankers partner the global RM. The wallets are the largest in the bank, but the coordination cost is also the highest, and the return depends on genuine cross-sell depth.
Mid-caps and SMEs are the scale segments
In the middle of the market, the relationship is regional and lending-led. The entry product is credit, typically a revolving credit facility, term loans and investment finance. Cash management, trade and basic hedging are then built around it. Credit uses rating models and delegated authority so that decisions are fast, and bespoke committees are reserved for larger exposures.
For small businesses, one RM may cover hundreds of names, or clients may be pooled and served through a service centre and digital channels. Products are standardised, onboarding and credit decisions are automated wherever possible, and the economics work only through scale. Profit per client is small, so the cost to serve decides everything.
Financial institutions and sponsors are two specialised worlds
Financial institutions are covered by a dedicated FIG team, because their needs and their risk are different. Banks, insurers and asset managers are counterparties as much as clients. They buy correspondent banking, which gives another bank access to payments and clearing in your currency, along with repo, securities lending and custody. Their credit risk is modelled with financial-institution logic based on regulatory capital and interconnectedness rather than corporate scorecards.
Private-equity sponsors are covered on a different principle. The relationship is with the fund, but the business comes from each of its portfolio companies. One sponsor can generate a steady flow of acquisition financings, refinancings and exits. Sponsors also buy subscription lines, short-term facilities secured on the fund's undrawn commitments from its investors. Coverage must be fast and always ready to mandate, because sponsor deals cluster around fundraising, acquisitions and exits.
FIG coverage needs counterparty-risk systems and network-level anti-money-laundering controls, because correspondent banking is one of the highest-risk activities in financial crime. Sponsor coverage needs to model fund structures, investor commitments and a web of special-purpose vehicles that own the portfolio companies. Both stretch a standard corporate client model in different directions, and both are places where a CRM configured for ordinary corporates quietly breaks.
Sector and geography are the second matrix
Coverage can be organised by industry or by place
Within coverage itself there is a second choice. A bank can organise its bankers by sector — technology, energy, healthcare, consumer, industrials, real estate — or by geography, by region and country.
Sector coverage brings deep industry knowledge. Sector bankers know the value chains, the cycles and the competitors, and they can bring ideas that generic coverage cannot: a consolidation play, an energy-transition financing, a supply-chain restructuring. Sector coverage dominates at the top of the market, because board-level credibility with a multinational is earned through industry insight.
Geographic coverage brings proximity. Regional and country teams know local decision-makers, local markets and local regulation. For mid-caps and domestic clients, being close and responsive matters more than deep sector specialism. Even sector-led global clients need local teams to deliver in each country, so geography is how coverage physically reaches the client.
Most banks run both at once
In practice the large banks run a sector-by-geography grid. A German automotive supplier might be covered by the industrials sector banker and by the Germany country team, coordinated by a global RM. The grid gives both expertise and proximity, but it raises exactly the same question as the coverage-product matrix: who leads, and whose target does the revenue count towards?
The economics of a relationship explain why coverage exists
Capital is the scarce resource
To understand why banks care so much about cross-sell, you need one regulatory idea. When a bank lends, it must hold a cushion of its own equity capital against the risk that the loan is not repaid. The size of the cushion depends on the bank's risk-weighted assets, or RWA: the loan amount scaled by a weight that reflects how risky the borrower is. A bank must hold equity equal to a percentage of its RWA; including the various buffers that supervisors require, large European banks typically manage to around 11 to 13 per cent in common equity.
Equity is expensive, because shareholders expect a return on it. A bank's shareholders might expect 10 to 12 per cent a year. Every euro of capital tied up in a loan must therefore earn at least that, or the loan destroys shareholder value even if it is repaid in full.
Return on equity is the test every relationship must pass
The simplest version of the test divides the revenue a relationship generates by the capital it consumes. Banks call the minimum acceptable return the hurdle rate. In practice they refine the calculation by deducting costs and expected losses and by using economic rather than regulatory capital, an approach called RAROC that Part 2 develops in full. For now, the simple ratio is enough to see why the loan alone fails.
This is the entire commercial logic of coverage
The example explains behaviour that otherwise looks strange. Banks offer revolving credit facilities to large companies at margins that barely cover their costs, because the facility buys a seat at the table. They then pursue the ancillary business — the cash, trade, hedging and capital-markets mandates that follow — with great persistence, because that is where the return lives. A relationship that only ever borrows will, at the annual review, be challenged to grow, repriced or gently exited.
French banks steer coverage on return on normative equity (RoNE), a version of return on equity in which each business line is allocated capital according to a standard formula, typically a fixed percentage of its risk-weighted assets, rather than its share of the group's actual book equity. The allocation runs under the European capital rules (the Capital Requirements Regulation and Directive) supervised by the European Central Bank for the largest banks, with the Autorité de contrôle prudentiel et de résolution (ACPR) as the national supervisor. A lending-only relationship that fails the hurdle is challenged at the annual account review, the revue de compte.
Corporates bank with several banks by design
Multi-banking is the norm, not the exception
A corporate client rarely has only one bank. Large companies deliberately spread their business for three reasons: resilience, so that the failure or retreat of one bank does not cut off their credit; pricing tension, so that banks compete for each mandate; and diversified access, because no single bank can provide every product in every country at the best terms.
In France this pattern is called multi-bancarisation, and it is unusually pronounced: even mid-sized companies commonly keep three to five banking relationships. The best a bank can usually hope for is to be the lead relationship bank, the banque de référence or banque principale, with the deepest relationship and the largest share of wallet.
The bank panel and the request for proposal
Large companies formalise multi-banking through a bank panel: a list of banks with which they agree to do business, often the same banks that participate in their syndicated revolving credit facility. One or two panel banks earn lead or coordinator status. The others compete for mandates as they arise.
Many mandates are awarded through a request for proposal, or RFP: a formal tender in which the client describes its needs and invites banks to submit written proposals. Cash management, large facilities and hedging programmes are frequently tendered this way. Coverage must therefore always be ready to pitch, and must know which RFPs are coming before they arrive.
Banks that compete only on price get commoditised
A bank that wins mandates purely by quoting the lowest margin trains its clients to treat it as a commodity supplier. The RM's job is to win on something else: insight into the client's business, reliability of execution, speed of credit decision and being the first call when something goes wrong. Price still matters, but it should be the tie-breaker rather than the argument.
| Competitor type | Where it is strong | How it competes with a universal bank |
|---|---|---|
| Domestic universal banks | Deepest local relationships, full product shelf | Head-to-head for lead-bank status |
| Global investment banks | Capital markets, M&A, cross-border | Compete for advisory, bond and equity mandates and large syndicated deals |
| Regional and mutual banks | Dense local networks, fast decisions | Win mid-cap and SME lead relationships on proximity |
| Private credit funds and direct lenders | Speed, flexibility, appetite for leverage | Take leveraged and mid-market lending away from banks |
| Payments and fintech providers | User experience, specialised flows | Nibble at transaction-banking margins |
Competing on relationship rather than price demands intelligence. The CRM and the client 360 must show where the bank stands on each client's panel, which products the client buys from competitors, which facilities mature when, and which tenders are expected. When that data is missing, the RM walks into a pitch knowing less about the client's banking than the client's treasurer does. That is a direct, measurable cost of poor client data, and it is often the argument that gets client-data programmes funded.
Work these through before reading on. Full solutions are in the Answer Key.
1.1 Share of wallet
A client's estimated annual banking wallet is 20 million euros. Your bank earns 2.4 million from lending, 0.6 million from cash management and 0.2 million from hedging. Calculate your overall share of wallet. The client's lending wallet is estimated at 5 million; calculate your share of it and state in one sentence what the comparison tells the RM.
Total revenue is 2.4 + 0.6 + 0.2 = 3.2 million euros, so share of wallet is 3.2 / 20 = 16 per cent. In lending, 2.4 / 5 = 48 per cent. The comparison tells the RM that the bank is the client's main lender but a marginal provider of everything else: the relationship is capital-heavy and low-return, and the account plan should target transaction banking and hedging, where the bank holds less than a fifth of the remaining 15 million euro wallet.
1.2 Return on capital
A 200 million euro facility earns 0.35 per cent a year and generates risk-weighted assets equal to 45 per cent of its amount. The bank holds capital equal to 12 per cent of risk-weighted assets and has a 12 per cent hurdle. Calculate the return on capital of the facility alone, and the additional annual revenue from capital-light products needed to reach the hurdle, assuming those products consume no further capital.
Risk-weighted assets are 45% × 200 = 90 million euros. Capital is 12% × 90 = 10.8 million. Revenue is 0.35% × 200 = 0.7 million. Return on capital is 0.7 / 10.8 = about 6.5 per cent. To reach 12 per cent, revenue must be 12% × 10.8 = 1.296 million, so the relationship needs about 0.6 million euros of additional capital-light revenue a year.
1.3 Segment placement
For each client, name the segment and coverage model you would assign and give the decisive reason: a family-owned food manufacturer with 1,200 employees and 400 million euros of turnover, operating in France and Spain; a listed telecommunications group with operations in 25 countries; a mid-market private-equity fund that has bought four companies in two years; a regional insurer.
The food manufacturer is a mid-cap, an ETI in French terms, covered regionally with a lending-led model; the decisive reason is its size and its domestic, two-country footprint. The telecommunications group is a multinational, covered by a global RM with a sector team; the decisive reason is its 25-country footprint and its need for the full product shelf. The private-equity fund belongs to sponsors coverage; the decisive reason is that the relationship is with the fund and the business comes from its portfolio companies. The regional insurer belongs to FIG coverage; the decisive reason is that a financial institution is a counterparty with its own products and its own credit models.
1.4 Reading the matrix
A cash-management mandate worth 0.8 million euros a year is won after the RM introduced the product specialist to the client's treasurer. Explain why the bank's internal scorecards might show 1.6 million euros of revenue while its published accounts show 0.8 million, and why the bank tolerates the difference.
The cash-management product line books 0.8 million in its profit and loss, and the coverage team receives full shadow credit of 0.8 million in its scorecard, so the scorecards total 1.6 million. The published accounts record the revenue once. The bank tolerates the difference because rewarding both sides for the same revenue keeps them collaborating; the double count is reconciled away in group reporting and must not be presented as bank revenue.
Explain to a friend who runs a small restaurant why a bank would happily lend to a large company at a margin that barely covers its costs. Use the restaurant's own economics if it helps: which dish on the menu is priced to bring people in, and which items actually make the money? If you find yourself saying "risk-weighted assets", start again.
What does the RM own outright, and what does the RM only convene?
The RM owns the client relationship, the account plan, the wallet target and first-line KYC; the RM convenes, but does not control, credit, product execution and operations.
Name the six stages of the corporate banking value chain in order.
Origination, credit, structuring, booking, servicing and monitoring.
Why do banks deliberately count the same revenue in two scorecards?
To keep coverage and product motivated to collaborate on the same client; the double count exists only in internal scorecards.
What three reasons lead corporates to bank with several banks at once?
Resilience, pricing tension and diversified access to products and credit.
Why is transaction banking worth more to a relationship than its revenue alone suggests?
It is capital-light and recurring, it makes the relationship hard to switch, and its flows give an early view of the client's financial health.
In one sentence, why does a loan on its own often fail the return-on-capital test?
A loan consumes a lot of capital and earns a thin margin, so its revenue alone rarely covers the return shareholders expect on that capital.
1
What does the relationship manager own outright?
Credit is approved by an independent function; the RM only convenes it.
The RM owns the relationship and its plan.
Right. Credit, product execution and operations are convened, not controlled.
The RM convenes these teams but does not control them.
Why do banks count the same revenue in both coverage and product scorecards?
It is deliberate.
Right. The double count exists only in internal scorecards.
Reported revenue is counted once; only internal scorecards double count.
No regulation asks for it.
Which sequence describes the corporate banking value chain in this module?
That is a markets chain.
Those are the later stages only.
Right.
The order is wrong.
2The End-to-End Flow
A coverage relationship is not a transaction with a start and an end. It is a loop that runs for years. This part follows one client around that loop, from a name on a target list to a renewed facility and a refreshed account plan, and looks closely at the gates where relationships are won, delayed or lost.
Hold the whole loop in your head before the detail
Six stages, and the last feeds back into the third
The coverage lifecycle has six stages. The bank prospects and originates: it identifies targets and secures a first meaningful conversation. It onboards the client, including the know-your-customer checks that make the client legally bankable. It builds an account plan that links the client's strategy to a product pipeline and a revenue target. It pitches and structures solutions with product partners. It wins a mandate and hands execution to the product team. And it serves, monitors and reviews the relationship, which produces a refreshed account plan and restarts the loop at stage three.
| Stage | What happens | Key output | Who leads | Typical duration |
|---|---|---|---|---|
| 1. Prospect and originate | Target screening, idea-led calling, referrals | A qualified, engaged prospect | RM, sector team | Months, sometimes years |
| 2. Onboard and KYC | Entity set-up, due diligence, screening, initial rating | A bankable client with limits | Onboarding, KYC, Risk | Weeks to months |
| 3. Account planning | Strategy, needs, position and actions | A living account plan | RM | Annual cycle, updated continuously |
| 4. Pitch and structure | RFPs, deal-team formation, solution design, pricing | A proposal and pre-approved terms | RM with product and credit | Days to weeks per opportunity |
| 5. Mandate and hand-off | Approvals, documentation, booking | An executed, booked deal | Product, legal, operations | Weeks |
| 6. Serve and review | Service, monitoring, KYC refresh, renewal | A reviewed relationship and new plan | RM, Risk, service teams | Continuous, with an annual review |
What changes hands at each stage
Three ideas hold the loop together. First, every hand-off is a data and control boundary: coverage originates, Risk assesses, product structures and executes, operations books and services, and the client must be correctly identified at each step. Second, the account plan is the spine of the relationship, the document that links the client's strategy to what the bank will try to sell and earn. Third, reviews close the loop: the annual credit review, the relationship review and the KYC refresh all feed the next planning cycle. Coverage is never finished.
Step 1 — Origination is a competition for the client's attention
Access is the scarce resource
The most valuable thing in coverage is not capital or product; it is a chief financial officer's time. A large company is approached by dozens of banks. Its CFO and treasurer will meet only those they expect to learn something from. The RM's real asset is being worth a meeting, and that is earned with insight rather than persistence.
Finding the right names
Origination starts with target identification. The coverage team screens the market for companies that fit the bank's risk appetite, its sector strategy and its cross-sell ambitions, and builds a named, prioritised list. The bank is not looking for every company that might borrow; it is looking for companies whose whole wallet it could eventually serve at an acceptable risk.
The approach is idea-led. A first call that says "can we bank you?" is wasted. A first call that says "your 2027 bonds are trading wide of your peers and the private-placement market could refinance them two years early at a lower cost" gets a meeting. The idea might concern a refinancing window, an acquisition angle, a currency exposure or a working-capital problem visible in the accounts.
Much origination comes through referrals. Existing clients introduce suppliers and customers. Product partners in M&A or capital markets meet companies before coverage does. Lawyers, accountants and private-equity sponsors make warm introductions. A well-connected coverage team mines all of these systematically.
From suspect to engaged
Banks grade prospects on a simple ladder so that the pipeline means something.
| Grade | Definition | What moves it to the next grade |
|---|---|---|
| Suspect | A name that might fit the bank's strategy; unqualified | Basic screening against appetite and sector focus |
| Prospect | Fits strategy and appetite; a plausible need identified | Confirming the need and reaching a decision-maker |
| Qualified | Need confirmed, decision-maker reachable, timing known | A first substantive meeting |
| Engaged | First meeting held; a real dialogue and a live opportunity | A mandate or an onboarding request |
Origination lives in the CRM: target lists, call plans, call reports and the top of the opportunity pipeline. Leading banks now layer propensity models and external data on top — company financials, news, regulatory filings, legal-entity identifier links — to score which prospects to call and with which idea. In France, the business registry (SIREN and SIRET numbers), filings available through the national business register and the Banque de France's company rating database (FIBEN) feed prospect qualification and financial pre-screening.
Step 2 — Onboarding turns a won prospect into a bankable client
Five things must happen before the first euro is lent
Winning a client's agreement to do business is not the same as being able to do business. Before the bank can open an account, lend or trade, five things must happen.
The client must be captured in the systems of record through account opening: accounts are opened, a client identifier is assigned and an RM is named. The legal entities must be registered with their hierarchy: parent, subsidiaries, special-purpose vehicles, authorised signatories and the mandates that say who may do what. KYC, or know your customer, must be completed: due diligence, identification of beneficial owners, screening and a risk rating. Risk must assign an initial credit rating and indicative limits, the credit envelope for the group. And product enablement must set up the channels the client will actually use: cash accounts, electronic banking, trade lines and the market documentation needed for hedging, such as an ISDA master agreement for derivatives.
Onboarding is the client's first real experience of the bank
Onboarding is where many relationships are damaged before they have earned anything. It is slow, document-heavy and spread across many systems and teams. A client that agreed to move its cash management to the bank in March and is still sending documents in June has already formed a view of the bank's competence. A smooth onboarding, by contrast, builds trust and shortens time to revenue.
Onboarding is also expensive for the bank. Remediating incomplete files and re-papering clients when rules change are among the largest operational costs in any corporate bank.
Onboarding orchestrates more systems than any other coverage step: the CRM, where the prospect lives; an onboarding or case-management platform; KYC and screening engines; document management; the entity and hierarchy store; and account and limit set-up. These are rarely integrated end to end, so the same client reference data is keyed in repeatedly, which is the classic source of delay, error and duplicate records. Client lifecycle management platforms such as Fenergo, or workflows built on Pega, exist precisely to bind this chain together. Part 4 opens the box.
KYC means peeling an onion until you reach real people
What KYC is and why the law requires it
Know your customer means that the bank must be able to prove it knows exactly who it is dealing with and where the money comes from, before the relationship starts and throughout its life. It is the law's way of making banks the first line of defence against money laundering, terrorist financing and sanctions evasion. In the European Union the framework comes from successive anti-money-laundering directives, now being replaced by a directly applicable regulation, and from the standards of the Financial Action Task Force.
The core of KYC is customer due diligence, or CDD. It has four building blocks.
| Building block | What the bank must establish | Typical evidence |
|---|---|---|
| Entity identification | That the legal entity exists, its legal form, registration, address and legal representatives | Registry extract (in France, the Kbis), legal-entity identifier |
| Ownership structure | The full ownership chain: parents, holding companies, trusts, nominees, often across jurisdictions | Shareholder registers, organisation charts, constitutional documents |
| Ultimate beneficial owner | The natural persons who ultimately own or control the client, and verification of their identity | Beneficial-ownership registers, identity documents, declarations |
| Purpose and source of funds | The intended activity, the expected flows and the legitimate origin of the client's wealth and funds | Business description, financial statements, expected transaction profile |
The beneficial owner is the hardest part
An ultimate beneficial owner, or UBO, is a natural person who ultimately owns or controls the client. The usual threshold is ownership of more than 25 per cent, held directly or indirectly, but control by other means counts too: a shareholders' agreement, the power to appoint the board, or a casting vote. Where no one meets the ownership test and no one controls the entity by other means, European rules fall back on the senior managing official, such as the chief executive.
Finding the UBO is like peeling an onion. The bank must look through each layer of ownership — a holding company owned by a family holding, owned in turn by a trust — until it reaches people, and it must verify every layer with documents.
UBO discovery is, technically, a graph traversal: entities are nodes, ownership and control links are edges, and the bank must walk the graph from the client to every natural person above a threshold. The edges live in different registries, and each must be backed by a document. Every entity in the structure needs a verified record, and the whole graph must be stored so that a change at any node can be detected. That is why entity-data and document-management systems sit at the heart of KYC, and why straight-through onboarding for complex corporate groups remains rare.
In France, customer due diligence obligations sit in the Monetary and Financial Code, transposing the European directives, with the ACPR as anti-money-laundering supervisor for banks. Beneficial owners of companies must be declared to the registre des bénéficiaires effectifs kept by the commercial court registries. Suspicious-activity reports go to Tracfin, the French financial intelligence unit. At European level, a single Anti-Money Laundering Regulation and a new supervisory authority, AMLA, based in Frankfurt, were adopted in 2024; the regulation applies from July 2027 and will harmonise rules that each member state currently transposes in its own way. Since a 2022 judgment of the Court of Justice of the European Union, public access to beneficial-ownership registers has been restricted, although obliged entities such as banks keep their access.
Screening, risk rating and perpetual KYC keep the file alive
Three kinds of screening
Once the bank knows who the client and its owners are, it checks them against three kinds of list. Sanctions screening compares the entity, its owners and its counterparties with the sanctions lists published by the European Union, the United Nations and the United States Treasury's Office of Foreign Assets Control (OFAC). A true sanctions match is a hard block, with no exceptions. Politically exposed person screening flags owners or directors who hold, or recently held, prominent public functions, or who are close to someone who does; such clients are not forbidden, but they require enhanced due diligence and senior sign-off. Adverse-media screening scans news and databases for links to fraud, corruption, crime or other reputational red flags.
The money-laundering risk rating
The bank then combines everything it knows into a money-laundering and terrorist-financing risk rating. The main factors are geography, where the client and its flows are located; sector, because some industries carry higher inherent risk; products, because cross-border payments and trade finance are riskier than a domestic term loan; the complexity of the ownership structure; and the screening results. The rating decides how much due diligence is required, who must approve the client and how often the file is reviewed.
From periodic review to perpetual KYC
The traditional model reviewed each client on a fixed cycle according to its risk rating: every year for high-risk clients, every three years for medium and every five for low. The weakness is obvious: a client whose owner is sanctioned the day after a review is not looked at again for years.
The newer model, perpetual KYC or pKYC, is continuous and event-driven. A change in ownership, a new adverse-media item, an update to a sanctions list or unusual account activity reopens the file automatically. Effort follows risk: high-risk clients are monitored intensively, low-risk clients are touched only when something changes, and names are re-screened every time the lists are updated.
Screening engines match names fuzzily against watchlists, and most of the hits they generate are false positives that an analyst must clear by hand. Tuning match thresholds to reduce false positives without missing true matches is one of the biggest cost levers in financial crime operations, and one of the most mature use cases for machine learning. Perpetual KYC adds a further requirement: event feeds from registries, news providers and transaction monitoring must be wired into case management, and the entity model must be clean enough that a change on one subsidiary triggers a reassessment of the whole group.
Step 3 — The account plan is the spine of the relationship
Four questions, one living document
The account plan is the document in which the RM sets out what the bank knows about a client and what it intends to do about it. Every good account plan answers four questions.
Who is the client? Its strategy, financial position, industry standing, ownership, decision-makers and the group of people who actually decide which banks to use, often called the buying centre.
What does it need? Financing for growth, acquisitions or refinancing; hedging of currency, rate or commodity risk; working-capital and liquidity management; access to capital markets; advice.
What is our position? The bank's current share of wallet by product, the products the client uses, which competitors are present and how strongly, how much credit appetite Risk has for the group, and what risks the relationship carries.
What is the plan? Prioritised product actions for the next twelve months, each with an owner, a timeline and a revenue target.
What makes an account plan actually work
A plan that works is owned and alive: the RM owns it and updates it continuously rather than once a year. It is built on data — real wallet, exposure and revenue figures — rather than memory and optimism. It is co-written with product partners, who commit to the actions they will deliver. And it is governed: presented at the annual client review, with progress tracked against the revenue target.
The pipeline is the coverage team's forward book
Every opportunity carries a probability
The pipeline is the list of live opportunities across the coverage team's clients, each with a product, an expected revenue and a stage. Each stage carries a probability that the opportunity will convert into booked revenue. A common scale is 10 per cent for an identified opportunity, 25 per cent once qualified, 50 per cent once a proposal has been made, 75 per cent when the client has given a verbal mandate, and 100 per cent when the deal is won and booked.
Multiplying each opportunity's revenue by its stage probability and summing gives the weighted pipeline: the team's expected forward revenue, which management compares with budget.
What the pipeline steers
The pipeline does four jobs. It produces a weighted revenue forecast to compare with budget. It allows conversion analytics: win rates and stage-to-stage conversion show which stages leak and where the team needs coaching. It informs resource and capital planning, showing where to deploy scarce banker time, balance sheet and product support. And over time it measures forecast accuracy, comparing what was forecast with what was earned.
The pipeline lives in the CRM as opportunity records carrying product, value, stage and probability. Its value depends entirely on hygiene. Stale, optimistic or missing opportunities produce a fictional forecast, and poor forecast accuracy almost always means poor pipeline discipline rather than bad luck. This is a strong use case for machine assistance: nudges to update records after a meeting, flags for opportunities that have not moved in months, and probabilities adjusted by what has historically converted for similar clients and products.
Step 4 — The pitch mobilises the whole bank behind one opportunity
From opportunity to proposal
An opportunity becomes a pitch through four moves. A trigger opens it: a client need, an upcoming maturity, or a formal RFP for a financing, a hedge or a cash-management tender. The RM then forms a deal team, convening the relevant product partners plus credit and, where needed, legal. The team shapes a solution: structure, terms and value proposition, differentiated on insight, certainty of execution and the relationship rather than on price alone. Finally it delivers a proposal, a pitch book or a written RFP response, which wins or loses the right to execute.
Who joins the deal team
| Member | Role in the pitch |
|---|---|
| RM and coverage | Convenes the team, owns the client relationship and the overall proposition |
| Product specialists | Structure and price their part: loans, debt capital markets, cash, trade, markets, M&A |
| Credit and Risk | Confirm appetite and indicative terms before anything is promised to the client |
| Analysts | Build the financial model, the pitch book and the RFP content |
| Legal and compliance | Flag structural, documentation and conduct constraints early |
Assembling a pitch pulls client data, financials, market data and previous materials from many systems, largely by hand. It is one of the biggest time sinks for junior coverage bankers. Drafting the first version of a pitch or an RFP response with generative AI, grounded in the client 360 and a governed library of previous proposals, is one of the highest-value automation use cases in coverage. Part 5 returns to it.
Pricing passes through the RAROC gate
The question behind every price
Before a price reaches the client, the bank asks one question: does this deal earn enough for the risk and the capital it consumes? The tool that answers it is RAROC, risk-adjusted return on capital.
RAROC = (revenue − operating costs − expected loss − tax) / economic capital
Each element has a precise meaning. Revenue is the margin and fees the deal earns. Operating costs are the cost of originating and servicing it. Expected loss is the average credit loss the bank expects to suffer over time, calculated as the probability of default (PD) multiplied by the loss given default (LGD) multiplied by the exposure at default (EAD). Economic capital is the capital the bank needs to absorb unexpected losses — the bad years rather than the average year.
If RAROC is below the hurdle rate, the deal is repriced, restructured or declined — or justified by the ancillary business that lifts the return of the relationship as a whole.
The four pre-approval steps
Pricing a deal involves four steps in sequence. The product team proposes an indicative price, and the pricing tool computes RAROC both for the deal and for the whole relationship. Risk performs credit pre-clearance, confirming appetite, rating and indicative terms so that coverage does not over-promise. The deal is checked against capital and limits, to confirm that its risk-weighted assets fit within the client's limits, the portfolio's concentration limits and the bank's capital budget. Only then does coverage receive approval to quote a firm price to the client.
Two kinds of capital appear in pricing and are easily confused. Regulatory capital is what the supervisor requires, computed from risk-weighted assets under the Basel rules as implemented in the European Capital Requirements Regulation. Economic capital is the bank's own estimate of what it needs to absorb unexpected losses at a chosen confidence level, typically calibrated to the bank's target credit rating. Pricing engines often compute both and use whichever is binding. Since the final Basel III reforms began applying in the European Union on 1 January 2025, the output floor progressively limits how far internal models can reduce risk-weighted assets below the standardised approach, which raises regulatory capital for some low-risk corporate lending and therefore puts pressure on exactly the thin-margin facilities that open relationships.
A pricing engine must pull the client's existing exposure, its rating, the funding cost from Treasury, the expected-loss parameters from Risk and the ancillary revenue from finance, and compute a whole-relationship return in close to real time. When that data is fragmented, pricing falls back on rules of thumb, which means under-pricing risk on some deals and pricing good business away on others. Accurate deal pricing is therefore one of the most direct pay-offs of a clean client 360.
Step 5 — Winning the mandate starts the revenue argument
The mandate and the hand-off
When the client selects the bank, coverage secures the final internal approvals: the formal credit approval, the agreed pricing and confirmation that KYC is complete. Execution then passes to the product line — loan booking, the debt capital markets syndicate, cash-management implementation — while coverage stays accountable for the client. Legal documents are negotiated, the facility is booked, limits are activated and funds are drawn: a control-heavy process that the lifecycle module covers stage by stage.
Who owns the revenue?
As soon as revenue is booked, the matrix asks its permanent question. Coverage originated the client; product structured the deal. Both claim the revenue, and both need to be paid for it, or they stop collaborating.
Banks resolve this with revenue attribution rules: a governed rulebook that assigns credit by role, product and effort. Most banks credit the full revenue to both the coverage and the product scorecards — the double count introduced in Part 1 — while some split it, for example 50/50 or according to who originated the idea.
Step 6 — Serve, monitor and renew, then start again
Serving and monitoring
After the mandate, the relationship must be served. Queries, waiver requests, amendments and escalations are handled day to day, and service quality defends the relationship far more effectively than any pitch. Once a year the bank conducts an annual relationship and credit review, assessing relationship health, share of wallet, the credit rating and limits; in French banks this is the revue de compte. The KYC file is refreshed, periodically or on events. And the bank runs early-warning monitoring, watching for signs of credit deterioration and coordinating with Risk on watchlists and remedial action.
Renewing and deepening
Maturing facilities must be renewed or refinanced, which is the natural moment to reprice and rescope the relationship. The RM pursues the next cross-sell target identified in the plan, deepening the wallet towards lead-bank status. And the RM defends the relationship, anticipating competitors' moves and protecting the sticky transaction-banking flows that anchor it. The outputs of the review feed a refreshed account plan, and the loop returns to stage three.
| Early-warning signal | Where the bank sees it | What coverage should do |
|---|---|---|
| Covenant headroom shrinking | Quarterly compliance certificates | Discuss with the CFO before a breach; involve Risk early |
| Rising utilisation of the revolving facility | Loan servicing data | Ask why; liquidity stress often shows here first |
| Slower incoming payments, more outgoing urgency | Cash-management flows | Treat as a leading indicator of working-capital strain |
| Rating downgrade or widening bond spreads | Market data, rating agencies | Review exposure and limits; prepare the relationship conversation |
| Management or auditor changes | News, filings, the client | Understand the reason; update KYC if ownership or control changes |
Serving and reviewing a client depend on the client 360 reflecting reality: current exposure, live revenue, the KYC status, open service cases and early-warning signals in one view. When these sit in separate systems, the annual review is assembled by hand from stale extracts and early-warning signals arrive late. A genuinely current 360 is what makes proactive coverage possible, and it is the reason transaction-banking flow data is so prized as a monitoring input.
Work these through before reading on. Full solutions are in the Answer Key.
2.1 Beneficial ownership
Company X is owned 40 per cent by Holding A and 60 per cent by Holding B. Holding A is owned 80 per cent by Mr P and 20 per cent by Ms Q. Holding B is owned in equal thirds by three unrelated individuals, R, S and T. Using a threshold of more than 25 per cent, identify the beneficial owners on ownership grounds, and state what further question the bank must ask before concluding.
Mr P owns 40% × 80% = 32 per cent of Company X and is a beneficial owner on ownership grounds. Ms Q owns 8 per cent. R, S and T each own 60% × one third = 20 per cent, below the threshold. The bank must then ask about control: Holding B controls Company X with 60 per cent of the votes, so the bank needs to know whether anyone controls Holding B — through a shareholders' agreement, board appointment rights or acting in concert among R, S and T. If someone does, they are a beneficial owner by control; if no one is identified by ownership or control at any point, the fallback is the senior managing official.
2.2 Weighted pipeline
A coverage banker has four opportunities: a 2.0 million euro bond mandate at the proposal stage, a 0.6 million hedging programme at the identified stage, a 1.0 million syndicated loan with a verbal mandate, and a 0.4 million trade facility that is qualified. Using the stage weights in this part, compute the weighted pipeline. The banker's remaining budget for the year is 2.0 million; is the pipeline sufficient?
Bond mandate: 2.0 × 50% = 1.00. Hedging: 0.6 × 10% = 0.06. Syndicated loan: 1.0 × 75% = 0.75. Trade facility: 0.4 × 25% = 0.10. Weighted pipeline: 1.91 million euros, against a remaining budget of 2.0 million. The pipeline is not sufficient, and because weighted pipelines tend to be optimistic, a banker would normally want cover well above budget; the banker needs to add new opportunities, not only convert existing ones.
2.3 RAROC
A 100 million euro facility earns a 1.10 per cent margin and no fees. Operating costs are 0.25 million a year, probability of default 0.50 per cent, loss given default 45 per cent, economic capital 7 per cent of exposure and tax 25 per cent. Compute RAROC and say whether it passes a 12 per cent hurdle. Then compute the margin at which it would exactly meet the hurdle.
Revenue is 1.10% × 100 = 1.10 million. Expected loss is 0.50% × 45% × 100 = 0.225 million. Pre-tax profit is 1.10 − 0.25 − 0.225 = 0.625 million; after tax, 0.469 million. Economic capital is 7.0 million. RAROC is 0.469 / 7.0 = about 6.7 per cent, which fails a 12 per cent hurdle. To meet the hurdle, after-tax profit must be 12% × 7.0 = 0.84 million, so pre-tax profit must be 1.12 million and revenue 1.12 + 0.25 + 0.225 = 1.595 million: a margin of about 1.60 per cent.
2.4 Revenue attribution
Three product lines each earn 1.0 million euros from the same client this year, and the coverage team receives full shadow credit on all of it. What does the bank report, what do the internal scorecards total, and what single control would stop the difference from distorting bonuses?
The bank reports 3.0 million euros. The internal scorecards total 6.0 million: 3.0 in the three product lines and 3.0 in coverage. The single control that prevents distortion is to size the bonus pool from the bank's actual, reconciled profit and loss rather than from the sum of scorecards, so that the double count allocates the pool but does not inflate it.
2.5 The pitch
The client's treasurer asks the RM, during an informal lunch, whether the bank could commit 250 million euros at a margin of 0.90 per cent for a new facility. Credit has not been consulted. Write the RM's two-sentence answer.
"That is exactly the kind of facility we would like to provide, and I will come back to you within a few days with terms our credit committee has seen. I would rather give you a number we can stand behind than one we might have to revise."
Explain beneficial ownership to someone who has bought a flat through a property company. Why might their bank want to know who owns the company that owns the flat, and why would "the company" not be an acceptable answer?
Which stage of the lifecycle does the loop return to after the annual review, and why not to the beginning?
Stage three, account planning, because the client is already onboarded; prospecting and onboarding happen once, while planning, pitching, mandate and review repeat.
Name the five onboarding steps that must happen before a client can transact.
Account opening, legal-entity set-up, KYC and customer due diligence, initial credit rating and limits, and product enablement.
Give the formula for expected loss and name each term.
Expected loss = probability of default × loss given default × exposure at default.
What triggers a file review under perpetual KYC that would not trigger one under periodic KYC?
An event: a change of ownership, new adverse media, a sanctions-list update or unusual activity, at any point between scheduled reviews.
What four questions does a good account plan answer?
Who is the client, what does it need, what is our position, and what is the plan.
Why is revenue attribution described as a data problem before it is a political one?
Because a fair split requires mapping every booked euro to the right client group, product and owner, and fragmented client identifiers make that mapping wrong before anyone negotiates.
2
After the annual review, where does the coverage loop return, and why?
Right. Prospecting and onboarding happen once; planning, pitching, mandate and review repeat.
Prospecting happens once.
The loop repeats.
KYC is refreshed periodically, but the loop returns to planning.
Which is NOT one of the five onboarding steps before a client can transact?
One of the five.
Right. The steps are account opening, legal-entity set-up, KYC and due diligence, initial rating and limits, product enablement.
One of the five.
One of the five.
How is expected loss calculated?
That is a spread.
That is a revenue measure.
Right.
The terms multiply.
3Roles, Actors, and Interactions
Coverage is a team sport played across two organisations: the bank and the client. This part names every player on both sides, explains what each one wants, and shows where their incentives line up and where they collide.
The RM conducts, but many specialists play
Six roles surround the client inside the bank
| Role | What they do | What they want from the relationship |
|---|---|---|
| Relationship manager | Owns the client, the account plan and the revenue target; orchestrates everyone else | Wallet growth, a healthy client, a clean risk record |
| Senior banker or sector head | Provides seniority, board-level access and industry credibility for the largest clients | Strategic mandates, franchise reputation |
| Product specialists | Structure and execute loans, bond issues, cash management, trade finance, hedging, M&A | Product revenue and volume, executable deals |
| Credit officer | Independently assesses creditworthiness, sets and polices limits | Sound exposures within appetite, no surprises |
| Coverage analyst | Prepares pitches, financial analysis, account plans and call reports | Learning, visibility, manageable hours |
| Client service and operations | Onboarding, KYC, documentation and day-to-day servicing | Complete files, accurate instructions, realistic deadlines |
Accountability without authority
The core tension of the role is that the RM is accountable for the client but controls almost none of the resources that serve it. Credit sits with Risk. Execution sits with product. Onboarding sits with operations and financial security. Coverage is therefore a role of influence rather than authority. An RM who cannot persuade a credit officer, motivate a product team or get an onboarding file prioritised will lose clients regardless of how good their relationships are. Internal navigation is not a distraction from the job; it is half of the job.
The RM is measured on depth, not just volume
Four skills make an RM
Relationship and trust come first: access to decision-makers, credibility and being the first call. Financial fluency comes next: the ability to read the client's balance sheet, strategy and risks well enough to spot a need before the client articulates it. Orchestration is the third: mobilising product, credit and operations without authority over any of them. And commercial drive completes the set: originating, cross-selling, defending and growing the wallet.
How the RM is measured
| KPI family | Typical measures | What it is meant to encourage |
|---|---|---|
| Revenue | Total client revenue and net new revenue against budget | Origination and growth |
| Quality of revenue | Share of wallet, return on equity or RAROC of the relationship | Depth over volume; capital discipline |
| Breadth | Cross-sell ratio: products used per client | Moving clients up the ladder |
| Risk and conduct | Complete KYC files, no limit breaches, no suitability findings | Owning risk as first line |
In French banks these measures are expressed through produit net bancaire, part de portefeuille and risk and conduct scorecards, and increasingly through sustainable-finance origination targets aligned with the bank's climate commitments.
A day in the life
An RM's time falls into four buckets: time with clients, in meetings, calls and idea generation; deal work, in deal teams, pitches, pricing and approvals; internal navigation, through credit, product, operations and committees; and administration, writing call reports and account plans, re-keying data and updating the CRM.
The client is not one actor either
The buying centre
Inside the client, banking decisions are made by a small group of people with different priorities, often called the buying centre. An RM who knows only the CFO knows half the client.
| Client role | Optimising for | What they need from the bank | Where it collides with the bank |
|---|---|---|---|
| Chief executive | Strategy, growth, shareholder value | Strategic ideas, M&A access, certainty of funding for big moves | Wants commitment before the bank has completed credit work |
| Chief financial officer | Cost of capital, financial flexibility, rating | Competitive pricing, a robust capital structure, capital-markets access | Pushes margins down; plays banks against each other |
| Group treasurer | Liquidity, operational reliability, risk control | Cash visibility, efficient payments, hedging execution | Judges the bank on service failures the RM may never hear about |
| Board and shareholders | Governance, long-term value, control | Discretion, sound advice | Family or sponsor shareholders can override management priorities |
| Procurement | Cost and process | Comparable, priced RFP responses | Turns relationship business into a price tender |
| Accounting and tax | Correct booking and compliance | Accurate statements, documentation | Little visibility to coverage until something breaks |
The treasurer is the most underrated relationship
RMs naturally gravitate towards the CFO, who decides on large financings. But the group treasurer runs the daily relationship with every bank: payments, cash positions, hedging trades, guarantees. The treasurer experiences the bank's service quality directly and forms a view that shapes the CFO's choices at the next panel review. A bank that wins the CFO and neglects the treasurer is often surprised to lose the cash-management tender.
Product partners engage the client through coverage
Six product families sit behind the RM
| Product family | Examples | Capital and margin profile | Role in the relationship |
|---|---|---|---|
| Lending and credit | Revolving facilities, term loans, bridge and acquisition finance | Capital-heavy, thin margin | Door-opener that buys the relationship |
| Transaction banking | Cash management, payments, accounts, liquidity | Capital-light, recurring fees | The sticky daily relationship |
| Trade finance | Letters of credit, guarantees, supply-chain finance | Moderate capital, fee plus margin | Working capital and international trade |
| Markets and hedging | Foreign-exchange, interest-rate and commodity hedging | Capital-light, high margin | Monetises the client's real exposures |
| Debt capital markets | Bonds, private placements, commercial paper, ratings advice | Fee-based, event-driven | Access to public funding |
| Equity and advisory | Equity issuance, M&A advice, strategic advice | High fees, lumpy, prestigious | Top of the value ladder |
The pattern is the ladder from Part 1 expressed as a catalogue. Products rise from capital-heavy and thin-margined to capital-light and fee-rich. The RM uses the balance sheet as an entry ticket and earns the return on the products above it.
When and how product engages
Coverage decides when to bring a product partner to a client, and which one, and stays in the room. Product owns the technical solution and its own profit and loss. The best RMs bring partners in early, with a genuine client insight, rather than as a product push, so that the client experiences one coordinated bank rather than six competing sales pitches. A client who receives separate calls from the bank's foreign-exchange desk, its cash-management team and its trade-finance specialists in the same week, none of whom know about the others, has learned that the bank's matrix is not working.
Products that involve derivatives or investment services are sold under conduct rules that do not apply to a loan. Under MiFID II, a bank selling hedging to a corporate must classify the client, assess the appropriateness or, where advice is given, the suitability of the product, respect the product's target market as defined in the bank's product-governance process, and document the rationale. In France these rules are supervised by the Autorité des marchés financiers alongside the ACPR. The practical consequence for coverage is that the RM may open the conversation about a client's currency exposure, but the recommendation and the sale must run through a process that can show the product fits a real, documented need.
Different segments get fundamentally different coverage
| Segment | Coverage model | Primary products | Credit approach | Systems intensity |
|---|---|---|---|---|
| Multinationals and large caps | Global RM plus sector team | Full shelf including capital markets and M&A | Bespoke, committee-heavy | Very high: global 360 |
| Large corporates | Dedicated RM | Loans, cash, trade, hedging, debt capital markets | Rating model plus committee | High |
| Mid-caps | Regional coverage | Loans, cash, trade, basic hedging | Scorecard plus delegated limits | Medium, with scale effects |
| SMEs | Portfolio or digital | Standardised lending and payments | Automated scorecard | High, through automation |
| Financial institutions | Specialised FIG team | Correspondent banking, repo, custody, markets | Financial-institution models | High, specialised |
| Public sector and sponsors | Public-sector or sponsors desk | Loans, bonds, fund financing | Sovereign or fund-specific | Medium to high |
One bank runs several coverage operating models in parallel. A mid-cap needs efficient, scorecard-driven coverage; a multinational needs bespoke, globally coordinated coverage; a small business needs automation. Systems and processes that fit one segment fail another, which is why "one CRM configuration for everybody" rarely works and why coverage transformation programmes usually proceed segment by segment.
The RM is the first line of defence
Three lines, three roles
Banks organise risk management in three lines of defence. The first line is the business: the RM and the product teams own and manage the risks they originate — credit quality, KYC completeness, conduct and the suitability of what they sell. They see risk first and are responsible for it first. The second line is Risk and Compliance: independent functions that set the risk appetite, ratings and limits, police anti-money-laundering, sanctions and conduct, and challenge the first line. The third line is Internal Audit, which gives independent assurance to the board's audit committee that the first two lines actually work.
Why coverage is first line
The RM sits closest to the client and therefore sees risk first: a deteriorating credit, a suspicious flow, an unsuitable product being pushed. Making coverage explicitly accountable for risk, rather than treating it as a sales force that Risk must police, is central to modern conduct and financial-crime frameworks. The tension is real: the same person who is paid to grow revenue must own its risk. Good banks manage that tension through the KPIs in this part, which put risk and conduct quality alongside revenue, and through a credit process in which the RM proposes and an independent officer decides.
Coverage bankers sit on the private side of the wall
The RM knows things the markets do not
An RM who covers a listed company, or a private company with listed bonds, will regularly receive inside information: an acquisition being prepared, a profit warning about to be issued, a refinancing that will move the bond price. Under the EU Market Abuse Regulation, such information must not be used for trading or disclosed outside the normal exercise of one's job.
This places coverage on the private side of the bank's information barriers, alongside M&A advisers and capital-markets originators, and separated from the public side — sales, trading and research. The barriers, often called Chinese walls, are an organisational and technical control: separate teams, restricted physical and system access, watch and restricted lists maintained by Compliance, and a formal wall-crossing procedure when a public-side person, such as a trader needed to price a hedge for an acquisition, must be brought over.
Why this matters for coverage systems
The CRM holds call reports that can contain inside information. A note saying "CFO confirmed they are in exclusive talks to buy a US competitor" is price-sensitive. Coverage systems must therefore apply entitlements by role and deal, log access, and keep public-side users out. A CRM that is searchable by the whole bank is a market-abuse incident waiting to happen. The ECM module covers the wall in depth from the capital-markets side; here, it is enough to know that coverage is on the private side and that its data must be treated accordingly.
Work these through before reading on. Full solutions are in the Answer Key.
3.1 Map the buying centre
A listed retailer is preparing a cash-management tender. Name the three people in the client whose support the RM most needs, and for each say what they will judge the bank on.
The group treasurer, who will judge the bank on operational reliability, visibility of cash and the quality of implementation. The CFO, who will judge it on cost, on the strategic value of the relationship and on whether the bank also supports the group's financing. Procurement, which will judge it on a comparable, complete and well-priced RFP response. The chief executive matters for the overall relationship but rarely decides a cash-management tender.
3.2 A first-line decision
During a site visit, an RM learns that a mid-cap client has started selling equipment to a distributor in a country subject to EU sectoral sanctions. Nothing has yet appeared in screening. What should the RM do, in what order, and why is waiting for the screening engine not an option?
The RM should record what was learned immediately and escalate it the same day to Compliance or financial security, without tipping off the client about any suspicion; then support the review of the client's flows and contracts with the distributor and update the KYC file. Waiting for screening is not an option because screening only matches names against lists; it cannot know that a new commercial activity exposes the client to sectoral sanctions, and the RM, as first line, is the only person who knows.
3.3 The uncoordinated bank
A large-cap client's treasurer complains that she received three separate calls from the bank in one week about foreign-exchange hedging, supply-chain finance and a new payments platform. Diagnose the cause in the operating model and propose two fixes.
The cause is product teams contacting the client directly without coordination through coverage, and a CRM that does not show who is talking to the client about what. Two fixes: require product outreach to large-cap clients to be planned in the account plan and logged in the CRM before contact; and give the treasurer one named coordinator, normally the RM, who brings specialists in together around a single conversation about the client's needs.
3.4 The wall
A coverage banker wants to ask the rates trading desk for an indicative price on a large interest-rate swap that would hedge an acquisition financing that is not public. Describe the steps that must happen before the trader is told anything about the deal.
The coverage banker must go to Compliance first. Compliance decides whether a wall-crossing is justified, adds the trader to the insider list for the deal, and briefs the trader on their obligations. Only then is the trader told about the transaction, and only what is needed to price the hedge. The trader is then restricted from trading in related instruments until the information is public or no longer relevant.
Explain to a teenager who has a part-time job in a shop why the person who sells the most at a bank is also the person responsible for making sure the bank does not take on dangerous customers. What would go wrong if the two jobs were given to different people who never spoke to each other?
Why is coverage described as a role of influence rather than authority?
Because the RM is accountable for the client but controls none of the resources — credit, product execution, onboarding — that serve it, and must persuade rather than direct.
Name the four KPI families on which an RM is typically measured.
Revenue, quality of revenue (share of wallet and return), breadth (cross-sell) and risk and conduct.
Why is the group treasurer often the most underrated relationship?
Because the treasurer runs the daily relationship with every bank and forms a view of service quality that shapes the CFO's decisions, yet RMs tend to focus on the CFO.
Describe the three lines of defence and where coverage sits in them.
The first line is the business, including coverage, which owns the risks it originates; the second line is Risk and Compliance, which sets rules and challenges; the third line is Internal Audit, which gives independent assurance.
Why must CRM access be restricted by role and deal?
Because call reports and opportunities can contain inside information, which must be protected from public-side staff under the Market Abuse Regulation.
3
Why is coverage a role of influence rather than authority?
No.
Right. The RM must persuade rather than direct.
Not the point.
Seniority is not the reason.
Which is NOT one of the four KPI families of an RM?
One of the four.
One of the four.
Right. The families are revenue, quality of revenue, breadth and risk and conduct.
One of the four.
Why is the group treasurer often the most underrated relationship?
The reverse.
They meet them daily.
Right. RMs tend to focus on the CFO.
Not their role.
4Systems, Data, and IT
Coverage looks like a business of people and conversations. Underneath, it runs on a stack of systems that must agree on one deceptively simple fact: who the client is. This part opens each box in the stack and explains why that agreement is so hard to reach.
The coverage stack is six blocks around one client
Ask the clean question first: what does coverage need from its systems?
Strip coverage down and it needs six capabilities. It needs a record of the relationship: who the client is, who the bank talks to, what was said, which opportunities are live. It needs to onboard and keep clients compliant. It needs a single, trusted master record of each client and its legal structure. It needs to know the client's credit rating, limits and exposure. It needs to price deals and measure profitability. And it needs management information that shows the whole book.
| Block | Purpose | Typical platforms |
|---|---|---|
| CRM | System of record for clients, contacts, calls, pipeline and account plans | Salesforce Financial Services Cloud, Microsoft Dynamics 365, DealCloud, in-house tools |
| KYC and onboarding (CLM) | Due diligence, beneficial owners, screening, case management, perpetual KYC | Fenergo, Pega-based workflows, in-house; screening data from LSEG World-Check, Dow Jones and others |
| Client master data (MDM) | Master client and entity data, legal hierarchies, golden records | Informatica, Reltio and others, often heavily customised |
| Credit and limits | Rating engines, limit management, exposure aggregation, risk-weighted assets | In-house rating tools, Moody's Analytics, core banking and risk platforms |
| Pricing and RAROC | Deal pricing, profitability, capital allocation, hurdle testing | Mostly in-house, sometimes vendor modules |
| Wallet analytics and MIS | Client profitability, revenue by product, share of wallet, dashboards | Data warehouse plus Tableau, Power BI, Qlik |
The integration problem is the real problem
These six blocks were built at different times, by different teams, for different segments. The client exists in all of them, usually under different identifiers. Coverage productivity, KYC speed and wallet analytics all depend on stitching the blocks together, which is why client-data programmes in banks rarely end. The rest of this part opens each block and then the seams between them.
The CRM is the system of engagement
The data model is five linked objects
| Object | What it represents | Key fields |
|---|---|---|
| Account | The client entity or group, with its hierarchy | Legal name, identifiers, segment, sector, owning RM, parent account |
| Contact | The people at the client and their roles | Name, title, role in the buying centre, consent and communication preferences |
| Opportunity | A live deal | Product, expected revenue, stage, probability, expected close date, deal team |
| Activity or call report | A meeting, call or email and its content | Date, participants, notes, next actions, sensitivity flag |
| Account plan | Strategy, wallet, product priorities and targets | Wallet estimate by product, objectives, actions, owners, revenue target |
Contacts and opportunities hang off accounts; activities and plans reference both. Done well, the CRM is the single source of truth for the relationship: anyone entitled to see a client can find out who covers it, what was last discussed and what the bank is trying to do.
Adoption matters more than platform choice
A CRM is only as good as the data RMs put into it. If logging calls and updating opportunities feels like bureaucracy, RMs stop doing it, the pipeline goes stale and the 360 becomes fiction. Banks that succeed with CRM invest at least as much in adoption as in configuration: mobile capture, automatic logging of meetings and emails, call reports drafted by AI from notes or recordings, and dashboards that give RMs something useful back for the data they enter.
KYC platforms orchestrate, they do not own
What client lifecycle management platforms do
Client lifecycle management platforms, usually shortened to CLM, run the onboarding and KYC workflow. They do four things. They orchestrate the end-to-end onboarding journey, coordinating tasks across teams, systems and jurisdictions. They manage cases: every onboarding and every refresh is a case with tasks, statuses, service-level targets and a full audit trail for the regulator. They integrate screening engines, calling the sanctions, PEP and adverse-media services and routing hits to analysts. And they manage documents, collecting, storing and expiring KYC evidence and requesting fresh documents when old ones lapse.
The prize is "KYC once, use everywhere"
The CLM platform is an orchestration layer, not a data owner. It must talk to the CRM, which knows who the client is commercially; to the master-data hub, which holds the legal hierarchy; to the screening engines; to document management; and to account and limit set-up downstream. The integration is the project.
The prize is reuse. A large group may bank with the same institution in ten countries and through five product lines. If each product and each country collects its own KYC, the client sends the same documents ten times and the bank pays for ten reviews. Reusing verified KYC and beneficial-ownership data across the group, subject to local rules, is the single biggest efficiency lever in onboarding.
Reuse depends on a shared entity model. If the CLM platform in France and the one in Germany hold the same subsidiary under different identifiers, neither can see that the other has already verified it. KYC reuse is therefore not primarily a workflow problem but a master-data problem, which is why the next section matters so much.
The client 360 is the hardest problem in corporate banking
A client is a hierarchy, not a record
A single corporate client is really a hierarchy: an ultimate parent, dozens of subsidiaries, special-purpose vehicles and joint ventures, across many countries and currencies. Helion, with 38 legal entities in 12 countries, is a modest example; a CAC 40 group may have several hundred.
The bank needs to know, about that hierarchy, several things at once. It needs to know the legal entity: who signs and who is liable. It needs the risk group: where its aggregate exposure lies. It needs the economic group: the total wallet and revenue. And it needs the coverage view: who in the bank owns the relationship.
What the 360 must pull together
| Data domain | Content | Typical source system |
|---|---|---|
| Reference and hierarchy | Legal entities, ownership, identifiers, KYC status, sector codes, ratings | MDM, CLM, credit |
| Exposure and limits | Drawn and undrawn facilities, limits, collateral, risk-weighted assets by risk group | Credit and limits, loan systems |
| Revenue and wallet | Fees and margin by product, trends, share of wallet, profitability | Finance, product systems, analytics |
| Activity and interactions | Calls, meetings, pipeline, mandates, service cases, documents | CRM, service desk |
| External intelligence | Financial statements, news, market data, ESG data, agency ratings | Data vendors, public registries |
Each domain lives in a different system with a different client key. Reconciling them into one trustworthy view — and keeping it current across tens or hundreds of thousands of entities — is the single biggest data-management challenge in corporate banking, and the foundation that analytics, AI and straight-through processing all depend on.
Four hierarchies describe the same group, and they never match
Four valid answers to "what is this client?"
The same corporate group is modelled four different ways for four different purposes.
The legal hierarchy answers "who signs and who is liable?" It is the true ownership and legal structure — parent, holding companies, subsidiaries, special-purpose vehicles, branches — anchored on registrations and legal-entity identifiers. It is used for contracts, KYC and legal recourse.
The risk group, or obligor group, answers "where is our exposure?" It groups entities that are economically interdependent, so that if one fails the others are likely to follow, for the purpose of aggregate exposure and large-exposure limits. It can cross legal boundaries — a guarantor outside the legal group, for example — and it is owned by Risk.
The economic group answers "what is the total wallet?" It includes everything that belongs to one client commercially, for revenue, profitability and share of wallet. It follows the money rather than the law, and it is owned by coverage and finance.
The coverage view answers "who owns the relationship?" It assigns entities and countries to RMs and teams. It reflects the bank's organisation rather than the client's structure, and it changes every time the bank reorganises.
Why they never match
A subsidiary can be legally owned by one parent, risk-grouped with a guarantor, counted commercially under a different brand, and covered by a third team in a third country. That gives four valid hierarchies, four owners and four systems. The client 360 must hold all four and the mappings between them. Naive designs that force a single hierarchy on everyone fail, because whichever hierarchy they choose is wrong for three of the four purposes.
The risk group is not a matter of internal preference. The European Capital Requirements Regulation limits a bank's exposure to a single client or a group of connected clients to 25 per cent of its Tier 1 capital, and defines connection both by control and by economic interdependence — the situation where the failure of one would probably cause funding or repayment difficulties for the other. The European Banking Authority has published guidelines on how to identify such interdependence. Getting the risk group wrong is therefore a regulatory breach, not merely an analytical error, and it is one of the places where the client 360 has a direct supervisory consequence.
Identifiers anchor the golden record
The anchors
Master data starts with identifiers that are external to the bank and therefore shared with the outside world.
LEI names the legal entity globally
The Legal Entity Identifier is a 20-character code defined by the ISO 17442 standard that uniquely identifies a legal entity worldwide. It is issued by accredited local operating units under the oversight of the Global LEI Foundation. Beyond identification, the LEI system includes "level 2" data on direct and ultimate accounting parents, which lets a bank rebuild part of a client's legal tree from a public source. Entities that trade derivatives or issue securities in the EU must have one, so most large corporates do; many smaller ones do not.
SIREN and SIRET name French entities and establishments
In France, the SIREN number, nine digits, identifies a legal unit, and the SIRET, fourteen digits, identifies each of its establishments. They are allocated by INSEE and appear on every official document. For French clients they are the domestic anchor, and the national business register provides rich, authoritative data behind them.
Internal identifiers are the problem MDM must solve
Each system in the bank — CRM, credit, cash, markets, trade — mints its own client key. A single subsidiary may have five or six internal identifiers. This is the fragmentation that master data management exists to resolve.
Other anchors cover slices of the universe
VAT numbers, D-U-N-S numbers from Dun & Bradstreet, ISINs for entities that issue securities, Banque de France identifiers and national registry numbers in other countries each cover part of the entity universe. None covers all of it.
Master data management turns scattered records into one golden source
Master data management, or MDM, aims to maintain one authoritative, de-duplicated record per real-world entity — the golden record — that every system references. It has four components. Entity resolution uses deterministic rules and fuzzy or machine-learning matching to decide whether two records describe the same entity. Data stewardship and governance assign owners, quality rules, survivorship logic (which source wins when two disagree) and workflows to fix and approve records; these are people as much as technology. Distribution publishes the golden record to consuming systems through APIs and events. And lineage keeps a traceable link from every attribute back to its source.
A useful analogy is the contacts app on a phone. MDM is the bank's contacts app for companies: one clean entry per real entity, so that every other application stops keeping its own messy copy.
Linking LEIs to SIREN numbers and to internal identifiers is one of the most common and highest-value data-quality workstreams in French bank transformation. The LEI database carries the business-register identifier as a field, and INSEE publishes the SIRENE database as open data, so much of the linking can be automated. The hard residue is entities without an LEI, foreign entities without a national identifier the bank holds, and the legacy records with neither, which need human stewardship.
Wallet analytics and profitability engines turn the 360 into decisions
Four engines sit on top of the client 360
Client profitability computes a full-relationship profit and loss per client group: revenue by product, minus the cost of capital, funding and operations. It answers whether the relationship earns its capital. Share-of-wallet analytics estimate the client's total banking spend and the bank's captured share by product, and reveal white space and cross-sell targets. RAROC and capital allocation compute risk-adjusted returns by deal and by client and gate pricing. Underneath all three sits cross-silo revenue aggregation, the hard engine that pulls revenue from loans, cash, trade, markets and investment banking, maps it to the right client group and reconciles it.
Garbage in, spreadsheets out
These engines are only as good as the client 360 and the revenue feeds beneath them. The killer difficulty is aggregation across silos and entities: product systems book revenue against their own client keys, so mapping every fee and every basis point of margin to the right economic group, net of internal transfers and without double counting, is a genuine data-engineering feat. When the numbers are wrong, RMs stop trusting them and fall back on their own spreadsheets, and the bank ends up with as many versions of client profitability as it has coverage teams.
Pipeline dashboards are the visible tip of the data stack
What coverage management sees
| Dashboard view | What it shows | Who uses it |
|---|---|---|
| Weighted pipeline and forecast | Expected revenue by product, team and stage against budget | RMs, desk heads, business management |
| Forecast accuracy | Forecast against actual by RM and product | Desk heads, for coaching |
| Coverage productivity | Revenue per RM, cross-sell ratio, meetings, time to onboard, win rates | Business management |
| Portfolio and risk | Wallet share, concentration, rating migration, KYC status heatmaps | Desk heads, Risk, business management |
The RM sees their own book and next actions, ideally on a phone between meetings. The desk head manages the pipeline, coaches on leaky stages and reallocates effort and balance sheet. The chief operating officer of the coverage business steers budget, risk-weighted assets, productivity and headcount.
Management information exposes, ruthlessly, whatever data quality lies beneath it. A dashboard fed by a stale CRM and unreconciled revenue shows confident but wrong numbers, which erodes trust faster than having no dashboard at all. The Basel Committee's principles for risk data aggregation and risk reporting, known as BCBS 239, were written for risk reporting, but supervisors increasingly expect their disciplines — clear ownership, lineage, reconciliation and timeliness — wherever management relies on data to steer, and coverage MIS is increasingly built to that standard.
Integration architecture: the target and the reality
How the systems should connect
In the target architecture, systems exchange data through APIs: governed services that answer requests in real time, such as "give me this entity", "give me this group's exposure" or "give me this client's revenue this year". Changes propagate through event streaming: when the master-data hub updates an entity, it publishes an event, and every subscribing system updates automatically. The hub distributes the golden record, and consumers subscribe rather than keeping private copies. A data platform, warehouse or lake, consolidates everything for the 360, wallet analytics and MIS.
What the reality usually looks like
In practice most banks still run decades of point-to-point interfaces and nightly batch files between pairs of systems. Staff re-key client data from one screen to another, which is slow, error-prone and the root of duplicate records. There is no shared client identifier, so joins between systems are approximate and reconciliation is a permanent manual effort. And data moves in batch, not real time, so the 360 and the dashboards are always slightly out of date.
| Pattern | How it works | Strength | Weakness |
|---|---|---|---|
| Batch file | Nightly or monthly extract sent from one system to another | Simple, robust, cheap to build | Always stale; failures discovered the next morning |
| Point-to-point API | One system calls another directly | Real-time for that pair | Becomes an unmanageable web as systems multiply |
| Event streaming | Systems publish changes; subscribers react | Real-time, decoupled, scalable | Requires discipline on event design and a shared entity model |
| Golden-source hub | MDM publishes authoritative records for all | One version of the truth | Hard to build; needs governance as much as technology |
Regulation shapes every system in the chain
Coverage technology is not free to do whatever is convenient
| Regime | What it requires | Where it bites in coverage systems |
|---|---|---|
| Anti-money-laundering rules (EU directives, the 2024 AML Regulation, FATF standards) | Customer due diligence, beneficial owners, screening, suspicious-activity reporting | CLM workflows, screening engines, document retention, audit trails |
| Sanctions (EU, UN, US OFAC) | No dealings with designated parties; screening of clients and flows | Real-time screening, list management, hard blocks |
| MiFID II and MiFIR | Client classification, suitability and appropriateness, product governance, record keeping | CRM classification fields, suitability records, recorded communications |
| GDPR | Lawful basis, minimisation, retention, data-subject rights, transfer rules for personal data | CRM contacts, UBO data, cross-border 360 |
| Capital Requirements Regulation and Directive | Capital, risk-weighted assets, large exposures, connected-client groups | Credit and limits, risk group hierarchy, pricing engines |
| Market Abuse Regulation | Protection of inside information, insider lists, wall-crossing | CRM entitlements, call-report sensitivity flags, access logs |
| DORA and BCBS 239 | ICT risk management, resilience testing, third-party risk; risk-data aggregation | Architecture, vendor and cloud contracts, data lineage |
Regulation is not a back-office concern for coverage. It decides how fast a client can be onboarded, which products can be sold to whom, what data can be centralised and where, and how risk exposure must be aggregated. Compliance is designed into coverage processes and systems from the start, not added afterwards.
Work these through before reading on. Full solutions are in the Answer Key.
4.1 Which hierarchy?
For each question, name the hierarchy that answers it: whether the bank may lend another 50 million euros to a subsidiary without breaching a large-exposure limit; who must sign a facility agreement; which RM receives credit for a new foreign-exchange mandate in Singapore; whether the group's share of wallet has risen this year.
The large-exposure question is answered by the risk group. Who signs is answered by the legal hierarchy. Which RM gets credit is answered by the coverage view. Share of wallet is answered by the economic group.
4.2 Design the golden record
List the minimum attributes and cross-references a golden record for a corporate legal entity should carry so that the CRM, the CLM platform, the credit system and the finance system can all use it. Justify each in a few words.
A sound minimum: a bank-wide master identifier, which every system can reference; the legal name and legal form, for contracts and KYC; the external identifiers — LEI, national registry number such as SIREN, VAT number — for matching and reporting; the registered address and country of incorporation, for jurisdiction and sanctions; the immediate and ultimate parent links, for the legal hierarchy; the risk-group and economic-group memberships, for exposure and revenue; the KYC status and risk rating with review date, so no system transacts with an unverified client; the sector code; the owning RM, for coverage; cross-references to each source system's local identifier, so the record can be joined back; and lineage and last-verified dates for each attribute.
4.3 Survivorship
The CRM says a client's registered address is in Lyon; the KYC file, refreshed last month from the business register, says it moved to Paris; the loan system still has Lyon. Which source should win for the golden record, why, and what should happen to the other two?
The KYC file should win for the registered address, because it was refreshed recently from the authoritative public source, the business register. The golden record is updated and the change is published to the CRM and the loan system, which should subscribe to the golden record rather than keep their own copies. The loan system's legal documentation may need a notice of change, and the discrepancy should be logged as a data-quality issue so that its cause — the absence of distribution from the master record — is fixed.
4.4 The stale dashboard
A desk head notices that the weighted pipeline on the dashboard has not changed in three weeks although the team has won two mandates. Name three possible causes, at least one in data and one in behaviour.
Possible causes include: RMs not updating opportunity stages in the CRM after winning (behaviour); a failed or stopped batch feed from the CRM to the data warehouse (data and integration); the won deals booked under client identifiers that do not map to the accounts in the pipeline (master data); or the dashboard filtering on a stage definition that changed (configuration). The desk head should check the feed's last successful run first, because that is quickest to rule out.
4.5 KYC reuse
A group already onboarded in France asks to open accounts for its German subsidiary. Describe what the German team can reuse from the French file and what it must still do itself, and name the data condition that makes reuse possible at all.
The German team can reuse the verified group structure, the beneficial-ownership analysis, the screening results on the group and its owners, and the documents already held, subject to their validity dates. It must still complete any German-specific requirements, verify the subsidiary itself — its registration, representatives and signatories — and assess the subsidiary's own expected activity and risk. Reuse is possible only if the subsidiary and its parent are linked to the same golden-record entities as the French file, so that the German process can find and trust what France has verified.
Explain to someone who has moved house why a bank can end up with three different addresses for the same company and not know which is right. Use their own experience: which organisations still send letters to the old address, and why?
Name the six blocks of the coverage technology stack.
CRM; KYC and onboarding (CLM); client master data (MDM); credit and limits; pricing and RAROC; wallet analytics and MIS.
What are the four hierarchies, and which function owns each?
Legal, owned by legal and KYC functions; risk or obligor group, owned by Risk; economic group, owned by coverage and finance; coverage view, owned by coverage management.
What is a golden record, and what are the four components of master data management?
One authoritative, de-duplicated record per real-world entity that all systems reference; MDM's components are entity resolution, stewardship and governance, distribution, and lineage.
What does an LEI identify, and what extra information does its "level 2" data provide?
A legal entity, globally and uniquely; level 2 data adds its direct and ultimate accounting parents.
Why does a client lifecycle management platform depend on master data to deliver KYC reuse?
Because reuse requires every country and product to recognise the same entity; without a shared entity identity, each process cannot see what another has already verified.
Give two reasons a CRM is subject to regulatory constraints that a spreadsheet of sales leads in another industry would not be.
Contact records are personal data under the GDPR, and call reports can contain inside information protected by the Market Abuse Regulation; client classification and suitability records are also required under MiFID II.
4
Which hierarchy is owned by Risk?
Owned by coverage and finance.
Owned by coverage management.
Owned by legal and KYC.
Right. The legal hierarchy is owned by legal and KYC, the economic group by coverage and finance.
What is a golden record?
No.
No.
No.
Right.
Which block is NOT part of the coverage technology stack?
One of the six.
Right. The six blocks are CRM, KYC/CLM, client master data, credit and limits, pricing and RAROC, wallet analytics and MIS.
One of the six.
One of the six.
5Risks, Pain Points, and Transformation
Coverage creates value by concentrating knowledge of the client in one place. The same concentration creates its risks. This part maps where coverage goes wrong, looks at real failures that cost banks billions, and sets out how banks are re-inventing the coverage model.
A MECE map of where coverage risk lives
Six risks, two roots
Before the case studies, a clean taxonomy. Coverage risk falls into six categories that do not overlap and between them cover the field.
| Risk | What goes wrong | Typical consequence |
|---|---|---|
| KYC and onboarding drag | Slow, document-heavy onboarding and refresh | Lost mandates, delayed revenue, frustrated clients |
| Financial crime | A client or its flows are used for laundering, terrorist financing or sanctions evasion | Fines, enforcement, licence risk, reputational damage |
| Credit and concentration | Exposure grows faster than the bank's understanding of the group | Unexpected losses, large-exposure breaches |
| Conduct and mis-selling | Pressure to cross-sell pushes unsuitable products | Client harm, remediation, regulatory action |
| Key-person dependency | The relationship lives in one RM's head and contacts | Lost knowledge and lost clients when the RM leaves |
| Manual, low-value work | Call reports, pitch books, spreadsheet plans, re-keying | Less client time, more errors, higher cost |
Almost every one of these traces back to two roots: fragmented client data and manual process. Fix the data foundation and automate the low-value work, and most of the other risks shrink. That is the transformation thesis in one sentence.
How to tell a good relationship from a large one
Size is not quality
A large relationship is not necessarily a good one. A client that generates 20 million euros of revenue but consumes 400 million euros of capital, demands bespoke service from five teams and carries complex financial-crime risk may be worth less to the bank than a mid-cap that generates 3 million euros from cash management with almost no capital.
Five tests a coverage head applies
Does it earn its capital? Relationship RAROC above the hurdle, or a credible plan to get there. Is the revenue diversified? A relationship that depends on one product, or one lumpy advisory fee, is fragile. Is it sticky? Transaction-banking flows anchor relationships; pure lending relationships move at the next refinancing. Is the risk understood? A complete KYC file, a correctly built risk group and early-warning monitoring that actually runs. Is it institutional? More than one person in the bank knows the client, and more than one person at the client knows the bank.
Three real failures show three ways a relationship goes wrong
Nothing teaches faster than watching it break
The cases below are well documented in public sources. Each illustrates a different coverage failure: a financial-crime failure in a client portfolio, a concentration failure in a fast-growing relationship, and an aggregation failure in which the bank did not see its own exposure.
Danske Bank, Estonia, 2007 to 2015 — when the client book was never really known
What happened
Danske Bank's Estonian branch built a large business with non-resident clients, many of them companies registered in the United Kingdom and elsewhere whose owners were in Russia and other former Soviet states. A report commissioned by the bank in 2018 found that roughly 200 billion euros of payments had flowed through the non-resident portfolio between 2007 and 2015, a large part of it suspicious. The bank's chief executive resigned, and in 2022 the bank pleaded guilty in the United States and agreed to pay about 2 billion dollars to US authorities.
Root cause
The non-resident portfolio was profitable, ran on separate systems from the rest of the group, and was covered by local teams under weak oversight. Beneficial owners were often not identified, red flags raised by correspondent banks and whistle-blowers were not acted on quickly, and the group did not have an integrated view of who the branch's clients really were.
The lesson
KYC is not an onboarding checkbox; it is the first line's continuing responsibility for knowing the client. A portfolio that runs on its own systems, outside the group's client data, is a portfolio the group does not know. Coverage incentives that reward revenue without equal weight on the quality of the client book produce exactly this outcome.
Greensill Capital, 2021 — when a relationship grew faster than understanding
What happened
Greensill Capital presented itself as a supply-chain finance provider. It packaged short-term receivables into notes that were sold, notably, to supply-chain finance funds managed by Credit Suisse, which held about 10 billion dollars of assets when they were frozen in March 2021. Greensill collapsed that month after its credit insurance was not renewed. Much of the exposure turned out to be concentrated on companies linked to one industrial group, and some of the financed receivables were "prospective" — invoices for sales that had not yet happened.
Root cause
The relationship was valued for its growth and fees. The concentration on a small number of connected obligors was not aggregated properly, the nature of the underlying assets was not examined deeply enough, and reliance was placed on insurance that could be withdrawn. Several institutions, across asset management, banking and insurance, each saw part of the picture.
The lesson
Understanding what a client actually does, and who the ultimate obligors are, is a coverage responsibility, not only a risk one. Connected-client grouping — the risk hierarchy of Part 4 — exists precisely to prevent concentrated exposure from hiding behind many names. Supply-chain finance is a legitimate and valuable product, covered in its own module; the failure was in how the relationship was understood, not in the product itself.
Archegos Capital Management, 2021 — when the bank could not see its own exposure
What happened
Archegos, a family office, built enormous leveraged positions in a small number of stocks through total return swaps with several prime brokers. When the prices of those stocks fell in March 2021, Archegos could not meet margin calls. The banks that sold the collateral fastest limited their losses; those that moved slowly lost heavily. Credit Suisse lost about 5.5 billion dollars and Nomura nearly 3 billion.
Root cause
Each bank saw only its own slice of the client's positions, and swaps do not require public disclosure of the underlying holdings, so no bank saw the full concentration. Inside Credit Suisse, the independent report into the losses found that risk signals had been visible for months, that limits had been breached repeatedly, and that the business and risk functions had failed to act on them. The client was valued for its fees, and the relationship was not managed as a risk.
The lesson
Archegos was a prime-brokerage client rather than a corporate borrower, but the coverage lesson is universal. A bank must aggregate its full exposure to a client across desks, products and legal entities, and must act on what the aggregate shows. A client 360 that includes revenue but not exposure, or exposure on one product line but not another, is a tool for growing the relationship blind.
Six transformation themes are reshaping coverage
Technology here is about time and risk, not just efficiency
| Theme | What it means | Why it matters for coverage |
|---|---|---|
| Digital onboarding and perpetual KYC | Straight-through, reusable, continuously monitored KYC with digital document capture | Removes the largest single friction and shortens time to revenue |
| Unified client 360 | Master data plus a data platform giving one trusted view of each group | The foundation for profitability, pricing, risk aggregation and AI |
| AI-assisted coverage | Generative AI for call reports, pitch preparation, research summaries; next-best-action models | Gives RMs time back and sharpens prioritisation |
| Coverage productivity tools | Modern CRM, mobile access, integrated pipeline and pricing | Less administration, better forecasting |
| ESG and sustainable finance | ESG data in the 360, green and sustainability-linked lending, transition advice | A new revenue pool, a credit factor and a regulatory requirement |
| Platform and ecosystem models | APIs to the client's treasury systems, embedded finance, partnerships | Keeps the bank relevant as corporate treasury digitises |
Each theme earns its place the same way: by moving the cost curve or the risk profile, not by putting a nicer surface on the same manual work. That filter is the one to apply to any coverage transformation proposal.
Generative AI arrives in coverage, with governance close behind
Where it genuinely helps
| Use case | What it does | The gain |
|---|---|---|
| Call-report generation | Drafts the call report and CRM update from notes or a recording | Removes the most hated administrative chore; improves CRM data |
| Next-best-action and propensity | Predicts which product a client is likely to need next and ranks actions by expected value | Focuses scarce RM time |
| Research and pitch copilots | Summarises filings, news and internal research; drafts first-cut pitches and RFP answers grounded in the 360 | Hours saved per pitch; more consistent quality |
| KYC and screening assistance | Triages screening hits, summarises adverse media, drafts KYC narratives for analyst review | Lower false-positive cost, faster onboarding |
| Meeting preparation | Prepares a briefing before each meeting and drafts follow-ups | Better meetings, faster follow-through |
| Portfolio alerts | Flags at-risk relationships, white space, expiring facilities and early-warning signals | Proactive rather than reactive coverage |
The risks are specific
Generative models can produce fluent text that is wrong, and a confident error in a credit memo or a client proposal is a liability. Every serious deployment keeps a human in the loop and treats the output as a draft. Models must be grounded in the bank's own data — the 360, the document library — rather than inventing facts. Confidential client information, including inside information from the private side, must not leave the bank's controlled environment. And decisions that significantly affect a client must remain explainable and attributable to a person.
The EU AI Act entered into force on 1 August 2024 and applies in phases. Prohibited practices and the AI-literacy obligation applied from February 2025, and the obligations on general-purpose AI models from August 2025; the high-risk regime is scheduled to follow, although its timing has been the subject of proposals for postponement, so the current dates should be checked before relying on them. Most coverage use cases — drafting, summarising, prioritising — are not in the high-risk categories, but creditworthiness assessment of natural persons is, which matters for small-business coverage where the owner's personal credit is assessed. French banks also operate under CNIL guidance on AI and personal data, and under their own model-risk governance. The practical result is that coverage AI is deliberately assistive, with a human decision-maker on record.
ESG has become a coverage conversation
A new product pool
Sustainable finance has created products that coverage now sells. Green and social loans finance eligible projects, such as renewable energy or energy efficiency, under market principles published by the Loan Market Association and its US and Asian counterparts. Sustainability-linked loans do something different: the proceeds can be used for any purpose, but the margin steps up or down according to whether the borrower meets agreed sustainability performance targets, such as a reduction in emissions intensity. Transition advice helps clients plan decarbonisation and the financing it requires, and often leads to green bond issuance.
A credit factor and a data problem
ESG increasingly feeds credit appetite, ratings and pricing: a company exposed to carbon-intensive activities without a credible transition plan carries risks that a lender must assess. And ESG creates a new data domain in the client 360. Emissions, taxonomy alignment and sustainability-linked loan targets must sit beside the financials, from vendors whose methodologies differ and from client self-reporting of variable quality. Sustainability-linked margins depend on verifying the client's performance every year, which is a new reporting and control workflow. And banks must aggregate ESG data across the whole book for their own disclosures, including financed emissions.
French banks are among the world's leading arrangers of green and sustainable financing, and operate under the EU Taxonomy, the Corporate Sustainability Reporting Directive, the Sustainable Finance Disclosure Regulation for their asset-management arms, and the ECB's supervisory expectations on climate-related and environmental risks. The scope and timing of corporate sustainability reporting were reduced and postponed by the EU's 2025 "Omnibus" simplification package, which affects how much reported data banks can expect from mid-sized clients. ESG data availability for the ETI segment is therefore likely to remain patchy, and coverage teams often collect it directly.
The target operating model is built in three phases
Sequence matters more than any single tool
Coverage transformation that works is sequenced. The first phase builds the data foundation: master data and a golden source, entity resolution, a unified client 360, an entity API and an event backbone. Nothing durable is built without it. The second phase redesigns the processes on top: digital onboarding and perpetual KYC, straight-through workflows, integrated pipeline and pricing, and KYC reuse across products and entities. The third phase drives adoption and intelligence: a modern CRM, generative-AI copilots, next-best-action models, ESG data and dashboards, landed with training and change management so that RMs actually use them.
Measuring whether it worked
| Metric | What it measures | Direction of travel |
|---|---|---|
| Time to onboard | Days from mandate to first transaction | Down, sharply |
| KYC reuse rate | Share of new accounts using already-verified KYC | Up |
| RM client time | Share of RM time spent with clients | Up |
| Share of wallet | Captured share of client banking spend | Up |
| Forecast accuracy | Weighted pipeline against actual revenue | Up |
| Duplicate entity rate | Share of entities with more than one active record | Down |
Work these through before reading on. Full solutions are in the Answer Key.
5.1 Classify the failure
For each situation, name the risk category from the taxonomy in this part: an RM leaves for a competitor and three clients follow within six months; a client's new minority shareholder turns out to be on an EU sanctions list eight months after the change; a sales team sells complex interest-rate structures to mid-caps with no matching debt; a bank discovers that four apparently separate borrowers share a guarantor.
The departing RM is key-person dependency. The sanctioned shareholder is financial crime, made worse by a periodic rather than event-driven KYC process. The complex rate structures are conduct and mis-selling. The shared guarantor is credit and concentration risk, specifically a failure to build the connected-client group.
5.2 The good relationship
Client A generates 12 million euros of revenue, 85 per cent from lending, with RAROC of 8 per cent. Client B generates 3 million euros, 70 per cent from cash management and hedging, with RAROC of 22 per cent. Using the five tests in this part, explain which relationship is better and what the account plan for each should aim at.
Client B is the better relationship on four of the five tests: it earns well above its capital, its revenue is diversified, it is sticky because it is built on cash management and hedging, and it consumes little capital. Client A is larger but fails the capital test and is concentrated in lending, which moves at refinancing. The plan for A should aim to cross-sell transaction banking and hedging or reprice the lending; the plan for B should aim to grow carefully, adding lending only where it deepens the relationship without diluting the return, and to make the relationship institutional rather than dependent on one RM.
5.3 Triage the AI use cases
Rank these by value and by risk, and say which you would deploy first: drafting call reports from RM notes; automatically approving low-risk KYC refreshes; drafting RFP responses from a library of previous answers; recommending credit limits for mid-caps.
Call-report drafting is high value and low risk, because a human reviews every draft and it improves CRM data: deploy it first. RFP drafting from a governed library is high value and moderate risk: deploy second, with review. Automatically approving KYC refreshes is moderate value and high risk, because it removes the human from a regulated decision: use AI to triage and pre-draft instead. Recommending mid-cap credit limits is high risk, touches model-risk governance and, where personal credit is involved, the AI Act's high-risk category: keep it advisory, under formal model validation.
5.4 Sequence the programme
A bank wants to launch a generative-AI meeting-preparation tool for RMs in six months. Its client data sits in five systems with no shared identifier. Advise the programme sponsor in three sentences.
The tool will only be as good as the client data it summarises, and today it would brief RMs from five inconsistent sources. Start in parallel with a minimal entity-resolution layer linking the five systems for the pilot segment, and launch the tool for that segment only once its briefings can be trusted. Measure the pilot on RM time saved and briefing accuracy before extending it.
5.5 Carry it forward
Name one idea from each of Parts 2, 3 and 4 that explains a root cause in one of the three case studies.
From Part 2: KYC and beneficial-ownership identification, whose weakness is at the heart of Danske Bank. From Part 3: the RM as first line of defence, whose failure to act on visible risk signals marked Archegos. From Part 4: the risk-group hierarchy and exposure aggregation, whose absence let concentration hide in Greensill and Archegos.
Explain to a relative who has never worked in finance why a bank could lose billions on a client it thought it knew well. Use Archegos, and avoid the words "swap", "prime broker" and "margin call".
What are the two roots of most coverage risk?
Fragmented client data and manual process.
Name the five tests that distinguish a good relationship from a large one.
Does it earn its capital; is the revenue diversified; is it sticky; is the risk understood; is it institutional.
Which coverage failure does each of the three case studies illustrate?
Danske Bank: not knowing the clients (KYC and integration). Greensill: concentration on connected obligors not grouped or understood. Archegos: exposure not aggregated or acted on.
What is the difference between a green loan and a sustainability-linked loan?
A green loan finances eligible green projects; a sustainability-linked loan can fund any purpose but its margin moves with the borrower's performance against sustainability targets.
In what order should the three phases of coverage transformation be delivered, and why?
Data foundation, then process redesign, then adoption and intelligence, because each layer depends on the one beneath it.
5
What are the two roots of most coverage risk?
No.
No.
Right.
Not the roots identified.
Which case illustrates concentration on connected obligors that were not grouped?
KYC and integration.
Greensill is.
Exposure not aggregated or acted on.
Right. Danske is not knowing the clients; Archegos is exposure not aggregated or acted on.
Which test distinguishes a good relationship from a merely large one?
Not a test.
Size is what the test looks past.
Right. The others: diversified revenue, stickiness, understood risk, institutional relationship.
Not a test.
Module Summary — The Six Ideas That Matter Most
If you retain only a handful of things, retain these. Each compresses a whole part into a single, load-bearing idea.
Coverage owns the client and orchestrates the whole bank
The RM is the conductor: accountable for the relationship, reliant on Credit, product and operations for everything the client actually receives. The role is one of influence rather than authority, exercised inside a matrix in which coverage and product share — and deliberately double count — the same revenue. Segmentation is the operating model: the same bank runs bespoke global coverage for multinationals, regional lending-led coverage for mid-caps and automated coverage for small businesses, and no single process or system configuration fits all three.
The loan buys the relationship; cross-sell earns the return
Capital-heavy lending is the entry ticket. On its own a revolving facility often fails the return-on-capital test; the cash management, trade, hedging and capital-markets business sold around it is what lifts the relationship above the hurdle. Share of wallet is the scoreboard, RAROC is the gate every price must pass, and transaction banking is the rung that makes a relationship sticky.
The lifecycle is a loop, and KYC is its slowest gate
Prospect, onboard, plan, pitch, mandate, serve and review — then plan again. The account plan is the spine and the weighted pipeline is the forward book. Onboarding and KYC are where relationships are most often damaged, because the bank must peel the ownership onion down to real people, screen everyone involved and keep the file alive under perpetual KYC.
The RM is first line, on the private side
The person paid to grow revenue also owns its risk: credit quality, KYC completeness and the suitability of what is sold. Compliance can only review what the first line records. And because coverage regularly holds inside information, its people and its data sit behind the information barrier, with entitlements, logs and wall-crossing procedures to match.
Client data is the hardest problem
One economic group is many legal entities, represented in four hierarchies — legal, risk, economic and coverage — that never match and are owned by different functions. External identifiers such as the LEI and SIREN anchor a golden record that master data management maintains through entity resolution and stewardship. Profitability, pricing, risk aggregation, KYC reuse and AI all inherit the quality of that foundation.
Transformation is foundation, then process, then adoption
Most coverage risk has two roots: fragmented client data and manual process. Real failures show the cost of not knowing clients, not grouping exposure correctly and not seeing it in aggregate. The transformation themes — digital onboarding, the unified 360, AI-assisted coverage, ESG — succeed when they move cost or risk, and only in sequence: a shiny tool on fragmented data surfaces the fragmentation more visibly.
Process Cheat Sheet — The Relationship on One Page
A single reference for the coverage lifecycle. Everything below is developed in full in Parts 2 to 4.
| Stage | Timing | Core activity | Key output | Lead and systems |
|---|---|---|---|---|
| 1. Prospect and originate | Months | Target screening, idea-led calling, referrals | Qualified, engaged prospect | RM, sector team · CRM, external data |
| 2. Onboard and KYC | Weeks to months | Entity set-up, CDD, UBO, screening, initial rating, product enablement | Bankable client with limits | Onboarding, KYC, Risk · CLM, screening, MDM, credit |
| 3. Account planning | Annual, continuous | Who, needs, position, plan | Living account plan | RM with product · CRM, wallet analytics |
| 4. Pitch and structure | Days to weeks | Deal team, solution, RFP response, pre-clearance, RAROC | Proposal, approval to quote | RM, product, credit · CRM, pricing engine |
| 5. Mandate and hand-off | Weeks | Approvals, documentation, booking, revenue attribution | Executed, booked deal | Product, legal, operations · loan and product systems |
| 6. Serve and review | Continuous, annual review | Service, monitoring, KYC refresh, renewal | Reviewed relationship, refreshed plan | RM, Risk, service · 360, early-warning tools |
Key numbers to remember
| Metric | Rule of thumb |
|---|---|
| Banking relationships per French mid-cap | Commonly three to five |
| Beneficial-ownership threshold | More than 25 per cent, or control by other means |
| Periodic KYC review cycle (traditional) | About 1, 3 or 5 years by risk band |
| Capital held against risk-weighted assets | Around 11 to 13 per cent common equity for large European banks, including buffers |
| Typical hurdle rate | About 10 to 12 per cent, close to the cost of equity |
| Large-exposure limit | 25 per cent of Tier 1 capital per client or connected group |
| Pipeline stage weights (common scale) | 10 / 25 / 50 / 75 / 100 per cent |
| ETI definition (France) | 250 to 4,999 employees, turnover up to 1.5 billion euros or balance sheet up to 2 billion euros |
| LEI | 20 characters, ISO 17442 |
| SIREN / SIRET | 9 digits / 14 digits |
| RM time with clients | Often well under half, per industry surveys |
The one-sentence version
Coverage is the bank's single front door for a corporate client: one banker who knows the client, uses the balance sheet to open the relationship, grows it profitably by bringing the rest of the bank behind it, owns its risk as the first line of defence, and depends on a client record that is far harder to keep straight than it looks.
Sources and further reading
Regulatory positions and market facts stated in this module reflect sources available in mid-2026. Regulation in this field changes on a scale of months, so treat every date below as a fact with a shelf life and verify it before relying on it in client work.
How to read this section
Where the module states a legal obligation, it names a specific instrument, and the reference is given below so that it can be checked at source rather than taken on trust. Where the module gives a number that is not a market convention, its origin is identified. Figures explicitly flagged as illustrative — the Helion case throughout, and the wallet, pricing and pipeline examples — are invented to teach a mechanic and have no source, by design.
Primary legislation referenced
| Short name used in the module | Instrument | Relevance |
|---|---|---|
| AML directives | Directive (EU) 2015/849, as amended by Directive (EU) 2018/843 | Current customer due diligence and beneficial-ownership obligations |
| AML Regulation | Regulation (EU) 2024/1624 | Single EU rulebook on AML/CFT, applying from 10 July 2027 |
| AMLA Regulation | Regulation (EU) 2024/1620 | Establishes the EU Anti-Money Laundering Authority |
| Capital Requirements Regulation | Regulation (EU) No 575/2013, as amended by Regulation (EU) 2024/1623 | Capital, risk-weighted assets, large exposures, connected clients |
| MiFID II | Directive 2014/65/EU | Client classification, suitability, product governance |
| Market Abuse Regulation | Regulation (EU) No 596/2014 | Inside information, insider lists, market soundings |
| GDPR | Regulation (EU) 2016/679 | Personal data in CRM, KYC and the client 360 |
| DORA | Regulation (EU) 2022/2554 | ICT risk, resilience testing, third-party risk; applies from 17 January 2025 |
| AI Act | Regulation (EU) 2024/1689 | Risk-tiered obligations for AI systems |
| CSRD | Directive (EU) 2022/2464 | Corporate sustainability reporting |
| French AML provisions | Code monétaire et financier, articles L561-1 and following | French transposition of the AML directives |
Dated facts and where they come from
The final Basel III reforms began applying in the EU on 1 January 2025 through Regulation (EU) 2024/1623, with the output floor phased in over the following years.
The EU AML package was adopted in 2024. The AML Regulation applies from 10 July 2027; AMLA is based in Frankfurt and began operating in 2025.
The Court of Justice of the European Union ruled on 22 November 2022 (joined cases C-37/20 and C-601/20) that general public access to beneficial-ownership registers was incompatible with the Charter of Fundamental Rights; access for obliged entities continues.
BNP Paribas agreed in 2014 to pay about 8.9 billion dollars to US authorities in connection with violations of US sanctions, as published by the US Department of Justice.
Danske Bank's Estonian branch: the bank's own 2018 investigation report estimated about 200 billion euros of non-resident payments between 2007 and 2015; in December 2022 the bank pleaded guilty in the United States and agreed to pay about 2 billion dollars, as published by the US Department of Justice and the SEC.
Credit Suisse's supply-chain finance funds, managed with Greensill Capital, held about 10 billion dollars when they were suspended in March 2021; Greensill entered insolvency the same month.
Archegos: Credit Suisse reported a loss of about 5.5 billion dollars and Nomura about 2.9 billion. The independent report commissioned by Credit Suisse was published in July 2021.
The AI Act entered into force on 1 August 2024, with phased application from February 2025. Proposals to adjust the timing of the high-risk obligations were under discussion in 2025–2026; the European Commission maintains the authoritative timeline.
Where to go deeper
For the regulatory material, the consolidated texts on EUR-Lex are authoritative, and the supervisors' own guidance is more usable: the European Banking Authority for customer due diligence risk factors, connected clients and loan origination; the ECB and the ACPR for supervisory expectations in France; the AMF for conduct rules on markets products; the CNIL for personal data; and the Financial Action Task Force for international AML standards and its guidance on beneficial ownership.
For identifiers and master data, the Global LEI Foundation's documentation explains the LEI and its parent-relationship data, and INSEE documents the SIRENE database. For sustainable lending, the Loan Market Association publishes the Green Loan Principles and the Sustainability-Linked Loan Principles.
For the case studies, the primary sources are the published investigation reports — Bruun & Hjejle's report for Danske Bank, and the Paul, Weiss report for Credit Suisse on Archegos — together with the settlement announcements of the US Department of Justice. They are long, but they are the best available accounts of how coverage, risk and data failures combine in practice.