There is no “nature desk” in a bank. Nature and social issues reach a banker through a control function that sits beside the credit process and has a say in it: environmental and social risk management (ESRM), sometimes named ESG risk, sustainability risk or, in project finance, the E&S advisory team. At Meridian Bank it is a team of about 25 analysts attached to the Risk division, with three or four specialists embedded in the sectors where the files are heaviest: mining and metals, energy, agriculture and commodities, infrastructure. Their output is an opinion attached to a credit application, and their power is procedural: under Meridian’s rules, a file in a sensitive sector cannot go to credit committee without their sign-off or a documented escalation.
Who the “clients” are. The internal clients are the coverage bankers and the product desks (corporate lending, project finance, commodity trade finance, DCM, export finance). The external counterparties are the borrowers’ sustainability and procurement teams, independent environmental and social consultants, satellite-monitoring and supply-chain data vendors, certification bodies, and the non-governmental organisations that publish the allegations the team has to verify.
How it earns money. It does not. ESRM is a cost centre: with a fully loaded cost of about EUR 150,000 per analyst (a Meridian planning assumption), the team costs under EUR 4m a year. It protects revenue in three ways that never appear on a profit and loss line. It avoids credit losses on clients whose licence to operate, market access or supply chain breaks. It avoids the reputational and legal cost of being named as the financier of a deforestation or forced-labour case. And it keeps the bank eligible for mandates where the sponsor, the export credit agency or the development bank requires Equator Principles or IFC Performance Standards compliance. The product desks pay for the team through cost allocation, which is one reason bankers ask it for speed.
The adjacent functions. Coverage owns the client relationship and the first conversation. KYC and financial-crime compliance run sanctions, adverse-media and politically exposed person checks, which overlap with human-rights screening more than either team admits. Credit risk owns the rating and the recommendation. Legal drafts the undertakings. The reputational risk committee, chaired at Meridian by the chief risk officer with the heads of CIB, compliance and communications, takes the cases that policy cannot decide. The sustainability reporting team in Finance (see E06) consumes the same client data for the bank’s own disclosures.
A realistic week. Monday: triage of new files from the pipeline meeting; perhaps forty transactions, of which six fall under a sector policy and two need a full review. Tuesday: call with a palm-oil refiner’s sustainability head to go through its supplier grievance log line by line. Wednesday: reading the independent consultant’s environmental and social due diligence on a wind farm with a critical-habitat question (see B06). Thursday: an NGO report names a client; the team has 48 hours to give the coverage banker a factual note and a position. Friday: annual review of thirty existing clients against the exclusion list, and the monthly meeting where policy exceptions are logged. In November 2026 a good share of the week also goes to one date: 30 December 2026, when the EU Deforestation Regulation starts to apply to large and medium-sized companies (section 3.3).
WHY IT MATTERS ON THE DESK — A student joining lending, trade finance or DCM will not write the environmental and social opinion, but will be the one who fills in the screening form, chases the client’s data and explains to the client why a covenant on traceability sits next to the leverage covenant. Knowing what the reviewer needs saves weeks on a deal.
Before adding nature or people, recall how a bank decides to lend. The process has five documents, and every later section of this module hooks onto one of them.
IN ONE SENTENCE — Environmental and social risk management does not add a sixth document; it adds a section to each of the five.
The financial test behind all of it is whether the borrower can absorb a shock and still service its debt. For that the analyst looks at the cushion between cash generation and debt (leverage, interest cover) and at how fast that cushion shrinks when something goes wrong. The contribution of nature and social analysis is to enlarge the list of “something”.
The case of this module is an agricultural commodity trader, so its economics must be understood first (module B13 treats commodity finance in full). A trader buys physical goods from producers or intermediaries, moves them, sometimes processes them, and sells them to industrial buyers. Three features drive credit analysis.
Thin margins on large turnover. A trader earns a few per cent of the value of what it moves. Revenue is a poor measure of size; gross margin and EBITDA are the relevant quantities.
Self-liquidating debt. Most of a trader’s debt finances inventories that are hedged and can be sold within days, known as readily marketable inventories (RMI), and receivables from buyers. Analysts therefore compute an RMI-adjusted leverage: net debt minus RMI, divided by EBITDA. Only the “core” debt has to be repaid from earnings.
Secured, short-dated bank lines. The typical instrument is a borrowing-base facility: a revolving credit whose available amount is recalculated every month, or every week, as a percentage (the advance rate) of eligible inventories and receivables. What counts as eligible is defined in the contract. Section 5.7 will use exactly that definition to carry an environmental condition.
INTUITION — A trader is a lorry driver paid a small fee per tonne carried. The cargo is worth a fortune, but it is not his, and his income depends on being allowed to unload at destination. If the customs officer at the port of arrival refuses the cargo, the fee is lost, and so is the fuel.
FORMULA
is the volume traded in tonnes; the average price per tonne; the gross margin as a share of revenue; the operating expenses; the net debt left after deducting readily marketable inventories.
The simulator below uses a generic trader, not the client of the case: 5 million tonnes at EUR 600 a tonne is EUR 3,000m of revenue; a 3.0% gross margin gives EUR 90m; after EUR 50m of operating expenses, EBITDA is EUR 40m and, with EUR 150m of core debt, adjusted leverage is 3.75x. Lower the gross margin by half a point and EBITDA falls to EUR 25m: leverage jumps to 6.0x.
INTERACTIVE LAB — A trader’s cushion: how a small margin shock moves leverage
Compute revenue, gross margin, EBITDA and RMI-adjusted leverage of a generic commodity trader and see how sensitive leverage is to the gross margin.
| Parameter you can move | Range | Default |
|---|---|---|
| Volume traded | 1 to 10 Mt | 5 Mt |
| Average price | 200 to 1500 EUR/t | 600 EUR/t |
| Gross margin on revenue | 1 to 5 % | 3 % |
| Operating expenses | 10 to 100 EUR m | 50 EUR m |
| Core net debt (after RMI) | 0 to 400 EUR m | 150 EUR m |
| Result at default settings | Value |
|---|---|
| Revenue | 3,000 EUR m |
| Gross margin | 90.0 EUR m |
| EBITDA | 40.0 EUR m |
| RMI-adjusted leverage | 3.75 x |
| EBITDA as share of revenue | 1.33 % |
Try it. Lower the gross margin from 3.0% to 2.5%: a shock of half a point of revenue takes leverage from 3.75x to 6.0x. Keep this elasticity in mind for section 5.4.
PITFALL — Judging a trader on total debt to EBITDA. Unadjusted leverage of 12x or 15x is normal when inventories are financed by banks, and says little. The analyst’s questions are whether the inventories are really marketable and hedged, and how much margin stands between the core debt and a bad year.
Internal rating models for large corporates combine a quantitative score with a qualitative overlay: market position, management quality, regulatory environment, concentration of customers and suppliers. Nature and social factors have always been there under other names (“regulatory risk”, “supply risk”, “litigation”). What has changed since about 2020 is that supervisors ask banks to identify them explicitly (EBA Guidelines on the management of ESG risks, applicable since 11 January 2026 for large institutions; see A02) and that regulations such as the EUDR have turned a reputational question into one of market access with a date attached. The rest of the module follows that chain: from ecosystems and people to a cash flow, from the cash flow to a PD, and from the PD to the bank’s capital.
IN ONE SENTENCE — A company depends on nature when it needs an ecosystem service to operate, and has an impact on nature when its activity changes the state of an ecosystem; a bank inherits both through its clients.
Ecosystem services are the flows that nature provides free of charge: provisioning services (water, timber, fish, genetic material), regulating services (pollination, flood protection, climate and rainfall regulation, soil quality) and cultural services. The five direct drivers of nature loss identified by the IPBES global assessment are land- and sea-use change, direct exploitation, climate change, pollution and invasive species. Climate is thus one of five drivers, and carbon accounting captures only part of the picture: a soy field emits little carbon once it is established, but it has replaced a savannah.
INTUITION — Think of a restaurant. It depends on the water supply, and it has an impact on the neighbours through noise and waste. The landlord’s risk comes from both: no water, no business; too many complaints, no licence. A bank is the landlord of many restaurants.
The two directions map onto the double materiality of A01. Dependencies create mostly physical risk (a drought cuts the harvest and the river-borne logistics at once). Impacts create transition risk (a regulation closes a market to products linked to deforestation) and liability risk (a court holds a company responsible). There is also a systemic component with no climate equivalent in the standard toolkit: the collapse of an ecosystem is local, non-linear and cannot be offset by an improvement elsewhere. A tonne of CO2 is the same everywhere; a hectare of forest is not.
The supervisory headline number comes from the ECB: nearly 75% of euro-area banks’ corporate lending goes to firms that are highly dependent on at least one ecosystem service (ECB analysis first published in 2023 and repeated by Frank Elderson in a speech of 9 March 2026). The figure is striking and has to be read with care: “highly dependent on at least one service” is a low bar, and the number is a dependency count, not a loss estimate.
TNFD. The Taskforce on Nature-related Financial Disclosures published its recommendations in September 2023: 14 recommended disclosures under the four pillars inherited from the climate taskforce (governance, strategy, risk and impact management, metrics and targets). By 16 October 2026, 836 organisations in 57 countries had committed to report in line with it. The TNFD has wound down its own technical programme during 2026 and handed the baton to the standard setters: on 22 October 2026, during the COP17 biodiversity conference in Yerevan, the ISSB opened a consultation on a nature-related Practice Statement that draws on the TNFD framework and would sit on top of IFRS S1 and S2 (comments due by 19 February 2027). In the EU, sustainability reporting under the CSRD covers nature through ESRS E2 to E5 (pollution, water, biodiversity and ecosystems, circular economy).
LEAP. The assessment method proposed by the TNFD has four steps and is what an ESRM analyst actually does, whether or not the bank publishes anything:
| Step | Question | Output for a banker |
|---|---|---|
| Locate | Where does the client touch nature? | List of sites and sourcing regions, flagged if in sensitive areas |
| Evaluate | Which dependencies and impacts? | Heat map by activity and ecosystem service |
| Assess | Which risks and opportunities follow? | Financial effect: revenue, cost, capex, access to market |
| Prepare | What to do and to disclose? | Conditions, covenants, engagement plan, reporting |
The “Locate” step is the novelty for bankers used to climate work. Greenhouse gases can be estimated from a sector average; nature cannot, because the same activity is harmless in one place and destructive in another. Location data (plots, mills, mines, water basins) therefore become credit information.
ENCORE. For the “Evaluate” step, the free ENCORE database (maintained by Global Canopy, UNEP FI and UNEP-WCMC) rates, for each economic activity, the materiality of its dependencies on ecosystem services and of its pressures on nature, from very low to very high. Banks use it to build a first heat map of a portfolio by sector code. It is a screening tool at sector level: two soy traders receive the same ENCORE rating whatever their sourcing practices.
Footprint metrics. To add nature impacts up across a portfolio, as tonnes of CO2 equivalent do for climate, practitioners use an intactness metric. The most common is the mean species abundance (MSA), developed with the GLOBIO model.
FORMULA
is the number of native species in the reference (undisturbed) state; the current abundance of species ; its abundance in the reference state. The minimum caps each ratio at 1, so that a species that thrives on disturbance does not compensate for those that vanish. MSA is 1 for an intact ecosystem and 0 for a fully destroyed one.
A related metric from life-cycle assessment is the potentially disappeared fraction of species (PDF): the share of species lost, whatever happens to the abundance of the others. Both are then multiplied by an area, giving MSA.km² or PDF.m².yr: 1 MSA.km² lost is equivalent to one square kilometre of pristine nature turned into a car park. Commercial tools (the Global Biodiversity Score of CDC Biodiversité, the Corporate Biodiversity Footprint of Iceberg Data Lab, the Biodiversity Footprint for Financial Institutions) apply such coefficients to companies’ activities, and the Partnership for Biodiversity Accounting Financials (PBAF) publishes a standard on how a financial institution attributes the result to its loans and investments.
The simulator takes a plot with five native species. Reference abundances are 40, 20, 10, 8 and 2 individuals; today they are 30, 20, 2, 0 and 5. The capped ratios are 0.75, 1, 0.2, 0 and 1, so MSA is 2.95 / 5 = 0.59: the plot has lost 41% of its intactness. One species of five has gone, so PDF is 20%.
INTERACTIVE LAB — Mean species abundance of a plot
Compute MSA and PDF from current abundances against a fixed reference state (40, 20, 10, 8 and 2 individuals for species A to E).
| Parameter you can move | Range | Default |
|---|---|---|
| Species A today (reference 40) | 0 to 60 | 30 |
| Species B today (reference 20) | 0 to 40 | 20 |
| Species C today (reference 10) | 0 to 20 | 2 |
| Species D today (reference 8) | 0 to 16 | 0 |
| Species E today (reference 2) | 0 to 10 | 5 |
| Result at default settings | Value |
|---|---|
| Mean species abundance | 0.590 |
| Intactness lost (1 - MSA) | 41.0 % |
| Potentially disappeared fraction | 20.0 % |
Try it. Raise species E to 10 (five times its reference): MSA does not move, because each ratio is capped at 1. Then set species C to 0: PDF rises to 40% and MSA falls to 0.55.
PITFALL — Reading MSA as a measure of biological richness. A desert and a rainforest can both score 1. MSA measures distance from the undisturbed state, which is why a footprint in MSA.km² must be read with the location beside it: one square kilometre lost in a biodiversity hotspot is not the same loss as one in a wheat plain, although the metric says so.
EUDR. Regulation (EU) 2023/1115 forbids placing on the EU market, or exporting from it, seven commodities and listed derived products (cattle, cocoa, coffee, oil palm, rubber, soya, wood) unless they are (i) deforestation-free, meaning produced on land not deforested after the cut-off date of 31 December 2020, (ii) produced in accordance with the laws of the country of production, and (iii) covered by a due diligence statement. The operator that first places the product on the market must collect the geolocation of every plot of production, assess the risk and file the statement in the Commission’s information system before customs clearance. Member States must set maximum fines of at least 4% of the operator’s annual EU turnover, and goods can be confiscated.
The calendar has moved twice. The original date of 30 December 2024 became 30 December 2025, then Regulation (EU) 2025/2650 of 19 December 2025 set 30 December 2026 for large and medium-sized operators and 30 June 2027 for micro and small ones. The same amendment concentrated the obligation on the operator that first places the product on the market: downstream companies no longer file their own statements, and micro and small primary producers in low-risk countries make a one-off simplified declaration. The Commission’s simplification review, published on 4 May 2026, concluded that the regulation would not be reopened and that the dates stand; it came with updated guidance and a draft delegated act adjusting the product list. The country benchmarking adopted on 22 May 2025 classes four countries as high risk (Belarus, Myanmar, North Korea, Russia), 140 as low risk, and the rest as standard risk, including Brazil, Indonesia and Malaysia. The class sets the share of operators that authorities must check each year (1%, 3% or 9%) and gives access to simplified due diligence for low-risk origins.
Banks are not operators under the EUDR and carry no direct obligation. They are exposed through their clients: a cargo that cannot be cleared is collateral that cannot be sold in its intended market.
CSDDD. Directive (EU) 2024/1760 on corporate sustainability due diligence obliges very large companies to identify, prevent, mitigate and remedy adverse human-rights and environmental impacts in their own operations, their subsidiaries and their chain of activities. The Omnibus I directive, published in the Official Journal on 26 February 2026, narrowed it sharply. As it stands in November 2026:
| Item | CSDDD after Omnibus I |
|---|---|
| Scope, EU companies | More than 5,000 employees and EUR 1.5bn of net worldwide turnover |
| Scope, non-EU companies | More than EUR 1.5bn of net turnover in the EU |
| Dates | Transposition by 26 July 2028; application from 26 July 2029 |
| Method | Scoping on reasonably available information; focus on direct partners |
| Enforcement | Fines capped at 3% of worldwide turnover; no EU-wide civil liability regime |
The obligation to adopt a climate transition plan has been removed from the directive. For banks, two points matter. First, financial undertakings above the thresholds are covered for their own operations and their suppliers, but the services they provide to clients (lending, underwriting) stay outside the chain of activities, and the clause that would have had the Commission reconsider that exclusion has been deleted. Second, the bank’s large clients are in scope, and their due diligence systems are what an ESRM analyst will review.
Other texts complete the picture: the EU Forced Labour Regulation (EU) 2024/3015, which bans products made with forced labour from the EU market from 14 December 2027; national laws such as the French duty-of-vigilance law of 2017 and the German supply-chain act; and, behind all of them, the UN Guiding Principles on Business and Human Rights (2011) and the OECD Guidelines for Multinational Enterprises, with the OECD’s 2019 guidance on due diligence for corporate lending and securities underwriting. Under these soft-law texts a bank can be “directly linked” to an adverse impact through its client and is expected to use its leverage; that expectation does not depend on the CSDDD.
The Equator Principles (current version EP4, in effect since 1 October 2020) are a risk management framework adopted by banks for project-related finance. They apply to project finance and project finance advisory where total capital costs are USD 10m or more, to project-related corporate loans where the total loan and the bank’s own commitment each reach USD 50m with a tenor of at least two years, and to related bridge loans, refinancing and acquisition financing. Projects are categorised A (significant adverse impacts), B (limited) or C (minimal). Ten principles follow: assessment, applicable standards, management system and action plan, stakeholder engagement, grievance mechanism, independent review, covenants, independent monitoring, reporting. In “designated countries” (a list of 35 as at August 2026) host-country law is the reference; elsewhere the reference is the IFC Performance Standards.
The eight Performance Standards (2012) are the common language of the profession: PS1 assessment and management system; PS2 labour and working conditions; PS3 resource efficiency and pollution prevention; PS4 community health, safety and security; PS5 land acquisition and involuntary resettlement; PS6 biodiversity conservation; PS7 indigenous peoples; PS8 cultural heritage. Two notions recur in credit files. Under PS6, a project in natural habitat must achieve no net loss of biodiversity where feasible, and one in critical habitat a net gain. Under PS7, certain projects need the free, prior and informed consent (FPIC) of affected indigenous peoples: consent given without coercion, before the decision, on full information. FPIC is a process with evidence (meetings, agreements, a record of dissent), and the consultant’s report must say whether it was achieved, which a lender cannot verify from its office.
PITFALL — Believing that the Equator Principles cover the bank’s exposure to nature and social risk. They cover project-related finance. A general-purpose revolving facility to a trader, a miner or an agribusiness group is outside their scope, and that is where most of the exposure sits. For those, the bank relies on its own sector policies. Membership has also become less universal: the four largest US banks left the association in early 2024.
Sector policies and exclusion lists are the bank’s home-made law. Meridian’s agricultural commodities policy, typical of European banks, has three layers: exclusions (no financing of clients that convert primary forest, high conservation value areas or peatland, or that use forced or child labour without remediation); conditions (membership of a recognised certification scheme such as RSPO for palm oil, a public no-deforestation, no-peat, no-exploitation commitment, a traceability target with a date); and escalation triggers (credible allegations, loss of certification, litigation).
On the product side, the social equivalent of green finance is the social bond or loan: a use-of-proceeds instrument (see A04) whose proceeds finance projects with positive social outcomes for a defined target population. The ICMA Social Bond Principles (current edition June 2025) and the Social Loan Principles of the LMA, APLMA and LSTA (updated 26 March 2025) share four components: use of proceeds, project evaluation and selection, management of proceeds, reporting. Typical categories are affordable basic infrastructure, access to essential services (health, education, financial services), affordable housing, employment generation including through SME and microfinance lending, food security and socio-economic advancement. The market was built by public issuers: the EU’s SURE programme alone issued EUR 98.4bn of social bonds between 2020 and 2022 to fund short-time work schemes. For a bank, the largest social use of proceeds is often its own book: loans to SMEs in disadvantaged regions, social housing, hospitals, refinanced by a social bond of the treasury (see C15).
Financial inclusion is the oldest social theme in banking: access to an account, to payments, to credit and to insurance for people and small firms that the standard offer leaves out. It appears in target populations (women entrepreneurs, smallholders, unbanked households) and in sustainability-linked KPIs, for example the number of smallholders given access to pre-harvest finance.
Just transition means that the cost of the environmental transition should not fall on those least able to bear it: workers of a coal plant that closes, farmers excluded from a supply chain because they cannot produce a geolocation file. The reference text is the ILO’s 2015 guidelines. It matters in this module for a precise reason: a traceability requirement can be met either by helping smallholders to comply or by dropping them. The second route is cheaper and deforestation-free on paper, and it moves the social problem out of sight.
Three product families are developed in C14 and only placed on the map here. Blue finance applies the use-of-proceeds format to the ocean and water economy (sustainable fisheries, port pollution control, wastewater treatment), following a practitioner’s guide published in 2023 by ICMA, IFC and UN bodies; ICMA added a guide on sustainable bonds for nature in June 2025. Debt-for-nature swaps refinance a sovereign’s expensive debt with cheaper, credit-enhanced funding and redirect part of the saving to conservation; Belize (2021), Ecuador for the Galápagos (2023) and Gabon (2023) are the reference deals. Biodiversity credits are certificates representing a measured gain or avoided loss of biodiversity. The voluntary market remains very small; the International Advisory Panel on Biodiversity Credits published a framework for high-integrity markets in October 2024, the European Commission a roadmap towards nature credits in July 2025, and England has run a compliance scheme since February 2024 that requires a 10% biodiversity net gain from developers. For the lender of this module, the relevant point is negative: none of these instruments can be used by a borrower to offset deforestation in its supply chain under the EUDR, which knows no compensation mechanism.
The lifecycle below is that of an environmental and social review attached to a corporate credit, from the first screening to the annual monitoring. Elapsed time for a full review is three to six weeks; the constraint is nearly always the client’s data.
INTERACTIVE LAB — Environmental and social review of a corporate credit, from screening to monitoring
Follow the file from the coverage banker’s first form to the annual review, with the document produced at each step.
| # | Who | Does what | Deliverable |
|---|---|---|---|
| 1 | Coverage banker | Fills in the E&S screening form at pipeline stage: sector code, countries of operation and sourcing, transaction type | Screening form with automatic flags (sector policy, Equator Principles scope, high-risk country) |
| 2 | KYC and compliance | Runs sanctions, adverse-media and litigation searches on the group and its main subsidiaries | KYC file with controversy list |
| 3 | ESRM analyst | Triage: assigns a review level (none, standard, enhanced) from the flags and the controversy list | Triage decision logged in the credit workflow |
| 4 | ESRM analyst | Sends the sector questionnaire and data request: sourcing volumes by origin, traceability, certifications, grievance log, policies | Client questionnaire and data room index |
| 5 | Client sustainability and procurement teams | Return data; hold a due diligence call with coverage and ESRM | Completed questionnaire, traceability report, call minutes |
| 6 | ESRM analyst | Locates sourcing areas, cross-checks against satellite deforestation alerts, official embargo and forced-labour lists, certification databases | Location and exposure note |
| 7 | ESRM analyst | Scores inherent risk and management quality; quantifies the financial scenario with the credit analyst | E&S opinion with residual risk rating and proposed conditions |
| 8 | Credit analyst | Integrates the scenario in the rating overlay and the sensitivity section; computes RAROC | Credit application with E&S section |
| 9 | Reputational risk committee | Decides on files rated high residual risk or deviating from policy | Committee minute: approve, approve with conditions, decline |
| 10 | Credit committee | Approves the transaction with the E&S conditions as conditions of approval | Credit decision |
| 11 | Legal and loan structuring | Negotiate conditions precedent, information undertakings, eligibility criteria and the action plan schedule with the borrower and the agent | Facility agreement and E&S action plan |
| 12 | Loan agency and middle office | Check conditions precedent before first drawing; collect periodic certificates and borrowing-base reports | CP checklist, compliance certificates |
| 13 | ESRM analyst and coverage banker | Annual review and event-driven review after any allegation; track action plan milestones | Monitoring report; waiver or escalation request if a milestone is missed |
Screening and triage (steps 1 to 3). The screening form is the control on which everything rests: a file that is mis-coded never reaches the analyst. Meridian’s form derives the flags from three fields (NACE code, countries, product), so that the banker cannot answer “no” to “is this a sensitive sector?”. The triage rule is written in the policy: any agricultural commodity trader with sourcing in a standard- or high-risk country gets an enhanced review.
Data collection (steps 4 and 5). The questionnaire asks for facts with units, not for policies: tonnes by origin, share traceable to the plot, share certified and under which supply-chain model, number of grievances opened and closed. The call matters as much as the paper; the analyst is testing whether the people in front of him know their supply chain.
Verification (step 6). The analyst never audits farms. He triangulates: the client’s mill or cooperative list against public deforestation alerts and concession maps; supplier names against official lists (in Brazil, the federal register of employers found using labour in conditions analogous to slavery, known as the “lista suja”, and the environmental agency’s embargoed areas); certificates against the scheme’s database. Vendors sell this as a service; the judgement stays in the bank.
Opinion and credit integration (steps 7 and 8). The opinion is two pages: findings, residual rating, conditions. The step that most banks still do poorly is the hand-over to the credit analyst, so that the finding becomes a number in the sensitivity table and not a paragraph nobody reads. Section 5 does it in full.
Decision (steps 9 and 10). The reputational risk committee does not approve credit; it decides whether the bank is willing to be associated with the client and under which conditions. The credit committee then decides on the risk and the return.
Documentation and monitoring (steps 11 to 13). Conditions that are not in the facility agreement do not exist. In a syndicated facility they have to be accepted by the other lenders, which caps their ambition at what the syndicate will carry, unless Meridian obtains a side letter. Monitoring is where the process is most often weak: milestones slip, the analyst has moved to another file, and nobody wants to call a default over a traceability percentage.
Question. Who is the client, what is asked of Meridian, and what does the file look like before any environmental or social analysis?
Data (case assumptions). Cerrado Atlantic Commodities SA (CAC) is a privately held trader based in Geneva. It originates four commodities and sells them to crushers, refiners, chocolate makers and roasters, through CAC Europe BV in Rotterdam for its European customers. In November 2026 CAC is renewing its two-year, EUR 800m syndicated borrowing-base revolving facility, and the coordinating bank invites Meridian to take EUR 100m. Signing is planned for mid-December, two weeks before the EUDR starts to apply to CAC Europe BV.
| Commodity | Volume (t) | Price (EUR/t) | Gross margin | Sold in the EU |
|---|---|---|---|---|
| Soy (beans and meal) | 4,000,000 | 380 | 2.5% | 35% |
| Palm oil | 800,000 | 900 | 3.0% | 40% |
| Cocoa beans | 120,000 | 5,000 | 4.0% | 70% |
| Coffee | 150,000 | 4,000 | 4.4% | 60% |
Operating expenses are EUR 55m. Total debt is EUR 700m, of which EUR 520m finances readily marketable inventories, leaving a core net debt of EUR 180m that costs 6%. Equity is EUR 450m. Meridian’s internal rating is BB (PD 1.00%), under a scorecard in which RMI-adjusted leverage below 3.5x is consistent with BB, between 3.5x and 4.5x with BB- (PD 1.60%), between 4.5x and 5.5x with B+ (PD 2.50%) and above with B (PD 4.00%).
Calculation. Revenue is volume times price for each line: EUR 1,520m for soy, EUR 720m for palm oil, EUR 600m for cocoa and EUR 600m for coffee, EUR 3,440m in total. EU sales are 35% of 1,520 plus 40% of 720 plus 70% of 600 plus 60% of 600, that is 532 + 288 + 420 + 360 = EUR 1,600m, or 46.5% of revenue. Gross margin is 38.0 + 21.6 + 24.0 + 26.4 = EUR 110.0m, EBITDA EUR 55.0m, adjusted leverage 180 / 55 = 3.27x, and interest cover on core debt 55 / 10.8 = 5.09x.
INTERACTIVE LAB — CAC at a glance: revenue, EU exposure and credit ratios
Rebuild the financial profile of the trader from its four commodity lines and read the two ratios the rating depends on.
| Parameter you can move | Range | Default |
|---|---|---|
| Soy volume | 2 to 6 Mt | 4 Mt |
| Cocoa price | 2500 to 9000 EUR/t | 5000 EUR/t |
| Operating expenses | 40 to 80 EUR m | 55 EUR m |
| Core net debt (after RMI) | 100 to 300 EUR m | 180 EUR m |
| Result at default settings | Value |
|---|---|
| Revenue | 3,440 EUR m |
| Revenue from EU sales | 1,600 EUR m |
| EU share of revenue | 46.5 % |
| Gross margin | 110.0 EUR m |
| EBITDA | 55.0 EUR m |
| RMI-adjusted leverage | 3.27 x |
| Interest cover on core debt (6%) | 5.09 x |
| Total volume | 5.07 Mt |
Try it. Move the cocoa price to EUR 9,000 a tonne: a commodity that is 2% of the tonnage becomes a quarter of revenue. Size in tonnes and size in euros give different pictures, and section 5.3 adds a third one, size in hectares.
Result and reading. On financials alone this is a standard BB trader, comfortably inside its grade. Two facts stand out for what follows. The EU is 46.5% of sales, and EUR 1,600m of EU turnover puts CAC above the EUR 1.5bn threshold that will bring non-EU companies within the CSDDD from July 2029. And the product mix is exactly the EUDR list: every line of the table is a regulated commodity. The screening form returns three flags (agricultural commodities policy, standard-risk sourcing countries, adverse media), so the file goes to enhanced review. The Equator Principles do not apply: the facility is not project-related.
Question. Which part of CAC’s volume could be linked to deforestation or conversion, and how much of it can the company trace?
Data. The ESRM analyst starts with a heat map by commodity, built from ENCORE-type sector ratings, the EUDR country classes and Meridian’s policy, on a scale of 1 (low) to 5 (very high). The inherent score of a line is the higher of its nature and social scores.
| Commodity | Main origins | Nature issue | Social issue | Inherent |
|---|---|---|---|---|
| Soy | Brazil (Mato Grosso, Matopiba) | Conversion of Cerrado savannah, forest edge | Land conflicts, labour on farms | 4 |
| Palm oil | Indonesia, Malaysia | Forest and peat conversion | Forced labour of migrant workers, FPIC | 5 |
| Cocoa | Côte d’Ivoire, Ghana | Encroachment in classified forests | Child labour, farmer income | 5 |
| Coffee | Brazil, Vietnam, Ethiopia | Local forest loss, water | Seasonal labour | 3 |
CAC’s questionnaire gives, for each commodity, the share of volume traceable to the plot of production (with geolocation) and the analyst estimates, from mill, silo and cooperative locations, the share coming from high-risk landscapes, meaning municipalities or districts with recorded deforestation or conversion since 2020.
| Commodity | Traceable to plot | From high-risk landscapes |
|---|---|---|
| Soy | 55% | 30% |
| Palm oil | 60% | 25% |
| Cocoa | 45% | 40% |
| Coffee | 65% | 10% |
Calculation. The quantity of interest is the volume that is both from a high-risk landscape and untraceable; for that volume, nobody can say whether it is clean. With no better information, the first estimate assumes that traceability is the same inside and outside high-risk landscapes:
FORMULA
is the unverified high-risk volume of commodity ; its volume; the share from high-risk landscapes; the share traceable to plot; a concentration factor, equal to 1 when traceability is as good in high-risk areas as elsewhere and above 1 when the untraceable volume concentrates there.
With : soy 4,000,000 × 30% × 45% = 540,000 t; palm oil 800,000 × 25% × 40% = 80,000 t; cocoa 120,000 × 40% × 55% = 26,400 t; coffee 150,000 × 10% × 35% = 5,250 t. The total is 651,650 t, or 12.9% of the 5.07 Mt traded. Valued at each commodity’s price it represents EUR 430.2m, 12.5% of revenue. Weighted by revenue, 56.0% of CAC’s business is traceable to plot.
INTERACTIVE LAB — Unverified volume from high-risk landscapes
Combine origin and traceability data to estimate, by commodity, the volume that is both from a high-risk landscape and untraceable to the plot.
| Parameter you can move | Range | Default |
|---|---|---|
| Concentration of untraceable volume in high-risk areas (k) | 1 to 2 | 1 |
| Soy traceable to plot | 0 to 100 % | 55 % |
| Cocoa traceable to plot | 0 to 100 % | 45 % |
| Result at default settings | Value |
|---|---|
| Soy: unverified high-risk volume | 540,000 t |
| Palm oil: unverified high-risk volume | 80,000 t |
| Cocoa: unverified high-risk volume | 26,400 t |
| Coffee: unverified high-risk volume | 5,250 t |
| Total unverified high-risk volume | 651,650 t |
| Share of total tonnage | 12.85 % |
| Value of that volume | 430.2 EUR m |
| Share of revenue | 12.51 % |
| Revenue-weighted share traceable to plot | 56.05 % |
Try it. Set k to 1.5: if the untraceable tonnes concentrate where the forest is being cleared, the unverified volume rises by half, to 977,475 t. The client’s average traceability figure cannot tell you which case you are in.
Result and reading. One tonne in eight cannot be shown to be clean. The pitfall lies in the assumption . Traceability is weakest where intermediaries aggregate the crop of many small farms, and those are the expanding frontiers. With the figure becomes 977,475 t. The analyst therefore asks CAC for traceability by origin, not the company-wide average, and that request becomes an information undertaking in section 5.6.
PITFALL — Accepting “traceable to the mill” or “traceable to the cooperative” as traceability. The EUDR requires the geolocation of the plots. A palm-oil mill buys fruit within a radius of several tens of kilometres from hundreds of suppliers; a cocoa cooperative may resell beans bought outside its membership. The questionnaire must name the level.
Question. How much nature does CAC’s sourcing occupy, and what would Meridian report as its financed share?
Data. The analyst uses a land-use footprint, the simplest of the MSA approaches: the area needed to grow the volume, multiplied by the intactness lost on that land. Yields and MSA loss coefficients are case assumptions of the right order of magnitude (intensive annual crops retain about 10% of the original abundance; perennial and partly shaded crops more): soy 3.5 t/ha and a loss of 0.9; palm oil 3.6 t of oil per hectare and 0.8; cocoa 0.5 t/ha and 0.7; coffee 1.5 t/ha and 0.7.
FORMULA
is the static footprint in MSA.km² once areas are converted from hectares (100 ha = 1 km²); the yield in tonnes per hectare; the mean species abundance remaining on land under crop ; the attribution factor is the PCAF convention for unlisted companies (see A03): the bank’s outstanding amount over the borrower’s equity plus debt.
Calculation. Soy: 4,000,000 / 3.5 = 1,142,857 ha, or 11,429 km², times 0.9 = 10,286 MSA.km². Palm oil: 2,222 km² and 1,778 MSA.km². Cocoa: 120,000 / 0.5 = 240,000 ha, or 2,400 km², times 0.7 = 1,680 MSA.km². Coffee: 1,000 km² and 700 MSA.km². The total is 17,051 km² of sourcing area and 14,443 MSA.km² of footprint. Meridian would have EUR 60m drawn on average (60% of EUR 100m) against EUR 1,150m of equity and debt: an attribution factor of 5.22% and a financed footprint of 753.6 MSA.km², or 12.6 MSA.km² per EUR million lent.
INTERACTIVE LAB — Land-use biodiversity footprint of the sourcing and the bank’s attributed share
Convert traded volumes into hectares and MSA.km², then attribute the footprint to Meridian’s drawn exposure with the PCAF factor.
| Parameter you can move | Range | Default |
|---|---|---|
| Meridian’s average drawn amount | 0 to 100 EUR m | 60 EUR m |
| Cocoa yield | 0.3 to 1.2 t/ha | 0.5 t/ha |
| Intactness lost under soy (1 - MSA) | 0.5 to 1 | 0.9 |
| Annual conversion rate in high-risk landscapes | 0 to 3 % | 1 % |
| Result at default settings | Value |
|---|---|
| Sourcing area | 17,051 km² |
| Footprint: soy | 10,286 MSA.km² |
| Footprint: palm oil | 1,778 MSA.km² |
| Footprint: cocoa | 1,680 MSA.km² |
| Footprint: coffee | 700 MSA.km² |
| Total static footprint | 14,443 MSA.km² |
| Attribution factor | 5.22 % |
| Footprint attributed to Meridian | 753.6 MSA.km² |
| Attributed footprint per EUR m drawn | 12.56 MSA.km² |
| Cocoa share of footprint | 11.6 % |
| Soy: hectares exposed to conversion each year in unverified high-risk supply | 1,543 ha |
Try it. Double the cocoa yield to 1.0 t/ha: its footprint halves. Yield improvement on existing land is the main lever by which a trader’s programme can reduce pressure on forests, and it is also a farmer-income lever.
Result and reading. Three lessons. First, the ranking changes with the unit: cocoa is 2.4% of the tonnage and 11.6% of the footprint, because its yield is low. Second, this is a static footprint: it describes land already converted, much of it decades ago, and says nothing about whether CAC’s purchases cause new loss. The dynamic part is what the EUDR targets. As an order of magnitude, the 540,000 t of unverified high-risk soy occupy about 154,000 ha; if 1% of such landscapes is converted each year (a case assumption), about 1,543 ha a year of new conversion sit somewhere in that unverified supply. Third, the attribution is fragile: a trader does not farm, so the same hectares are also in the footprint of the farmers and of their own lenders, and the result depends on drawings at the reporting date. The figure is useful to rank clients inside a portfolio and to follow one client over time, and it cannot be added to the footprints of other links of the chain.
WHY IT MATTERS ON THE DESK — The footprint will not decide the loan. It tells the analyst where to look (cocoa deserves as much attention as palm oil despite its small tonnage) and gives the bank a number for its own nature reporting. The credit decision turns on the next stage.
Question. What happens to CAC’s credit profile when the EUDR applies on 30 December 2026?
Data. From that date CAC Europe BV may place on the EU market only volumes for which it can file a due diligence statement with plot geolocation. CAC can direct its traceable volumes to Europe and sell the rest elsewhere, but not perfectly: lots are blended in silos and tanks, and segregated logistics do not exist on every route. CAC’s head of logistics estimates that 80% of the traceable volume can be routed to the EU. Two further assumptions come from the due diligence call: running costs of compliance (data collection from suppliers, segregation, verification, statements) of 0.5% of EU revenue, and a discount of 3% on volumes that have to be diverted from the EU to less demanding markets, where they compete with every other trader’s non-compliant tonnes.
Calculation. For each commodity, the gap is the EU share of sales minus the routable traceable share, if positive:
FORMULA
is the volume that CAC wants to sell in the EU and cannot; the EU share of sales; the share traceable to plot; the routable fraction of traceable volume.
Soy: 80% × 55% = 44%, above the 35% sold in the EU; no gap. Palm oil: 48% against 40%; no gap. Cocoa: 80% × 45% = 36% against 70%; the gap is 34% of 120,000 t, or 40,800 t, worth EUR 204m. Coffee: 52% against 60%; the gap is 8% of 150,000 t, or 12,000 t, worth EUR 48m. Revenue that must be diverted is EUR 252m, 15.75% of EU sales. The cost of compliance is 0.5% × 1,600 = EUR 8.0m; the diversion discount is 3% × 252 = EUR 7.56m. EBITDA falls from EUR 55.0m to EUR 39.44m, a drop of 28.3%. Adjusted leverage rises to 180 / 39.44 = 4.56x, which on Meridian’s scorecard is B+ and a PD of 2.50%.
INTERACTIVE LAB — EUDR market-access stress on the trader’s EBITDA and rating
Turn the traceability gap into diverted volumes, lost margin and compliance cost, then into leverage and the PD of the scorecard.
| Parameter you can move | Range | Default |
|---|---|---|
| Share of traceable volume that can be routed to the EU | 50 to 100 % | 80 % |
| Running compliance cost | 0 to 1.5 % of EU revenue | 0.5 % of EU revenue |
| Price discount on diverted volume | 0 to 10 % | 3 % |
| Cocoa traceable to plot | 0 to 100 % | 45 % |
| Fine | 0 to 4 % of EU turnover | 0 % of EU turnover |
| Result at default settings | Value |
|---|---|
| Cocoa volume that cannot enter the EU | 40,800 t |
| Coffee volume that cannot enter the EU | 12,000 t |
| Revenue to be diverted | 252.0 EUR m |
| Share of EU revenue | 15.75 % |
| Compliance cost | 8.00 EUR m |
| Diversion discount | 7.56 EUR m |
| EBITDA after stress | 39.44 EUR m |
| Change in EBITDA | -28.3 % |
| RMI-adjusted leverage after stress | 4.56 x |
| Scorecard PD after stress | 2.50 % |
Try it. Raise cocoa traceability to 90%: the cocoa gap closes, EBITDA recovers to EUR 45.6m and leverage to 3.95x. Then set the fine to 4% of EU turnover: EUR 64m, more than a year of EBITDA.
Result and reading. The stress costs two rating notches, and almost all of it comes from one commodity that is 2.4% of the tonnage. Three remarks put the number in proportion.
The running case is not a catastrophe: the EBITDA stays positive, the inventories remain marketable, and half of the hit (the compliance cost) is a permanent cost of doing business that competitors bear too, so part of it will be passed on in prices. The 3% discount is a judgement; a two-tier market for compliant and non-compliant cocoa is expected by traders but had not been observed in a stable way when this module was written.
The tail is another matter. A fine at the legal maximum of 4% of EU turnover is EUR 64m, and EBITDA becomes negative at EUR -24.56m. More plausible than the maximum fine is a detained cargo: a vessel of cocoa blocked in Amsterdam is pledged collateral that cannot be sold where it was meant to be.
And the stress is avoidable. With cocoa traceability at 90%, the cocoa gap disappears, EBITDA is EUR 45.56m and leverage 3.95x. The whole credit question thus becomes: how fast can CAC close the gap, and how can Meridian make sure it does?
The dependency side. The stress above is an impact story: CAC’s sourcing harms forests, a regulation responds, a market closes. LEAP also asks what CAC depends on. The answer is rainfall and soil in a handful of regions. Soy yields in central Brazil rely on a rainy season that the forest itself helps to generate by recycling moisture inland; cocoa in West Africa is sensitive to dry spells and to the loss of shade and pollinators. For a trader the transmission is indirect: a poor harvest means lower volumes through fixed logistics, counterparties that default on forward contracts, and price spikes that swell margin calls on hedges and inventory financing needs at the same time. CAC’s liquidity in such a year is exactly the headroom of its borrowing base, which section 5.6 will test. The analyst records the dependency in the opinion without putting a number on it: concentration by origin (how much of the soy comes from one state) is the indicator followed, and it enters the credit application as a supply-concentration remark of the classical kind. The irony is worth stating in committee: the conversion that the impact analysis looks for in CAC’s supply chain is also what erodes, over time, the rainfall on which CAC’s own volumes depend.
WHO DOES WHAT — The ESRM analyst owns the inputs of this table (traceability, routing, the legal reading of the EUDR). The credit analyst owns the outputs (EBITDA, leverage, the notch). They sign the sensitivity section of the credit application together; at Meridian this joint signature has been a requirement of the credit policy since the EBA guidelines on ESG risk management entered into application.
Question. Deforestation has a regulation and a date. How does the analyst treat the social findings, and how do nature and social findings combine into one rating of the client?
Data. The adverse-media search and the questionnaire return three issues. In cocoa, child labour on farms supplying two cooperatives from which CAC buys; CAC has a monitoring and remediation system covering 40% of its direct supply. In palm oil, an NGO report alleging recruitment-fee debt and passport retention at a plantation supplying a mill on CAC’s list, and an unresolved land dispute with a community that says it never gave its consent. In soy, one indirect supplier that appeared on the Brazilian “lista suja” in 2025, since suspended by CAC.
The UN Guiding Principles rank human-rights impacts by severity, made of scale (how grave), scope (how many people) and irremediability (can it be put right), and only then by likelihood. Severity is judged from the standpoint of the person affected, not from the size of the company’s exposure.
Meridian’s method produces a residual score in two steps. The inherent score is the revenue-weighted average of the commodity scores of section 5.2. The management score is the average of five pillars rated from 0 to 100 on evidence: policy and governance (CAC: 80), traceability (56, the revenue-weighted figure of section 5.2), grievance mechanism and remediation (35), certification (30), independent verification (20).
FORMULA
is the residual score; the inherent score on a scale of 1 to 5; the management score between 0 and 1; the coefficient 0.6 is Meridian’s cap on mitigation credit: even perfect management leaves 40% of the inherent risk, because the bank cannot verify everything. Bands: below 2 low; 2 to 3 medium (approve with standard conditions); 3 to 4 high (reputational risk committee and time-bound action plan); above 4 outside appetite.
Calculation. Inherent: (4 × 1,520 + 5 × 720 + 5 × 600 + 3 × 600) / 3,440 = 4.21. Management: (80 + 56 + 35 + 30 + 20) / 5 = 44.2%. Residual: 4.21 × (1 − 0.6 × 0.442) = 3.09. The client is in the high band.
INTERACTIVE LAB — From inherent risk and management quality to a residual E&S rating
Score the client on five evidence-based management pillars and see which band of Meridian’s policy the residual score falls in.
| Parameter you can move | Range | Default |
|---|---|---|
| Policy and governance | 0 to 100 /100 | 80 /100 |
| Traceability to plot | 0 to 100 /100 | 56 /100 |
| Grievance mechanism and remediation | 0 to 100 /100 | 35 /100 |
| Certification | 0 to 100 /100 | 30 /100 |
| Independent verification | 0 to 100 /100 | 20 /100 |
| Maximum mitigation credit | 30 to 90 % | 60 % |
| Result at default settings | Value |
|---|---|
| Inherent score (1 to 5) | 4.21 |
| Management score | 44.2 % |
| Residual score | 3.09 |
| Band (1 low, 2 medium, 3 high, 4 outside appetite) | 3 |
Try it. Raise only ‘Policy and governance’ to 100: the residual score moves from 3.09 to 2.99, just enough to cross into the medium band. A score that sits next to a threshold invites this kind of optimisation, which is why the committee reads the four evidence pillars (traceability, remediation, certification, verification) and not the band alone.
Result and reading. A score of 3.09 sends the file to the reputational risk committee. The score has a known weakness, and Meridian’s policy corrects it with an override: a weighted average can dilute a severe human-rights impact in a large, clean business line. Any severe impact that is verified and not under remediation leads to escalation, whatever the score. Here the palm-oil allegation triggers it independently of the arithmetic. The question put to CAC is then precise: what did you do when you learned of it? A trader that suspends the supplier the same day protects itself, and leaves the workers where they were. The guiding principles ask for the use of leverage to obtain remediation first, and for disengagement only as a last resort.
PITFALL — Treating a grievance log with zero entries as good news. A mechanism that receives nothing is unknown to the people it is meant to serve, or distrusted by them. The analyst asks for the number of cases opened, the time to close them and one case file in full.
Question. How does Meridian turn findings into contract terms that change the borrower’s behaviour and protect the lender?
Data and design. The analyst and the structurer build the package from four instruments, in increasing order of bite.
Conditions precedent. Before the first drawing: delivery of CAC’s EUDR due diligence procedure and evidence that CAC Europe BV is registered in the EU information system; delivery of the supplier list for palm oil and cocoa to the first aggregation point.
Information undertakings. Quarterly traceability report by commodity and by origin (the answer to the problem of section 5.2); notification within ten business days of any EUDR enforcement action, detained cargo or credible allegation of forced or child labour; annual independent verification of the traceability figures.
An environmental and social action plan, annexed to the agreement, with dated milestones: cocoa traceable to plot at 70% by 30 June 2027 and 90% by 31 December 2027; child-labour monitoring and remediation extended to all direct cocoa supply by end 2027, with smallholders supported to comply and not delisted; a documented remediation process for the palm-oil case with a progress report in six months; soy traceable to farm at 75% for indirect supply by end 2028. A missed milestone is not an event of default; it opens a 90-day remedy period, then a drawstop on new utilisations.
Eligibility criteria in the borrowing base. Inventory destined for the EU without a due diligence statement reference, and inventory bought from suppliers on Meridian’s and the agent’s agreed exclusion lists, stop counting as eligible collateral.
Calculation. CAC’s borrowing base has EUR 520m of hedged inventories with an advance rate of 80% and EUR 380m of receivables at 85%: 416 + 323 = EUR 739m of availability, against EUR 480m drawn on average across the syndicate (60% of EUR 800m). If 10% of the inventory fails the new eligibility test, the base falls to 0.8 × 468 + 323 = EUR 697.4m, a reduction of EUR 41.6m, and headroom over drawings is still EUR 217.4m. The criterion would only constrain drawings if 62.3% of the inventory became ineligible.
INTERACTIVE LAB — An E&S eligibility criterion in a borrowing base
Recompute the availability of the facility when part of the inventory fails the environmental and social eligibility test.
| Parameter you can move | Range | Default |
|---|---|---|
| Hedged inventories | 200 to 800 EUR m | 520 EUR m |
| Receivables | 100 to 600 EUR m | 380 EUR m |
| Advance rate on inventories | 50 to 90 % | 80 % |
| Advance rate on receivables | 50 to 95 % | 85 % |
| Inventory failing the E&S eligibility test | 0 to 100 % | 10 % |
| Amount drawn (whole syndicate) | 0 to 800 EUR m | 480 EUR m |
| Result at default settings | Value |
|---|---|
| Borrowing base before the E&S criterion | 739.0 EUR m |
| Borrowing base after the E&S criterion | 697.4 EUR m |
| Reduction in availability | 41.6 EUR m |
| Headroom over drawings | 217.4 EUR m |
| Ineligible share at which drawings exceed the base | 62.3 % |
Try it. Raise drawings to EUR 680m, the level of a price spike: with 10% of inventory ineligible, headroom shrinks to EUR 17.4m. The criterion bites when the borrower needs liquidity most, which is why CAC’s treasurer will negotiate it hard.
Result and reading. The eligibility criterion is the most elegant of the four tools, because it needs no judgement and no waiver: a cargo without a statement reference simply stops generating availability, exactly as an unhedged or uninsured cargo would. It also aligns the lender’s interest with the regulation, since such a cargo is impaired collateral anyway. Its limit is that it only reaches what the bank finances and can observe; the child-labour question cannot be handled through a collateral report, hence the action plan.
In syndication, the coordinating bank accepts the conditions precedent, the information undertakings and the eligibility criterion for all lenders. The action plan is judged too detailed by two lenders; it survives in a lighter form, with Meridian’s dated milestones in a side letter that gives Meridian information rights but no separate drawstop.
Question. What does the committee decide, and what would make it change its mind?
Decision. The reputational risk committee approves the participation with the conditions of section 5.6, and with a rule for the annual review: if the residual score has not returned to the medium band within 18 months, Meridian does not renew. Credit committee approves EUR 100m at BB, with the EUDR scenario recorded as the main downside and a watch on the cocoa line. The RAROC of the hold is computed line by line in section 6.2: 12.85% before monitoring costs and 12.38% after, against a hurdle of 12%, and 6.41% if the B+ scenario materialises.
Replay in December 2027 (case assumptions). CAC reports cocoa traceability at 88% against the 90% milestone, total revenue-weighted traceability at 80%, a grievance mechanism with 37 cases opened and 29 closed, certification extended, and a first independent verification report. The analyst re-rates the five pillars at 90, 80, 60, 50 and 60. The management score is 68%, the residual score 4.21 × (1 − 0.6 × 0.68) = 2.49: medium band. With cocoa at 88%, the routable traceable share is 70.4%, just above the 70% sold in the EU; only the coffee gap remains. The missed milestone (88% against 90%) opens the remedy period; Meridian records it and does not act on the side letter, since the trajectory is credible and the gap immaterial. The palm-oil case is less tidy: the supplier refunded recruitment fees to part of the workforce, the land dispute is before a court, and the analyst keeps the override flag open.
A check the regulation does not ask for. Before closing the review, the analyst asks one more question: how did cocoa traceability rise from 45% to 88% in a year? CAC’s supplier list shows that the number of cooperatives fell from 61 to 48. For the 13 that left, CAC documents that nine were mapped and moved to another buyer’s programme and four were dropped because their members could not be geolocated in time. Those four represent about 2,000 farmers who now sell to buyers with no EUDR constraint, at a lower price. Nothing in the facility agreement was breached, and the indicator improved. This is the just-transition problem of section 3.5 in a concrete form, and the reason why the action plan said “supported to comply and not delisted”. Meridian asks CAC for a re-inclusion plan for the four cooperatives and records the point in the monitoring report. An indicator can be met in a way that defeats its purpose; the annual review has to look at how it was met.
CASE — The honest summary for the committee: the instrument that changed CAC’s behaviour was the regulation and the loss of market access, with Meridian’s conditions aligned on it. On the social issues, where no regulation with a date existed, progress was slower and the bank’s leverage as a 12.5% lender in a syndicate was limited.
| Stage | Key figure | Value |
|---|---|---|
| 5.1 | Revenue; EU sales; EU share | EUR 3,440m; EUR 1,600m; 46.5% |
| 5.1 | EBITDA; adjusted leverage; interest cover | EUR 55.0m; 3.27x; 5.09x |
| 5.2 | Revenue-weighted traceability to plot | 56.0% |
| 5.2 | Unverified high-risk volume (k = 1) | 651,650 t; 12.9% of tonnage; EUR 430.2m |
| 5.2 | Same with k = 1.5 | 977,475 t |
| 5.3 | Sourcing area; static footprint | 17,051 km²; 14,443 MSA.km² |
| 5.3 | Attribution factor; financed footprint | 5.22%; 753.6 MSA.km² (12.6 per EUR m) |
| 5.3 | Cocoa: share of tonnage; share of footprint | 2.4%; 11.6% |
| 5.4 | EU volume gap: cocoa; coffee | 40,800 t; 12,000 t |
| 5.4 | Diverted revenue; share of EU sales | EUR 252m; 15.75% |
| 5.4 | Compliance cost; diversion discount | EUR 8.0m; EUR 7.56m |
| 5.4 | Stressed EBITDA; leverage; PD | EUR 39.44m (-28.3%); 4.56x; 2.50% |
| 5.4 | With cocoa traceable at 90% | EBITDA EUR 45.56m; 3.95x |
| 5.4 | Maximum fine (4% of EU turnover) | EUR 64m; EBITDA EUR -24.56m |
| 5.5 | Inherent; management; residual score | 4.21; 44.2%; 3.09 (high) |
| 5.6 | Borrowing base before and after criterion | EUR 739.0m; EUR 697.4m |
| 5.6 | Headroom; break-even ineligible share | EUR 217.4m; 62.3% |
| 5.7 | Residual score after action plan | 2.49 (medium) |
| 6.2 | RAROC: base; with monitoring; B+ scenario | 12.85%; 12.38%; 6.41% |
Meridian’s exposure is a EUR 100m commitment drawn at 60% on average. Under CRR3 the undrawn EUR 40m receives a credit conversion factor of 40%, so exposure at default is 60 + 0.4 × 40 = EUR 76m. Meridian uses the foundation IRB approach. For a facility secured by receivables and commodity inventories, the regulatory LGD is lower than the 40% of a senior unsecured claim, after haircuts on the collateral; Meridian’s collateral team arrives at a blended LGD of 30% for this structure (case assumption). With a PD of 1.00% and a maturity of 2.5 years, H.irbCorporateRW gives a risk weight of 61.54%, hence RWA of EUR 46.77m for the facility. Ancillary business adds EUR 2.5m of RWA. Allocated capital at 13% is EUR 6.41m.
Nothing in this calculation refers to deforestation or child labour. CRR3 contains no risk-weight adjustment for environmental or social factors (see E02). They can only enter through the inputs: the PD, if the rating overlay reflects the scenario of section 5.4, and the LGD, if the collateral value is haircut for market-access risk. A cargo that cannot be sold in the EU is worth less as collateral, and Meridian’s collateral policy has no haircut for that yet; the eligibility criterion of section 5.6 does the job contractually.
Assumptions. Drawn margin 200 bp; commitment fee 35% of the margin, or 70 bp; participation fee 30 bp spread over two years; FTP liquidity premium for two-year funding 12 bp on drawn amounts; liquidity buffer equal to 10% of the undrawn commitment, carried at 40 bp; cost-income ratio 40% on lending revenues and 60% on ancillary revenues; ancillary revenues (foreign exchange, commodity hedging, cash management) EUR 1.0m a year.
| Line | Calculation | EUR m |
|---|---|---|
| Drawn margin | 60 × 2.00% | 1.200 |
| Commitment fee | 40 × 0.70% | 0.280 |
| Participation fee | 100 × 0.30% / 2 | 0.150 |
| FTP liquidity premium | 60 × 0.12% | -0.072 |
| Liquidity buffer carry | 4 × 0.40% | -0.016 |
| Net lending revenue | 1.542 | |
| Ancillary revenue | 1.000 | |
| Operating costs | 40% × 1.542 + 60% × 1.000 | -1.217 |
| Expected loss | 1.00% × 30% × 76 | -0.228 |
| Pre-tax result | 1.097 | |
| After tax (25%) | 0.823 | |
| Allocated capital | 13% × 49.27 | 6.406 |
| RAROC | 0.823 / 6.406 | 12.85% |
The deal clears the 12% hurdle, thanks to the ancillary business, as usual for a relationship facility.
What the environmental and social layer changes. Three things, of very different sizes.
Cost. The enhanced monitoring (a share of the satellite-alert subscription, the analyst’s time on quarterly reports, the annual review of the verification report) is estimated at EUR 40,000 a year. Charged to the deal, it lowers RAROC to 12.38%. Most banks do not charge it, which hides the cost and makes the review look free.
The downside scenario. If the stress of section 5.4 materialises and CAC migrates to B+ (PD 2.50%), the risk weight rises to 81.44%, expected loss to EUR 0.570m, capital to EUR 8.37m, and RAROC falls to 6.41% with monitoring costs (6.77% without).
The probability of that scenario. This is where the action plan earns its keep, and the estimate is a judgement. Suppose the analysts consider that, without conditions, CAC has a 25% chance of ending in the B+ state within the life of the facility, and 10% with the conditions and monitoring of section 5.6. The probability-weighted RAROC is 0.75 × 12.85 + 0.25 × 6.77 = 11.33% without the plan and 0.90 × 12.38 + 0.10 × 6.41 = 11.78% with it. The plan adds 0.45 point of expected return, net of its cost.
INTERACTIVE LAB — RAROC of Meridian’s EUR 100m hold, with and without the E&S action plan
Work out revenues, costs, expected loss, RWA and RAROC of the hold, then weigh the base case and the downgrade scenario with and without the conditions.
| Parameter you can move | Range | Default |
|---|---|---|
| Drawn margin | 100 to 350 bp | 200 bp |
| Average utilisation | 0 to 100 % | 60 % |
| PD, base case | 1 / 1.6 / 2.5 / 4 | 1 % |
| PD, downgrade scenario | 1.6 / 2.5 / 4 | 2.5 % |
| LGD | 20 to 45 % | 30 % |
| Ancillary revenue | 0 to 2 EUR m | 1 EUR m |
| Annual E&S monitoring cost | 0 to 150 EUR k | 40 EUR k |
| Probability of downgrade without action plan | 0 to 50 % | 25 % |
| Probability of downgrade with action plan | 0 to 50 % | 10 % |
| Result at default settings | Value |
|---|---|
| Exposure at default | 76.0 EUR m |
| Net lending revenue | 1.542 EUR m |
| Operating costs (before monitoring) | 1.217 EUR m |
| Expected loss | 0.228 EUR m |
| F-IRB risk weight | 61.54 % |
| RWA including ancillary business | 49.27 EUR m |
| Allocated capital | 6.406 EUR m |
| Pre-tax result | 1.097 EUR m |
| RAROC, base case, no monitoring cost | 12.85 % |
| RAROC, base case, with monitoring cost | 12.38 % |
| RAROC, downgrade, with monitoring cost | 6.41 % |
| RAROC, downgrade, no monitoring cost | 6.77 % |
| Risk weight in the downgrade scenario | 81.44 % |
| Expected loss in the downgrade scenario | 0.570 EUR m |
| Capital in the downgrade scenario | 8.37 EUR m |
| Probability-weighted RAROC without action plan | 11.33 % |
| Probability-weighted RAROC with action plan | 11.78 % |
| Two-year lifetime ECL in the downgrade scenario | 1.126 EUR m |
Try it. Set both probabilities to 10%: the plan then only adds cost. The case for conditions rests entirely on the belief that they change the odds, which nobody can measure on one client. Then set LGD to 40%, as if the collateral were worthless: RAROC falls to 9.1%.
PITFALL — Presenting the 0.45 point as a result. It is the product of two probabilities chosen by the analysts. The table is useful because it shows what must be believed for the conditions to pay for themselves (here, that they bring the probability of the bad state from 25% to below about 18%), not because the output is a fact.
The funding of the drawn part is charged at the two-year liquidity premium, EUR 72,000 a year, and the undrawn part obliges Treasury to hold a buffer of high-quality liquid assets, carried at EUR 16,000. Neither depends on the environmental or social profile of the borrower. A nature link exists on the liability side only if Meridian’s Treasury issues a use-of-proceeds bond whose framework includes sustainable agriculture; a general-purpose facility to a trader is not an eligible asset under such a framework, whatever the borrower’s progress (see C15). Commodity price spikes, which may follow harvest failures, increase drawings on such facilities across the whole book at once: this is the liquidity face of the nature dependency.
The facility is held to collect and its cash flows are solely payments of principal and interest; it is measured at amortised cost. The conditions of section 5.6 do not change the contractual cash flows, unlike a sustainability-linked margin (see B03 and E05). The loan sits in stage 1, with a 12-month expected credit loss of EUR 0.228m. A downgrade from BB to B+ is two notches; under Meridian’s staging rules this is a significant increase in credit risk, and the exposure moves to stage 2 with a lifetime loss allowance. On a two-year horizon and a flat PD of 2.50%, cumulative default probability is 1 − 0.975² = 4.94% and the allowance about EUR 1.126m, five times the stage 1 amount. Where the model cannot capture a risk such as the EUDR for a whole sector, banks book a management overlay on the portfolio; supervisors accept overlays if they are documented and temporary.
Green Asset Ratio and BTAR. CAC is a Swiss group outside the scope of the CSRD, so the exposure cannot enter the numerator of the Green Asset Ratio; and trading is not a Taxonomy-eligible activity in any case (see E06). The loan contributes nothing to Meridian’s green ratios, even if CAC reaches full traceability.
Pillar 3 ESG. The EBA templates are built around climate: sector exposures, financed emissions, physical risk. The exposure appears under its NACE sector with its financed emissions. There is no nature or human-rights template; the qualitative tables on environmental and social risk are where Meridian describes its sector policies and the process of section 4.
Financed emissions. Under PCAF, Meridian attributes 5.22% of CAC’s emissions. For an agricultural trader the dominant item is scope 3 upstream, including land-use change emissions from the crops purchased, which depend directly on the deforestation question (see A03 and E07). A client with poor traceability reports those emissions with sector averages, at the worst data quality scores.
CSRD and nature reporting. In Meridian’s own sustainability statement, exposures such as CAC feed the materiality assessment for biodiversity (ESRS E4) and for workers in the value chain and affected communities (ESRS S2 and S3). If Meridian reports on nature in the TNFD format, it will publish the share of its lending in sectors with high dependencies and impacts, and possibly a financed footprint such as the 753.6 MSA.km² of section 5.3, with the caveats given there.
Liability. Meridian is outside the EUDR and, for its clients, outside the CSDDD chain of activities. Its exposure is reputational and, in some jurisdictions, legal under national vigilance laws: in France, NGOs have brought an action against a large bank under the 2017 duty-of-vigilance law over the financing of a Brazilian meat company. The documented process of section 4 is the bank’s defence file.
| Risk | Control | Owner |
|---|---|---|
| File not flagged at screening | Flags derived automatically from sector code, countries, product | Coverage; ESRM for form design |
| Client data unverified | Triangulation with satellite alerts, official lists, certification databases; independent verification undertaking | ESRM |
| Average traceability hides high-risk origins | Reporting by commodity and origin; concentration factor in the scenario | ESRM, credit analyst |
| Market access lost (EUDR) | Scenario in credit application; eligibility criterion in borrowing base | Credit risk, structuring |
| Impaired collateral (detained cargo) | Notification undertaking; insurance and title checks; ineligibility | Loan agency, collateral management |
| Severe human-rights impact diluted in a score | Override rule; escalation to reputational risk committee | ESRM, committee chair |
| Smallholders excluded to meet a KPI | Action plan wording; supplier-count review at annual review | ESRM, coverage |
| Conditions lost in syndication | Minimum conditions agreed with coordinator before commitment; side letter | Structuring, legal |
| Milestones not monitored | Milestones in loan system with alerts; annual review sign-off | Middle office, ESRM |
| Bank’s public claims exceed practice | Review of external statements against policy exceptions log | Compliance, communications |
Red-flag checklist.
Dated facts (as of November 2026).
What changed recently. The direction of travel in the EU since 2025 has been simplification: later dates, higher thresholds, fewer companies. The market-access mechanism of the EUDR survived intact, which makes it the provision with the clearest credit effect. At the same time nature moved from voluntary frameworks towards standard setters (ISSB, ISO) and supervisors.
Open debates.
INTERACTIVE LAB — Self-assessment: nature and social issues for bankers
Check the concepts, the regulation and the calculations of the module.
Q1. A brewery needs clean groundwater and discharges effluent into a river. In TNFD vocabulary these are respectively:
Answer: B. The need for water is a dependency on an ecosystem service; the discharge changes the state of nature and is an impact. Risks are what follows financially.
Q2. What do the four letters of LEAP stand for?
Answer: A. Locate the interface with nature, Evaluate dependencies and impacts, Assess risks and opportunities, Prepare to respond and report.
Q3. A plot has four native species with reference abundances 10, 10, 10, 10 and current abundances 5, 10, 0, 30. Its MSA is:
Answer: C. Capped ratios are 0.5, 1, 0 and 1; their mean is 2.5 / 4 = 0.625. The species at 30 is capped at 1.
Q4. From which date does the EUDR apply to large and medium-sized operators?
Answer: D. Regulation (EU) 2025/2650 set 30 December 2026 for large and medium-sized operators and 30 June 2027 for micro and small ones.
Q5. The EUDR cut-off date, after which land must not have been deforested, is:
Answer: A. Products must come from land not deforested after 31 December 2020.
Q6. After Omnibus I, which EU company is within the scope of the CSDDD?
Answer: C. The thresholds were raised to 5,000 employees and EUR 1.5bn of net worldwide turnover; 1,000 employees and EUR 450m is the CSRD threshold.
Q7. Meridian lends under a general-purpose borrowing-base facility to a trader. Do the Equator Principles apply?
Answer: C. EP4 covers project finance, project-related corporate loans and related products. General corporate facilities are governed by the bank’s own policies.
Q8. Under IFC Performance Standard 6, a project located in critical habitat must achieve:
Answer: B. No net loss applies in natural habitat where feasible; critical habitat requires net gains.
Q9. In the case, why does cocoa drive the EUDR stress although it is 2.4% of the tonnage?
Answer: B. The gap is the EU share minus the routable traceable share: 34% of the volume, worth EUR 204m of the EUR 252m to be diverted.
Q10. CAC’s EBITDA falls from EUR 55.0m to EUR 39.44m with core debt of EUR 180m. Adjusted leverage moves from:
Answer: A. 180 / 55 = 3.27x and 180 / 39.44 = 4.56x. The 12.7x figure would use total debt, which ignores readily marketable inventories.
Q11. Under the UN Guiding Principles, the severity of a human-rights impact is assessed by:
Answer: B. Severity combines how grave the impact is, how many people are affected and whether it can be put right, from the standpoint of the people affected.
Q12. Which contract term makes a cargo without a due diligence statement reference stop generating availability?
Answer: D. Eligibility criteria define which inventories and receivables count in the borrowing base; the effect is automatic at each recalculation.
Q13. Meridian’s drawn exposure is EUR 60m; CAC has EUR 450m of equity and EUR 700m of debt. The attribution factor is:
Answer: C. 60 / (450 + 700) = 5.22%, the PCAF convention for unlisted companies.
Q14. Why does the CAC facility add nothing to Meridian’s Green Asset Ratio even if CAC becomes fully deforestation-free?
Answer: C. Exposures to companies not subject to the CSRD are excluded from the numerator, and the activity is not covered by the Taxonomy.
A coffee and cocoa trader sells 200,000 t of coffee at EUR 3,500/t (EU share 50%, traceable to plot 50%) and 80,000 t of cocoa at EUR 6,000/t (EU share 60%, traceable 40%). 90% of traceable volume can be routed to the EU. EBITDA is EUR 30m, core debt EUR 100m. Compliance costs 0.6% of EU revenue; diverted volume sells at a 4% discount. Compute the stressed EBITDA and leverage.
Solution. Coffee: routable traceable share 0.9 × 50% = 45%; gap 5% × 200,000 = 10,000 t, worth EUR 35m. Cocoa: 0.9 × 40% = 36%; gap 24% × 80,000 = 19,200 t, worth EUR 115.2m. Diverted revenue EUR 150.2m. EU revenue: 50% × 700 + 60% × 480 = 350 + 288 = EUR 638m. Compliance cost 0.6% × 638 = EUR 3.83m; discount 4% × 150.2 = EUR 6.01m. Stressed EBITDA 30 − 3.83 − 6.01 = EUR 20.16m. Leverage rises from 100 / 30 = 3.33x to 100 / 20.16 = 4.96x.
A rubber processor sources 300,000 t of natural rubber at a yield of 1.2 t/ha, with an MSA loss of 0.75. It has EUR 300m of equity and EUR 500m of debt. A bank has EUR 40m outstanding. Compute the static footprint, the bank’s share and the footprint per EUR million.
Solution. Area: 300,000 / 1.2 = 250,000 ha = 2,500 km². Footprint: 2,500 × 0.75 = 1,875 MSA.km². Attribution: 40 / 800 = 5%. Financed footprint: 93.75 MSA.km², or 2.34 MSA.km² per EUR million. If the processor buys from smallholders financed by a local bank, those hectares are also in that bank’s footprint: the figures of different lenders along a chain must not be added.
A bank holds EUR 50m of a fully drawn one-year loan to a BB agribusiness client: net margin over FTP 220 bp, cost-income 40%, LGD 40%, risk weight 82.06% at PD 1.00% and 124.1% at PD 4.00% (both given). Monitoring an action plan costs EUR 30,000. Without the plan, the probability of a downgrade to B is 20%; with it, 8%. Should the bank ask for the plan?
Solution. Revenue 50 × 2.20% = EUR 1.10m; costs EUR 0.44m. Base: expected loss 1% × 40% × 50 = EUR 0.20m; pre-tax 0.46; after tax 0.345; capital 13% × 0.8206 × 50 = EUR 5.334m; RAROC 6.47%. With monitoring: pre-tax 0.43; after tax 0.3225; RAROC 6.05%. Downgrade: expected loss 4% × 40% × 50 = EUR 0.80m; capital 13% × 1.241 × 50 = EUR 8.067m; without monitoring, pre-tax −0.14, after tax −0.105, RAROC −1.30%; with monitoring, pre-tax −0.17, RAROC −1.58%. Expected RAROC without plan: 0.8 × 6.47 + 0.2 × (−1.30) = 4.92%. With plan: 0.92 × 6.05 + 0.08 × (−1.58) = 5.44%. The plan adds about half a point; and the stand-alone loan is far below the 12% hurdle in every case, which is the more important finding.
1. What is the difference between a nature-related dependency and an impact, and why does a credit analyst care about both? A dependency is an ecosystem service the company needs, such as water or pollination; an impact is a change the company causes in the state of nature. Dependencies produce physical risk: the service fails and volumes or costs suffer. Impacts produce transition and liability risk: a regulation, a customer or a court reacts. Both end in the same place, the borrower’s cash flow, by different routes and on different horizons, so the analyst needs both lists before building a scenario.
2. A client tells you it is “EUDR-ready”. What do you ask? I ask which legal entity places the goods on the EU market, since that operator carries the obligation. Then the share of EU-bound volume with plot-level geolocation, by commodity and by origin, and whether the due diligence statements have been tested in the information system. I ask how traceable volumes are physically routed to Europe and what happens to the rest. Finally I ask how suppliers that cannot comply are treated, because dropping them solves the indicator and creates a social issue.
3. Why are the Equator Principles not enough for a bank’s nature and social risk? They apply to project-related finance above defined thresholds, where the asset and its location are known. Most exposure to deforestation or labour abuse arises in general corporate lending and trade finance to companies with diffuse supply chains, which is outside their scope. Banks therefore need sector policies with exclusions, conditions and escalation rules. Membership is also no longer universal since the large US banks left in 2024.
4. How would you bring a deforestation finding into a credit application? As a scenario with numbers, in the sensitivity section, co-signed by the E&S analyst and the credit analyst. I would estimate the volume that loses market access, the margin lost and the compliance cost, then show EBITDA, leverage and the implied rating. I would add the tail events, such as a detained cargo or a fine, and say which contract terms address each. A paragraph of narrative in an annex does not influence the decision.
5. What is FPIC and how does a lender check it? Free, prior and informed consent is the agreement of affected indigenous peoples, obtained without coercion, before the decision and with full information; IFC Performance Standard 7 requires it in defined circumstances. The lender cannot verify it directly. It relies on an independent consultant who reviews the process and its evidence: who was consulted, through which representative institutions, what was agreed, and what dissent was recorded. The lender then makes the commitments a covenant and monitors the grievance mechanism.
6. A relationship manager says the E&S conditions will lose the mandate. How do you answer? I would first separate what is required by regulation from what is our preference: an eligibility criterion that mirrors the EUDR costs a compliant client nothing. I would show the scenario in which the conditions protect our collateral and our RAROC. I would check what the coordinator and other lenders ask, because the client compares. And if the client refuses information rights on its most material risk, that refusal is itself a credit signal that belongs before the committee.
7. What are the limits of a biodiversity footprint in MSA.km²? It measures intactness relative to an undisturbed state and ignores where the loss happens, so a hectare in a hotspot counts like any other. It relies on average coefficients by land use and on yields, so two companies with different practices can look the same. The static footprint describes past conversion and says little about ongoing damage. Attribution to a lender double counts along the value chain. It is useful for ranking and for following a client over time.
8. Is there a business case for a bank to do this work? Partly, and it should be stated honestly. The work does not reduce capital, since prudential rules have no adjustment for it, and it costs analysts and data. Its value lies in avoiding tail losses on clients that lose market access, in protecting collateral, and in limiting reputational and legal exposure. In the case of this module the conditions add under half a point of expected RAROC, on assumptions nobody can verify. The stronger argument is that supervisors now expect the bank to show how such risks enter its decisions.
Regulation
Standards and frameworks
Supervisory and policy papers
Academic and reference works