Treasury is the bank’s own bank. It raises the money that the business lines lend, keeps enough liquid assets to survive a run, manages the interest-rate risk that the balance sheet creates, and charges each business for the funding and liquidity it consumes through funds transfer pricing (FTP). At Meridian Bank, Group Treasury reports to the Chief Financial Officer and sits next to, but separately from, the trading floor. Its counterparties are the bank’s own desks inside, and debt investors, rating agencies, central banks and resolution authorities outside.
Where it sits. Meridian’s Treasury has about 70 people in five teams. Funding and issuance (eight people) writes the funding plan, chooses instruments and tenors, mandates the bookrunners and is the issuer in every bond Meridian sells; it works with DCM FIG, which executes the deals (see C03). ALM and FTP (fifteen) models deposits and prepayments, measures interest-rate risk in the banking book (IRRBB) and publishes the FTP curve every day. Liquidity management (twenty) runs the cash position, the liquidity buffer of high-quality liquid assets (HQLA) and the regulatory ratios. Collateral management (ten) mobilises assets for repo, central bank operations and covered bonds. Debt investor relations (four), shared with Finance, maintains the dialogue with bond investors and rating agencies and publishes the green bond reports. The sustainability team contributes one or two people to the Green Bond Committee, which Treasury chairs.
How it earns money. Treasury is not a fee business. Its result is the difference between what it charges the business lines through FTP and what funding actually costs, plus the result of its own positions (the HQLA portfolio, the interest-rate gap it is allowed to run). In most banks Treasury is run as a cost centre with a small target result, because a Treasury that tries to make large profits is either taking risk or overcharging the businesses. The money at stake is large in absolute terms. Meridian raises about EUR 8bn to EUR 10bn of term wholesale funding a year at spreads of 40 bp to 120 bp over mid-swaps: every basis point saved on the whole programme is worth about EUR 0.9m a year per vintage, and the vintages accumulate. The liquidity buffer of about EUR 50bn costs tens of millions a year in negative carry.
The team and adjacent functions. The Group Treasurer chairs the funding meeting and sits on the Asset and Liability Committee (ALCO), chaired by the CFO, which approves the funding plan, the FTP methodology, the liquidity risk appetite and the IRRBB limits. Risk (the second line, in a liquidity risk team) sets and monitors limits and runs the internal stress tests of the internal liquidity adequacy assessment process (ILAAP). Finance produces the regulatory reports (LCR monthly, NSFR quarterly, asset encumbrance, funding plans) and the sustainability statement. Resolution planning, often inside Treasury, talks to the Single Resolution Board (SRB) about MREL. The rating agency analysts read the funding plan as closely as the supervisor does.
A realistic week (the last week of November 2026).
IN ONE SENTENCE — Treasury decides how much the bank borrows, in which form and at what price, keeps the liquid assets that let it survive a run, and charges the cost to the businesses that cause it.
A commercial bank lends long and borrows short. Its loans run for years; most of its deposits can leave tomorrow. Treasury’s work starts from this mismatch. The table below is Meridian’s funding structure at 30 September 2026 (case assumptions, consistent with the total assets of EUR 310bn used in A02).
| Liabilities and equity | EUR bn | Behaviour | Cost over swaps |
|---|---|---|---|
| Customer deposits | 175.0 | Sticky in aggregate, callable individually | Below swaps |
| Covered bonds | 28.0 | Term, secured on mortgages | 38 bp at 5 years |
| Senior preferred (SP) | 9.0 | Term, unsecured | 78 bp |
| Senior non-preferred (SNP) | 9.5 | Term, subordinated MREL | 112 bp |
| Tier 2 and AT1 | 4.72 | Capital | 190 bp and more |
| Short-term wholesale (CP, CD) | 10.0 | Rolls every 1 to 12 months | Near €STR |
| Repo and trading liabilities | 35.0 | Secured, mostly overnight to 3 months | Near €STR |
| Equity and other | 38.78 | Permanent | Cost of equity |
Deposits fund 56% of the balance sheet; the loan-to-deposit ratio is about 109% (customer loans of EUR 190bn against EUR 175bn of deposits), so Meridian needs term wholesale funding for the rest and for the liquidity buffer. Term wholesale (covered, SP, SNP, capital) is EUR 51bn, about 16% of the balance sheet. That proportion is typical of a large European universal bank.
Why wholesale funding exists. Deposits come with the franchise and cannot be increased quickly without paying up. Wholesale funding fills the gap, matches the tenor of long assets, and, for SNP and capital, meets resolution requirements. Each instrument has a different price because it has a different risk for the investor (see C03, section 2.2, for the ranking).
The funding plan is the document in which Treasury states how much term funding the bank will raise in the coming year, in which instruments, currencies and tenors, and why. It is built in September and October, approved by ALCO and the board in December, and reported to the supervisor with a three-year horizon under the EBA guidelines on harmonised definitions and templates for funding plans (EBA/GL/2019/05).
FORMULA
is the term issuance needed in the year; the term debt maturing or called in the year; the growth of customer loans; the growth of customer deposits; the planned change in the liquidity buffer; the funding already raised in advance (pre-funding) in the previous year.
INTUITION — A household that must repay a car loan next year, plans to buy a bigger car, and expects its salary to rise a little, knows how much it must borrow by subtracting the salary increase from the sum of the repayment and the new purchase. A bank does the same with billions, and with a rule that it must always keep three months of expenses in the drawer.
The total is then split by instrument according to three constraints. The MREL and subordination requirement fixes the minimum amount of SNP or capital (C03 showed that subordination is Meridian’s binding constraint). The cost favours covered bonds, the cheapest term funding, but the encumbrance they create (mortgages pledged to covered bond holders are no longer available to unsecured creditors) and rating agency views limit them. Diversification of investors and maturities, finally, spreads issuance across instruments, currencies and months, so that no single market closure stops the bank. A plan also contains the redemption profile (no more than 10% to 15% of term debt maturing in any one year is a common rule of thumb), the target average tenor, and the contingency options if markets close.
FTP is the internal price at which Treasury lends to the business lines and borrows from them. A loan desk borrows from Treasury at the swap rate of the loan’s interest-rate profile plus a liquidity premium for the term; a deposit desk is paid the swap rate of the deposit’s behavioural maturity plus a credit for the stable funding it brings. The difference between client rates and FTP is the commercial margin; the difference between FTP and actual cost is Treasury’s result. The method is governed by internal policy, guided by the CEBS guidelines on liquidity cost benefit allocation (2010) and by the supervisor’s review of internal governance.
The liquidity premium is the price of term funding over the risk-free swap rate. It is not the spread of any one bond but the blended spread of the marginal funding mix Treasury would use to fund a new five-year asset.
FORMULA
is the five-year liquidity premium; the weight of instrument in the marginal funding mix (summing to 1); the five-year spread over mid-swaps of instrument ; the premium for tenor years; the slope in bp per year.
INTERACTIVE LAB — From the marginal funding mix to Meridian’s FTP liquidity premium
Derive the five-year liquidity premium from the funding mix and compare the FTP curve with the curves at which Meridian actually issues.
| Parameter you can move | Range | Default |
|---|---|---|
| Weight of covered bonds in the mix | 0 to 80 % | 50 % |
| Weight of senior preferred | 0 to 80 % | 25 % |
| Covered bond spread, 5 years | 0 to 100 bp | 38 bp |
| Senior preferred spread, 5 years | 30 to 200 bp | 78 bp |
| Stable deposits, value against swaps | -80 to 20 bp | -34 bp |
| Tenor of the asset | 1 to 10 years | 5 years |
| Result at default settings | Value |
|---|---|
| Weight of stable deposits (balance) | 25 % |
| Five-year liquidity premium | 30.0 bp |
| Slope of the FTP curve | 6.00 bp per year |
| Liquidity premium at the chosen tenor | 30.0 bp |
Try it. Move 15 points of weight from covered bonds to senior preferred, as a bank would after a rating downgrade of its covered programme: the whole FTP curve shifts up and every loan in the bank becomes more expensive.
With half the marginal five-year funding from covered bonds at 38 bp, a quarter from senior preferred at 78 bp and a quarter from stable deposits valued at 34 bp below swaps, the liquidity premium is 30.0 bp at five years, or 6 bp per year of tenor: the course’s convention. The bank’s issuance curves are steeper than the FTP curve at long tenors and flatter at short ones (in the widget they are drawn with a simple shape around their five-year point, for illustration).
PITFALL — SNP and Tier 2 are not in the FTP mix. They are issued to meet resolution and capital requirements, not to fund a particular asset, so their extra cost (84 bp for Meridian’s SNP over the 30 bp FTP) is allocated by RWA as an MREL charge, not by tenor. A bank that puts SNP in the liquidity curve overcharges long, low-risk assets such as mortgages and undercharges short, high-RWA ones.
WHY IT MATTERS ON THE DESK — Every loan price in the bank starts from this curve. One basis point on the five-year FTP changes the margin of every new five-year loan, which is why business heads fight over the weights and why Treasury documents them for ALCO.
Two ratios bind Treasury. The Liquidity Coverage Ratio (LCR), in Article 412 of the CRR and detailed in Commission Delegated Regulation (EU) 2015/61, requires a stock of unencumbered HQLA at least equal to net cash outflows over a 30-day stress. The Net Stable Funding Ratio (NSFR), introduced by CRR2 (Regulation (EU) 2019/876) and applicable since June 2021, requires available stable funding at least equal to required stable funding over one year.
HQLA come in three levels. Level 1: central bank reserves, bonds of EU sovereigns and of multilateral development banks and international organisations listed in the regulation, some promotional banks guaranteed by a state, and extremely high-quality covered bonds (the latter with a 7% haircut and a cap of 70% of the buffer). Level 2A: high-quality covered bonds and corporate bonds of the top credit quality step, with a 15% haircut. Level 2B: other corporate bonds, shares and certain securitisations, with haircuts of 25% to 50%. Level 2 may not exceed 40% of the buffer, and Level 2B 15%.
FORMULA
, , are the values of each HQLA level after haircuts; is Level 2B after its cap; the stressed outflows over 30 days; the inflows, capped at 75% of outflows. This is a simplified form of the caps, ignoring the unwinding of short-term secured transactions that the regulation requires.
INTERACTIVE LAB — Meridian’s Liquidity Coverage Ratio at 30 September 2026
Compute the LCR with haircuts and caps, and see how much surplus Treasury holds above the regulatory minimum and above its own target.
| Parameter you can move | Range | Default |
|---|---|---|
| Central bank reserves | 0 to 40 EUR bn | 20 EUR bn |
| Level 1 sovereign bonds | 0 to 40 EUR bn | 21 EUR bn |
| Level 1 supranational and agency bonds | 0 to 15 EUR bn | 5 EUR bn |
| Level 1 covered bonds (market value) | 0 to 10 EUR bn | 3 EUR bn |
| Level 2A bonds (market value) | 0 to 10 EUR bn | 2 EUR bn |
| Level 2B corporate bonds (market value) | 0 to 10 EUR bn | 1 EUR bn |
| Stressed outflows, 30 days | 10 to 80 EUR bn | 44 EUR bn |
| Inflows, 30 days | 0 to 40 EUR bn | 8 EUR bn |
| Internal LCR target | 100 to 180 % | 130 % |
| Result at default settings | Value |
|---|---|
| Level 1 after haircuts | 48.79 EUR bn |
| HQLA after haircuts and caps | 50.99 EUR bn |
| Net outflows | 36.00 EUR bn |
| LCR | 141.64 % |
| Surplus over 100% | 14.99 EUR bn |
| Surplus over the internal target | 4.19 EUR bn |
Try it. Raise stressed outflows from EUR 44bn to EUR 50bn, as if corporate deposits were reclassified as operational-only in part: the LCR falls to about 121%, below Meridian’s 130% target, and Treasury must buy more HQLA, which costs carry.
Meridian holds EUR 50.99bn of HQLA after haircuts against EUR 36.0bn of net outflows: an LCR of 141.64%. The surplus over the 100% minimum is EUR 14.99bn, but over Meridian’s own 130% target only EUR 4.19bn. The internal target is what Treasury manages to, because a bank that publishes an LCR close to 100% invites questions from investors and the supervisor.
The carry cost. HQLA yield less than the term funding that pays for them. Meridian’s convention is a net carry cost of 40 bp on the buffer. On the securities part of the buffer and the reserves this is tens of millions a year, charged to the businesses that cause the outflows: undrawn corporate lines, operational deposits, derivatives collateral.
NSFR in one paragraph. Term debt with more than one year to run counts 100% as available stable funding; retail deposits 90% to 95%; corporate deposits 50%. Loans with more than one year to run require 65% (low-risk-weight mortgages) to 85% (other loans) of stable funding; Level 1 HQLA almost nothing. Meridian’s NSFR is above 120% (case assumption); the 2027 plan must not let it fall below 115%.
WHY IT MATTERS ON THE DESK — The liquidity buffer is Treasury’s largest single asset and its largest single cost. Any policy on it, green or not, is judged first on its effect on the LCR, its carry and its liquidity in a stress.
An asset is encumbered when it is pledged or otherwise unavailable to unsecured creditors: mortgages in a covered bond pool, bonds lent in repo, loans pledged to the central bank. The encumbrance ratio (encumbered assets plus collateral received and reused, over total assets plus collateral received) is reported to the supervisor and disclosed in the Pillar 3 templates on encumbered and unencumbered assets. A high ratio weakens senior unsecured creditors and rating agencies notice it.
The Eurosystem collateral framework lists the assets a bank can pledge to borrow from the central bank. After the end of reinvestment in the asset purchase programmes (the Eurosystem no longer reinvests maturing principal), central bank funding is again a lender-of-last-resort tool rather than a source of cheap term money. The deposit facility rate stood at 2.50% after the decision of 29 October 2026, which matches the 2.50% swap level used in this course. Reserves at the central bank therefore earn about the risk-free rate; their carry cost is the spread of the funding that pays for them.
WHO DOES WHAT — The collateral team keeps a daily inventory of every asset that could be mobilised: bonds in the HQLA portfolio, retained covered bonds and securitisations, and credit claims (loans) pre-positioned at the central bank. The total is the counterbalancing capacity, the first line of the internal stress test.
Treasury also manages the interest-rate risk created when fixed-rate loans are funded with floating or shorter liabilities. It hedges new fixed-rate bonds with receive-fixed swaps, so that a new SNP at 3.625% becomes floating funding at about €STR plus a spread, and monitors the economic value and net interest income sensitivities against the supervisory outlier tests. This module does not develop IRRBB (see E04); it matters here only because long-dated green assets, such as 25-year mortgages and renewable project loans, are the longest assets on the balance sheet.
Sustainability reaches Treasury through three doors. As an issuer, the bank sells green, social or sustainability bonds (and, at some banks, green deposits or green commercial paper) backed by a portfolio of eligible loans. As an investor, it decides what goes into the liquidity buffer and the banking-book bond portfolio. As a price-setter, it decides whether any funding advantage reaches the businesses through the FTP curve. None of the three changes the regulatory ratios: the LCR, the NSFR, MREL and the leverage ratio are label-blind. What changes is the work behind the numbers: a register of eligible assets, a framework, reports, an investment policy, and an FTP rule that must be defensible in front of ALCO and the auditors.
IN ONE SENTENCE — A bank’s green bond is ordinary funding with an earmarked asset portfolio and a reporting promise; Treasury’s job is to keep the portfolio large enough, the promise true and the economics honest.
Banks almost always use the portfolio approach: the bond proceeds are not traced to particular loans but matched, in aggregate, against a pool of eligible assets whose outstanding must stay at least equal to the green bonds outstanding. The governing standard for most deals is the ICMA Green Bond Principles (latest edition June 2025), with their four components: use of proceeds, process for project evaluation and selection, management of proceeds, and reporting, plus the recommendation of a framework and an external review. For banks, ICMA’s guidance on the portfolio approach and its Guidance Handbook (updated June 2026) are the practical references.
The European Green Bond Standard (Regulation (EU) 2023/2631, applicable since 21 December 2024) is stricter. Proceeds must finance Taxonomy-aligned assets, including the technical screening criteria, do-no-significant-harm tests and minimum safeguards; the issuer publishes a factsheet reviewed by an external reviewer registered with ESMA; allocation and impact reports follow. Financial institutions may use a portfolio approach in which the aligned assets must cover the EuGBs outstanding, and assets that stop being aligned because the Taxonomy criteria change can remain in the portfolio for a transition period. Uptake among banks is limited: according to ING research (27 May 2026), only eight banks out of almost 160 sustainable bond issuers had used the EuGBS format since its introduction, EuGBS bonds were 6% of bank green supply in 2025 and 11% in 2026 to late May. The main obstacle is data: proving full Taxonomy alignment of mortgages or of non-EU loans is hard (C03 section 5.2 shows Meridian’s aligned share at 68.47%).
The register. The pool lives in a register maintained by Treasury with the sustainability team: one line per loan, with the eligibility category, the evidence (energy performance certificate, project documentation), the Taxonomy flag, the origination date, the outstanding amount and a flag recording which bond programme it supports. The register is refreshed monthly from the loan systems and reviewed quarterly by the Green Bond Committee.
Coverage and over-allocation. Coverage is the pool divided by green bonds outstanding. A coverage below 100% means over-allocation: more green bonds than green assets, so part of the proceeds is “unallocated”. Frameworks say how unallocated proceeds are held (usually cash or HQLA, excluding fossil-fuel issuers) and most banks hold an internal buffer above 100% so that amortisation, prepayments, loan sales or a methodology change do not push them into over-allocation. Meridian’s buffer is 115%.
PITFALL — Over-allocation does not breach a bond covenant: green bond documentation does not make the label a default event, and the bond stays MREL-eligible. The damage is reputational and commercial. It shows in the next allocation report, investors with green mandates ask questions, and the bank’s next green deal prices with fewer orders.
How large are bank pools compared with their bonds? An IEEFA study (September 2026) of 47 European banks found about EUR 240bn of green bonds outstanding, about 5% of their market debt securities funding and less than 1% of their total assets on average. Some banks had more than 75% of their eligible pools unallocated, while about a quarter reported pools almost fully allocated. The same study notes that green buildings (mostly mortgages) receive about two-thirds of allocated proceeds on average but contribute a small share of reported avoided emissions, while renewable energy receives about one fifth and contributes most of them. For a Treasury, that is a reminder that the pool is what the balance sheet holds, not what an impact maximiser would choose.
Some banks offer corporates green deposits: a term deposit whose amount is matched against eligible green assets, with an annual report. The deposit is a normal deposit for the LCR and NSFR. It consumes the same pool as the green bonds, so it must be registered against the pool and counted in coverage; otherwise the same mortgage backs a bond and a deposit. There is no market standard comparable to the Green Bond Principles for deposits, so banks rely on their own frameworks. Meridian does not offer green deposits in 2026, partly for that reason.
Regulation is label-blind: a green Bund is Level 1 like its conventional twin, with the same 0% haircut; a green corporate bond is Level 2A or 2B according to its rating, not its label. There is no green supporting factor in the LCR or the NSFR, and the EBA has not proposed one. What banks do is voluntary and takes three forms.
The constraints are financial. Green lines are usually smaller than conventional benchmarks and trade with wider bid-offer spreads, which matters for the “liquidity value” of the buffer in a stress; investment policies therefore cap green lines at a share of each issue. The green tilt costs a yield give-up roughly equal to the greenium of the bonds bought, a few basis points for sovereigns (see the green Bund in C03). And a tilt into sovereign and supranational green bonds does not change the bank’s Green Asset Ratio, because sovereigns, central banks and supranational issuers are excluded from the GAR denominator (see A02 and E06).
The Eurosystem has two climate tools that touch Treasury. First, after the end of reinvestments, the climate tilt of corporate bond purchases has no new purchases to apply to; its effect now runs through the shrinking existing portfolios. Second, and more directly, a climate factor in the collateral framework became effective on 15 June 2026 for marketable assets issued by non-financial corporations. It reduces the collateral value of each asset, after the normal haircut, according to an uncertainty score built from the sector’s sensitivity to a transition shock, the issuer’s emissions, targets and disclosure, and the residual maturity. On 24 July 2026 the ECB announced the extension to non-financial corporate credit claims, expected at the earliest by the end of 2027, and stated that the maximum additional reduction would be 5% for both bonds and credit claims. The ECB expects the immediate effect on banks to be limited, because borrowing from the Eurosystem is low and corporate bonds are a small part of pledged collateral. For Treasury the effect is on counterbalancing capacity, not on the LCR.
The measured greenium on bank bonds is a few basis points at issue, often within the measurement error of the issuer’s curve (C03 section 5.4). Academic studies agree on the order of magnitude: Zerbib (2019) found a small negative premium of about 2 bp on average; Larcker and Watts (2020) found essentially none in matched US municipal bonds; Fatica, Panzica and Rancan (2021) found a smaller premium for financial institutions than for supranationals and corporates. For Treasury the real question is how much of the bank’s total funding the greenium touches, which section 5.5 quantifies.
For Treasury the “deal” is a yearly cycle: the funding plan, the issues that carry it out, the reports that close it. The green programme is threaded through the same cycle.
INTERACTIVE LAB — Meridian’s annual funding and green programme cycle
Follow the funding plan from business forecasts to issuance, reporting and the next FTP update, and see where the green programme enters.
| # | Who | Does what | Deliverable |
|---|---|---|---|
| 1 | Finance (planning) and business lines | Submit three-year balance sheet and RWA forecasts by business line (September) | Budget balance sheet |
| 2 | ALM, Treasury | Model deposit behaviour, prepayments and the maturity ladder; compute the term funding need | Funding gap by year |
| 3 | Resolution planning, Treasury | Project MREL and subordination ratios against the SRB decision; size SNP and capital needs | MREL plan |
| 4 | Liquidity management | Set the HQLA buffer size and composition against the LCR target and the ILAAP stress, including the ESG investment policy | Buffer plan and investment policy |
| 5 | Green Bond Committee (Treasury, Sustainability, Risk, Legal) | Project the eligible pool for three years and fix the green issuance capacity at the 115% buffer | Green capacity memo |
| 6 | Group Treasurer | Draft the funding plan: amounts by instrument, currency, tenor and month; green share; contingency options | Funding plan draft |
| 7 | ALCO, then Board Risk Committee | Approve the plan, the FTP methodology and the green rebate policy (December) | Approved funding plan |
| 8 | Finance (regulatory reporting) | Report the three-year funding plan to the supervisor under the EBA funding plan guidelines | Funding plan templates |
| 9 | Treasury and DCM FIG | Mandate bookrunners and execute issues in market windows; Green Bond Committee confirms coverage before each green deal | Priced bonds, final terms |
| 10 | Liquidity management | Invest proceeds and any unallocated green proceeds in the buffer under the exclusion policy | Buffer positions |
| 11 | ALM | Update the FTP curve with new issuance costs and the green rebate rule | FTP grid |
| 12 | Treasury, Sustainability, external reviewer | Produce the annual allocation and impact report with limited assurance | Green bond report |
| 13 | Debt investor relations | Present the funding plan, MREL position and green programme to investors and rating agencies | Fixed income investor presentation |
Forecasts (step 1). The plan is only as good as the business forecasts. Each business line sends a balance sheet and RWA forecast; Treasury challenges them, because lending businesses tend to forecast growth generously and deposit businesses tend to forecast stability. The budget balance sheet is signed off by the CFO.
Behavioural modelling (step 2). ALM converts contractual maturities into behavioural ones: how long sight deposits stay, how fast mortgages prepay, how much of a revolving facility is drawn. The output is a maturity ladder and the funding gap for each of the next three years. The models are validated by Risk and reviewed by the supervisor in the ILAAP.
MREL (step 3). The SRB sets the MREL and subordination requirements in its yearly decision; resolution planning projects the ratios using the RWA forecast (C03 section 5.1 shows the calculation for Meridian at end-2027). This fixes the floor for SNP and capital issuance.
Buffer plan (step 4). Liquidity management sets the size of the buffer from the LCR target and the internal stress test, and its composition from the investment policy approved by ALCO, which now includes the ESG rules: green share target, exclusions, issuer limits.
Green capacity (step 5). The Green Bond Committee projects the pool forward, net of amortisation, prepayments, planned loan sales and known methodology changes, and states how much green issuance the pool can carry at the 115% buffer in every year of the plan. Treasury does not plan a green deal that would break the buffer in any of the three years.
Plan, approval and reporting (steps 6 to 8). The Group Treasurer drafts the plan with a base case and a contingency (for example, replacing a planned SP with covered bonds or a central bank operation if the unsecured market closes). ALCO approves it with the FTP methodology; the board risk committee approves the liquidity risk appetite. Finance reports the plan to the supervisor in the templates of the EBA guidelines, which ask for projected assets, liabilities, pricing and the reliance on public sector funding.
Execution (steps 9 and 10). Issues follow the windows: January and early September are the busiest for European banks, quiet periods before results limit issuance, and Treasury avoids competing with peers’ benchmark days. Before each green deal the Green Bond Committee confirms coverage in writing; the bookrunners’ due diligence includes a call with Treasury and the sustainability team on the pool. Proceeds go to the general treasury account; any unallocated amount is tracked and held under the framework’s rules.
FTP and reporting (steps 11 to 13). ALM updates the FTP curve monthly from secondary spreads and new issues. The allocation report, published once a year with the external reviewer’s limited assurance, shows the pool by category, the allocation of each bond, the share of refinancing, any unallocated amount and impact metrics. Investor relations presents the plan and the green programme in the fixed income investor presentation after the annual results; a funding plan that is credible and well executed narrows the bank’s spreads more than any label.
WHO DOES WHAT — The Group Treasurer owns the funding plan. The Head of Liquidity owns the buffer and its investment policy. The Head of ALM owns the FTP curve. The chair of the Green Bond Committee (at Meridian, the deputy treasurer) owns the pool register and signs every coverage confirmation. Risk owns the limits and the stress test, and Internal Audit reviews the register every two years.
The case follows Meridian’s Treasury through the preparation of the 2027 funding plan in late November 2026, one week after the bank priced its EUR 1bn green SNP (C03). All Meridian figures are case assumptions, consistent with modules A02 and C03: total assets EUR 310bn, TREA of EUR 124bn projected for end-2027, an eligible green pool of EUR 5,100m against EUR 3,500m of green bonds outstanding after the SNP, a 115% coverage buffer, and a 3 bp FTP rebate on green loans. Market and regulatory facts are real and dated. The five stages answer the five questions the Group Treasurer must settle before ALCO: how much to raise and in what; how much of it can be green; what to do with the buffer; which green format; and whether the rebate survives.
Question. What is the 2027 term funding need, how is it split by instrument, and what does it cost?
Data (case assumptions). Term debt maturing or called in 2027: covered bonds EUR 3.5bn, senior preferred EUR 2.0bn, and the EUR 1.25bn SNP called in January 2027. Customer loans are forecast to grow by EUR 6.0bn (from EUR 190bn) and deposits by EUR 3.5bn (from EUR 175bn). The buffer is held flat. The green SNP of November 2026 pre-funds EUR 1.0bn of the 2027 need. C03 left a subordination shortfall of EUR 0.262bn at end-2027 after the green SNP. New-issue spreads over mid-swaps for five years, including the new issue concession: covered bonds 40 bp, senior preferred 83 bp, SNP 117 bp; a green senior preferred is assumed to price 2 bp inside a conventional one. ALCO’s draft split: covered EUR 4.0bn, conventional SP EUR 3.0bn, green SP EUR 0.75bn, and SNP for the balance.
INTERACTIVE LAB — Meridian’s 2027 funding plan: need, split and spread cost
Compute the 2027 term funding need, the SNP balance after the other instruments, the annual spread cost of the plan and the green share.
| Parameter you can move | Range | Default |
|---|---|---|
| Covered bonds maturing | 0 to 8 EUR bn | 3.5 EUR bn |
| Senior preferred maturing | 0 to 6 EUR bn | 2.0 EUR bn |
| SNP called | 0 to 3 EUR bn | 1.25 EUR bn |
| Growth of customer loans | -5 to 15 EUR bn | 6.0 EUR bn |
| Growth of customer deposits | -10 to 15 EUR bn | 3.5 EUR bn |
| Change in liquidity buffer | -5 to 10 EUR bn | 0 EUR bn |
| Pre-funded in 2026 (green SNP) | 0 to 3 EUR bn | 1.0 EUR bn |
| Covered bonds to issue | 0 to 8 EUR bn | 4.0 EUR bn |
| Conventional SP to issue | 0 to 8 EUR bn | 3.0 EUR bn |
| Green SP to issue | 0 to 2 EUR bn | 0.75 EUR bn |
| Covered new-issue spread | 0 to 100 bp | 40 bp |
| SP new-issue spread | 30 to 200 bp | 83 bp |
| Greenium on green SP | 0 to 10 bp | 2 bp |
| SNP new-issue spread | 50 to 250 bp | 117 bp |
| Subordination shortfall to close (C03) | 0 to 3 EUR bn | 0.262 EUR bn |
| Customer loans, start of year | 100 to 300 EUR bn | 190 EUR bn |
| Customer deposits, start of year | 100 to 300 EUR bn | 175 EUR bn |
| Result at default settings | Value |
|---|---|
| Gross term need before pre-funding | 9.25 EUR bn |
| Term issuance needed in 2027 | 8.25 EUR bn |
| SNP to issue (balance) | 0.50 EUR bn |
| SNP above the subordination shortfall | 0.238 EUR bn |
| Annual spread cost of 2027 issuance | 52.825 EUR m |
| Average new-issue spread | 64.03 bp |
| Annual greenium saving in the plan | 0.150 EUR m |
| Greenium saving as share of the vintage spread cost | 0.284 % |
| Green share of 2027 issuance | 9.09 % |
| Green share including the pre-funded green SNP | 18.92 % |
| Loan-to-deposit ratio, start | 108.57 % |
| Loan-to-deposit ratio, end | 109.80 % |
Try it. Set deposit growth to EUR 0.5bn, as in a year when savers move money to money market funds: the need jumps by EUR 3bn and, with the other lines fixed, falls on SNP, the most expensive instrument. In practice Treasury would add covered bonds instead, which is what the contingency plan is for.
Calculation. Maturities and calls total EUR 6.75bn; loans grow EUR 2.5bn faster than deposits; the gross need is EUR 9.25bn. The green SNP already raised EUR 1.0bn, so 2027 issuance is EUR 8.25bn. With EUR 4.0bn of covered bonds, EUR 3.0bn of conventional SP and EUR 0.75bn of green SP, the balance is EUR 0.5bn of SNP, EUR 0.238bn more than the subordination shortfall: a cushion against RWA growth above forecast.
Result. The annual spread cost of the 2027 vintage is EUR 52.825m, an average of 64.03 bp over mid-swaps. The green SP saves EUR 0.15m a year. Green issuance is 9.09% of 2027 issuance, and 18.92% of the year’s programme if the pre-funded green SNP is included. The loan-to-deposit ratio rises from 108.57% to 109.80%.
Reading. The plan is driven by three things the green label does not touch: maturities, the gap between loan and deposit growth, and MREL. The split is driven by cost (covered bonds at 40 bp against 83 bp for SP) against encumbrance and investor diversification. The green question comes last: which of the bonds the bank must issue anyway can carry a label? The answer depends on the pool (section 5.2). The greenium saving is 0.284% of the vintage’s spread cost.
CASE — In Meridian’s draft, the green SP is placed in September 2027, after the Green Bond Committee has seen the half-year pool, rather than in January when the unsecured market is deepest. Treasury accepts a possibly weaker window to issue only against a pool it has verified.
The calendar. Treasury spreads the EUR 8.25bn over the year: EUR 1.5bn of covered bonds and EUR 1.5bn of SP in January, when investors have new money to invest and the market is deepest; EUR 0.5bn of SNP in March, after the annual results and the SRB’s new MREL decision; EUR 1.5bn of covered bonds in May; the green SP of EUR 0.75bn and EUR 1.0bn of covered bonds in September; and EUR 1.5bn of SP in October. No month carries more than EUR 3bn, so that a closed market in any one month delays the plan by weeks, not quarters. The contingency plan names the substitutes: retained covered bonds pledged to the Eurosystem, a larger covered bond if the unsecured market is closed, and a reduction of the buffer towards the 130% target, which frees up to EUR 4.19bn (section 2.4).
PITFALL — “Green issuance of 19%” and “9%” describe the same plan. Investor presentations often count pre-funding in the year that suits the message. The funding plan reported to the supervisor follows the settlement date of each bond; use that convention and say so.
Question. Can the pool carry the EUR 0.75bn green SP in 2027 and the planned green refinancings of 2028 and 2029 at the 115% buffer, and what would cause over-allocation?
Data (case assumptions). The pool is EUR 5,100m at end-2026 (C03), of which EUR 2,600m green mortgages. New eligible loans of EUR 1,000m a year and amortisation and prepayments of EUR 600m a year. Green bonds outstanding: EUR 3,500m at end-2026. The 2023 green SP of EUR 1,000m matures in 2028 and the 2024 green SP of EUR 750m in 2029; Treasury plans to refinance each in green. Three adverse events are tested: a securitisation of green mortgages that removes them from the balance sheet, the carve-out of mortgages with estimated EPCs (the EUR 390m flagged in November, 15% of the green mortgages), and slower origination.
FORMULA
is the pool at end of year ; new eligible loans; amortisation and prepayments; assets sold; the share of green mortgages carved out; the coverage; green bonds outstanding; the buffer (115%); the largest new green bond Treasury can issue in 2027 while staying above the buffer in every year, where is green bonds outstanding excluding that new bond.
INTERACTIVE LAB — The eligible pool and coverage of green bonds, 2026 to 2029
Project the eligible pool against green bonds outstanding, find the green capacity for 2027, and test the events that cause over-allocation.
| Parameter you can move | Range | Default |
|---|---|---|
| Pool at end-2026 | 2000 to 10000 EUR m | 5100 EUR m |
| Green mortgages in the pool | 0 to 6000 EUR m | 2600 EUR m |
| New eligible loans per year | 0 to 2500 EUR m | 1000 EUR m |
| Amortisation and prepayments per year | 0 to 2000 EUR m | 600 EUR m |
| Green mortgages securitised in 2027 | 0 to 2000 EUR m | 0 EUR m |
| Mortgages carved out in 2027 (estimated EPCs) | 0 to 50 % of green mortgages | 0 % of green mortgages |
| Green bonds outstanding, end-2026 | 0 to 8000 EUR m | 3500 EUR m |
| New green bond in 2027 | 0 to 2000 EUR m | 750 EUR m |
| Green bonds maturing 2028 | 0 to 2000 EUR m | 1000 EUR m |
| Green refinancing 2028 | 0 to 2000 EUR m | 1000 EUR m |
| Green bonds maturing 2029 | 0 to 2000 EUR m | 750 EUR m |
| Green refinancing 2029 | 0 to 2000 EUR m | 750 EUR m |
| Coverage buffer | 100 to 150 % | 115 % |
| Result at default settings | Value |
|---|---|
| Pool end-2027 | 5,500 EUR m |
| Pool end-2029 | 6,300 EUR m |
| Coverage end-2026 | 145.71 % |
| Coverage end-2027 | 129.41 % |
| Coverage end-2028 | 138.82 % |
| Coverage end-2029 | 148.24 % |
| Largest 2027 green bond keeping the buffer in every year | 1,282.61 EUR m |
| Unallocated proceeds end-2027 | 0 EUR m |
Try it. Add only the securitisation of EUR 800m: coverage at end-2027 falls to about 110%, below the buffer, although the bank has not lost a single green loan. Treasury’s rule is that no green asset sale is approved without a coverage test.
Result, base case. The pool grows by EUR 400m a year, to EUR 5,500m at end-2027 and EUR 6,300m at end-2029. With the green SP, coverage is 129.41% at end-2027, 138.82% at end-2028 and 148.24% at end-2029, all above 115%. The binding year is 2027: the largest new green bond that keeps the buffer in all three years is EUR 1,282.61m. The EUR 750m green SP fits with room to spare.
Result, carve-out only. If the EUR 390m of mortgages with estimated EPCs leave the pool, coverage at end-2027 is 120.24% and the 2027 capacity falls to EUR 943.48m. The green SP still fits.
Result, combined stress. With the carve-out, a securitisation of EUR 800m of green mortgages and origination falling to EUR 700m a year, the pool is EUR 4,010m at end-2027. With the green SP, coverage would be 94.35%: EUR 240m of proceeds would be unallocated. Even without the new bond, coverage would be 114.57%, just below the buffer; the 2027 capacity is negative (minus EUR 13.04m).
Reading. Over-allocation does not come from a lack of green lending but from balance sheet management decisions taken elsewhere in the bank: a securitisation decided by the capital team to release RWA, a data clean-up decided by Finance. The Green Bond Committee’s three-year test is the control that links them. Treasury’s response in the stress case would be, in order: postpone the green SP and issue it conventional; ask the retail bank to finish the EPC evidence project (the carve-out is temporary if certificates are found); and exclude green mortgages from the securitisation, or include the securitisation in the programme as a separate green securitisation with its own pool (see C04).
Who does what when coverage falls. The register owner in Treasury runs the projection every quarter. If projected coverage falls below 125% in any year, the Green Bond Committee is convened within two weeks; below 115%, green issuance is suspended and the capital team must obtain the committee’s opinion before any sale of green assets; below 100%, investor relations prepares the disclosure of unallocated proceeds for the next report, and liquidity management earmarks the corresponding amount of Level 1 bonds, outside fossil-fuel issuers, in the buffer. These triggers are written into the framework’s internal procedure, which the external reviewer reads when it gives its opinion.
INTUITION — The pool is a reservoir and the green bonds are the pipes drawing from it. Rain (new loans) fills it slowly; evaporation (amortisation) empties it slowly; a neighbour opening a sluice (a loan sale) or the surveyor redrawing the waterline (a methodology change) can empty it overnight. The 115% buffer is the minimum level below which no new pipe is connected.
PITFALL — Do not test coverage only at the issue date. A pool that covers today may not cover in eighteen months, after a planned sale. The issuance decision must use the projected pool for the whole period until the next refinancing.
Question. The Chief Sustainability Officer proposes that the securities part of the buffer reach a 15% green share by end-2027 and that fossil-fuel issuers be excluded from its corporate part. What does it cost, what does it do to the LCR, to capital and to the bank’s reported metrics?
Data (case assumptions, buffer of section 2.4 at market value). Securities buffer: Level 1 sovereign bonds EUR 21bn, Level 1 supranational and agency (SSA) bonds EUR 5bn, Level 1 covered bonds EUR 3bn, Level 2A EUR 2bn, Level 2B corporate bonds EUR 1bn; central bank reserves EUR 20bn. Green holdings today: sovereign EUR 1.2bn, SSA EUR 0.8bn, covered EUR 0.2bn. The switch into green bonds is made 60% in sovereigns and 40% in SSA, at a yield give-up of 2 bp (a cautious figure against the 1 bp twin greenium of the green Bund in C03, because some lines bought are less liquid) and a one-off transaction cost of 4 bp (bid-offer on the sale and the purchase). Fossil-fuel issuers represent EUR 0.3bn of the Level 2B book; they would be replaced by Level 1 SSA bonds yielding 60 bp less. Risk weights under the standardised approach of CRR3 (the buffer is outside the IRB scope at Meridian): sovereign and listed multilateral development banks 0%, covered bonds of the top step 10%, Level 2A 20%, Level 2B (BBB corporates) 75%. Meridian also pledges EUR 4bn of non-financial corporate bonds to the Eurosystem; the climate factor reduces their collateral value by 1.5% on average (case assumption, inside the 5% maximum announced by the ECB).
INTERACTIVE LAB — Green tilt and exclusions in the HQLA buffer: cost, LCR, capital
Measure the annual cost, the LCR effect, the capital effect and the collateral effect of the CSO’s buffer proposal.
| Parameter you can move | Range | Default |
|---|---|---|
| Sovereign bonds | 0 to 40 EUR bn | 21 EUR bn |
| SSA bonds | 0 to 15 EUR bn | 5 EUR bn |
| Level 1 covered bonds | 0 to 10 EUR bn | 3 EUR bn |
| Level 2A bonds | 0 to 10 EUR bn | 2 EUR bn |
| Level 2B corporate bonds | 0 to 10 EUR bn | 1 EUR bn |
| Central bank reserves | 0 to 40 EUR bn | 20 EUR bn |
| Green sovereign held | 0 to 10 EUR bn | 1.2 EUR bn |
| Green SSA held | 0 to 5 EUR bn | 0.8 EUR bn |
| Green covered held | 0 to 3 EUR bn | 0.2 EUR bn |
| Green share target | 0 to 40 % of securities | 15 % of securities |
| Share of the switch into sovereigns | 0 to 100 % | 60 % |
| Yield give-up on green bonds | 0 to 10 bp | 2 bp |
| One-off transaction cost | 0 to 20 bp | 4 bp |
| Level 2B fossil-fuel issuers excluded | 0 to 1 EUR bn | 0.3 EUR bn |
| Yield lost replacing them with SSA | 0 to 200 bp | 60 bp |
| Net outflows, 30 days | 10 to 60 EUR bn | 36 EUR bn |
| Corporate bonds pledged to the Eurosystem | 0 to 15 EUR bn | 4 EUR bn |
| Average climate factor reduction | 0 to 5 % | 1.5 % |
| Result at default settings | Value |
|---|---|
| Securities buffer | 32.0 EUR bn |
| Green share today | 6.875 % |
| Amount to switch into green | 2.60 EUR bn |
| Annual yield give-up of the tilt | 0.52 EUR m |
| One-off transaction cost | 1.04 EUR m |
| Annual cost of the exclusion | 1.80 EUR m |
| Total annual cost of the policy | 2.32 EUR m |
| LCR before | 141.64 % |
| LCR after | 142.06 % |
| RWA of the securities buffer, before | 1.450 EUR bn |
| RWA of the securities buffer, after | 1.225 EUR bn |
| Capital released (13% of RWA) | 29.25 EUR m |
| Required return on the capital released (12%) | 3.51 EUR m |
| Collateral value lost to the climate factor | 60.0 EUR m |
Try it. Set the target to 25%: the switch more than doubles to EUR 5.8bn and Treasury would have to buy green lines that trade by appointment. That is where the liquidity value of the buffer, not its cost, becomes the binding constraint.
Result. The securities buffer is EUR 32bn and today 6.875% green. Reaching 15% means switching EUR 2.6bn into green bonds: EUR 0.52m a year of yield give-up and EUR 1.04m of one-off transaction costs. The exclusion of fossil-fuel issuers costs more, EUR 1.8m a year, because Level 2B corporates yield much more than the SSA bonds that replace them. Total annual cost: EUR 2.32m. The LCR rises slightly, from 141.64% to 142.06%, since the replacement bonds are Level 1 with no haircut. The buffer’s RWA fall from EUR 1.45bn to EUR 1.225bn, releasing EUR 29.25m of allocated capital. The climate factor removes EUR 60m from the collateral value of the bonds pledged to the Eurosystem.
Reading. Four points for ALCO.
The green tilt is cheap; the exclusion is the expensive part. Swapping conventional sovereigns for green twins costs a couple of basis points. Excluding high-yielding corporates costs real carry, partly compensated by lower capital: the EUR 29.25m of capital released, at the 12% hurdle, is worth EUR 3.51m a year of required return, more than the EUR 1.8m of carry lost. Seen through RAROC, the exclusion is close to neutral; seen through net interest income, it is a cost. Both readings must be shown.
The LCR does not care. The green tilt is between Level 1 assets with a 0% haircut: no effect. The small improvement comes from the exclusion, not from the label.
The metrics barely move. Sovereign and supranational bonds are outside the GAR denominator, so the tilt does not change the GAR. It appears in the bank’s own disclosures on its investment portfolio and in the green bond report as a home for unallocated proceeds. The exclusion reduces financed emissions of the bond portfolio, which Meridian measures with the PCAF method for listed corporate bonds; the sovereign bonds carry PCAF sovereign emissions that do not change with the label.
Liquidity first. Investment policy caps any single green line at 10% of its issue size and keeps 80% of the securities buffer in bonds with an issue size of at least EUR 3bn. The CSO’s 15% passes these caps; 25% would not without buying smaller, less liquid lines.
WHO DOES WHAT — The proposal is a change to the investment policy of the liquidity portfolio, so it goes to ALCO with a memo from the Head of Liquidity, an opinion from Liquidity Risk on the stress value of the new lines, and an opinion from the sustainability team on the exclusion list. The portfolio managers execute the switch over a quarter, avoiding month-ends, when liquidity in green lines is thinnest.
PITFALL — Calling the buffer “green” because 15% of its securities are green bonds is a disclosure risk. The buffer is held for liquidity; the label describes the issuers’ use of proceeds, not the bank’s. Report the share as a fact, not as a financing claim.
Question. DCM suggests replacing the EUR 750m green SP by a EUR 750m green covered bond backed by green mortgages: cheaper, and with a deep investor base. What does Meridian gain and give up?
Data (case assumptions). Green SP at 81 bp (83 bp less 2 bp greenium); green covered bond at 39 bp (40 bp less 1 bp greenium). Meridian’s covered bond programme is managed at 25% overcollateralisation for its rating, so each euro of covered bond encumbers EUR 1.25 of mortgages. Green mortgages earmarked to a green covered bond are removed from the pool supporting senior green bonds, at the same 115% buffer. Encumbered assets today: EUR 55bn (covered bond pools, repo, central bank pledges) on total assets of EUR 310bn. Taxonomy-aligned green mortgages: 62% of EUR 2,600m (C03). End-2027 pool EUR 5,500m (section 5.2); TREA EUR 124bn.
INTERACTIVE LAB — Green senior preferred or green covered bond: cost, encumbrance, pool and MREL
Compare the two formats for the EUR 750m green deal on annual cost, encumbrance ratio, pool coverage and remaining green capacity, Taxonomy room and MREL.
| Parameter you can move | Range | Default |
|---|---|---|
| Size of the green deal | 250 to 2000 EUR m | 750 EUR m |
| SP new-issue spread | 30 to 200 bp | 83 bp |
| Greenium on SP | 0 to 10 bp | 2 bp |
| Covered new-issue spread | 0 to 100 bp | 40 bp |
| Greenium on covered | 0 to 10 bp | 1 bp |
| Programme overcollateralisation | 5 to 50 % | 25 % |
| Encumbered assets today | 0 to 150 EUR bn | 55 EUR bn |
| Total assets | 100 to 600 EUR bn | 310 EUR bn |
| Pool end-2027 before the deal | 2000 to 10000 EUR m | 5500 EUR m |
| Senior green bonds outstanding before the deal | 0 to 8000 EUR m | 3500 EUR m |
| Coverage buffer | 100 to 150 % | 115 % |
| Green mortgages | 0 to 6000 EUR m | 2600 EUR m |
| Mortgages Taxonomy-aligned | 0 to 100 % | 62 % |
| TREA end-2027 | 80 to 160 EUR bn | 124 EUR bn |
| Result at default settings | Value |
|---|---|
| Annual spread cost, green SP | 6.075 EUR m |
| Annual spread cost, green covered | 2.925 EUR m |
| Annual saving of the covered format | 3.150 EUR m |
| Saving as a multiple of the SP greenium | 21.0 x |
| Additional mortgages encumbered | 937.5 EUR m |
| Encumbrance ratio with green SP | 17.70 % |
| Encumbrance ratio with green covered | 18.00 % |
| Green mortgages earmarked to the covered bond | 862.5 EUR m |
| Senior pool coverage, SP option | 129.41 % |
| Senior pool coverage, covered option | 132.50 % |
| Remaining green capacity, SP option | 532.61 EUR m |
| Remaining green capacity, covered option | 532.61 EUR m |
| Aligned mortgages left after earmarking (EuGB room) | 749.5 EUR m |
| Total MREL ratio contributed by the SP | 0.605 pp of TREA |
Try it. Set the overcollateralisation to 10%: encumbrance rises less, but the rating agencies would likely lower the covered bond rating, and the 40 bp spread would no longer hold.
Result. The covered format costs EUR 2.925m a year against EUR 6.075m for the SP: a saving of EUR 3.15m a year, 21 times the greenium of the SP. It encumbers EUR 937.5m more mortgages and raises the encumbrance ratio from 17.70% to 18.00%. Both formats use the same green capacity: coverage of senior green bonds is 129.41% with the SP and 132.50% with the covered bond (whose EUR 862.5m of earmarked mortgages leave the senior pool), and the remaining capacity at the buffer is EUR 532.61m in both cases. After earmarking, EUR 749.5m of aligned mortgages remain, enough for the covered bond itself to be issued as a European Green Bond if Meridian wished. The SP would add 0.605 points of TREA to the total MREL ratio; the covered bond adds nothing.
Reading. The pool is indifferent to the format, because Meridian applies the same buffer to both; the decision lies elsewhere. In favour of the covered bond: cost, a 3.15m saving every year, and a natural fit, since mortgages are the most Taxonomy-ready part of the pool. Against it: encumbrance (rating agencies and senior investors watch it, and Meridian’s risk appetite caps the ratio at 20%), and MREL, because C03 left only a EUR 0.122bn total MREL shortfall before the SNP but a growing balance sheet will need senior eligible debt in 2028. ALCO’s decision: keep the green SP in 2027 and plan the first green covered bond in 2028, when the conventional covered maturity is larger. The lesson for students: in a bank, the format question (secured or unsecured, senior or subordinated) is worth ten times the label question.
WHY IT MATTERS ON THE DESK — On the desk, the green label is decided after the format, never before. A DCM banker who pitches “a green deal” without saying which layer of the liability stack it sits in has not done the work.
Question. Meridian grants a 3 bp FTP rebate on loans in the eligible pool (B02). C03 showed that in 2026 the greenium could pay for only about 2.06 bp of it, leaving a subsidy of EUR 0.48m a year. With the 2027 plan, is the rebate self-financing, how much does green funding lower the bank’s blended cost of funds, and what should ALCO decide?
Data (case assumptions). Green bonds outstanding at end-2027: EUR 3,500m at an average greenium of 3 bp (C03) and the new EUR 750m green SP at 2 bp. Pool at end-2027: EUR 5,500m (section 5.2). Running costs of the programme (allocation and impact reporting, external review, register maintenance, data) EUR 0.20m a year, shared by all bonds. Policy rebate 3 bp. Term wholesale funding at end-2027, from the plan of section 5.1: covered EUR 28.5bn, SP EUR 10.75bn, SNP EUR 9.75bn, capital EUR 4.72bn, so EUR 53.72bn. Total funding (balance sheet) at end-2027: EUR 316bn.
FORMULA
is the annual saving from greeniums; and the amount and greenium of green bond ; the eligible pool; the rebate the saving can pay on the whole pool; the programme’s running costs; the rebate net of those costs; the reduction of the blended cost of term wholesale funding if the saving were spread over all of it instead.
INTERACTIVE LAB — Is the green FTP rebate self-financing? Saving, rebate and Treasury’s net result
Compute the greenium saving, the self-financing rebate gross and net of programme costs, Treasury’s subsidy under the policy rebate, and the effect of the saving on the blended cost of funds.
| Parameter you can move | Range | Default |
|---|---|---|
| Green bonds outstanding before 2027 | 0 to 8000 EUR m | 3500 EUR m |
| Their average greenium | 0 to 10 bp | 3 bp |
| New green bond 2027 | 0 to 2000 EUR m | 750 EUR m |
| Its greenium | 0 to 10 bp | 2 bp |
| Eligible pool | 1000 to 12000 EUR m | 5500 EUR m |
| Programme running costs | 0 to 1 EUR m per year | 0.2 EUR m per year |
| Policy rebate | 0 to 5 bp | 3 bp |
| Term wholesale funding | 20 to 120 EUR bn | 53.72 EUR bn |
| Total funding (balance sheet) | 100 to 600 EUR bn | 316 EUR bn |
| Five-year FTP liquidity premium | 0 to 80 bp | 30 bp |
| Result at default settings | Value |
|---|---|
| Annual greenium saving | 1.200 EUR m |
| Average greenium on green bonds | 2.824 bp |
| Self-financing rebate (gross) | 2.182 bp |
| Self-financing rebate net of running costs | 1.818 bp |
| Annual cost of the policy rebate | 1.650 EUR m |
| Treasury subsidy before running costs | 0.450 EUR m |
| Treasury subsidy after running costs | 0.650 EUR m |
| FTP for an eligible, non-aligned loan (premium less net rebate) | 28.2 bp |
| Green bonds as share of term wholesale funding | 7.91 % |
| Saving spread over term wholesale funding | 0.223 bp |
| Saving spread over total funding | 0.0380 bp |
Try it. Raise the pool to EUR 8,000m with the same bonds, as if Meridian’s green lending grew faster than its green issuance: the self-financing rebate falls to 1.5 bp. A rebate on the whole pool is only self-financing if the bank issues green almost as fast as it lends green.
Result. In 2027 the green bonds save EUR 1.2m a year, an average greenium of 2.824 bp. Spread over the EUR 5,500m pool, that pays a rebate of 2.182 bp, or 1.818 bp net of the programme’s running costs. The 3 bp policy rebate costs EUR 1.65m a year: Treasury subsidises green lending by EUR 0.45m a year before running costs and EUR 0.65m after. The second test reproduces C03’s 2026 figures: 2.059 bp self-financing and EUR 0.48m of subsidy. Green bonds are 7.91% of term wholesale funding; if their saving were spread over all of it, the blended cost would fall by 0.223 bp; over the whole balance sheet, by 0.038 bp.
Reading. Does green funding lower Meridian’s cost of funds? Yes, by about four hundredths of a basis point on the total funding, a rounding error on the FTP curve. The rebate is therefore not a pass-through of a market benefit but mostly an internal allocation choice. ALCO has three honest options:
Meridian’s ALCO chooses option 1 for 2027 with a review clause: the rebate is kept at 3 bp only for loans that are Taxonomy-aligned (to push the aligned share up, the data problem of C03) and falls to the net self-financing level for loans that are only framework-eligible. This rule gives the incentive to the assets that will make a European Green Bond possible later.
What the desks see. For a five-year loan, the corporate lending desk’s FTP liquidity premium is 30 bp. Under ALCO’s rule, a loan admitted to the pool and Taxonomy-aligned is charged 27 bp; a loan that is only framework-eligible is charged 30 bp less the net self-financing rebate of 1.818 bp, rounded in the FTP grid to 28.2 bp; a loan outside the pool pays 30 bp. The rebate is fixed at drawdown for the life of the drawing, like the rest of the FTP, and stops if the loan leaves the pool (for example after a declassification, see B02). Each quarter ALM reports to ALCO the volume of loans at each rebate level and the cost to Treasury, so that the subsidy stays visible.
INTUITION — A household that gets a 1% discount at one supermarket because it shops there with a loyalty card does not get poorer or richer by spending that 1% on vegetables rather than on biscuits. The discount is real but small, and how the family spends it is a choice, not a law of nature.
PITFALL — “Our green bonds allow us to offer cheaper green loans” is a claim that the numbers rarely support. The greenium saving is a few hundred thousand euros a year against a green pool of billions. If the rebate exceeds the self-financing level, the claim should read “we choose to subsidise green loans from our general funding”.
Question. What will Meridian publish about the programme in 2027, and what will investors ask?
Data. The figures of sections 5.1 to 5.5.
The allocation report (published with the annual report, with limited assurance by the external reviewer) shows, for end-2026: the pool by category (EUR 5,100m), green bonds outstanding (EUR 3,500m), coverage (145.71%), no unallocated proceeds, the share of refinancing (close to 100% in a portfolio approach), the age profile of the pool, the Taxonomy-aligned share (68.47%, C03), and impact metrics (renewable capacity financed, estimated avoided emissions, energy performance of the mortgage book) with their methods. In 2027 the report adds a forward note on the EPC evidence project, because an analyst who sees EUR 390m of mortgages flagged as “estimated EPC” in the register extract will ask.
The encumbrance disclosure (Pillar 3 templates on encumbered and unencumbered assets, median of four quarters) shows the effect of the covered bond programme. The green SP adds nothing to it; a green covered bond would (section 5.4).
The fixed income investor presentation shows the funding plan (EUR 8.25bn for 2027, split by instrument), the MREL position, the LCR and NSFR, and one slide on the green programme: pool, coverage, three-year capacity, planned green issuance. Investors ask four questions, every year: Is the pool growing faster than the green bonds? What happens to coverage if you securitise mortgages? How much of the pool is Taxonomy-aligned, and when will you issue a European Green Bond? Does the label change anything in resolution? The answers are in sections 5.2, 5.2, C03 5.2, and C03 5.6 (no).
The path to a European Green Bond. The question that investors ask most is when Meridian will issue an EuGB. Treasury’s answer in the 2027 presentation is conditional: once the aligned share of mortgages reaches about 80% (the EPC evidence project and physical climate risk assessments), the aligned pool would cover the whole green programme (C03 section 5.2), and the first EuGB would most naturally be a green covered bond backed by aligned mortgages, planned for 2028 (section 5.4). The register already separates aligned from eligible assets, so the factsheet’s portfolio test could be produced from it.
WHO DOES WHAT — Debt investor relations keeps a log of investor questions after each roadshow and sends it to the Green Bond Committee. In Meridian’s 2026 log, the most frequent question was about the age of the pool, not its size: investors increasingly want to know how much of the green funding finances new loans.
| Stage | Key figure | Value |
|---|---|---|
| 5.1 Plan | Gross need; issuance 2027 | EUR 9.25bn; EUR 8.25bn |
| 5.1 Plan | SNP balance; cushion over subordination gap | EUR 0.5bn; EUR 0.238bn |
| 5.1 Plan | Spread cost of the vintage; average | EUR 52.825m; 64.03 bp |
| 5.1 Plan | Greenium saving; share of spread cost | EUR 0.15m; 0.284% |
| 5.1 Plan | Green share (2027; with pre-funding) | 9.09%; 18.92% |
| 5.1 Plan | Loan-to-deposit ratio | 108.57% to 109.80% |
| 5.2 Pool | Coverage 2027, 2028, 2029 | 129.41%, 138.82%, 148.24% |
| 5.2 Pool | 2027 green capacity at 115% | EUR 1,282.61m |
| 5.2 Pool | Combined stress: coverage; unallocated | 94.35%; EUR 240m |
| 5.3 Buffer | Switch to reach 15% green | EUR 2.6bn |
| 5.3 Buffer | Annual cost (tilt; exclusion; total) | EUR 0.52m; 1.8m; 2.32m |
| 5.3 Buffer | LCR before and after | 141.64%; 142.06% |
| 5.3 Buffer | Capital released; climate factor loss | EUR 29.25m; EUR 60m |
| 5.4 Format | Annual saving of covered vs SP | EUR 3.15m (21 x greenium) |
| 5.4 Format | Encumbrance ratio SP vs covered | 17.70%; 18.00% |
| 5.4 Format | Remaining green capacity (both) | EUR 532.61m |
| 5.5 FTP | Greenium saving 2027; average greenium | EUR 1.2m; 2.824 bp |
| 5.5 FTP | Self-financing rebate gross; net | 2.182 bp; 1.818 bp |
| 5.5 FTP | Subsidy at 3 bp (before; after costs) | EUR 0.45m; EUR 0.65m |
| 5.5 FTP | Effect on cost of term wholesale; total | 0.223 bp; 0.038 bp |
The thread through the case: the funding plan, the format and the buffer are financial decisions worth millions a year; the green label is worth a few hundred thousand euros and depends on a pool that other parts of the bank can shrink. Treasury’s contribution to sustainability is to keep the pool honest, the reporting accurate and the subsidy visible.
The bonds Meridian issues consume no capital; they are liabilities. Treasury’s capital consumption comes from its assets and its risks. Credit risk on the HQLA portfolio is small: sovereign and listed multilateral development bank bonds carry 0%, top-rated covered bonds 10%, Level 2A 20% and BBB corporate bonds 75% under the CRR3 standardised approach, so the EUR 32bn securities buffer carries EUR 1.45bn of RWA, falling to EUR 1.225bn after the exclusion of section 5.3. Counterparty credit risk and CVA arise on the swaps that hedge fixed-rate issues; collateralised under CSAs with central clearing for most of them, they are modest. IRRBB is a Pillar 2 risk, capitalised through the SREP requirement rather than Pillar 1. Market risk on the buffer is in the banking book; under IFRS 9 the FVOCI portfolio’s fair value changes flow into CET1 through other comprehensive income, which is why Treasury hedges the duration of the buffer with swaps.
The leverage ratio is a stronger constraint on the buffer than RWA: every euro of HQLA, including central bank reserves, is leverage exposure. At a 3.5% internal leverage allocation (C11), EUR 50bn of buffer uses about EUR 1.75bn of Tier 1. Green or not, the buffer’s capital cost is its size.
MREL is Treasury’s capital-like constraint. The excess cost of SNP over the FTP curve is charged to businesses by RWA. With the 2027 plan, SNP outstanding rises to EUR 9.75bn, so the charge rises from C03’s 6.27 bp to 6.60 bp of RWA (section 6.2).
The funding programme has no RAROC of its own; Treasury’s result is a cost-centre result. Its economic effect on the bank’s returns passes through the FTP charged to the business lines. We therefore work out the RAROC of the representative green loan used in C03, with the 2027 parameters.
Data (case assumptions, from C03). EUR 100m five-year term loan to a BBB-rated renewable developer, fully drawn; margin 110 bp over EURIBOR; FTP liquidity premium 30 bp; upfront fee 40 bp spread over five years; cost-income ratio 40% on revenues before rebate; F-IRB, PD 0.20%, LGD 40%, M 2.5 years. MREL charge: SNP EUR 9.75bn at 84 bp over FTP on TREA of EUR 124bn. Rebate options: none, the net self-financing level of 1.818 bp, the gross self-financing level of 2.182 bp, and the policy rebate of 3 bp.
INTERACTIVE LAB — RAROC of a green loan under the 2027 FTP options
Compute the RAROC of a EUR 100m green loan with the 2027 MREL charge and each rebate option, and the margin needed to reach the 12% hurdle.
| Parameter you can move | Range | Default |
|---|---|---|
| Loan amount | 10 to 300 EUR m | 100 EUR m |
| Margin over EURIBOR | 50 to 300 bp | 110 bp |
| FTP liquidity premium | 0 to 80 bp | 30 bp |
| Upfront fee (spread over 5 years) | 0 to 100 bp | 40 bp |
| PD | 0.03 to 4 % | 0.2 % |
| SNP outstanding | 0 to 15 EUR bn | 9.75 EUR bn |
| SNP spread over FTP | 0 to 200 bp | 84 bp |
| TREA | 80 to 160 EUR bn | 124 EUR bn |
| Net self-financing rebate | 0 to 5 bp | 1.818 bp |
| Gross self-financing rebate | 0 to 5 bp | 2.182 bp |
| Policy rebate | 0 to 5 bp | 3 bp |
| Result at default settings | Value |
|---|---|
| F-IRB risk weight | 39.02 % |
| RWA | 39.02 EUR m |
| Revenues before rebate | 0.880 EUR m |
| Operating costs | 0.352 EUR m |
| Expected loss | 0.080 EUR m |
| Allocated capital | 5.072 EUR m |
| MREL charge per euro of RWA | 6.605 bp |
| MREL charge on the loan | 0.0258 EUR m |
| RAROC, no rebate | 6.24 % |
| RAROC, net self-financing rebate | 6.51 % |
| RAROC, gross self-financing rebate | 6.57 % |
| RAROC, policy rebate | 6.69 % |
| RAROC gap between policy and net self-financing rebate | 0.17 points |
| Distance to the hurdle with no rebate | 5.76 points |
| Margin for 12% with no rebate | 174.9 bp |
| Margin for 12% with the policy rebate | 169.9 bp |
Try it. Compare the two lines on the chart: the policy rebate shifts RAROC up by less than half a point at every margin. The 12% hurdle is reached at about 170 bp with the rebate and 175 bp without; the label does not make an unprofitable loan profitable.
Line by line. Revenues: net margin over FTP of 80 bp on EUR 100m, EUR 0.80m, plus EUR 0.08m of upfront fee a year: EUR 0.88m. Costs at 40%: EUR 0.352m. Expected loss: 0.20% x 40% x EUR 100m = EUR 0.08m. Risk weight 39.02%, RWA EUR 39.02m, allocated capital EUR 5.072m. The 2027 MREL charge is 6.605 bp of RWA, EUR 0.0258m on this loan. RAROC: 6.24% with no rebate; 6.51% with the net self-financing rebate; 6.57% with the gross self-financing rebate; 6.69% with the 3 bp policy rebate. The margin needed for 12% is 174.9 bp without rebate and 169.9 bp with it. The second test reproduces C03’s 2026 figures (6.26% and 6.71%), the difference coming only from the extra EUR 0.5bn of SNP in the 2027 plan.
How the sustainable feature changes it. By less than half a point of RAROC, whatever the rebate design. The choice ALCO debated in section 5.5, policy rebate against net self-financing rebate, is worth 0.18 point; the distance to the hurdle is 5.76 points. The loan’s profitability is decided by its margin and the client’s ancillary business (B01), not by the green funding.
The green SP and SNP are ordinary term liabilities. For the NSFR they count 100% as available stable funding while their residual maturity exceeds one year, 50% between six months and one year, and nothing after. For the LCR they generate an outflow only in their last 30 days. A green covered bond is the same for the issuer; for the holder it may be Level 1 HQLA. Nothing in either ratio depends on the label.
The FTP effects are those of section 5.5: the greenium lowers the blended cost of term wholesale funding by 0.223 bp, a figure ALM would not even round into the published curve; the rebate is a separate line in the FTP grid, applied to loans flagged in the pool. The HQLA carry charged to businesses (40 bp on the buffer attributable to their outflows) rises marginally with the buffer policy of section 5.3, because its EUR 2.32m annual cost is part of the buffer’s carry; Meridian chose to keep it in Treasury’s result rather than charge it out, on the grounds that the policy is a bank-wide decision.
Unallocated proceeds. In the stress of section 5.2, EUR 240m of proceeds would be unallocated and held in the buffer under the exclusion policy. The cost is the difference between the green SP’s cost and the yield of Level 1 assets, about the SP spread on that amount, and the reputational cost of reporting it.
Issued bonds. Under IFRS 9 issued green bonds are financial liabilities at amortised cost, with fees and discount in the effective interest rate. Fixed-rate issues are hedged with receive-fixed swaps in fair value hedges (individual or, for portfolios, under the IAS 39 carve-out that EU banks still use for macro hedging). The green label has no accounting consequence: there is no contingent feature linked to a sustainability outcome, unlike a sustainability-linked bond (C02), whose step-up raises the questions addressed by the IASB amendments to the classification and measurement of financial instruments of May 2024, effective from 1 January 2026.
HQLA portfolio. The buffer is typically held in a “hold to collect and sell” business model and measured at fair value through other comprehensive income (FVOCI), with expected credit losses on stage 1 for almost all positions. Green and conventional bonds of the same issuer are classified the same way; both pass the SPPI test. The sale of conventional bonds for the green switch realises OCI gains or losses into profit or loss and must be consistent with the business model (sales are expected in a “collect and sell” model).
Disclosures. IFRS 7 requires liquidity risk disclosures (maturity analysis), and the notes on debt securities issued show the green bonds within their seniority class. The covered bond pools appear in the disclosure of assets pledged as collateral.
Green Asset Ratio and BTAR. The funding programme does not change the GAR: it is computed on assets. The buffer tilt does not change it either, because sovereign and supranational exposures are outside the denominator (section 5.3). What links Treasury to the GAR is data: the pool register and the GAR must agree on which loans are aligned; a bank that issued a EuGB while deferring its detailed Taxonomy KPIs would face an obvious inconsistency, and the simplified Taxonomy reporting forbids that combination.
Pillar 3 ESG. Under the current templates the pool appears in the Taxonomy-alignment templates (stopped for large institutions under the EBA’s final draft of 22 June 2026, not yet adopted) and, for assets not aligned but supporting mitigation, in the template on other climate change mitigating actions, where banks also report the green bonds they hold. The buffer’s corporate bonds appear in the transition risk templates by sector.
Financed emissions (PCAF). Meridian’s liabilities create no financed emissions. Its HQLA holdings do: corporate bonds under the PCAF listed equity and corporate bond method, sovereign bonds under the PCAF sovereign debt method. The exclusion of section 5.3 lowers the first; the green tilt into sovereign bonds does not change the second, because PCAF attributes country emissions by the size of the exposure, not by the use of proceeds of the bond.
CSRD statement. Meridian’s sustainability statement describes the green funding programme under its policies and actions on climate change mitigation, and the buffer’s ESG policy under its investment policies. Both must be consistent with the allocation report and the Pillar 3 figures, which is the job of Finance’s sustainability reporting team (E06).
| Risk | What it looks like | Control | Owner |
|---|---|---|---|
| Over-allocation | Coverage below 100% after a sale or carve-out | Three-year projected coverage at 115%; no green deal or asset sale without a coverage test | Green Bond Committee |
| Double counting | Same mortgage behind a green SP, a green covered bond and a green deposit | One programme flag per asset in the register | Treasury (register owner) |
| Data quality | Estimated EPCs, missing project evidence | Actual-or-estimated field; audit sample; carve-out rule | Sustainability team, Internal Audit |
| Eligible versus aligned confusion | Pool called “Taxonomy-aligned” | Separate fields and wording review of all publications | Sustainability team, Legal |
| Funding concentration | Too much maturing in one year; reliance on one investor type | Maturity ladder limits; diversification targets | Group Treasurer, Liquidity Risk |
| Buffer liquidity | Green lines too small to sell in a stress | Issue-size floors and line caps in the investment policy | Head of Liquidity, Liquidity Risk |
| Encumbrance | Covered bond growth weakens unsecured creditors | Encumbrance limit (20%) in risk appetite | ALCO |
| Hidden subsidy | Rebate above the self-financing level presented as market benefit | Annual self-financing test in the FTP policy | Head of ALM, Finance |
| Misleading claims | “Green buffer”, “green funding lowers loan prices” | Disclosure committee review | Investor relations, Compliance |
| Collateral value | Climate factor reduces counterbalancing capacity | Collateral inventory recalculated after each ECB update | Collateral management |
Red-flag checklist for a reviewer of a bank’s green funding programme.
PITFALL — The most common weakness in practice is not fraud but drift: the pool was right at issue, then a sale, a methodology change and a slowdown in green origination erode coverage between two annual reports. A quarterly register review with a forward test is the control that catches drift.
What changed recently. The climate factor turned a long-discussed idea into a live, if small, collateral haircut. The CMDI reform changes the ranking of senior preferred debt against deposits, which matters more for bank spreads than any label. The narrowing of the CSRD scope makes it harder for banks to collect evidence from borrowers, which slows the growth of aligned pools.
Open debates.
INTERACTIVE LAB — Treasury, funding and liquidity: check your understanding
Test the mechanics of funding plans, FTP, LCR, pool management and green pass-through.
Q1. Which constraint fixed the minimum amount of SNP in Meridian’s 2027 plan?
Answer: B. C03 left a EUR 0.262bn subordination shortfall; SNP closes it. The label played no role.
Q2. Why are SNP and Tier 2 left out of the FTP liquidity curve?
Answer: C. Charging them by tenor would overprice long low-RWA assets such as mortgages.
Q3. A EUR 750m green SP replaces a conventional SP at a 2 bp greenium. What is the annual saving?
Answer: D. 750m x 2 bp = EUR 0.15m a year.
Q4. In the LCR, a green Bund compared with a conventional Bund of the same maturity has:
Answer: A. The LCR is label-blind.
Q5. Which event caused the largest single fall in Meridian’s pool in the combined stress?
Answer: C. The sale removed EUR 800m at once, against EUR 390m for the carve-out and EUR 300m a year for slower origination.
Q6. Why does tilting the buffer towards green sovereign bonds not raise the Green Asset Ratio?
Answer: B. The GAR covers exposures to undertakings and households; sovereigns, central banks and supranationals are excluded.
Q7. What did the Eurosystem climate factor change from 15 June 2026?
Answer: C. It reduces collateral value after normal haircuts, scaled by an uncertainty score.
Q8. Meridian’s green covered bond option saves EUR 3.15m a year against the green SP. What does it cost in balance sheet terms?
Answer: A. At 25% overcollateralisation, EUR 750m of covered bonds encumbers EUR 937.5m of mortgages; it also adds nothing to MREL.
Q9. The self-financing rebate is defined as:
Answer: D. It is what the saving can pay on every euro of the pool.
Q10. Spread over Meridian’s total funding, the 2027 greenium saving lowers the cost of funds by about:
Answer: A. EUR 1.2m over EUR 316bn is 0.038 bp.
Q11. Under the CMDI reform published in April 2026, senior preferred bonds:
Answer: B. General depositor preference puts all deposits above ordinary senior unsecured claims.
Q12. Which statement about over-allocation of a bank green bond is correct?
Answer: C. Green labels carry no contractual consequence; the effect is reputational and in reporting.
Q13. Why does the exclusion of fossil-fuel issuers from Level 2B cost more carry than the green tilt?
Answer: A. 60 bp of yield lost on EUR 0.3bn is EUR 1.8m a year, against 2 bp on EUR 2.6bn for the tilt.
Q14. Which document reports a bank’s three-year funding plan to its supervisor?
Answer: C. The EBA guidelines harmonise definitions and templates for funding plans.
A bank has EUR 5.0bn of term debt maturing next year. Loans grow by EUR 4.0bn and deposits by EUR 5.0bn; Treasury wants a buffer EUR 1.0bn larger; EUR 0.5bn was pre-funded. It plans EUR 2.0bn of covered bonds at 35 bp and EUR 1.5bn of SP at 70 bp, and SNP at 105 bp for the balance. Compute the need, the SNP amount, the annual spread cost and the average spread.
Solution. bn. Deposits growing faster than loans reduce the need. SNP = 4.5 − 2.0 − 1.5 = EUR 1.0bn. Spread cost: 2.0bn x 35 bp = EUR 7.0m; 1.5bn x 70 bp = EUR 10.5m; 1.0bn x 105 bp = EUR 10.5m; total EUR 28.0m a year. Average: 28.0 / 4.5bn = 62.22 bp. If the bank labelled the SP green at a 2 bp greenium, it would save EUR 0.3m a year, about 1.1% of the vintage’s spread cost.
A bank’s pool is EUR 3,000m against EUR 2,200m of green bonds, with no green maturities over the next three years. It originates EUR 600m of eligible loans a year, amortisation is EUR 450m a year, and it will sell EUR 300m of green loans in year 1. Its buffer is 120%. What is the largest green bond it can issue in year 1? What would coverage be with a EUR 500m bond?
Solution. Pool: year 1, 3,000 + 600 − 450 − 300 = EUR 2,850m; year 2, EUR 3,000m; year 3, EUR 3,150m. Capacity each year: 2,850 / 1.2 − 2,200 = EUR 175m; 3,000 / 1.2 − 2,200 = EUR 300m; 3,150 / 1.2 − 2,200 = EUR 425m. The binding year is year 1: EUR 175m, below benchmark size, so the bank should not issue a green benchmark in year 1. With EUR 500m, coverage in year 1 would be 2,850 / 2,700 = 105.56%: no over-allocation, but below the buffer, and one more sale would push it below 100%. The bank could instead issue in year 3, or exclude green loans from the sale.
A bank has EUR 2,000m of green bonds at a 2.5 bp greenium and EUR 500m at 1.5 bp, an eligible pool of EUR 4,000m, running costs of EUR 0.15m a year and a policy rebate of 2 bp. Its term wholesale funding is EUR 30bn. Compute the saving, the self-financing rebate gross and net, the subsidy and the effect on the cost of term wholesale funding.
Solution. Saving: 2,000 x 2.5 bp + 500 x 1.5 bp = EUR 0.5m + EUR 0.075m = EUR 0.575m. Gross self-financing rebate: 0.575 / 4,000 = 1.4375 bp; net of costs: (0.575 − 0.15) / 4,000 = 1.0625 bp. Policy cost: 4,000 x 2 bp = EUR 0.8m; subsidy EUR 0.225m before costs and EUR 0.375m after. Spread over term wholesale funding the saving is 0.575 / 30,000 = 0.19 bp. The 2 bp rebate is almost twice the net self-financing level: it is mainly a subsidy.
1. Why does a bank issue green bonds if the greenium is so small? Because it must issue the bonds anyway, for funding and MREL, and the label is a way to sell part of them to a wider and stickier investor base. Green books tend to be more granular, with dedicated funds that hold to maturity, which supports secondary performance and the next deal. The label also forces the bank to build a register of green assets that it needs for its Taxonomy and Pillar 3 reporting. The saving itself is a few hundred thousand euros a year per benchmark, and I would never present it as the main reason.
2. Walk me through a bank’s funding plan. Start from term debt maturing and called in the year, add the growth of loans net of the growth of deposits, add any planned increase in the liquidity buffer, and subtract what was pre-funded. Then split the total by instrument: the MREL and subordination requirements fix the minimum SNP and capital, cost favours covered bonds, and encumbrance limits and diversification set the rest. Finally spread it by currency, tenor and month, with a contingency plan if the unsecured market closes. The green question comes last: which of these bonds can the pool support?
3. What is over-allocation and how do you prevent it? It is having more green bonds outstanding than eligible assets, so part of the proceeds is unallocated. It usually comes from asset sales, methodology changes or amortisation outpacing new green lending, not from a lack of green business. The control is a register with programme flags, an internal coverage buffer above 100%, and a forward test: no green bond is issued, and no green asset sale is approved, if projected coverage falls below the buffer in any year to the next refinancing. If it happens, the bank discloses the unallocated amount and holds it in liquid assets that exclude fossil-fuel issuers.
4. Should the liquidity buffer hold green bonds? It can, within limits, because green sovereign and supranational bonds are Level 1 like their twins and cost only a basis point or two of yield. The limits are liquidity: green lines are smaller and trade less, so the investment policy should cap them by issue size and share of issue. I would also be careful with the message: the buffer exists for liquidity, the tilt does not change the GAR, and calling the buffer green would be misleading. Exclusions of fossil-fuel corporates are a bigger decision, because they cost real carry.
5. Does green funding lower a bank’s cost of funds? Measurably, no. A few basis points on green bonds that are less than a tenth of term wholesale funding move the blended cost by a fraction of a basis point, and the effect on total funding is a few hundredths of a basis point. Covered bonds against senior unsecured, or the timing of issuance, are worth far more. The honest statement is that green funding diversifies the investor base, which may lower spreads indirectly over time.
6. How should a green FTP rebate be set? First compute the self-financing level: the annual greenium saving, net of the programme’s running costs, divided by the eligible pool. A rebate at that level is a genuine pass-through; anything above it is a subsidy that Treasury pays from general funding. A subsidy can be a legitimate strategic choice, but it should be labelled as such in the FTP policy and its effect on origination should be monitored. I would also consider targeting it at Taxonomy-aligned loans, where the bank’s data effort pays off.
7. Green senior or green covered bond? The covered bond is much cheaper, often by 40 bp or more, and mortgages are the easiest assets to prove Taxonomy-aligned, so it could even be a European Green Bond. Against it, it encumbers mortgages, which weakens unsecured creditors and is capped by risk appetite, and it does not count towards MREL. The pool is consumed in both cases if the bank applies the same buffer. So the choice is a funding-structure choice first; the label follows.
8. What does the ECB climate factor mean for a bank treasurer? Since 15 June 2026 the Eurosystem reduces the collateral value of non-financial corporate bonds by an additional amount, at most 5%, depending on the issuer’s exposure to transition risk and the bond’s maturity. It does not affect the LCR, but it reduces the counterbalancing capacity of the collateral pool, which matters for the internal stress test. The effect is small today because banks borrow little from the Eurosystem and pledge few corporate bonds. The ECB plans to extend it to corporate credit claims at the earliest by end-2027, which will matter more for banks that pre-position loans.
Standards and market guidance
Regulation and supervision
Central bank
Market data and research
Academic