The equity derivatives business of Meridian Bank sits in Global Markets, next to rates, FX, credit and commodities. It makes prices on anything whose value depends on shares or equity indices: listed futures and options, over-the-counter options and variance swaps, total return swaps, and the structured notes and certificates that private banks and insurers sell to their clients. In the eurozone the French banks built this business in the 1990s and 2000s and still lead it; the French savings market, where structured products are held in life-insurance contracts and securities accounts, is its largest outlet in Europe.
The clients fall into three groups. Distributors (private banks, retail networks, independent financial advisers and life insurers) buy structured notes to place with their own clients; they want a payoff they can explain, a coupon that stands out and a story. Institutional investors (asset managers, pension funds, hedge funds) trade index futures, options, total return swaps and quantitative investment strategies, increasingly with an ESG or climate constraint in their mandate. Corporates use equity derivatives occasionally, for share buy-backs, employee plans or stakes.
How the desk earns money. On flow products (futures, listed options, delta-one swaps) the bank earns a bid-offer spread and financing spread measured in basis points of notional, on very large volumes. On structured products it earns a structuring margin embedded in the issue price, typically one to two points of notional, and keeps the residual risks of the hedge on its book. The note issued to Hauteville in our case leaves Meridian about EUR 1.2m of margin on EUR 100m, while the private bank receives EUR 4m of distribution fees. The structuring margin is booked when the trade is done, subject to the valuation controls of section 6, and is earned against risks that last up to ten years.
WHO DOES WHAT — The structurer designs the payoff and the underlying with the salesperson, runs the pricer and writes the term sheet. The exotic trader prices the risk that cannot be bought in the market (volatility skew, correlation, dividends, gap risk) and runs the hedge. The delta-one trader buys and sells the index exposure through futures, swaps or baskets of shares, and finances them. The index engineering and QIS team writes index rules and systematic strategies and deals with the external index administrator. Distributor sales owns the relationship with Hauteville’s product selection team. Product governance and compliance approve the product, define its target market and check the marketing material. Product control checks the model and the day-one profit; market risk sets the desk’s sensitivity limits.
Adjacent functions: Treasury charges and credits the funding raised by notes (see E04); the debt issuance team runs Meridian’s base prospectus and its issuance programme; the sustainable finance team checks every sustainability claim against the bank’s frameworks (see A04, E10); the legal team signs off the final terms and the PRIIPs key information document.
A realistic week in mid-November 2026. Monday: the structurer reprices Hauteville’s draft term sheet after long-dated implied volatility moved. Tuesday: the index team receives the external administrator’s quarterly rebalancing file for the Paris-aligned index; two names leave the index after a fossil-fuel revenue update. Wednesday: the product governance committee reviews the target market of the note, including the client’s sustainability preferences. Thursday: the delta-one desk rolls its December futures hedge into March. Friday: product control questions the volatility skew used for a bespoke index, which has no listed options, and asks the desk to justify the day-one profit.
IN ONE SENTENCE — An equity derivative is a contract whose cash flows depend on the future price of a share or an index, and a structured product is a bond whose redemption is one of these contracts.
Delta-one products have a value that moves one for one with the underlying: index futures, total return swaps, exchange-traded funds, certificates that track an index. They transfer exposure without optionality. The price is a forward, and the debate between buyer and seller is about financing and dividends.
Options give a right, not an obligation. A call pays the rise above a strike, a put the fall below it. Their value depends on volatility, which is quoted, traded and hedged as a separate risk.
Exotic options add path dependence: barriers that activate or cancel the option if the index touches a level, digital payments, early redemption on observation dates, worst-of baskets. They cannot be bought in the listed market and stay on the dealer’s book.
Structured products wrap a package of options and a zero-coupon bond into a security issued by a bank. The investor lends money to the issuer and receives a payoff linked to the underlying. The investor therefore bears two risks: the market risk of the formula and the credit risk of the issuer.
The price of an equity future or a total return swap rests on a no-arbitrage argument. A dealer who sells an index forward for delivery at can buy the shares today, borrow the cash at the rate , and receive the dividends until delivery.
FORMULA
is the forward price for maturity ; the index level today; the continuously compounded funding rate (2.50% in this course); the expected dividend yield of the index; the maturity in years. A higher expected dividend lowers the forward.
INTUITION — Holding the shares pays the dividends; holding the forward does not. The forward buyer must be compensated by a lower price, exactly as a buyer of a flat with a tenant pays more than a buyer of an empty one who will have to wait for rent.
The difficulty is that is not known. Dividends are decided each year by boards. For a ten-year product, the dealer’s pricing depends on ten years of dividends that no one can predict. For the main indices, dividend futures exist (Eurex lists them on the EURO STOXX 50), but they are thin beyond five years, and for a bespoke index there is nothing to hedge with.
A decrement index (also called a synthetic dividend index) starts from a net total return index, in which dividends are reinvested, and deducts a fixed amount each day: either a percentage (5% a year is common) or a fixed number of index points (50 points a year is the other common convention).
FORMULA
is the decrement index level on day ; the net total return index; the decrement (5% a year); the number of calendar days since the previous level. Over a year the index behaves like the total return index less .
Because the total return index already contains the dividends, the forward of a decrement index no longer depends on dividends at all:
where is the contractual decrement. Dividend risk has been transferred from the dealer to the investor: if companies pay less than 5%, the decrement index falls behind the price index; if they pay more, it rises above it.
INTERACTIVE LAB — Price index or decrement index: the forward and the put
Compare the forward and an at-the-money put on a price index with an uncertain dividend and on a decrement index with a fixed one.
| Parameter you can move | Range | Default |
|---|---|---|
| Funding rate | 0 to 5 % | 2.5 % |
| Expected dividend yield of the price index | 0 to 6 % | 3.0 % |
| Decrement | 0 to 7 % | 5 % |
| Maturity | 1 to 10 years | 10 years |
| Volatility | 10 to 35 % | 18 % |
| Result at default settings | Value |
|---|---|
| Forward of the price index (spot 100) | 95.12 points |
| Forward of the decrement index (spot 100) | 77.88 points |
| At-the-money put on the price index | 18.99 % of spot |
| At-the-money put on the decrement index | 25.56 % of spot |
Try it. Set the decrement equal to the expected dividend yield: the two forwards coincide. The difference between the products is then only who bears the error on dividends.
At the default values the ten-year forward of the decrement index is 77.88, against 95.12 for the price index with a 3% dividend, and a ten-year at-the-money put is worth 25.56% of spot instead of 18.99%. A lower forward makes the puts that the investor sells in a structured product more valuable, which pays for a higher coupon. This is the first reason structurers like decrement indices; the second is that the dealer no longer carries dividend risk it cannot hedge; the third is that the index can be built on any basket, including a bespoke ESG one, without a dividend market.
PITFALL — A 5% decrement does not “cost” the investor 5% a year. It costs the difference between 5% and the dividends that the companies actually pay. On an index yielding 3%, the drag is about 2% a year against the price index that the client reads in the newspaper. Section 5.3 quantifies it for the case.
The autocallable (or “autocall”) is the most sold structured product in France. Its family name in French marketing is often Athena when coupons accumulate, or Phoenix when they are paid periodically. The version used in this module works as follows.
Decomposed, the investor is long a zero-coupon bond issued by Meridian, long a strip of digital options that pay the coupon if the index is above 100% on a date and the note has not been called before, and short a down-and-in put that becomes a normal put if the index ends below 60%. The investor sells protection against a crash and receives coupons in exchange.
INTERACTIVE LAB — What the investor receives: call year, redemption and annual return
Read the investor’s redemption and annual return for each possible call date, and for the scenario in which the note is never called.
| Parameter you can move | Range | Default |
|---|---|---|
| Annual coupon (snowball) | 4 to 16 % | 12.5 % |
| First observation date with index at or above 100% | 1 / 2 / 3 / 4 / 5 / 6 / 7 / 8 / 9 / 10 / never | 3 |
| Final index level if never called | 20 to 99 % of initial | 55 % of initial |
| Protection barrier at maturity | 40 to 80 % of initial | 60 % of initial |
| Result at default settings | Value |
|---|---|
| Redemption amount | 137.50 % of notional |
| Years held | 3 years |
| Investor’s annual return | 11.20 % |
Try it. Compare a call in year 1 (12.5% a year) with a call in year 10 (about 8.5% a year): the snowball coupon is simple interest, so its annual return falls with time.
WHY IT MATTERS ON THE DESK — The investor’s return depends on when the note is called, not only on whether the index rises. A salesperson who says “12.5% a year” without adding “if called in year 1” overstates the return; the PRIIPs key information document shows scenarios for this reason.
The issuer prices the note in three layers.
The hedge is dynamic. Meridian is short the note, so it buys the index in the quantity of the note’s delta (sensitivity to a 1% move, expressed in EUR of index), adjusts it daily, and manages vega (sensitivity to volatility) through listed or OTC options where they exist. The residual risks that cannot be hedged precisely (skew, gap risk near the barrier at maturity, correlation if several underlyings, dividends on a price index) are charged in the price and held in reserves.
The note is issued under Meridian’s base prospectus approved under the Prospectus Regulation (EU) 2017/1129; each issue has its own final terms. Because it is sold to retail clients, a PRIIPs key information document (Regulation (EU) No 1286/2014) is required: three pages with the summary risk indicator, performance scenarios and costs. The distributor and Meridian sign a distribution agreement that sets fees and the target market. The index is described in the prospectus and in an index rulebook published by the administrator, which is authorised or registered under the Benchmarks Regulation (EU) 2016/1011. The hedge is documented under ISDA (for OTC hedges) and exchange rules (for futures).
A QIS is a rules-based strategy (momentum, carry, low volatility, a systematic option-selling programme) wrapped in an index calculated by the bank or an external administrator, and delivered to institutional clients through a total return swap. The investor pays a fee embedded in the index and gets exposure to the strategy without running it. ESG versions apply exclusions or carbon constraints to the strategy’s universe. The index engineering team that builds the case index of this module is the same team that runs QIS.
IN ONE SENTENCE — In equity derivatives, the sustainable layer sits in the underlying (an ESG or climate index), in the use of the issuer’s proceeds (a green note), or in the investor’s mandate, and each makes a different claim.
ESG underlyings. The payoff is linked to an index that excludes, underweights or overweights companies on ESG criteria. The regulated categories in the EU are the EU Climate Transition Benchmark (CTB) and the EU Paris-aligned Benchmark (PAB); the rest is a wide family of “ESG”, “low carbon”, “SRI” and “ESG screened” indices whose methodology is disclosed but not standardised. The investor’s money does not go to the index companies: it is lent to the issuer, which hedges by buying index shares, but the shares would be bought by someone in any case.
Green use-of-proceeds notes. The issuer commits to allocate an amount equal to the note’s proceeds to eligible green assets under its green bond framework (see C01). The payoff can be linked to any underlying, ESG or not. The claim is about what the bank finances, not about the index.
Investor-side constraints. Many institutional clients run under PAB or CTB exclusions because of their own rules: a fund whose name includes “sustainable” or “environmental” must apply PAB exclusions to its investments under ESMA’s guidelines on fund names. For them, a total return swap or a future on a PAB index is a tool to manage exposure within the mandate. Listed ESG futures and options serve the same purpose.
IN ONE SENTENCE — A Paris-aligned Benchmark is an index whose carbon intensity is at least 50% below its investable universe at launch and falls by at least 7% a year thereafter, after excluding fossil-fuel-heavy companies.
The two labels were created by Regulation (EU) 2019/2089, which amended the Benchmarks Regulation, and their minimum standards are set by Commission Delegated Regulation (EU) 2020/1818. The core rules are the following.
| Requirement | CTB | PAB |
|---|---|---|
| Intensity reduction vs universe | at least 30% | at least 50% |
| Year-on-year self-decarbonisation | 7% on average | 7% on average |
| High-climate-impact sectors | at least universe weight | at least universe weight |
| Base exclusions | weapons, tobacco, norm violators | same |
| Fossil exclusions | none | coal 1%, oil 10%, gas 50%, power above 100 g/kWh 50% |
The fossil exclusions of the PAB apply to revenue shares: companies deriving 1% or more of revenues from hard coal and lignite, 10% or more from oil fuels, 50% or more from gaseous fuels, or 50% or more from electricity generation with an intensity above 100 gCO2e/kWh are excluded. The base exclusions (controversial weapons, cultivation and production of tobacco, violators of the UN Global Compact principles or the OECD Guidelines for Multinational Enterprises) apply to both labels.
The intensity measure is greenhouse gas emissions divided by enterprise value including cash (EVIC), scopes 1 and 2 plus scope 3 phased in by sector: energy and mining first, then transport, construction, buildings, materials and industrials, then all other sectors by the end of 2027. The high-climate-impact sector constraint, defined by NACE sections A to H and L (agriculture, mining, manufacturing, energy, water, construction, trade, transport, real estate), prevents the index from reaching its target simply by moving into banks and software.
INTUITION — A climate benchmark is a diet with a scale and a menu. The scale says the weight must be half the universe’s and must fall by 7% a year. The menu says certain foods are banned. The sector constraint says you may not reach the weight by eating only salad: you must keep eating proteins, but lighter ones.
INTERACTIVE LAB — The 7% trajectory: how far must a PAB index move from its universe?
See how the required gap between a PAB index and its universe widens when the real economy decarbonises more slowly than 7% a year.
| Parameter you can move | Range | Default |
|---|---|---|
| Reduction vs universe at launch | 30 to 60 % | 50 % |
| Universe intensity decline per year | 0 to 10 % | 3 % |
| Universe intensity at launch | 50 to 300 tCO2e per EUR m EVIC | 164 tCO2e per EUR m EVIC |
| Result at default settings | Value |
|---|---|
| Required reduction vs universe after 5 years | 59.5 % |
| Required reduction vs universe after 10 years | 67.2 % |
| Index intensity ceiling after 10 years | 39.7 tCO2e per EUR m EVIC |
Try it. Set the universe decline to 7%: the gap stays at 50%. At 0% it reaches about 76% after ten years: the index must keep moving away from the economy it is meant to represent.
The chart shows the main structural critique of the PAB design. The 7% trajectory is absolute: if the companies in the universe decarbonise at 3% a year, the index must deviate more and more from its universe, through higher concentration and tracking error, rather than through the companies changing. Index providers answer with “transition” variants that reward company targets, but the regulation measures the index’s intensity, not the companies’ progress.
PITFALL — Intensities are measured per euro of EVIC. When share prices rise, EVIC rises and intensity falls without any tonne being avoided. The delegated regulation requires an adjustment for enterprise value inflation in the trajectory. A desk that reads a drop in index intensity in a bull market as decarbonisation is reading the market, not the climate.
Regulation (EU) 2025/914 amended the Benchmarks Regulation and has applied since 1 January 2026. It removes most non-significant benchmarks from the scope of the regulation (a benchmark becomes significant above EUR 50bn of use), but keeps EU climate benchmarks in scope whatever their size: an index can be marketed as a CTB or PAB in the EU only if its administrator is on ESMA’s register, and law-firm summaries indicate that the names of these benchmarks must carry the label. Administrators that remain registered must still publish ESG disclosures for their benchmarks. For Meridian this means that the case index cannot be run informally by the trading desk: it needs an authorised administrator, a rulebook, an oversight function and a published methodology.
Sustainability preferences. Since 2 August 2022 (Delegated Regulation (EU) 2021/1253), an adviser must ask clients whether they want sustainable investments, and how much: a minimum proportion of Taxonomy-aligned investments, a minimum proportion of sustainable investments in the SFDR sense, or the consideration of principal adverse impacts (PAI). A product can only be recommended as meeting those preferences if it can show the corresponding data. For a structured note this is awkward: the investor holds a debt claim on Meridian, not shares of the index companies.
Product governance. Since 22 November 2022 (Delegated Directive (EU) 2021/1269), manufacturers and distributors must specify any sustainability-related objectives in the target market. ESMA’s 2023 guidelines on product governance clarify how. Meridian, as manufacturer, defines the target market; Hauteville, as distributor, refines it for its clients.
PRIIPs. The key information document has no sustainability section; ESG information goes in the “other relevant information” section and in the prospectus. The French regulators’ joint unit (AMF and ACPR) published in April 2025 a mapping of the French structured product market for 2021 to 2023, which noted that the share of products with ESG criteria had fallen markedly over the period, and in June 2026 an explanatory note warning that decrement indices transfer dividend risk to the investor and complicate comparison with standard indices.
Retail Investment Strategy. The political agreement of December 2025 adds a value-for-money test for PRIIPs, including structured notes. The Commission’s call for advice to ESMA of July 2026 states that final votes are expected in the fourth quarter of 2026, publication by January 2027 and application from July 2029.
SFDR. The SFDR does not apply to structured notes issued by a bank as such, but it applies to funds and insurance-based products that hold them: an Article 8 unit-linked fund of a life insurer buying the note will look at its underlying. The SFDR revision proposed by the Commission in November 2025 (new “transition”, “ESG basics” and “sustainable” categories with a 70% threshold) entered trilogue in October 2026, had produced no agreement by mid-November 2026 and is expected to apply around 2029.
Eurex, the main European derivatives exchange, lists futures and options on ESG versions of the STOXX, DAX and MSCI indices. Its own statistics give an order of magnitude: in the first quarter of 2025 its ESG index derivatives (equity and fixed income) traded 900,611 contracts, EUR 30.9bn of notional, with open interest peaking at EUR 6.8bn in February 2025. This is small next to the flagship EURO STOXX 50 contracts, and the liquidity is concentrated in a handful of products. For a bespoke PAB index there is no listed contract at all: the delta-one desk hedges with a basket of the constituents, a total return swap with another dealer, or a liquid proxy (a standard index future) and accepts the basis risk.
Four observations frame the case. First, ESG labels on structured products have become less common in France since 2022, as the AMF-ACPR mapping shows for 2021 to 2023, while distributors became more cautious about sustainability claims. Second, when ESG features remain, the underlying is usually an ESG-screened or climate version of a standard index, often with a decrement. Third, the regulated PAB and CTB labels are more common in institutional products (index funds, total return swaps) than in retail notes, because the labels fit the fund-name rules. Fourth, green use-of-proceeds notes are issued by several European banks under their green bond frameworks; the payoff is unchanged and the claim depends on the bank’s eligible asset pool.
A structured note for a distributor goes from first idea to issue in six to ten weeks. A bespoke index adds a parallel workstream with the index administrator, which usually sets the timetable.
INTERACTIVE LAB — From the private bank’s request to redemption
Follow the roles, documents and controls of a bespoke-index autocallable note, from the request to redemption.
| # | Who | Does what | Deliverable |
|---|---|---|---|
| 1 | Hauteville product selection team and Meridian distributor sales | Agree the client need: a ten-year climate-themed autocall for advised clients, EUR 100m, coupon above 10% | Request for proposal and indicative constraints |
| 2 | Equity structurer and index engineering team | Design the payoff and the Paris-aligned decrement index; run indicative pricing | Indicative term sheet and draft index methodology |
| 3 | Sustainable finance team | Check the index against Delegated Regulation 2020/1818 and the note against the bank’s sustainability claims policy | Sustainability review memo |
| 4 | External index administrator and its oversight committee | Accept the index, review the rulebook, register it under the Benchmarks Regulation | Index rulebook, benchmark statement and ESG disclosures |
| 5 | New product approval committee | Approve the product, the model and the target market, including sustainability-related objectives | Approval minutes and target market definition |
| 6 | Exotic trader, model validation and product control | Validate the pricing model, the volatility and skew inputs and the valuation reserves | Pricing and day-one profit sign-off |
| 7 | Legal, debt issuance team and Hauteville | Draft final terms, PRIIPs key information document and distribution agreement | Final terms, KID and signed distribution agreement |
| 8 | Hauteville advisers with compliance review | Market the note to clients after suitability and sustainability-preference checks | Approved marketing material and subscription orders |
| 9 | Exotic and delta-one traders | Fix the initial index level on the strike date and buy the delta hedge | Trade booking, hedge execution and risk report |
| 10 | Issuing and paying agent, Treasury | Settle the note against cash and credit the funding to the desk at FTP | Issued note and funding ticket |
| 11 | Index administrator, traders and product control | Calculate the index daily, rebalance quarterly, revalue the note and quote buy-back prices | Daily index levels, valuations and secondary prices |
| 12 | Calculation agent and product governance | Run the annual observation, pay or roll the note, review the target market and ESG data each year | Observation notice, redemption or annual review |
Request and design (weeks 1 to 2). Hauteville’s product selection team sends a request to three issuers with a brief: maturity, minimum coupon, capital at risk tolerance, theme. The distributor salesperson at Meridian translates it for the structurer, who runs indicative prices on several underlyings. Because the request is for a climate theme, the structurer brings in the index engineering team to propose a bespoke PAB index rather than an off-the-shelf ESG index. The indicative term sheet is labelled as such and is not a commitment.
Sustainability review (weeks 2 to 3). The sustainable finance team checks two things. First, that the index meets the PAB minimum standards, using the same data that the administrator will use. Second, that the marketing language matches the product: a note on a PAB index may say that the underlying meets the EU PAB standard; it may not say that the investor finances the transition, because the money is lent to Meridian. The review memo is attached to the approval file.
Index set-up (weeks 2 to 6). Meridian could be the administrator of its own indices, but that requires authorisation under the Benchmarks Regulation, an oversight function and conflict-of-interest controls, because the bank is also the hedger and the issuer. Many banks therefore use an external administrator for bespoke indices. The administrator’s oversight committee approves the rulebook; the benchmark statement and the ESG disclosures are published. The administrator, not Meridian, decides on exclusions when a company’s data change.
Approval (week 4). The new product approval committee (also called product governance committee) brings together trading, structuring, risk, compliance, legal, finance, operations and the sustainable finance team. It approves the payoff, the model, the hedging plan, the target market (knowledge and experience, ability to bear losses, risk tolerance, objectives including sustainability-related objectives, distribution strategy) and the negative target market.
Pricing control (week 4 to 5). Model validation has approved the Monte Carlo local-volatility model for decrement indices. Product control checks the inputs: the volatility surface of a bespoke index has no listed options, so it is derived from the parent index with an adjustment, which makes it an unobservable input and drives the accounting of section 6.4.
Documentation (weeks 4 to 6). The final terms under the base prospectus, the PRIIPs key information document (prepared by Meridian as manufacturer), and the distribution agreement (fees, target market, information flows, ESG data to be provided to Hauteville each year) are drafted in parallel.
Marketing (weeks 6 to 10). Hauteville’s advisers present the product to clients after a suitability assessment that includes their sustainability preferences. Meridian’s compliance reviews any marketing material bearing its name. Subscriptions are collected; the final size may differ from EUR 100m.
Strike and issue (week 10). On the strike date, the index closing level becomes the initial level. The traders buy the delta hedge, partly before the close to avoid moving the market. Settlement follows a few days later through the clearing systems; Treasury credits the funding to the desk.
Life of the note. The administrator calculates the index daily and rebalances it quarterly. The desk revalues the note daily; Meridian, as market maker, quotes a buy-back price to Hauteville’s clients who wish to sell before maturity, with a bid-offer spread that the distribution agreement caps at one point in normal markets (case assumption). Each year the calculation agent checks the observation level; the product governance team reviews whether the note still fits its target market, including if the index loses its label.
Question. What exactly does the private bank ask for, and what constraints does that set?
Data (case assumptions). Hauteville Private Bank is a Paris-based private bank with about 6,000 advised clients. Its product selection team wants a ten-year autocallable note of EUR 100m, to be placed in November and December 2026 with clients who hold securities accounts and life-insurance contracts. The brief contains five points.
Meridian’s issuer rating is A, its funding spread for ten-year senior preferred debt is 70 bp over swaps, and the swap curve is flat at 2.50%. The desk targets an issuer margin of 1.0% of notional.
Reading. Point 1 sets the index work, points 2 to 4 set the pricing, and point 5 sets the product governance problem. The desk already knows that a coupon above 10% on a ten-year autocall is hard to reach on a standard price index with today’s rates and volatility: hence the decrement, which stage 5.3 justifies.
Question. Starting from Meridian’s eurozone large-cap universe, what does it take to build a PAB index, and what does it cost in tracking error?
Data. For teaching, the universe is grouped into eleven sector blocks. Weights, intensities (scopes 1, 2 and phased-in scope 3, per EUR m of EVIC), dividend yields and Taxonomy-aligned revenue shares are case assumptions; flags follow Delegated Regulation 2020/1818.
| Block | Weight | Intensity | Flag |
|---|---|---|---|
| Oil and gas | 6% | 520 | PAB exclusion (oil) |
| Fossil-heavy utility | 5% | 780 | PAB exclusion (power) |
| Renewable-led utility | 4% | 140 | high impact |
| Cement | 5% | 1,100 | high impact |
| Autos | 8% | 180 | high impact |
| Industrials | 12% | 60 | high impact |
| Banks and insurers | 15% | 6 | low impact |
| Software and telecom | 14% | 15 | low impact |
| Consumer and luxury | 16% | 30 | high impact |
| Pharma | 10% | 20 | high impact |
| Tobacco | 5% | 40 | CTB and PAB exclusion |
Dividend yields in the same order: 5.5%, 5.0%, 4.0%, 3.0%, 4.5%, 2.5%, 6.0%, 1.5%, 2.0%, 2.5%, 6.5%. Taxonomy-aligned revenue shares: 1%, 8%, 75%, 5%, 25%, 20%, 0%, 1%, 0%, 0%, 0% (banks report a Green Asset Ratio, not aligned revenue, and are set at zero here). Risk model: one-factor, market volatility 17%, betas from 0.6 (tobacco) to 1.3 (autos) and specific volatilities from 9% to 18%.
Calculation, step 1: the universe. The weighted average carbon intensity (WACI) is the weight-averaged intensity:
FORMULA
is the weight of block in the portfolio ; its intensity in tCO2e per EUR m of EVIC. For the universe, .
The PAB target is half that: 82.1.
Step 2: exclusions. Oil and gas, the fossil-heavy utility and tobacco leave the index: 16% of the universe. If the remaining weights were simply scaled up, the high-climate-impact share would fall from 71% to 65.5% (55% out of 84%), which the PAB rules forbid. Meridian therefore rescales the high-impact group back to 71% and the low-impact group to 29%. Result: an intensity of 117.9, a reduction of only 28.2%. Exclusions alone do not make a PAB, because the sector constraint forces the index to keep cement, autos and the renewable utility.
Step 3: the tilt. Within each group, weights are tilted away from carbon-intensive blocks:
FORMULA
is the universe weight of block ; the tilt exponent; the group (high or low climate impact) to which belongs; the group’s total weight (71% or 29%). With weights are proportional to the universe; a higher moves weight towards low intensities. The index team solves for the smallest that reaches the target.
INTUITION — The tilt is a dimmer, not a switch. Turning it slightly reduces every heavy emitter a little instead of removing one name entirely, which keeps the index close to its universe.
INTERACTIVE LAB — Building the Paris-aligned index from the universe
Apply the label’s exclusions, the high-climate-impact-sector constraint and a carbon tilt, and read the index intensity, tracking error, dividend yield and Taxonomy alignment.
| Parameter you can move | Range | Default |
|---|---|---|
| Label | PAB / CTB | PAB |
| Required intensity reduction vs universe | 20 to 70 % | 50 % |
| Keep high-climate-impact sectors at universe weight | yes / no | yes |
| Market volatility (one-factor model) | 10 to 30 % | 17 % |
| Result at default settings | Value |
|---|---|
| Universe intensity | 164.2 tCO2e per EUR m EVIC |
| Intensity after exclusions and sector rescaling only | 117.9 tCO2e per EUR m EVIC |
| Reduction from exclusions and rescaling only | 28.2 % |
| Carbon tilt exponent needed | 0.204 |
| Index intensity | 82.1 tCO2e per EUR m EVIC |
| Tracking error vs universe | 1.88 % a year |
| Weight in high-climate-impact sectors | 71.0 % |
| Dividend yield of the universe | 3.56 % |
| Dividend yield of the index | 3.07 % |
| Taxonomy-aligned revenue, universe | 8.25 % |
| Taxonomy-aligned revenue, index | 8.66 % |
| Weight of cement | 3.54 % |
| Weight of consumer and luxury | 23.59 % |
| Weight of industrials | 15.36 % |
| Weight of banks and insurers | 16.34 % |
| High-impact weight if exclusions were simply rescaled | 65.5 % |
Try it. Switch the sector constraint off: the index reaches 50% with less tilt by moving into banks and software, and its high-climate-impact weight falls to about 60%, which the PAB rules forbid.
Result. A tilt of brings the intensity to 82.1, exactly 50% below the universe. Cement falls from 5% to 3.54%; consumer and luxury rises from 16% to 23.59%; industrials from 12% to 15.36%; banks from 15% to 16.34%. The tracking error against the universe is 1.88% a year. The index’s dividend yield is 3.07%, against 3.56% for the universe, because the high-yield oil, utility and tobacco blocks are gone. Its Taxonomy-aligned revenue share is 8.66%, against 8.25% for the universe.
Reading. Three lessons. First, most of the tracking error comes from the exclusions, not the tilt: at a 30% target the tracking error is about the same. Second, the sector constraint matters: without it the index would reach 50% with less tilt by drifting into banks and software, and its high-impact share would fall to 60.2%. Third, a PAB index is not a green index. Its Taxonomy-aligned share is barely higher than the universe’s: the PAB rules reduce carbon intensity per euro of enterprise value; they do not select companies that earn green revenues. The textbook approach, which minimises tracking error under a carbon constraint (Roncalli, chapter 11, see section 11), gives a lower tracking error than this tilt rule for the same reduction, but behaves the same way.
PITFALL — The 1.88% tracking error is measured against the universe with a one-factor model. Against the EURO STOXX 50, which a trader may use as a proxy hedge, the tracking error is higher because the universes differ. Use the right benchmark for the right question.
Question. Should the note be linked to the PAB index in net total return with a 5% decrement, or to its price version?
Data. The index’s expected dividend yield is 3.07% (stage 5.2). There are no listed dividend futures on a bespoke PAB index, and no listed options.
Calculation. The decrement index grows like the total return index less 5%; the price index grows like the total return index less the dividends actually paid. The gap between them is therefore each year:
FORMULA
is the decrement index and the price index at (both starting at 100); the decrement (5%); the dividend yield actually paid (3.07%); the horizon. With and , the ratio is 0.8245.
INTERACTIVE LAB — The cost of a 5% decrement on an index yielding 3.07%
Measure how far the decrement index falls behind the price index, and what the barrier and the autocall trigger mean in price-index terms.
| Parameter you can move | Range | Default |
|---|---|---|
| Dividend yield actually paid by the index | 0 to 7 % | 3.07 % |
| Decrement | 0 to 7 % | 5 % |
| Horizon | 1 to 10 years | 10 years |
| Protection barrier | 40 to 80 % of initial | 60 % of initial |
| Annual price return of the shares | -5 to 8 % | 0 % |
| Result at default settings | Value |
|---|---|
| Annual drag vs price index | 1.93 points |
| Cumulative underperformance vs price index | 17.55 % |
| Barrier expressed as a price-index level at the horizon | 72.77 % of initial |
| Price-index level needed for a call in year 1 | 101.95 % of initial |
| Decrement index at the horizon | 82.45 % of initial |
Try it. Set the dividend yield to 5%: the decrement is neutral. At 2% the 60% barrier becomes a 81% barrier in price-index terms after ten years.
Result. The drag is 1.93 points a year. Over ten years, with flat share prices, the decrement index ends at 82.45 while the price index is still at 100: an underperformance of 17.55%. The 60% barrier on the decrement index corresponds to a price-index level of 72.77% at maturity. To be called in year 1, the price index must be at 101.95, not 100.
Reading. For Meridian, the decrement removes a risk it cannot hedge: there is no dividend market on this index, and the dealer who sells autocalls on price indices is structurally exposed to falling dividends (stage 5.5 quantifies it). For the investor, the decrement turns uncertain dividends into a known deduction and finances a higher coupon (stage 5.4). The trade-off is acceptable only if it is explained in comparable terms. Hauteville’s advisers must be able to say: “the barrier is 60% of an index that loses about 2% a year against the market index you know; in market terms, it is close to 73% after ten years”. The French supervisors’ June 2026 note asks exactly for that kind of explanation.
WHY IT MATTERS ON THE DESK — Decrement indices are a large part of the French market, and the PAB version makes the drag worse: excluding oil, utilities and tobacco lowers the dividend yield (3.07% instead of 3.56%), so the same 5% decrement costs about half a point more each year. Sustainable underlyings and decrements interact, and the structurer must show it.
Question. What coupon can Meridian pay, given a 4.0% distribution fee and a 1.0% target margin?
Data. Swap rate 2.50%; Meridian funding spread 0.70%; decrement 5%; at-the-money volatility 18% for the decrement index (derived from the parent index, case assumption); skew modelled with a local volatility that rises as the index falls, with elasticity , capped at 50%; quarterly time steps; 20,000 simulated paths (10,000 antithetic pairs) with a fixed seed.
Calculation. The note’s value is the sum of two legs, both computed on the same simulated paths:
FORMULA
is the fair value of the note (in % of notional) for an annual coupon (in %); the present value of the redemption of 100 or of the index performance; the present value of 1% of coupon, which is paid times if the note is called in year ; the distributor fee (4.0); the issuer margin (1.0); the fair coupon. Cash flows are discounted at = 3.20%, the issuer’s funding curve.
Because the coupon enters linearly, one simulation is enough to solve for it: no iteration is needed.
INTERACTIVE LAB — Pricing the ten-year autocall on the decrement index
Value the note by Monte Carlo with a skewed volatility, solve for the fair coupon and split the issue price between fair value, distributor and issuer.
| Parameter you can move | Range | Default |
|---|---|---|
| Decrement | 0 to 7 % | 5 % |
| At-the-money volatility | 12 to 30 % | 18 % |
| Skew elasticity (volatility rises as the index falls) | 0 to 2 | 1 |
| Swap rate | 0 to 5 % | 2.5 % |
| Meridian funding spread | 0 to 2 % | 0.7 % |
| Protection barrier | 40 to 80 % of initial | 60 % of initial |
| Distributor fee (upfront) | 0 to 6 % of notional | 4 % of notional |
| Issuer margin targeted | 0 to 3 % of notional | 1 % of notional |
| Coupon offered | 5 to 16 % a year | 12.5 % a year |
| Expected dividend of the price index (for comparison) | 0 to 6 % | 3.07 % |
| Result at default settings | Value |
|---|---|
| Value of the redemption leg | 76.91 % of notional |
| Value of 1% of coupon | 1.4336 % of notional |
| Fair coupon at the targeted fees | 12.62 % a year |
| Fair coupon on the price index instead | 8.97 % a year |
| Fair coupon with no decrement (total return index) | 5.00 % a year |
| Fair value of the note at the offered coupon | 94.83 % of notional |
| Issuer margin at the offered coupon | 1.17 % of notional |
| Probability of call in year 1 (risk-neutral) | 43.75 % |
| Expected life | 4.21 years |
| Probability of capital loss at maturity (risk-neutral) | 19.93 % |
| Average loss when there is a loss | 76.8 % of notional |
Try it. Set the skew elasticity to 0 (flat volatility): the fair coupon jumps above 15%. Flat-volatility pricing overstates what an autocall can pay, because it underprices the digital calls the investor receives.
Result. The redemption leg is worth = 76.91% and 1% of coupon is worth = 1.4336%. The fair coupon is = (95 − 76.91) / 1.4336 = 12.62%. The structurer proposes a round 12.50%; the fair value of the note at that coupon is 94.83%, so the issue price of 100 splits into 94.83 of fair value, 4.00 for Hauteville and 1.17 for Meridian, EUR 1.17m on EUR 100m. Under the pricing measure, the probability of a call in year 1 is 43.75%, the expected life is 4.21 years, and the probability of a capital loss at maturity is 19.94%, with an average loss of 76.8% when it happens.
Reading. Four sensitivities explain the coupon.
PITFALL — A structurer who quotes a coupon from a flat-volatility model will win the mandate and lose money on the hedge. Product control’s check of the skew input is not bureaucracy: in this case it is worth more than three points of coupon.
Question. What does Meridian buy on the strike date, and what risk stays on its book?
Data. Notional EUR 100m; coupon 12.50%; same model as stage 5.4. Bumps: index ±1%, volatility +1 point, decrement −1 point. Tracking error of the PAB index against a proxy hedge: 1.88% a year (stage 5.2, used as a lower bound).
Calculation. Each sensitivity is a difference of two Monte Carlo values computed with the same random numbers, which removes most of the simulation noise:
is the delta in EUR of index per unit of index move, here expressed in EUR m; the note value in EUR m; the initial index level; the vega for one volatility point.
INTERACTIVE LAB — The dealer’s risks after the trade: delta, vega, dividends and basis
Measure what Meridian must buy to hedge the note and what remains on its book.
| Parameter you can move | Range | Default |
|---|---|---|
| Notional | 10 to 300 EUR m | 100 EUR m |
| Coupon | 5 to 16 % a year | 12.5 % a year |
| Decrement | 1 to 7 % | 5 % |
| At-the-money volatility | 12 to 30 % | 18 % |
| Skew elasticity | 0 to 2 | 1 |
| Swap rate | 0 to 5 % | 2.5 % |
| Meridian funding spread | 0 to 2 % | 0.7 % |
| Protection barrier | 40 to 80 % of initial | 60 % of initial |
| Tracking error of the index vs a proxy hedge | 0 to 5 % a year | 1.88 % a year |
| Result at default settings | Value |
|---|---|
| Fair value of the note | 94.83 EUR m |
| Index exposure to buy (delta) | 77.5 EUR m |
| Change in note value for +1 volatility point | -320 EUR k |
| Note value gain if dividends were 1 point lower (price-index version) | 2.98 EUR m |
| One-year standard deviation of a proxy hedge | 1.46 EUR m |
Try it. Lower the skew elasticity to 0: the delta changes, which is why the choice of volatility model is a risk in itself.
Result. On the strike date Meridian buys EUR 77.5m of index exposure (delta 77.5% of notional). The note loses EUR 320,000 of value for each additional point of volatility: Meridian, which is short the note, is therefore long vega by about EUR 320,000 per point. If the note had been written on the price index, a one-point fall in expected dividends would have raised the note’s value by EUR 2.98m, a loss for Meridian; on the decrement index this risk is with the investor. If the delta were hedged with a proxy future instead of the constituents, the one-year standard deviation of the hedging error would be at least EUR 1.46m (EUR 77.5m × 1.88%).
Reading. The delta-one desk executes the hedge. It has three choices. It can buy the constituents of the index in their weights, which is exact but requires trading around 50 to 100 real names at each quarterly rebalancing. It can enter a total return swap on the index with another dealer, which transfers the replication to that dealer at a financing spread. Or it can use liquid index futures and accept basis risk. With a EUR 1.46m standard deviation on a EUR 1.17m margin, the third choice is not acceptable for long: the desk uses futures for a few days and then moves to the constituent basket.
The vega position is the structural one. Every autocall sold by Meridian makes it longer volatility, because the investor has sold a put. Desks that sell large volumes of autocalls must sell volatility elsewhere, through listed options, variance swaps or other structured products. When many dealers do the same, long-dated implied volatility on European indices is pushed down: the structured product market shapes the volatility market. The decrement does not remove this; it only removes dividends from the list.
WHO DOES WHAT — Each morning the exotic trader receives the risk report of the autocall book: delta by underlying, vega by maturity and strike, the barrier exposure (“gap risk”) for notes whose index is close to 60% at maturity, and the P&L explained. For a bespoke index the report also shows the basis between the index and its hedge, and the expected trades of the next rebalancing, which the administrator announces a few days in advance.
Question. Hauteville also wants to offer the note to clients who asked for a minimum Taxonomy-aligned share of 20%. Does the PAB note qualify? What if Meridian issues it as a green note?
Data. Taxonomy-aligned revenue shares from stage 5.2: universe 8.25%, PAB index 8.66%. Meridian’s green bond framework (see C01) has an eligible green asset pool of EUR 7.2bn, of which 72% is Taxonomy-aligned (case assumption). Outstanding: EUR 5.5bn of green bonds and EUR 0.9bn of green structured notes before this issue.
Calculation. The ESG-underlying version is assessed by look-through: the investor’s exposure is the index, so its Taxonomy share is the index’s revenue share. The use-of-proceeds version is assessed by allocation: the proceeds of the note finance the pool, so its share is the pool’s aligned share times the share of proceeds allocated. The pool’s headroom is the difference between eligible assets and green instruments outstanding.
INTERACTIVE LAB — Which version meets a client’s Taxonomy preference?
Compare the Taxonomy-aligned share of the note by look-through to the index and by use of proceeds, test it against a client’s minimum, and check the headroom of Meridian’s green asset pool.
| Parameter you can move | Range | Default |
|---|---|---|
| Taxonomy-aligned revenue of the universe (stage 5.2) | 0 to 30 % | 8.25 % |
| Taxonomy-aligned revenue of the PAB index (stage 5.2) | 0 to 30 % | 8.66 % |
| Taxonomy-aligned share of Meridian’s eligible green pool | 0 to 100 % | 72 % |
| Share of proceeds allocated | 0 to 100 % | 100 % |
| Client’s minimum Taxonomy-aligned proportion | 0 to 50 % | 20 % |
| Meridian eligible green assets | 3 to 12 EUR bn | 7.2 EUR bn |
| Green bonds outstanding | 0 to 10 EUR bn | 5.5 EUR bn |
| Green structured notes outstanding before this one | 0 to 3 EUR bn | 0.9 EUR bn |
| This note | 0 to 500 EUR m | 100 EUR m |
| Result at default settings | Value |
|---|---|
| Taxonomy-aligned share, use-of-proceeds note | 72.0 % |
| Look-through gain of the PAB index over its universe | 0.41 points |
| ESG-underlying note meets the client minimum (1 = yes) | 0 |
| Use-of-proceeds note meets the client minimum (1 = yes) | 1 |
| Headroom in the green pool after this note | 0.70 EUR bn |
| Eligible assets / green instruments outstanding | 110.8 % |
Try it. Lower the client minimum to 5%: both versions qualify, which shows how much the answer depends on the threshold the client was led to choose.
Result. By look-through, the PAB note has a Taxonomy-aligned share of 8.66%, 0.41 points above the universe: it does not meet a 20% preference. A use-of-proceeds version has a share of 72% and qualifies. After this note, the green pool still covers outstanding green instruments at 110.8%, with EUR 0.70bn of headroom.
Reading. The two approaches answer different questions, and both are imperfect. Look-through measures what the payoff tracks, but the investor owns no share of these companies. Use of proceeds measures what Meridian finances, but the payoff is unchanged and the investor bears no green risk. How far a use-of-proceeds note can count towards a client’s Taxonomy preference depends on the data the manufacturer reports in the European ESG Template and on the distributor’s policy; practice differs between banks and supervisors have not settled it. Meridian’s approach, decided by its product governance committee, is to offer two lines of the same payoff: the PAB note for clients who asked for PAI consideration, and a green note for clients with a Taxonomy minimum, each with its own wording.
PITFALL — Combining both features (“a green note on a Paris-aligned index”) does not add up the claims. The marketing must still say, separately, what the index does and what the proceeds finance, and must not suggest that the client’s money is invested in the index companies.
Question. Who may buy this note, and what may be said about it?
Data and decisions. The product governance committee sets the following target market, which Hauteville will refine.
| Criterion | Positive target market | Negative target market |
|---|---|---|
| Client type | Retail and professional, advised | Execution-only retail |
| Knowledge | Informed: knows autocalls and barriers | No experience of capital-at-risk products |
| Ability to bear losses | Can lose all capital | Needs capital protection |
| Horizon | Up to 10 years | Needs liquidity within 2 years |
| Sustainability | Wants PAI on GHG considered | Asks for a Taxonomy minimum above 8% (PAB line) |
Reading. The KID shows a summary risk indicator, three performance scenarios plus a stress scenario, and costs: the entry cost includes the 4.00 distribution fee and Meridian’s 1.17 margin, which together explain why the fair value is 94.83. The sustainability wording is part of the approval. Allowed: “the note is indexed to an index that meets the minimum standards of the EU Paris-aligned Benchmark under Delegated Regulation (EU) 2020/1818”; “the index excludes companies above the PAB fossil-fuel revenue thresholds”. Not allowed: “your investment finances the energy transition”; “a green investment”; “carbon neutral”. The distribution agreement obliges Meridian to send Hauteville, each year, the index’s ESG disclosures and any change of label, and obliges Hauteville to report sales outside the target market.
WHO DOES WHAT — The compliance officer’s most common correction on ESG structured product brochures is the verb. “Invest in” suggests ownership of the index companies; “is indexed to” or “whose performance depends on” describes a debt claim on the issuer. One verb separates a correct brochure from a misleading one.
Question. What happens to the client, to Meridian’s book and to the label at the first observation date in December 2027?
Scenario A: the index is above its initial level. The note is called. Each client receives 112.5% of the amount invested, an annual return of 12.50% (section 2.4). Meridian pays EUR 112.5m, sells its delta hedge, and the remaining deferred margin is released to profit (section 6.4). The economic life was one year, not 4.21: the trade’s realised return on capital is far above the RAROC of the case. For Hauteville, a call is a commercial event: its advisers will propose a new note to the same clients. Supervisors watch this recycling, because each new note carries a new upfront fee; Hauteville’s product governance must show that the reinvestment serves the client, not the fee.
Scenario B: the index is 15% lower and a client asks to sell. The note is not called and continues. A client who needs liquidity asks Hauteville for a buy-back price. Meridian, as market maker, revalues the note with the same model at the new index level, a volatility that has probably risen, and its funding spread at that date, then applies its bid spread. The price will be well below 100: the client sells the put that has moved against him, and also pays back part of the fees that were embedded at issue. The distribution agreement caps the bid-offer spread at one point in normal markets, but not in stressed markets. A buy-back reduces Meridian’s funding and closes part of the hedge; Treasury treats repeated buy-backs as a liquidity outflow (section 6.3).
Scenario C: a constituent loses its eligibility. In spring 2027, a utility in the renewable-led block announces the acquisition of a gas-fired generation portfolio. Its revenues from electricity generated above 100 gCO2e/kWh pass 50%. The index administrator’s data provider flags the change; at the next quarterly review, the administrator removes the company and redistributes its weight according to the rulebook. Meridian’s delta-one desk sells the shares on the rebalancing date and buys the replacements; because the change is announced a few days in advance, other market participants may trade ahead of it, and the cost appears in the index level, borne by the investor. The note is unaffected legally, but Meridian sends the index’s updated ESG disclosure to Hauteville, as the distribution agreement requires.
Reading. The three scenarios show where responsibilities sit after issue. The investor bears market, dividend and rebalancing costs through the index. Meridian bears hedging, model and funding risks and must provide a fair secondary price. The administrator bears the integrity of the label. Hauteville bears the relationship with the client, including the decision to propose a new product after a call. A greenwashing complaint is most likely in scenario C, if a client learns from the press that “his” Paris-aligned index held a gas-power company for three months: the rulebook’s review frequency and the disclosures made at sale are the defence.
WHY IT MATTERS ON THE DESK — Product governance is not a step before issue; it is a duty that runs for the life of the product. ESMA’s product governance guidelines expect manufacturers to review their products periodically, and a change in the underlying’s ESG profile is precisely the kind of event that should trigger a review.
Question. Is the deal worth doing for Meridian at a 1.17% margin, once the costs of the sustainable features are counted?
Data. Margin EUR 1.17m (stage 5.4), recognised over the expected life of 4.21 years for the purpose of an annual return. Cost-income ratio 60% on markets revenue. Index administration fee and ESG data licences: EUR 200,000 a year for the index, allocated across EUR 600m of notes Meridian expects to have outstanding on it, so EUR 33,300 a year for this note. Market-risk RWA allocated to the note’s residual risks (vega, skew, gap and basis), from the desk’s VaR-based allocation: 3% of notional, EUR 3.0m. Operational-risk RWA: 15% of annual revenue times 12.5. No credit risk: the investor bears Meridian’s credit risk, and the hedges are collateralised. Funding at FTP equal to the 0.70% spread used in pricing, so no funding gain or loss.
INTERACTIVE LAB — Meridian’s RAROC on the note, with and without the sustainable features
Turn the day-one margin into an annual return on allocated capital and see how the index and ESG data costs change it.
| Parameter you can move | Range | Default |
|---|---|---|
| Issuer margin (stage 5.4) | 0.2 to 3 EUR m | 1.17 EUR m |
| Expected life (stage 5.4) | 1 to 10 years | 4.21 years |
| Notional | 10 to 300 EUR m | 100 EUR m |
| Market-risk RWA allocated | 0 to 10 % of notional | 3 % of notional |
| Cost-income ratio on markets revenue | 30 to 80 % | 60 % |
| Index administration and ESG data, per year | 0 to 500 EUR k | 200 EUR k |
| Expected notes outstanding on the index | 100 to 2000 EUR m | 600 EUR m |
| Result at default settings | Value |
|---|---|
| Revenue per year | 0.2779 EUR m |
| Index and ESG data cost allocated to the note | 33.3 EUR k a year |
| Operating costs per year | 0.2001 EUR m |
| of which cost-income charge | 0.1667 EUR m |
| Profit before tax | 0.0778 EUR m |
| Tax at 25% | 0.0195 EUR m |
| of which operational-risk RWA | 0.521 EUR m |
| Profit after tax | 0.0584 EUR m |
| RWA (market and operational) | 3.521 EUR m |
| Allocated capital (13%) | 0.458 EUR m |
| RAROC with sustainable features | 12.75 % |
| RAROC without index and ESG data costs | 18.21 % |
| Margin needed for the 12% hurdle | 1.116 EUR m |
Try it. Set the programme to EUR 100m, as if this were the only note on the index: the fixed costs of the index swallow the margin and the RAROC turns negative.
Result. Revenue per year is EUR 0.2779m. Operating costs are EUR 0.2001m (60% of revenue plus the EUR 33,300 of index and data costs). Profit before tax EUR 0.0778m; after tax EUR 0.0584m. RWA EUR 3.521m; allocated capital EUR 0.458m. RAROC 12.75%, just above the 12% hurdle. Without the index and data costs it would be 18.21%. The margin that just reaches the hurdle is EUR 1.116m.
Reading. The sustainable features cost more than five points of RAROC, entirely through fixed costs that are only bearable if the index carries a programme of notes. If this were the only note on the index, the RAROC would turn negative. This is why banks build a small number of reusable ESG indices rather than one per client, and why the structurer pushed for 12.50% instead of 12.62%: the twelve basis points of coupon given up are what keeps the deal above the hurdle.
| Item | Value | Stage |
|---|---|---|
| Universe intensity / PAB index intensity | 164.2 / 82.1 tCO2e per EUR m EVIC | 5.2 |
| Reduction from exclusions and rescaling only | 28.2% | 5.2 |
| Tilt exponent / tracking error | 0.204 / 1.88% | 5.2 |
| Dividend yield universe / index | 3.56% / 3.07% | 5.2 |
| Taxonomy-aligned revenue universe / index | 8.25% / 8.66% | 5.2 |
| Annual decrement drag / ten-year underperformance | 1.93 pts / 17.55% | 5.3 |
| 60% barrier in price-index terms | 72.77% | 5.3 |
| Fair coupon / offered coupon | 12.62% / 12.50% | 5.4 |
| Fair coupon on price index | 8.97% | 5.4 |
| Fair value / distributor fee / Meridian margin | 94.83 / 4.00 / 1.17 | 5.4 |
| Call in year 1 / expected life / loss probability | 43.75% / 4.21 years / 19.94% | 5.4 |
| Delta / vega | EUR 77.5m / EUR 320k per point | 5.5 |
| Dividend risk avoided / proxy-hedge risk | EUR 2.98m per point / EUR 1.46m | 5.5 |
| Taxonomy share: index look-through / use of proceeds | 8.66% / 72% | 5.6 |
| Green pool coverage / headroom | 110.8% / EUR 0.70bn | 5.6 |
| RAROC with / without sustainable costs | 12.75% / 18.21% | 5.9 |
| Break-even margin | EUR 1.116m | 5.9 |
A structured note does not create a credit exposure for the issuer: the investor lends to Meridian. The capital the trade consumes is of three kinds.
Market risk. The note and its hedges sit in the trading book. In November 2026 Meridian still computes its own-funds requirement for market risk under the pre-FRTB rules, because the EU postponed the Fundamental Review of the Trading Book twice; it applies from 1 January 2027. On 4 June 2026 the Commission adopted a delegated act with temporary adjustments, including a multiplier that offsets part of the capital increase for three years from 1 January 2027, subject to scrutiny by Parliament and Council. Under FRTB, an autocall book on a bespoke index raises specific issues. The standardised approach charges equity delta, vega and curvature by bucket and adds a residual risk add-on of 0.1% of notional for path-dependent payoffs such as autocalls (1% for exotic underlyings such as weather or longevity). An internal model must pass the risk factor eligibility test, and the volatility of a bespoke index with no traded options becomes a non-modellable risk factor charged separately. Our case allocates EUR 3.0m of market-risk RWA, 3% of notional, as the note’s share of a diversified book: desks manage these risks as portfolios, where the vega of one note offsets part of another’s.
Counterparty credit risk. The hedges are futures (cleared, with small exposure to the central counterparty) or total return swaps with other dealers under collateralised ISDA and credit support annexes. Their SA-CCR exposure and CVA charge are small and allocated at book level; the case ignores them.
Operational risk. Under the CRR3 standardised approach, the business indicator includes trading and fee income; with Meridian’s assumptions the case charges 15% of annual revenue, times 12.5, in RWA: EUR 0.521m.
There is no green supporting factor in CRR3, and the ESG features change none of these numbers: the capital of the PAB note is the capital of any autocall on a bespoke index.
From stage 5.9, per year in EUR m: revenue 0.2779 (the 1.17 margin spread over the 4.21-year expected life); operating costs 0.2001, of which 0.1667 at the 60% cost-income ratio and 0.0333 of index administration and ESG data; expected loss 0; profit before tax 0.0778; tax at 25%, 0.0195; profit after tax 0.0584. RWA 3.521 (3.000 market, 0.521 operational); allocated capital at 13%, 0.458. RAROC 12.75%, against a hurdle of 12%.
The sustainable feature changes the RAROC through costs, not through capital: index administration under the Benchmarks Regulation, ESG data licences for the exclusions and intensities, the sustainable finance team’s review and the extra product governance work. Without them the same margin would give 18.21%. It may change revenue indirectly: a credible PAB index gives access to distributors and insurers whose clients ask for climate themes, and it can be reused for total return swaps and QIS sold to institutional clients, which spreads the fixed cost. A bank that launches one bespoke ESG index per client destroys value; one that builds a short shelf of reusable indices does not.
NOTE — Recognising the margin over the expected life is a management convention used here to compute an annual return. The accounting of section 6.4 may recognise it differently, and the economic life varies with the market: if the index rises in year 1, the margin was earned in one year and the RAROC of the trade was much higher.
A structured note is a funding instrument. Meridian receives EUR 100m in cash and owes a payoff whose expected life is 4.21 years and maximum life ten years. Treasury pays the desk for this funding at the FTP rate of the note’s expected profile, and charges the desk for the funding of the hedge.
In the case the two cancel out: the price uses the same 70 bp spread as the FTP curve, and the cash raised pays for the EUR 77.5m of shares bought as delta, with the rest placed with Treasury. Three frictions remain. First, the life is uncertain: an early call after one year returns funding that Treasury had counted on for longer, and Treasury charges the desk for this optionality, often by using a shorter tenor than the expected life. Second, Meridian quotes buy-back prices: notes bought back before maturity are funding that leaves early. Third, the liquidity coverage ratio treats the note according to its earliest possible redemption and the buy-back commitment; the net stable funding ratio gives credit only to funding beyond one year.
Structured notes do not usually count towards Meridian’s MREL. The CRR excludes debt instruments with embedded derivatives from eligible liabilities unless their principal amount is known at issuance, fixed or increasing, and unaffected by the derivative; a note with 40% or more of principal at risk does not meet that condition. A green version does not change this.
Issuer’s side. Meridian issues a hybrid financial liability: a host debt instrument with embedded equity derivatives that are not closely related. Under IFRS 9 the bank can separate the embedded derivative or, more commonly for structured notes managed on a fair value basis, designate the whole note at fair value through profit or loss (fair value option). The change in fair value due to Meridian’s own credit risk is presented in other comprehensive income, so that the bank does not report a profit when its own credit worsens.
Day-one profit. The difference between the issue price net of fees (96.00) and the fair value (94.83) is the 1.17 margin. IFRS 9 (paragraph B5.1.2A) allows it to be recognised on day one only if fair value is evidenced by a quoted price or a valuation technique using only observable inputs. The volatility of a bespoke index with no listed options is not observable: the note is a Level 3 instrument under IFRS 13 and the margin is deferred and released over the life, or when the inputs become observable or the note is redeemed. In practice, product control may judge that the bespoke index’s volatility is close enough to the parent index’s observable surface; that is a judgement documented in the valuation policy.
Prudent valuation. Under the EBA’s prudent valuation standard, Meridian deducts additional valuation adjustments from CET1 for model risk, unobservable inputs and close-out costs on its Level 3 book. For a new bespoke index, the model-risk adjustment on skew is the largest.
Distributor fee. The 4.00 paid to Hauteville is a transaction cost of the liability; under the fair value option it is expensed immediately.
Hedges. Shares and futures held as delta hedges are measured at fair value through profit or loss; there is no hedge accounting, because both legs already go through profit or loss.
Green Asset Ratio. The note is a liability and the hedges are in the trading book: neither enters the GAR, whose denominator excludes the trading book. The trading book KPI and the fees and commissions KPI that would capture this business have been deferred to 2028 by the 2026 simplification of the Taxonomy disclosures (see A02, C02).
Green funding. The green version adds EUR 100m to Meridian’s green instruments outstanding, reported in its green bond allocation report: coverage 110.8% after this issue. The bank’s investors and rating agencies watch this coverage; if it fell below 100%, Meridian could no longer issue green notes without first growing the pool.
Pillar 3 ESG. The EBA ESG templates focus on the banking book. The equity derivatives business appears qualitatively, in the description of how the bank manages transition risk in its trading activities.
Financed and facilitated emissions. PCAF Part A covers listed equity held on the balance sheet, attributed by the ratio of holding to EVIC. The EUR 77.5m delta hedge is a listed equity holding; Meridian, like most banks, excludes trading-book hedges from financed emissions because the positions change daily and are held for the client, not as investments. If it included them, the PAB tilt would matter: the hedge of a PAB index has half the intensity of a universe hedge. Issuing the note is not a facilitation of the client’s financing in PCAF’s sense; no facilitated emissions are reported.
Product-level disclosure. Meridian provides Hauteville with the PAI data of the index through the European ESG Template, and the index administrator publishes the benchmark’s ESG statement. These are the data the distributor relies on for the sustainability-preference assessment.
| Risk | What can go wrong | Control | Owner |
|---|---|---|---|
| Model and skew | Coupon overpaid, hedge losses | Model validation, skew reserve, Level 3 review | Model validation, product control |
| Gap risk at barrier | Index near 60% at maturity | Barrier limits, shift reserves | Exotic trader, market risk |
| Basis and rebalancing | Proxy hedge drifts from index | Basket hedge, rebalancing calendar | Delta-one trader |
| Index data error | Wrong exclusion or intensity | Administrator oversight, data reconciliation | Index administrator, index team |
| Label loss | Index no longer meets PAB rules | Rulebook fallback, client notice | Index team, product governance |
| Conflict of interest | Bank designs, hedges and issues | External administrator, information barriers | Compliance |
| Mis-selling | Note sold outside target market | Suitability checks, sales reports | Hauteville, Meridian distribution |
| Greenwashing | Claim exceeds what product does | Wording rules, sustainability review | Sustainable finance team, compliance |
| Liquidity of buy-backs | Many clients sell in stress | Buy-back spread policy, funding plan | Trader, Treasury |
| Issuer credit | Investor exposed to Meridian | Clear disclosure in KID and prospectus | Legal, debt issuance |
Three of these risks deserve a comment. Model risk is the largest economic risk of an autocall desk, as stage 5.4 showed with three points of coupon between flat and skewed volatility. Label loss is specific to ESG underlyings: if data revisions make the index fail the PAB rules, or a regulatory change moves the thresholds, the note keeps running on an index that no longer carries its label. The rulebook must say what happens (methodology change, substitution, or continuation without label) and the distribution agreement must say who tells the client. Conflicts of interest arise because Meridian designs the index, hedges it and issues the note; a quarterly rebalancing that the desk can anticipate, or a methodology change that lowers the hedge cost, would harm investors. An external administrator with an independent oversight committee is the usual answer.
Greenwashing red flags in ESG structured products.
Regulation (dated).
Market data (with sources).
What changed recently. The BMR review of 2025 shifted supervision towards large and labelled benchmarks, which makes the PAB and CTB labels one of the few ESG features of an index still under direct EU rules. The French supervisors have increased their attention on decrement indices and fees (April 2025 mapping, June 2026 explanatory note). The Retail Investment Strategy will add a value-for-money test that will apply to structured notes once in force.
Open debates.
INTERACTIVE LAB — Equity derivatives, ESG indices and structured products: self-assessment
Check the mechanics of decrement indices, autocalls and climate benchmarks, and the regulatory and banking consequences.
Q1. With a funding rate of 2.5% and a 5% decrement, the ten-year forward of a decrement index starting at 100 is about:
Answer: A. F = 100 x exp((2.5% - 5%) x 10) = 77.88; the dividends of the underlying play no role.
Q2. Who bears dividend risk in a note on a 5% decrement index?
Answer: B. The decrement is fixed; if companies pay less than 5%, the index lags the price index and the investor bears the shortfall.
Q3. Which exclusion applies to a PAB but not to a CTB?
Answer: C. Fossil-fuel revenue exclusions (coal 1%, oil 10%, gas 50%, high-carbon power 50%) apply only to PABs; the other three apply to both.
Q4. The minimum intensity reduction of a PAB versus its investable universe is:
Answer: D. 50% for PABs and 30% for CTBs, then 7% a year of self-decarbonisation for both.
Q5. Why does the PAB rule on high-climate-impact sectors matter for index construction?
Answer: B. Exposure to NACE sections A to H and L must be at least that of the universe, so the reduction must come from choosing lower-intensity companies within those sectors.
Q6. In the case, a PAB index has a Taxonomy-aligned revenue share of 8.66% against 8.25% for its universe. What does that show?
Answer: D. A PAB cuts intensity per euro of EVIC; it does not select Taxonomy-aligned activities.
Q7. An autocall is called in year 4 with a 12.5% snowball coupon. The investor’s annual return is about:
Answer: A. The investor receives 150 after four years: 1.5^(1/4) - 1 = 10.7%.
Q8. Why does a flat-volatility model overstate the coupon of an autocall?
Answer: C. Downside skew makes calls more likely to be triggered and the put more valuable; in the case flat volatility gives 15.75% instead of 12.62%.
Q9. After selling autocalls, a dealer is typically:
Answer: D. The investor sells a put to the dealer; in the case Meridian gains about EUR 320,000 per volatility point on EUR 100m.
Q10. Which statement about a use-of-proceeds structured note is correct?
Answer: C. The green claim concerns the issuer’s use of funds, not the payoff.
Q11. Under IFRS 9, the day-one margin on a note valued with an unobservable bespoke-index volatility is:
Answer: B. Paragraph B5.1.2A requires deferral when fair value is not evidenced by observable inputs (Level 3).
Q12. Since 1 January 2026, under the amended Benchmarks Regulation, EU climate benchmarks:
Answer: A. Regulation (EU) 2025/914 removed most non-significant benchmarks from the scope but kept EU climate benchmarks in it.
Q13. Which fund-name terms trigger PAB exclusions under ESMA’s guidelines?
Answer: C. Environmental, impact and sustainability-related terms require PAB exclusions; transition, social and governance terms require CTB exclusions.
Q14. Meridian’s delta hedge of EUR 77.5m of index shares is excluded from its Green Asset Ratio because:
Answer: C. The GAR covers banking-book assets; trading-book positions are excluded from it.
An index pays a dividend yield of 2.6% a year. A note is linked to its 4% decrement version, with a 55% protection barrier at maturity in 8 years. The funding rate is 2.5%. (a) What is the eight-year forward of the decrement index and of the price index (both starting at 100)? (b) What is the cumulative underperformance of the decrement index against the price index after 8 years? (c) What price-index level does the barrier correspond to at maturity?
Solution. (a) Decrement index forward: . Price index forward: . (b) The drag is points a year; the ratio of the two indices after 8 years is , an underperformance of 10.60%. (c) The barrier is crossed when the decrement index is at 55, that is when the price index is at . The client who thinks in terms of the price index faces a 61.5% barrier, not 55%.
A pricer returns, for a ten-year autocall, a redemption leg worth 78.40% and a value of 1.52% per 1% of annual coupon. The distributor fee is 3.0% and the issuer targets 1.25%. (a) What is the fair coupon? (b) The structurer offers 11.25%. What is the fair value of the note and the issuer’s margin? (c) Explain why the coupon can be solved without re-running the simulation.
Solution. (a) The fair value must equal . The fair coupon is . (b) At 11.25%, the fair value is ; the issuer’s margin is , a quarter point more than targeted, paid for by the 0.16 point of coupon given up. (c) The coupon only scales cash flows that are already determined by the path (call date pays coupons), so the note value is linear in the coupon: one simulation gives both legs.
A EUR 50m autocall leaves the issuer a margin of EUR 0.9m and has an expected life of 3.0 years. Market-risk RWA is 4% of notional; operational-risk RWA is 15% of annual revenue times 12.5; the cost-income ratio is 60% and the note bears EUR 20,000 a year of index costs. Tax is 25% and capital is 13% of RWA. Compute the RAROC.
Solution. Revenue per year: EUR m. Costs: . Profit before tax: ; after tax: . RWA: market ; operational ; total . Capital: . RAROC: , well above the 12% hurdle, because the margin is a higher share of notional and the expected life shorter than in the case.
1. Why do structurers like decrement indices? They remove dividend risk from the issuer, because the forward of a decrement index depends only on the funding rate and the fixed decrement. They lower the forward, which makes the put the investor sells more valuable and allows a higher coupon. They also make it possible to build products on bespoke indices for which no dividend market exists. The cost is a drag for the investor equal to the decrement minus the dividends actually paid, which must be disclosed in comparable terms.
2. Walk me through the PAB minimum standards. The index’s greenhouse gas intensity, measured per euro of enterprise value including cash and including phased-in scope 3, must be at least 50% below its investable universe at launch. It must then fall by at least 7% a year on average. Exposure to high-climate-impact sectors must be at least that of the universe. The index excludes controversial weapons, tobacco, norm violators and companies above the fossil revenue thresholds: 1% for coal, 10% for oil, 50% for gas and 50% for high-carbon power generation.
3. How would you price an autocall? I would simulate the index under the pricing measure with a drift equal to the funding rate minus the decrement or dividend yield, and a volatility model that captures skew, such as local or stochastic volatility. On each path I record whether and when the note is called, and the redemption at maturity, then discount at the issuer’s funding curve. The value is linear in the coupon, so I solve for the coupon that equates the note value to the issue price less fees and margin. Then I check sensitivities to skew, decrement, rates and issuer spread, because they drive both the coupon and the hedge.
4. What risks remain with the dealer after selling an autocall? Delta is hedged with the underlying, but the dealer is left long vega, exposed to skew and to gap risk around the barrier at maturity, and, on a price index, to dividends. On a bespoke index there is also basis risk if the hedge uses proxies, and rebalancing risk at each index review. These risks are managed at book level, with limits by underlying and maturity and with valuation reserves, and they explain part of the margin.
5. A client wants a “green” autocall. What are the options and what can you claim? I can link the payoff to an ESG or climate index, ideally a registered PAB or CTB, and claim that the underlying meets those standards. Or the issuer can issue the note under its green bond framework, and claim that an amount equal to the proceeds finances eligible green assets. Neither lets me say that the client invests in green companies, since the client holds a claim on the issuer. The choice should follow the client’s stated sustainability preferences and be reflected in the target market.
6. Does an autocall on a PAB index meet a client’s Taxonomy preference? Usually not. By look-through, its Taxonomy-aligned share is the index’s aligned revenue share, which in practice is close to the universe’s, around 8% in our case. A PAB reduces carbon intensity, it does not select green revenues. The note may still meet a preference for considering principal adverse impacts on emissions, and a use-of-proceeds version may meet a Taxonomy minimum.
7. How is a structured note accounted for by the issuer? Most banks designate structured notes at fair value through profit or loss under IFRS 9, with changes due to own credit risk in other comprehensive income. The day-one margin can only be recognised immediately if the valuation uses observable inputs; with a bespoke index the volatility is unobservable and the margin is deferred. Prudent valuation adjustments are deducted from CET1 for model risk and unobservable inputs. The distributor fee is expensed at issuance under the fair value option.
8. Why would a bank not build one bespoke ESG index for each client? Each index has fixed costs: an authorised administrator, data licences, oversight, rulebook maintenance and disclosures. In our case these costs take more than five points off the RAROC of a EUR 100m note even when spread over EUR 600m of products, and make a single-note index loss-making. A small shelf of reusable indices also helps hedging, because notes on the same index net their risks. Finally, a proliferation of bespoke ESG indices increases conflict-of-interest and greenwashing risk.
EU regulation and standards
Supervisory and official publications
Market sources
Academic and textbook references