Module C01 ended on the evening of pricing: Helios Renewables had sold EUR 500m of green bonds, the book was allocated and the bonds were “free to trade”. From that minute until the bond matures eight years later, every investor who wants to buy or sell it calls a dealer, sends a request for quote on an electronic platform, or trades with another investor through an all-to-all protocol. The credit trading desk is the dealer. It sits in Global Markets, inside FICC (fixed income, currencies and commodities), next to rates trading, and it shares the floor with the syndicate desk and debt sales. It does not lend money to issuers and does not earn underwriting fees. It buys and sells bonds and credit derivatives with investors, holds the resulting inventory for hours, days or weeks, and hedges what it cannot sell.
Clients. Almost all are institutional: asset managers (benchmark-driven credit funds, ETFs, dedicated green bond funds), insurers and pension funds, which buy and hold but reposition at rating changes and quarter-ends, bank treasuries investing their liquidity buffers, hedge funds trading relative value and CDS-bond basis, and central banks’ and sovereign funds’ reserve managers. Green bond funds and portfolios with green allocation targets are a recognisable client segment: they buy in primary, rarely sell, and ask dealers for green paper they missed. Issuers themselves appear when they buy back bonds or exchange them.
How it earns money. Three sources, in this order of quality. First, bid-offer capture: the desk buys at its bid and sells at its offer, so every round trip earns part of the difference. On a liquid euro investment-grade bond the quoted bid-offer is a few basis points of spread, which is 20 to 40 cents per 100 of nominal for a bond of 5 to 10 years’ duration; the desk keeps only part of it after hedging and after the market moves. Second, carry: an inventory of bonds earns its credit spread every day, minus the cost of hedges and funding. Third, risk-taking: positioning ahead of expected moves, which the post-2014 Volcker-style constraints in the US and the bank’s own risk appetite in Europe keep small for a flow desk. Losses come from the mirror images: adverse selection (the client who sells to you often knows something), inventory that loses value before it can be sold, and hedges that do not move with the bonds.
The team and its neighbours. A euro investment-grade flow desk at a large European bank typically has a head of credit trading, a handful of traders each owning a sector “book” (for example utilities and energy, industrials and chemicals, financials), an index and CDS trader, and one or two people for electronic market making and portfolio trades, who run pricing algorithms that answer small requests automatically. Around them: credit sales, who own the client relationships and bring the orders; the syndicate desk, which allocates new issues and expects the trader to make a market in them afterwards; credit research or desk analysts, who write the morning comment on issuers; the repo desk, which finances long positions and borrows the bonds the desk sells short; market risk, which sets and monitors limits on CS01, jump-to-default and stress loss; product control, which marks the book to independent prices every evening and explains the P&L; compliance, which monitors best execution, information barriers and market abuse; and the sustainable finance team, consulted when a client asks whether a bond is green or a product is “ESG”.
A realistic week. On Monday at 7:30 the desk meets: overnight moves in US credit and index, new issues expected this week, the inventory, and axes (the bonds the desk wants to buy or sell). Traders send runs to sales: lists of bonds with indicative bid and offer spreads. From 8:00 requests for quote arrive on electronic platforms; the trader or the algorithm answers within seconds, typically against several competing dealers. At 9:00 a new issue is announced by another bank’s syndicate desk; the trader estimates where it should come against the issuer’s curve and calls sales clients who might switch out of old bonds into the new one. On Tuesday the desk is asked to show a price for EUR 25m of a green bond in a portfolio trade with 80 other bonds; it prices the whole list in an hour. On Wednesday market risk flags that the chemicals book is close to its CS01 limit after a client sale; the trader buys protection on the iTraxx index. On Thursday a rating agency downgrades an issuer and insurers sell; the desk widens its quotes and takes less than it is shown. On Friday product control presents the weekly P&L explain: how much came from new trades, carry, spread moves, rates and hedges, and what is unexplained. The rest of the module follows one such week, a bad one.
A corporate bond pays a fixed coupon annually (in euros) and its nominal at maturity. In the secondary market it is quoted in two ways that the desk translates constantly: a clean price per 100 of nominal, and a spread, the extra yield over a reference. In euros the reference for corporates is usually mid-swaps (the interpolated swap rate of the same maturity), and the spread is called the I-spread or, more loosely, the spread to swaps; government spreads and z-spreads are also used. This module uses spreads over mid-swaps, with swaps flat at 2.50% as in the rest of the course. Investors in investment-grade credit think in spread because the rate component is hedged or managed separately; dealers quote in spread for the same reason and settle in price.
IN ONE SENTENCE — A market-maker sells immediacy: it buys a bond the client wants to sell now, holds it, hedges it and resells it, and the bid-offer pays for that service and for the risk.
INTUITION — A second-hand car dealer buys your car today for less than he will sell it next month. The gap pays for the forecourt, for the risk that prices fall while the car sits there, and for the chance that you knew something about the gearbox that he did not.
The trader’s quote has two sides. The bid is where the desk buys, the offer where it sells; their difference is the bid-offer. Expressed in spread, a bid at 140 bp and an offer at 135 bp is a 5 bp bid-offer (the bid spread is wider because a wider spread means a lower price). Converted to price:
FORMULA
is the price change per 100 of nominal; the modified duration, which for a fixed-coupon bond is also its spread duration; the price as a fraction of 100; the spread change in decimal (5 bp is 0.0005). A 5 bp bid-offer on a bond of duration 6 at par is about 0.30 points, or EUR 30,000 on EUR 10m nominal.
How a trade happens. Most euro investment-grade trades in normal sizes are executed by request for quote (RFQ) on electronic platforms: the client asks several dealers for a price on a given size; each dealer answers; the client hits the best. The dealer learns whether it won and, on most platforms, how far it was from the winning price (the cover). Large or difficult trades go by voice through sales. Portfolio trading, in which a client asks for one price on a list of many bonds, has grown fast since 2020 and lets dealers price risk at the portfolio level. All-to-all protocols let investors trade with each other on the same platforms.
Inventory and axes. A desk that buys EUR 10m from a client now owns EUR 10m. It tells the market it wants to sell, through its axes shown to sales and to platforms, and it may skew its quotes: a slightly tighter offer to attract buyers, a slightly wider bid to avoid buying more. The time a position stays on the book is the holding period. During that time the bond’s spread moves; the expected revenue is fixed when the trade is done, the risk grows with the square root of time.
Shorting. A desk can sell a bond it does not own by borrowing it through the repo desk (a reverse repo or securities borrowing) and paying a borrowing fee. Short positions in a few bonds of an issuer are a common way to hedge a long position in another bond of the same issuer along its curve.
INTERACTIVE LAB — What a trade earns and what holding it risks
Convert a spread bid-offer into price and euros, and compare the revenue captured on a trade with the one-standard-deviation risk of holding the position for some days.
| Parameter you can move | Range | Default |
|---|---|---|
| Trade size | 1 to 50 EUR m | 10 EUR m |
| Quoted bid-offer | 1 to 20 bp of spread | 5 bp of spread |
| Spread duration of the bond | 1 to 12 years | 6.8 years |
| Share of half bid-offer kept after hedging | 0 to 100 % | 60 % |
| Daily spread volatility | 0.5 to 15 bp | 3 bp |
| Holding period | 1 to 20 days | 3 days |
| Result at default settings | Value |
|---|---|
| Bid-offer in price | 0.340 points per 100 |
| CS01 of the position | 6,800 EUR per bp |
| Expected revenue kept | 10,200 EUR |
| One-standard-deviation P&L over the holding period | 35,334 EUR |
| Revenue as a share of one-sigma risk | 28.9 % |
Try it. Set the holding period to 10 days and the volatility to 8 bp, as in a stressed week: the risk is about seventeen times the revenue. That is why dealers widen quotes and cut size when markets sell off.
The lab shows why market-making is a volume business. On EUR 10m of a duration-6.8 bond with a 5 bp bid-offer, the desk expects to keep about EUR 10,200 if it captures 60% of half the bid-offer. The one-sigma P&L of holding the position for three days at 3 bp of daily spread volatility is EUR 35,000. Any single trade is a bet with a small positive expectation; only many trades, short holding periods and hedges make the expectation dominate.
PITFALL — Students often treat the quoted bid-offer as revenue. It is the maximum; the desk earns it only if it can turn the position at the other side of the market before prices move. Realised capture on investment-grade flow is a fraction of the quoted half bid-offer, and it is negative on trades with better-informed clients.
A fixed-coupon bond carries two risks: rates and credit. The DV01 (price change for a 1 bp fall in yield) equals the CS01 (price change for a 1 bp tightening in spread) for a bullet bond, because both move the same yield. The credit trader does not want rate risk, so it neutralises the DV01 by selling government bond futures: in euros the Schatz (2 years), Bobl (5 years), Bund (10 years) and Buxl (30 years) contracts on Eurex, each with a notional of EUR 100,000 and a deliverable basket of German government bonds; or by paying fixed on interest rate swaps through the rates desk. The number of contracts is the bond’s DV01 divided by the DV01 of one contract. The hedge leaves the trader with swap spread or asset swap risk (bonds hedged with Bund futures are exposed to the spread between swaps and Bunds) and curve risk if the maturities differ.
A credit default swap (CDS) is insurance on an issuer’s default. The protection buyer pays a running premium on a notional; the protection seller pays the loss (notional minus recovery) if a credit event occurs (bankruptcy, failure to pay, and for some contracts restructuring). Since 2009 contracts trade with a standard fixed coupon (100 bp or 500 bp for the main European indices) and an upfront payment that equalises the value:
FORMULA
is the upfront paid by the protection buyer, as a fraction of notional (negative means the buyer receives it); the market spread; the fixed coupon; RPV01 the risky present value of 1 bp of premium paid quarterly; the maturity in years; the quarterly payment dates; the continuously compounded risk-free rate; the default intensity implied by a flat spread (the “credit triangle”); the assumed recovery rate, 40% by convention for senior unsecured.
A CDS index is a portfolio of single-name CDS traded as one contract. In Europe, iTraxx Europe (“Main”) contains 125 equally weighted investment-grade names and trades with a 100 bp coupon; iTraxx Crossover contains about 75 high-yield names with a 500 bp coupon; both roll every six months, on 20 March and 20 September, into a new series whose five-year maturity is the most liquid. Series 46 of Main started trading on 21 September 2026. Indices are cleared at central counterparties (ICE Clear Credit, LCH CDSClear) and are by far the most liquid credit instrument in Europe: a dealer can buy EUR 100m of protection in seconds when a bond may take days to sell.
INTUITION — Hedging a single bond with the index is like insuring your house by buying insurance on the whole street: it pays when the whole street burns, not when only your kitchen does.
INTERACTIVE LAB — From an index spread to the upfront, the RPV01 and the CS01
Price a standard-coupon CDS index trade from its quoted spread and see how the CS01 used for hedging is derived.
| Parameter you can move | Range | Default |
|---|---|---|
| Quoted index spread | 20 to 400 bp | 55 bp |
| Fixed coupon | 100 / 500 | 100 |
| Recovery rate | 0 to 80 % | 40 % |
| Risk-free rate | 0 to 6 % | 2.5 % |
| Maturity | 1 to 10 years | 5 years |
| Protection notional | 5 to 500 EUR m | 65 EUR m |
| Result at default settings | Value |
|---|---|
| RPV01 | 4.577 years |
| Upfront paid by the protection buyer | -2.059 % of notional |
| Upfront paid by the protection buyer | -1,338,664 EUR |
| CS01 of the protection (gain for +1 bp) | 30,305 EUR per bp |
Try it. Set the spread to 100 bp: the upfront is zero because spread and coupon are equal. Then set it to 300 bp with the 500 bp coupon, as for iTraxx Crossover.
At 55 bp, buying EUR 65m of five-year protection with a 100 bp coupon means receiving an upfront of about 2.06% (EUR 1.34m) and then paying 100 bp a year: economically, paying 55 bp a year. Each additional basis point of index spread makes the protection worth about EUR 30,305 more. That number, the hedge’s CS01, is what the trader compares with the CS01 of the bonds.
Basis. A bond and a CDS on the same issuer do not have the same spread. The CDS-bond basis is the CDS spread minus the bond’s spread over the risk-free curve (usually its z-spread or asset-swap spread). It is driven by funding (a bond must be financed, a CDS does not), the cheapest-to-deliver option in the CDS, supply and demand for protection and, for labelled bonds, any greenium: a green bond trading a few basis points tight to its issuer’s conventional curve shows a more positive basis against the issuer’s CDS, since the CDS references all senior bonds, green or not.
The desk’s risk is measured every evening by market risk in five numbers: DV01 by currency and tenor; CS01 by issuer, sector and rating; jump-to-default (JTD), the loss if an issuer defaulted overnight (nominal times one minus recovery, plus or minus the price’s distance to par); stress loss under predefined scenarios (for example, investment-grade spreads 50% wider, iTraxx Main doubling); and value at risk or expected shortfall for the internal models. Limits exist for each, set by the head of desk within the business line’s risk appetite and approved by market risk. Breaches are escalated the same day.
Spread beta. Single names do not move one-for-one with the index. The beta of an issuer’s spread to iTraxx Main, estimated by regressing daily spread changes, measures how many basis points the issuer moves for one basis point of the index. Defensive utilities typically have betas below one, cyclical and high-carbon industrials above one. Index hedges are sized on beta-weighted CS01.
P&L explain. Product control decomposes each day’s P&L into new trades (the difference between trade price and the evening’s mark: the realised bid-offer), carry and roll-down (the passage of time), rates moves, spread moves by issuer, hedge P&L, funding, fees, and the residual. A large unexplained residual is a control finding.
Bonds trade under the platform’s rulebook or a confirmation, and settle through Euroclear or Clearstream two business days after trade date. Repo is documented under the GMRA (Global Master Repurchase Agreement), securities lending under the GMSLA. CDS are documented under the ISDA Master Agreement with a Credit Support Annex for collateral and the 2014 ISDA Credit Derivatives Definitions; index trades incorporate the index’s annex published by the administrator (S&P Dow Jones Indices for iTraxx since S&P Global’s acquisition of IHS Markit). This module relies on public summaries for ISDA documents, which are member-only. On the regulatory side: MiFID II and MiFIR (best execution, transaction reporting to the regulator, post-trade publication), the Market Abuse Regulation (inside information, market manipulation), EMIR (derivatives reporting and clearing obligations, which apply to the main iTraxx indices), and the CSDR settlement discipline regime.
Nothing in the bond’s cash flows. A green bond of Helios ranks pari passu with its conventional bonds, defaults with them, and has the same recovery. Three things change around it.
The holder base. Dedicated green bond funds, impact mandates and investors with green allocation targets buy a large share in primary and rarely sell. This makes the free float smaller than the issue size and the bond harder to borrow, and it creates a pool of natural buyers when someone sells. Since May 2025, funds that use ESG or sustainability-related terms in their names must apply the exclusions of the EU Paris-aligned benchmarks under ESMA’s fund-names guidelines (applicable to new funds from 21 November 2024 and to existing funds from 21 May 2025), with an allowance for bonds issued under the EU Green Bond Standard. The logic for the trader is direct: a fund renamed or repositioned under these rules may have to sell an issuer’s conventional bonds while it can keep that issuer’s European Green Bonds, so two bonds of the same issuer can face different selling pressure.
The data. A dealer must know which bonds are labelled, under which standard (ICMA Green Bond Principles, EU Green Bond Standard, Climate Bonds certification), whether an external review exists, and whether the issuer is excluded from popular ESG indices. This comes from the data vendors’ green flags and index memberships (Bloomberg MSCI green bond indices, ICE and iBoxx ESG families), and it is far from uniform: a bond can be in one vendor’s green index and out of another’s.
The pricing variable. The greenium (see C01, section 3.5) is measured in primary as the difference between the green bond’s reoffer yield and the issuer’s conventional fair value; in secondary, as the difference between the green bond’s spread and the spread a conventional bond of the same issuer with the same maturity would have. The academic evidence summarised by Roncalli (Handbook, section 4.2.1, including the German twin Bunds, where the average was about −3 bp between 2020 and 2024) and the market measures cited in C01 (AFME’s ESG premium between 1.27 bp at end-March and 0.44 bp at end-June 2026) put the secondary greenium at zero to a few basis points. Stage 5.2 shows how fragile the estimate is.
A common prejudice on trading floors is that green bonds are less liquid because they are held by buy-and-hold investors. The evidence does not support it as a general rule. A December 2023 ECMI working paper by Molino, Prosperi and Zicchino on bonds from 2009 to 2022 found green bonds’ bid-ask spreads about 4 bp tighter than comparable conventional bonds of the same issuers on average, driven by sovereigns and supranationals; for corporates the effect was not significant overall, but was present for certified bonds, and the authors found a weakly significant larger premium for corporate green bonds in illiquid periods. A March 2026 working paper by Dulak and Wolff, using Euroclear settlement data from 2020 to 2025, found no systematic liquidity disadvantage for green bonds, higher trading volumes driven by more frequent trading, and differences that persisted in periods of market stress. Neither study proves causality, and both depend on proxies. The honest reading for a trader: do not price green bonds as less liquid by default; look at the specific bond’s float, borrowability and the dealer’s own axes.
iTraxx MSCI ESG Screened Europe. Launched in 2020 and cleared at LCH CDSClear from September 2020, it starts from the iTraxx Europe constituents and removes names that fail three MSCI screens: business-involvement exclusions (controversial and nuclear weapons, tobacco, thermal coal above a revenue threshold, unconventional oil and gas, and others), a minimum MSCI ESG rating, and the most severe controversy flag. Survivors are equally weighted; the index rolls with Main every March and September. Series 45 (effective 20 March 2026) had 92 names, a 1% coupon and a five-year maturity of 20 June 2031. Its liquidity is a small fraction of Main’s, so dealers who sell it to clients usually hedge with Main and carry the difference as basis risk.
ESG bond index futures. Eurex launched futures on the Bloomberg MSCI Euro Corporate SRI index and the Bloomberg MSCI Global Green Bond index in September 2021. They allow investors to gain or hedge exposure to an ESG credit benchmark in one trade; like other credit index futures they remain a niche next to government bond futures, and a dealer quoting them must be able to hedge the basket.
Total return swaps (TRS). A TRS pays the receiver the total return of a reference (a bond, a basket, an index such as an iBoxx MSCI ESG index or a green bond index) against a floating rate plus a spread. The dealer hedges by holding the reference bonds. For a green fund that cannot easily buy the cash bonds, a TRS on a green bond index is a way to get exposure; for the bank, it is an inventory held on behalf of the client, with the funding, capital and the label risk that come with it.
PITFALL — An ESG-screened index is not a climate hedge. It excludes some high-carbon names, so it moves less than Main when those names sell off; a desk that hedges a brown issuer’s bonds with the screened index is under-hedged exactly when transition risk materialises. Stage 5.5 quantifies this.
Do spreads price climate risk? Roncalli (Handbook, section 13.1.1) reviews the structural approach (Merton): a carbon price reduces the firm’s cash flow and its asset value, which lowers the distance to default and raises the spread. Empirical studies reviewed there find that higher carbon intensity is associated with a lower distance to default (Capasso and co-authors, 2020, on European firms) and that disclosure of transition risk is associated with wider CDS spreads, while disclosure of physical risk is not (Kölbel and co-authors, 2024). Sector membership dominates: utilities, materials and energy carry most of the exposure. For the trader, the practical consequences are three. Spread betas of high-carbon issuers to policy news are larger than their betas to the index; transition-risk events (a tighter carbon price path, the loss of free emission allowances, a lost subsidy) are idiosyncratic jumps that an index hedge does not cover; and the market’s pricing of climate risk varies over time, so a “climate premium” estimated in one period is not a constant.
Central banks are adding their weight. The Eurosystem decided on 29 July 2025 to introduce a climate factor in its collateral framework from the second half of 2026: the value assigned to marketable assets of non-financial corporations pledged as collateral can be reduced according to their exposure to transition risk, calibrated on sector data from the 2024 climate stress test, the issuer’s climate score and the residual maturity. A bond that is worth less as collateral is worth slightly less to banks and to the dealers who finance inventory through it.
Transparency. The 2024 review of MiFIR (Regulation (EU) 2024/791) rebuilt bond transparency. Since 2 March 2026 the amended RTS 2 (Delegated Regulation (EU) 2017/583 as amended) sets post-trade deferrals by trade size and bond liquidity, from 15 minutes for normal trades in liquid bonds to four weeks for the largest trades in less liquid ones, replacing a patchwork of national deferral regimes. A consolidated tape for bonds is coming: ESMA selected fairCT (Ediphy) as the EU’s bond tape provider in 2025; its authorisation was still in progress in April 2026 and market reports in mid-2026 pointed to a launch in early 2027. More transparency helps clients compare prices and makes dealers’ large trades visible sooner, which raises their cost of holding inventory.
Fund demand. The SFDR revision (Commission proposal of November 2025) would replace “Article 8 and 9” with three categories, Transition, ESG Basics and Sustainable, each requiring at least 70% of investments to meet the category’s criteria. The Council adopted its negotiating mandate on 24 June 2026, Parliament’s ECON committee its report on 10 September 2026 and the plenary its mandate on 5 October 2026; trilogues began in October 2026 without agreement by mid-November, and application is expected around 2029. The categories will shape which bonds funds must hold or exclude, and therefore the secondary flows.
Conduct. When sales describe a bond or a product as green, MiFID II’s fair, clear and not misleading standard and the product governance rules apply, and for advised clients the sustainability-preference rules. ESMA, EBA and EIOPA’s 2024 greenwashing reports flag secondary-market claims (index names, “ESG” labels on derivatives) as a risk area.
Prudential. Market-risk capital does not distinguish green from conventional. In the EU the FRTB applies from 1 January 2027, after two postponements; on 4 June 2026 the Commission adopted a delegated regulation that, among other relief, applies a multiplier scaling FRTB requirements back towards pre-FRTB levels until 31 December 2029. Trading-book exposures are excluded from the current Green Asset Ratio, and the Taxonomy KPIs for the trading book and for fees and commissions were postponed to 2028 by Delegated Regulation (EU) 2026/73 (see A02).
A secondary trade has no mandate letter and no prospectus, but it has a lifecycle as controlled as a primary deal, compressed into seconds for the trade itself and into days for its consequences. The flow below follows one inquiry: a green bond fund wants to sell EUR 25m of Vantoria’s green 2031 bond.
INTERACTIVE LAB — A secondary trade from inquiry to capital: who does what
Follow a client sale of a green bond through quoting, booking, hedging, financing, reporting, control and capital.
| # | Who | Does what | Deliverable |
|---|---|---|---|
| 1 | Credit sales | Receives the client’s inquiry by RFQ or chat and checks the client’s classification and any order instructions | Inquiry logged in the order management system with timestamp |
| 2 | Flow trader (chemicals book) | Checks inventory, axes, CS01 and JTD headroom, the issuer’s curve and the bond’s label and index flags | Price decision: bid spread and size |
| 3 | Electronic market-making algorithm or trader | Answers the RFQ against competing dealers; for small sizes the algorithm prices within limits set by the trader | Quote sent; trade won or lost, with cover price |
| 4 | Middle office | Captures the trade, enriches static data, sends the confirmation and books it in the risk system | Booked trade and confirmation |
| 5 | Regulatory reporting (APA and ARM) | Publishes the trade post-trade within the applicable deferral and reports the transaction to the regulator by the next working day | Post-trade publication and MiFIR transaction report |
| 6 | Flow trader | Hedges rate risk with Bobl futures and part of the credit risk with iTraxx Main protection | Hedge tickets and updated risk |
| 7 | Repo desk | Finances the long position in repo, or borrows bonds for any short | Repo ticket under the GMRA; borrow fee |
| 8 | Settlement operations | Settles delivery versus payment two business days after trade date at Euroclear or Clearstream and chases fails | Settled position; CSDR penalty reports if any |
| 9 | Product control | Marks the book to independent prices, verifies prices and explains the day’s P&L | Daily P&L and P&L explain; independent price verification report |
| 10 | Market risk | Monitors CS01, JTD, stress and expected-shortfall limits and escalates breaches | Daily limit report; escalation memo if breached |
| 11 | Sales and trader | Advertise the axe to green-mandated and conventional clients and resell the bonds | Axe list and offsetting trades |
| 12 | Compliance and sustainable finance team | Review any client communication describing the bonds or the axe as green or ESG | Approved wording or correction |
| 13 | Finance and regulatory capital | Compute FRTB own funds, the desk’s RWA and its allocated capital; feed Pillar 3 disclosures | Capital report and desk RAROC |
Inquiry and pre-trade checks. Sales receives the inquiry and checks that the client is set up (classification as professional client, legal entity identifier, settlement instructions). The trader’s decision is a quick mental model: where the issuer’s curve is, what the bond traded at last, the bond’s free float, whether the desk wants it (an axe to buy is a gift), the CS01 and JTD headroom on the chemicals book, and the bond’s flags: green label, under which standard, in or out of the main ESG indices. The last point is no longer a detail: a seller of a green bond may be a fund forced out by an index rule, and the buyers will be the green funds that are still allowed to hold it.
Quote and execution. On a platform the client sends the RFQ to several dealers and has a short time to execute. The winner is the best price; most platforms tell the losers the cover. Best execution under MiFID II is the client’s obligation towards its own clients and the dealer’s when it executes orders on behalf of a client; in RFQ trading the dealer is usually dealing on its own account, but its quote must not exploit inside information and its records must allow the trade to be reconstructed.
Booking and reporting. Middle office captures the trade in the front-office system, which feeds risk, P&L and settlement. Two regulatory outputs follow. Post-trade transparency: the trade is published through an approved publication arrangement (APA), immediately or after the deferral allowed for its size and liquidity under the amended RTS 2 applicable since 2 March 2026. Transaction reporting: the full details, including the client’s identity, go to the national regulator through an approved reporting mechanism by the end of the next working day, for market-abuse surveillance.
Hedging and financing. The trader neutralises the rate risk with futures and decides whether to hedge credit. The bonds are financed in repo: the bank sells them to a repo counterparty against cash and buys them back later, with a haircut that it must fund from its own resources at the Treasury’s transfer price. Short positions require borrowing the bond; a green bond held mostly by buy-and-hold funds can be expensive or impossible to borrow.
Settlement. Delivery versus payment two business days after trade date (the EU plans to move to one day in October 2027). Under the CSDR settlement discipline regime, a failed settlement incurs daily cash penalties.
Control. Every evening product control marks each bond to independent sources (composite prices from data vendors, platform trades, the tape when it arrives) and performs independent price verification monthly with stricter tolerances; illiquid bonds that cannot be verified attract prudent valuation adjustments deducted from capital. Market risk checks limits and runs stress tests. The P&L explain is the most useful control on a trading desk: a trader who repeatedly makes money that cannot be explained by risk factors is either very good or mis-marking.
Resale and communication. The bonds bought become an axe. Sales may tell green-mandated clients “we have Vantoria green 2031 to sell”. That sentence is a communication about a sustainability characteristic: the compliance and sustainable finance teams set the rules for what sales may say (the label and the standard, yes; “this bond reduces emissions by X”, only with the issuer’s published impact report as source).
Capital and reporting. Finance computes the market-risk own funds each day (each quarter for disclosure), allocates capital to the desk and reports RWA in Pillar 3. Section 6 does this for the case.
The case follows Meridian’s euro investment-grade flow desk through one week in which credit spreads widen sharply. The book is owned by two traders: the utilities trader, who makes markets in Helios Renewables (the issuer of C01, rated BBB, sector utilities) and the chemicals trader, who makes markets in Vantoria Chemicals, a fictional European chemicals producer rated BBB, carbon-intensive (steam crackers and ammonia), which issued a green bond in 2026 to finance the electrification of two crackers and a heat-pump programme. Both issuers have green and conventional bonds outstanding. All positions, spreads, betas and market moves are case assumptions chosen to be realistic for November 2026; they are not market data.
Question. What does the desk own, how much rate and credit risk does it carry, and how is that risk hedged?
Data. Swaps are flat at 2.50%. All bonds pay annual coupons and are valued on a coupon date (no accrued interest), so the clean price is the full price. Spreads are mid spreads over mid-swaps.
| Bond | Maturity (years) | Coupon | Spread (bp) | Position (EUR m) |
|---|---|---|---|---|
| Helios 2032 | 6 | 3.50% | 124 | +20 |
| Helios green 2034 | 8 | 3.95% | 137 | +40 |
| Helios 2036 | 10 | 4.00% | 156 | −10 |
| Vantoria green 2031 | 5 | 4.00% | 150 | +25 |
| Vantoria 2033 | 7 | 4.375% | 178 | +15 |
The Helios green 2034 is the C01 bond, issued at mid-swaps plus 147 bp. In the weeks after issue its spread has tightened by the 10 bp of new-issue premium and it now trades at plus 137 bp. The desk is long EUR 40m of it because it bought from investors who had been over-allocated and wanted to take profit. It is short EUR 10m of the Helios 2036, borrowed in repo, a curve hedge left over from a switch trade. Rate hedges: the trader sells Bobl futures against bonds of up to 7 years and Bund futures against the longer bonds; the DV01 per contract is assumed to be EUR 50 for the Bobl and EUR 100 for the Bund. Credit hedge policy (case assumption): hedge 60% of the beta-weighted CS01 with iTraxx Main five-year protection at 55 bp; spread betas to Main estimated over the last year are 0.8 for Helios and 1.4 for Vantoria.
Calculation. Each bond’s yield is the swap rate plus the spread; price and modified duration follow from the coupon and maturity; the CS01 is position times price times duration times 0.0001. For a bullet bond the DV01 equals the CS01.
FORMULA
is the nominal held (negative if short); the price per 100; the modified duration; the hedge ratio set by policy; the spread beta of issuer to iTraxx Main; the index CS01 per EUR 1m comes from the CDS lab of section 2.3 (EUR 30,305 for EUR 65m, so EUR 466 per million).
INTERACTIVE LAB — The Monday book: prices, CS01, rate hedge and index hedge
Rebuild the risk of the five positions and size the futures and iTraxx hedges from the policy hedge ratio and the spread betas.
| Parameter you can move | Range | Default |
|---|---|---|
| Mid-swap rate (flat) | 1 to 4 % | 2.5 % |
| Spread beta of Helios to iTraxx Main | 0.3 to 2 | 0.8 |
| Spread beta of Vantoria to iTraxx Main | 0.3 to 2.5 | 1.4 |
| Hedge ratio on beta-weighted CS01 | 0 to 100 % | 60 % |
| iTraxx Main five-year spread | 30 to 150 bp | 55 bp |
| DV01 of one Bobl future | 30 to 70 EUR | 50 EUR |
| DV01 of one Bund future | 60 to 140 EUR | 100 EUR |
| Result at default settings | Value |
|---|---|
| Price of Helios green 2034 | 100.542 per 100 |
| Modified duration of Helios green 2034 | 6.755 years |
| CS01 Helios 2032 | 10,491 EUR per bp |
| CS01 Helios green 2034 | 27,166 EUR per bp |
| CS01 Helios 2036 (short) | -8,063 EUR per bp |
| CS01 Vantoria green 2031 | 11,130 EUR per bp |
| CS01 Vantoria 2033 | 8,942 EUR per bp |
| Net CS01 Helios | 29,594 EUR per bp |
| Net CS01 Vantoria | 20,071 EUR per bp |
| Net CS01 of the book (= DV01) | 49,665 EUR per bp |
| Bobl futures to sell | 611 contracts |
| Bund futures to sell | 191 contracts |
| Beta-weighted CS01 | 51,775 EUR per bp |
| Index CS01 per EUR 1m of protection | 466.2 EUR per bp |
| iTraxx Main protection to buy | 66.6 EUR m |
| Annual spread carry of the bonds | 1,282.0 EUR k |
Try it. Raise Vantoria’s beta to 2.0: a carbon-intensive issuer whose spread reacts strongly to policy news needs a much larger index hedge, and the hedge becomes mostly a bet on the index.
Result. The book is long credit: a net CS01 of EUR 49,665 per bp, of which EUR 29,594 on Helios and EUR 20,071 on Vantoria. The Helios green 2034 alone is EUR 27,166, the largest line. The rate risk equals the CS01 and is neutralised by selling about 611 Bobl and 191 Bund futures. The beta-weighted CS01 is EUR 51,775; at the 60% policy and EUR 466 of CS01 per million of index protection, the trader should buy EUR 66.6m of iTraxx Main protection and buys EUR 65m, the nearest round amount. The bonds earn EUR 1.28m a year of spread over swaps on their nominal.
Reading. The green bond is not special in the risk system: it is a line of CS01 in the Helios curve, hedged with the same futures and the same index. What is special is how it was acquired (from over-allocated primary investors) and who will buy it back (green-mandated funds). The hedge leaves 40% of the beta-weighted credit risk open and all of the idiosyncratic risk: the desk is paid, through carry, to hold it.
WHO DOES WHAT — At 7:30 on Monday the utilities trader reads the risk report aloud in the morning meeting: “Helios, plus 29.6k CS01, of which 27k in the green 34s; axed to sell the green 34s, axed to buy back the 36s I am short.” Sales note the axes and call three green bond funds that missed the new issue. The chemicals trader adds that Vantoria’s CS01 is 20k against a limit of 25k, so he will not buy more Vantoria without selling something first. The head of desk checks the index hedge against the policy and asks product control to confirm that the green 34s are marked on trades, not on the issue spread.
Question. A client, a green bond fund, asks the trader whether the Helios green 2034 is “expensive” against Helios’s conventional curve. What is the secondary greenium, and how reliable is the number?
Data. Helios’s conventional bonds trade at mid spreads of 95 bp (3 years), 124 bp (6 years), 156 bp (10 years) and 163 bp (12 years); the green 2034 (8 years) at 137 bp. For Vantoria, conventional bonds trade at 140 bp (4 years) and 178 bp (7 years); the green 2031 (5 years) at 150 bp.
Calculation. Three methods, all common on desks and in the academic literature reviewed by Roncalli (bottom-up matching):
FORMULA
is the secondary greenium in bp; the green bond’s mid spread; the conventional spread estimated at the green bond’s maturity by the chosen method. Negative means the green bond trades tighter (richer) than its conventional curve.
INTERACTIVE LAB — One green bond, three greeniums: matched pair and curve fits
Estimate the secondary greenium of the Helios green 2034 by interpolation, linear fit and log fit, and see how the method and one data point move the answer.
| Parameter you can move | Range | Default |
|---|---|---|
| Helios conventional 3-year spread | 60 to 140 bp | 95 bp |
| Helios conventional 6-year spread | 90 to 170 bp | 124 bp |
| Helios conventional 10-year spread | 110 to 210 bp | 156 bp |
| Helios conventional 12-year spread | 120 to 220 bp | 163 bp |
| Helios green 2034 (8-year) spread | 100 to 200 bp | 137 bp |
| Vantoria conventional 4-year spread | 80 to 260 bp | 140 bp |
| Vantoria conventional 7-year spread | 100 to 300 bp | 178 bp |
| Vantoria green 2031 (5-year) spread | 80 to 280 bp | 150 bp |
| Result at default settings | Value |
|---|---|
| Helios conventional at 8 years, interpolated | 140.00 bp |
| Helios greenium, matched pair | -3.00 bp |
| Helios greenium, linear fit | 0.57 bp |
| Helios greenium, log fit | -5.54 bp |
| Average of the three Helios estimates | -2.65 bp |
| Range between highest and lowest Helios estimate | 6.11 bp |
| Vantoria greenium, matched pair | -2.67 bp |
| Value of the matched-pair Helios greenium on EUR 40m | 81,497 EUR |
Try it. Move the 12-year spread from 163 to 175 bp: the matched pair does not change, but both fitted greeniums move by several basis points. One illiquid long bond can create or erase a greenium.
Result. The matched pair gives a synthetic conventional spread of 140 bp and a greenium of −3.0 bp. The linear fit gives +0.6 bp (the green bond looks cheap), the log fit −5.5 bp. The average is −2.7 bp and the range between methods is 6.1 bp, twice the central estimate. On the desk’s EUR 40m, 3 bp is worth about EUR 81,500 of price. Vantoria’s green 2031 trades 2.7 bp inside its interpolated conventional curve.
Reading. The three methods agree that the greenium is small; they disagree about its sign. The linear fit is wrong for a concave curve: a straight line through 3- to 12-year points lies below the true curve in the middle, so the green bond looks wide. The log fit bends the other way. The matched pair is the most transparent but relies on two bonds only. Each method also ignores what the academic literature corrects for: coupon and price (the green bond’s 3.95% coupon at a price above par is less attractive to some investors than a lower-coupon bond below par), liquidity and age (the green bond is a recent benchmark; older bonds trade wider), and bid versus mid (marks taken on different sides). The trader’s answer to the fund is therefore “about 3 bp rich, give or take 3 bp”, which is the same message as C01’s at issue: real, small, and not a reason to buy or sell on its own.
PITFALL — Quoting a greenium to a client from a single method without the range is a misleading claim in both directions: it can sell a green bond as “cheap” when it is not, or justify a client’s exclusion of conventional bonds on spurious grounds. State the method, the comparison bonds, the date and the side of the market.
Question. Over five trading days credit sells off. What does the book lose, unhedged and hedged, and does hedging with the ESG-screened index instead of Main change the answer?
Data (case assumptions). A global risk-off move combines with news specific to Vantoria’s sector: the publication of tighter EU ETS benchmarks for steam crackers, which raises analysts’ estimates of the carbon cost of European petrochemical producers. iTraxx Main widens by 30 bp, from 55 to 85 bp; the iTraxx MSCI ESG Screened index, which excludes some of the most affected names, widens by 27 bp, from 53 to 80 bp. Bond spreads widen as follows, consistent with each issuer’s beta plus Vantoria’s sector news: Helios 2032 +22 bp, Helios green 2034 +22 bp, Helios 2036 +26 bp; Vantoria 2030 +52 bp, Vantoria green 2031 +48 bp, Vantoria 2033 +56 bp. Rates rally but the futures hedge is assumed to neutralise rate moves exactly (we ignore swap-spread and curve basis). The desk holds EUR 65m of Main protection. Carry over five days is ignored.
Calculation. Each bond is revalued at its new spread; the index protection is revalued with the upfront formula of section 2.3 at the new spread. The ESG alternative assumes the same EUR 65m notional on the screened index (same coupon and maturity convention).
INTERACTIVE LAB — The spread-widening week: book P&L unhedged, hedged with Main, hedged with the ESG-screened index
Revalue the five bonds and the index protection under the episode, scaled by the size of the index move, and measure the basis risk of hedging with an ESG-screened index.
| Parameter you can move | Range | Default |
|---|---|---|
| iTraxx Main widening | 0 to 60 bp | 30 bp |
| Index protection held | 0 to 150 EUR m | 65 EUR m |
| ESG-screened index move as share of Main move | 50 to 120 % | 90 % |
| Vantoria sector add-on (scales Vantoria’s move above beta) | 0 to 200 % | 100 % |
| Result at default settings | Value |
|---|---|
| P&L Helios 2032 | -229.1 EUR k |
| P&L Helios green 2034 | -592.2 EUR k |
| P&L Helios 2036 (short) | 206.9 EUR k |
| P&L Vantoria green 2031 | -527.1 EUR k |
| P&L Vantoria 2033 | -490.6 EUR k |
| P&L of the bonds (unhedged) | -1,632.1 EUR k |
| P&L of the Main protection | 898.1 EUR k |
| Net P&L hedged with Main | -734.0 EUR k |
| P&L of the ESG-screened protection | 810.7 EUR k |
| Net P&L hedged with the ESG-screened index | -821.5 EUR k |
| Extra loss from hedging with the ESG index | 87.4 EUR k |
| Helios greenium after the week (matched pair) | -5.00 bp |
| Vantoria greenium after the week (matched pair) | -8.00 bp |
| Loss avoided because green bonds widened less than their curves | 113.7 EUR k |
Try it. Set the Vantoria sector add-on to 0%: the episode becomes a pure market move and the hedged loss falls to about EUR 0.55m. The difference is the transition-risk jump that no index covers.
Result. Unhedged, the bonds lose EUR 1.63m in the week: EUR 0.61m on Helios (the short 2036 gains EUR 0.21m) and EUR 1.02m on Vantoria, although Vantoria is the smaller position. The Main protection gains EUR 0.90m, so the hedged book loses EUR 0.73m. Had the desk hedged with the same notional of ESG-screened protection, the hedge would have gained EUR 0.81m and the book would have lost EUR 0.82m: the “green” hedge costs EUR 87,000 more, because the index that excludes the most exposed names widened less. The green bonds widened less than their issuers’ conventional curves: the Helios matched-pair greenium moved from −3 to −5 bp and Vantoria’s from −2.7 to −8 bp. Had they moved with their curves, the book would have lost EUR 114,000 more.
Reading. Three lessons. First, the hedge did its job on the systematic part: the index protection offset more than half of the bonds’ loss. Two things remain. One is the 40% of beta-weighted risk that the policy leaves open, which the desk is paid to carry. The other is Vantoria’s idiosyncratic move, a transition-risk event (a regulatory change raising the carbon cost of one sector) that no index hedge covers: without the sector news the hedged loss would have been EUR 0.55m, so the news alone cost about EUR 0.19m. Section 3.4 said that high-carbon issuers carry jump risk to policy news; this is what it looks like in a P&L. Second, the ESG-screened index is the wrong tool for that risk: it hedges less precisely the names whose risk is climate-related. A trader who sells “ESG hedges” to clients should explain that basis. Third, the green bonds outperformed. This is consistent with a sticky, dedicated holder base and with the weak evidence of a larger liquidity premium for corporate green bonds in stress (ECMI, 2023), but it is one week of one case. It could go the other way: if green funds faced outflows, they would have to sell what they hold, which is green bonds. The trader treats the outperformance as a reason to keep green inventory in stress, not as a law.
WHO DOES WHAT — On Tuesday of that week market risk’s daily report shows the chemicals book above its stress-loss limit. The head of credit trading signs a temporary limit increase for 48 hours with a plan: sell half the Vantoria inventory, keep the Helios positions and the index hedge. The plan, the limit excess and its approval are recorded and reported to the market risk committee.
Question. Selling half the Vantoria inventory (EUR 12.5m of the green 2031 and EUR 7.5m of the 2033) at the end of the week: what does it cost in bid-offer, how many days does it take, and how much risk remains while the desk sells?
Data (case assumptions). Spreads at the end of the week: Vantoria green 2031 at 198 bp, Vantoria 2033 at 234 bp. Normal quoted bid-offer: 5 bp for both bonds. Stressed bid-offer: 12 bp for the green bond, which has natural buyers among green funds, and 15 bp for the conventional bond. Average daily market volume: EUR 25m and EUR 30m in normal times, EUR 10m and EUR 12m in the stressed week. The desk does not want to be more than 25% of the market’s volume (participation rate). Daily spread volatility in stress: 8 bp.
Calculation. The exit cost is half the bid-offer times the CS01 of the amount sold (the desk sells at the bid, while the book is marked at mid). The number of days is the amount divided by the participation rate times the daily volume. While the desk sells in equal daily slices, the position declines linearly, so the variance of the P&L over days is about one third of that of the full position held for days.
FORMULA
is the exit cost in euros; the bid-offer in bp; the CS01 of the amount to sell; the number of days to sell; the amount to sell; the participation rate; the market’s daily volume; the one-standard-deviation P&L during the exit; the daily spread volatility.
INTERACTIVE LAB — Exit cost and time to sell: green and conventional, normal and stressed
Compare the bid-offer cost, the days needed and the risk carried while liquidating half the Vantoria inventory, in normal and stressed markets.
| Parameter you can move | Range | Default |
|---|---|---|
| Share of the Vantoria inventory to sell | 10 to 100 % | 50 % |
| Stressed bid-offer, green 2031 | 3 to 40 bp | 12 bp |
| Stressed bid-offer, conventional 2033 | 3 to 40 bp | 15 bp |
| Normal bid-offer, both bonds | 2 to 15 bp | 5 bp |
| Stressed daily volume, green 2031 | 2 to 40 EUR m | 10 EUR m |
| Stressed daily volume, conventional 2033 | 2 to 40 EUR m | 12 EUR m |
| Participation rate | 5 to 50 % | 25 % |
| Daily spread volatility in stress | 2 to 20 bp | 8 bp |
| Result at default settings | Value |
|---|---|
| Green 2031 to sell | 12.50 EUR m |
| Conventional 2033 to sell | 7.50 EUR m |
| Exit cost, green, stressed | 32.5 EUR k |
| Exit cost, conventional, stressed | 32.2 EUR k |
| Exit cost, both bonds, normal bid-offer | 24.3 EUR k |
| Exit cost, both bonds, stressed | 64.7 EUR k |
| Days to sell, green, stressed | 5.00 days |
| Days to sell, conventional, stressed | 2.50 days |
| Days to sell, green, normal volume | 2.00 days |
| One-sigma P&L during the stressed exit (both bonds) | 85.3 EUR k |
Try it. Set both stressed bid-offers to 15 bp and the green bond’s volume to 6m: if green funds stop buying, the green bond becomes the harder one to sell.
Result. Selling EUR 12.5m of the green 2031 at a 12 bp bid-offer costs EUR 32,500; selling EUR 7.5m of the conventional 2033 at 15 bp costs EUR 32,200: together EUR 64,700, against EUR 24,300 at normal bid-offers, about 2.7 times more. At 25% of a stressed daily volume of EUR 10m, the green bond takes 5 days to sell (2 days at normal volumes); the conventional bond 2.5 days. While selling, the desk carries a one-sigma P&L of about EUR 85,000.
Reading. In this case the green bond is cheaper per euro of risk to sell (narrower bid-offer) but slower (smaller volume, because a large share is locked in buy-and-hold portfolios). That combination, more stable price but thinner flow, is what the 2023 and 2026 studies describe in different words. The liquidity cost is small compared with the week’s P&L (EUR 65,000 against a EUR 0.73m loss), which is typical: in investment grade, the main risk is the spread move, not the bid-offer. It is not small compared with the desk’s annual revenue on these bonds (stage 5.6), and it is the reason why the FRTB gives longer liquidity horizons to credit spread risk than to rates. The decision to sell at the bottom is a risk-management choice that the head of desk documents: the alternative, keeping the position and the excess, would have needed the market risk committee’s approval.
Question. From January 2027, what own funds does the Monday book require under the FRTB standardised approach, and what does the index hedge save?
Data. The sensitivities-based method (SBM) for credit spread risk on non-securitisations (BCBS MAR21, transposed in the CRR as amended by CRR3) assigns each position to a bucket: Helios to bucket 7 (investment grade, health care, utilities and professional services, risk weight 1.5%), Vantoria to bucket 4 (investment grade, basic materials, energy and industrials, 3.0%), the iTraxx Main protection to bucket 17 (investment-grade indices, 1.5%), treated as one position without look-through. Sensitivities are measured at the vertices 0.5, 1, 3, 5 and 10 years; we allocate each bond’s CS01 to the two vertices around its maturity in proportion to distance. Within a bucket the correlation between two sensitivities of the same issuer at different vertices is 65%. Across buckets the correlation is 5% between buckets 4 and 7 and 45% between bucket 17 and any sector bucket. The capital is the largest of three correlation scenarios: medium (as given), high (correlations times 1.25, capped at 100%) and low (the larger of twice the correlation minus 100% and 75% of it). The default risk charge (DRC) uses a loss given default of 75% for senior debt, a risk weight of 6% for BBB, and, for the index decomposed into its 125 constituents (neither Helios nor Vantoria is a member), an average risk weight of 4.8%, assuming 40% of names rated A (3%) and 60% BBB (6%). Long and short jump-to-default amounts net only for the same obligor; the remaining shorts are recognised through the hedge benefit ratio. GIRR (rates) is ignored because the futures hedge leaves only a small curve residual. We also leave aside the residual risk add-on and the EU’s temporary multiplier (section 6.1).
FORMULA
is the sensitivity at vertex , the change in value for a 1 bp spread increase divided by 0.0001 (negative for a long bond); the bucket’s risk weight; the weighted sensitivity; the within-bucket correlation; the bucket’s capital; the sum of the bucket’s weighted sensitivities; the cross-bucket correlation; the delta capital for credit spread risk.
is the jump-to-default amount, nominal plus the position’s mark-to-market gain or loss relative to par; the hedge benefit ratio, which reduces the recognition of short positions when they hedge long ones of other names.
INTERACTIVE LAB — FRTB standardised capital of the Monday book: credit spread delta and default risk
Compute the weighted sensitivities, bucket capitals, correlation scenarios and default risk charge of the five bonds and the index hedge, and see what the hedge saves.
| Parameter you can move | Range | Default |
|---|---|---|
| iTraxx Main protection held | 0 to 150 EUR m | 65 EUR m |
| Risk weight bucket 7 (Helios) | 0.5 to 5 % | 1.5 % |
| Risk weight bucket 4 (Vantoria) | 0.5 to 6 % | 3 % |
| Risk weight bucket 17 (IG index) | 0.5 to 5 % | 1.5 % |
| DRC risk weight of a BBB issuer | 3 to 15 % | 6 % |
| Average DRC risk weight of index names | 2 to 10 % | 4.8 % |
| Result at default settings | Value |
|---|---|
| Weighted sensitivity Helios, 5-year vertex | -2.889 EUR m |
| Weighted sensitivity Helios, 10-year vertex | -1.550 EUR m |
| Weighted sensitivity Vantoria, 5-year vertex | -4.948 EUR m |
| Weighted sensitivity Vantoria, 10-year vertex | -1.073 EUR m |
| Weighted sensitivity iTraxx protection | 4.546 EUR m |
| Bucket capital Helios | 4.071 EUR m |
| Bucket capital Vantoria | 5.704 EUR m |
| Delta capital, medium correlations | 5.445 EUR m |
| Delta capital, high correlations | 4.775 EUR m |
| Delta capital, low correlations | 6.042 EUR m |
| CSR delta capital (largest scenario) | 6.042 EUR m |
| Net long JTD (Helios + Vantoria) | 67.596 EUR m |
| Hedge benefit ratio | 58.1 % |
| Default risk charge | 2.696 EUR m |
| Own funds requirement (CSR + DRC) | 8.738 EUR m |
| RWA (12.5 x own funds) | 109.22 EUR m |
| Allocated capital (13% of RWA) | 14.20 EUR m |
| Own funds without the index hedge | 11.515 EUR m |
Try it. Set the index protection to zero, then to EUR 130m: capital first falls with the hedge, then rises again, because an over-hedge is itself a risk position in bucket 17.
Result. The weighted sensitivities are EUR −2.89m and −1.55m for Helios at the 5- and 10-year vertices, EUR −4.95m and −1.07m for Vantoria, and EUR +4.55m for the index protection. The Helios bucket requires EUR 4.07m and the Vantoria bucket EUR 5.70m, more than Helios on a smaller CS01 because its risk weight is twice as high. Across buckets the low-correlation scenario binds: CSR delta capital EUR 6.04m (EUR 5.45m in the medium scenario). The net long jump-to-default is EUR 67.6m; with a hedge benefit ratio of 58.1% the DRC is EUR 2.70m. Total own funds EUR 8.74m, RWA EUR 109.2m, allocated capital at 13% EUR 14.2m. Without the index hedge the own funds would be EUR 11.51m.
Reading. The book of EUR 110m of bonds consumes about EUR 14m of Meridian’s capital, 13% of its nominal, which is far more than the same bonds would consume as loans to investment-grade borrowers under the F-IRB (see B03). The standardised FRTB is deliberately conservative for credit spread risk: a 1.5% risk weight on a duration-7 bond is a 150 bp shock and a loss of about 10% of value. The low-correlation scenario binds because the hedge and the bonds offset: when the regulator assumes they are less correlated, the hedge is worth less. Nothing in the calculation looks at the green label. A green bond and a conventional bond of the same issuer and maturity have the same weighted sensitivity and the same jump-to-default; the 2.7 bp greenium changes the sensitivity by a fraction of a percent.
PITFALL — The vertex allocation used here (splitting each bond’s CS01 between the 5- and 10-year vertices by maturity) is a teaching shortcut. Banks compute sensitivities by bumping each vertex of the issuer’s credit curve separately, which gives slightly different numbers, especially for bonds with large coupons. The order of magnitude, and the conclusion that the label plays no role, do not change.
Question. Is the book worth its capital? What return does Meridian earn on the inventory and the client flow it supports, and what does the green label change?
Data (case assumptions, annualised). The book supports EUR 1.2bn a year of client trades in Helios and Vantoria bonds, on which the desk captures on average 10 cents per 100 of nominal after hedging. The inventory is the Monday book held on average through the year. Bonds are financed in repo at a rate equal to the swap rate, so they earn their spread over swaps on their nominal; the repo haircut of 5% on the EUR 100m of long positions is funded by Treasury at the 30 bp FTP liquidity premium; the EUR 10m short costs a 25 bp borrowing fee. The index protection costs its spread, 55 bp a year, on EUR 65m. Cost-income ratio for markets revenues: 60%. Expected loss: a one-year PD of 0.20% (BBB) for both issuers, a market loss given default of 60% (the 40% recovery convention of CDS), on the EUR 90m net long nominal. Tax 25%, capital 13% of RWA. The own funds come from stage 5.5. Ancillary revenue (DCM fees from the issuers) is set to zero in the stand-alone view, then added with EUR 2.5 of RWA per euro of revenue and a 60% cost ratio.
INTERACTIVE LAB — Revenue, cost, expected loss, capital and RAROC of the Helios–Vantoria book
Build the desk’s annual P&L line by line, compute RAROC, test the effect of inventory size, of the EU capital multiplier and of the green label, and find the ancillary revenue needed to reach the hurdle.
| Parameter you can move | Range | Default |
|---|---|---|
| Annual client volume | 200 to 3000 EUR m | 1200 EUR m |
| Capture per 100 traded | 2 to 25 cents | 10 cents |
| Average inventory as share of the Monday book | 25 to 150 % | 100 % |
| Own funds scaling (EU temporary multiplier) | 30 to 100 % | 100 % |
| FRTB own funds of the Monday book (stage 5.5) | 2 to 20 EUR m | 8.7379 EUR m |
| Ancillary revenue from the issuers | 0 to 6000 EUR k | 0 EUR k |
| Inventory in green bonds | Yes (case) / No: conventional bonds of same maturity | Yes (case) |
| Result at default settings | Value |
|---|---|
| Flow revenue (bid-offer capture) | 1,200.0 EUR k |
| Spread carry on the bonds | 1,282.0 EUR k |
| Cost of index protection | 357.5 EUR k |
| Haircut funding and borrow fees | 40.0 EUR k |
| Revenues | 2,084.5 EUR k |
| Operating costs (60%) | 1,250.7 EUR k |
| Expected loss | 108.0 EUR k |
| Pre-tax result | 725.8 EUR k |
| Tax at 25% | 181.45 EUR k |
| Net result after 25% tax | 544.35 EUR k |
| RWA | 109.22 EUR m |
| Allocated capital | 14,199 EUR k |
| RAROC | 3.83 % |
| Carry given up to the greenium | 18.67 EUR k |
| Ancillary revenue needed for 12% | 4,443 EUR k |
Try it. Halve the inventory: RAROC rises to about 6.4% because the flow revenue does not need all of the inventory. Back at 100%, set the multiplier to 50% and add EUR 2,000k of ancillary revenue: RAROC reaches about 14.8%. The franchise, not the book, clears the hurdle.
Result. Line by line, in EUR thousand a year: flow revenue 1,200; spread carry on the bonds 1,282; index protection −357.5; haircut funding and borrow fees −40; revenues 2,084.5. Operating costs 1,250.7; expected loss 108; pre-tax result 725.8; net result 544.35. Allocated capital EUR 14,199k on RWA of EUR 109.2m. RAROC 3.83%, a third of the 12% hurdle. If the EU’s temporary multiplier halved the own funds, RAROC would be 7.67%. Holding conventional bonds instead of the two green bonds would add EUR 18,700 of carry a year (the greenium given up) and lift RAROC to 3.87%. To reach 12% at full FRTB capital, the issuers would have to bring EUR 4.44m of ancillary revenue a year; with own funds halved, EUR 1.18m.
Reading. Three facts that any credit trader knows and few students expect. First, under the standardised FRTB, an investment-grade flow book barely earns its cost of capital on its own; banks use internal models where they can, keep inventory turning, and live on the relationship. Halving average inventory for the same client flow lifts RAROC to about 6.4%: the desk’s best lever is turnover, not position size. Second, Helios is a DCM client (C01): one bond mandate brings Meridian EUR 0.375m of fees, and the issuer expects its bookrunners to make markets in its bonds afterwards. Secondary market-making is part of the price of the primary franchise, and the RAROC of the relationship is what the bank manages (see C01, section 6.2). Third, the green label’s effect on the desk’s economics is tiny: EUR 18,700 of carry a year, 0.04 points of RAROC, and EUR 114,000 of loss avoided in one bad week of the case. It changes who the clients are more than what the book earns.
| Item | Value | Stage |
|---|---|---|
| Net CS01 of the book (= DV01) | EUR 49,665 per bp | 5.1 |
| Helios green 2034 CS01 | EUR 27,166 per bp | 5.1 |
| Futures sold (Bobl / Bund) | 611 / 191 | 5.1 |
| Beta-weighted CS01 | EUR 51,775 per bp | 5.1 |
| Index protection needed (60% policy) | EUR 66.6m (EUR 65m bought) | 5.1 |
| Index CS01 on EUR 65m at 55 bp | EUR 30,305 per bp | 2.3 |
| Helios greenium: matched pair / linear / log | −3.0 / +0.6 / −5.5 bp | 5.2 |
| Range between methods | 6.1 bp | 5.2 |
| Vantoria greenium, matched pair | −2.7 bp | 5.2 |
| Stress week: bonds unhedged | EUR −1.63m | 5.3 |
| Main protection P&L | EUR +0.90m | 5.3 |
| Net, hedged with Main / with ESG index | EUR −0.73m / −0.82m | 5.3 |
| ESG-index basis cost | EUR 87,000 | 5.3 |
| Net, hedged, without the sector news | EUR −0.55m | 5.3 |
| Greenium after the week (Helios / Vantoria) | −5 / −8 bp | 5.3 |
| Loss avoided by green outperformance | EUR 114,000 | 5.3 |
| Exit cost, half Vantoria, stressed / normal | EUR 64,700 / 24,300 | 5.4 |
| Days to sell green / conventional (stress) | 5 / 2.5 | 5.4 |
| CSR delta / DRC / own funds | EUR 6.04m / 2.70m / 8.74m | 5.5 |
| RWA / allocated capital | EUR 109.2m / 14.2m | 5.5 |
| Own funds without index hedge | EUR 11.51m | 5.5 |
| Revenues / pre-tax result | EUR 2,084.5k / 725.8k | 5.6 |
| RAROC: case / multiplier 50% / conventional inventory | 3.83% / 7.67% / 3.87% | 5.6 |
| Ancillary revenue for 12% (full / half own funds) | EUR 4.44m / 1.18m | 5.6 |
Which book. The bonds and the CDS are in the trading book because they are held with trading intent (to resell to clients and to hedge), under the CRR’s boundary rules: instruments held for market-making are presumed to be in the trading book, and moving them to the banking book is restricted. They therefore carry market-risk own funds, not credit-risk RWA; the issuer’s default is captured by the DRC instead of the IRB formula.
Which rules. Until 31 December 2026 the EU still applies the pre-FRTB standardised approach (specific risk by rating and maturity, general interest rate risk by maturity bands) or the bank’s approved internal model (VaR and stressed VaR plus the incremental risk charge). From 1 January 2027 the FRTB applies: the standardised approach of stage 5.5 (sensitivities, DRC and a residual risk add-on), or, for desks that pass the profit-and-loss attribution test and back-testing, the internal model approach based on expected shortfall with liquidity horizons (40 days for investment-grade corporate credit spreads and 60 days for high yield, against 10 days for euro interest rates), a default risk charge and a charge for non-modellable risk factors. The Commission’s delegated regulation of 4 June 2026 adds, for three years, a multiplier that scales FRTB own funds back towards pre-FRTB levels, relaxes the internal-model tests and phases in some standardised requirements, to keep EU banks in line with the UK and US, which have postponed their own implementation. The lab of stage 5.6 lets the multiplier vary: its actual value depends on each bank’s numbers.
Other capital. Operational risk under the CRR3 standardised approach, driven by the business indicator, which includes the trading book’s net P&L. Counterparty credit risk and CVA on the index CDS: cleared trades carry a small exposure to the CCP (2% risk weight for a qualifying CCP on trade exposures, plus default fund contributions), bilateral trades a full counterparty and CVA charge. Prudent valuation: the CRR requires additional valuation adjustments (for close-out costs, model risk, concentrated positions) deducted from CET1; a large position in a thinly traded bond, such as a green bond mostly locked in buy-and-hold portfolios, attracts a larger adjustment.
The stage 5.6 lab gives the desk’s annual economics:
| Line | Computation | EUR k |
|---|---|---|
| Flow revenue | EUR 1.2bn traded x 10 cents per 100 | 1,200.0 |
| Spread carry on bonds | Sum of nominal x spread over swaps | 1,282.0 |
| Index protection | EUR 65m x 55 bp | −357.5 |
| Haircut funding and borrow fee | 5% x EUR 100m x 30 bp + EUR 10m x 25 bp | −40.0 |
| Revenues | 2,084.5 | |
| Operating costs | 60% of revenues | −1,250.7 |
| Expected loss | 0.20% x 60% x EUR 90m | −108.0 |
| Pre-tax result | 725.8 | |
| Tax at 25% | −181.45 | |
| Net result | 544.35 | |
| Allocated capital | 13% x 12.5 x EUR 8.74m | 14,199 |
| RAROC | 544.35 / 14,199 | 3.83% |
The expected loss of a trading book is not a provision, since the bonds are marked to market; it is included because the carry includes compensation for default and a RAROC that ignored it would overstate the return. The green label changes three lines, all slightly. Carry falls by EUR 18,700 a year because green bonds trade a few basis points tight. Flow revenue may rise if green-mandated clients trade with the desk that holds green inventory and axes, but there is no reliable evidence that they pay wider bid-offers. Capital does not change: neither the FRTB nor the CRR distinguishes green from conventional bonds of the same issuer. The supervisory debate on a “green supporting factor” for capital has been settled by the EBA’s 2023 report on the prudential treatment of environmental risks, which found no evidence of lower risk justifying it and favoured incorporating environmental risk into existing instruments instead.
The inventory is financed mostly in repo. The repo rate for corporate bonds is close to the overnight rate plus a spread that depends on the collateral, and the counterparty applies a haircut: the bank funds that part from its own resources, charged by Treasury at the FTP rate for the tenor of the financing. Repo of investment-grade corporates is usually short term, so the funding cost is low but the rollover risk is real: in stress, haircuts rise and some counterparties stop accepting lower-rated collateral, as the Eurosystem’s climate factor may do at the margin for carbon-intensive issuers. Short positions need the bond to be borrowed, at a fee that rises when the bond is scarce; green bonds held by buy-and-hold funds that do not lend securities are often harder to borrow. For liquidity ratios, investment-grade corporate bonds can qualify as Level 2B high-quality liquid assets in the LCR, with a 50% haircut and conditions, only if they are under the control of the liquidity management function; trading inventory usually is not. The NSFR requires stable funding for a share of the trading inventory, which Treasury charges back to the desk. In the case these costs are a few tens of thousands of euros; on a large credit book they are a line the desk manages actively.
Under IFRS 9 the bonds and the CDS are held for trading and measured at fair value through profit or loss: every spread move goes straight to the income statement, which is why the stage 5.3 loss was a loss on the day. No hedge accounting is needed: hedged items and hedges are both at fair value. IFRS 13 classifies the marks by the observability of inputs: liquid bonds and indices are Level 1 or 2; a thinly traded bond marked from a curve or from comparable issuers can be Level 3, with additional disclosures. Profit on a trade at inception (day-one P&L) can be recognised only if the fair value is evidenced by observable market data. Short positions are financial liabilities held for trading. The label has no accounting effect.
Green Asset Ratio. Trading-book exposures are excluded from the GAR numerator and denominator. The trading-book KPI that was to show the share of Taxonomy-aligned trading activity was postponed to 2028 by Delegated Regulation (EU) 2026/73 (see A02). Meridian’s EUR 65m of green bonds in inventory therefore appear in no Taxonomy ratio.
Pillar 3 ESG. The EBA’s ESG disclosure templates focus on the banking book (exposures by sector, financed emissions, maturity, collateral energy efficiency); the trading book appears only where a template covers it explicitly.
Financed and facilitated emissions. PCAF’s Part A standard covers listed corporate bonds the bank holds on its balance sheet: attribution is the bond’s outstanding amount over the issuer’s enterprise value including cash (EVIC). Banks differ on whether they include short-term trading inventory, which changes daily; most exclude it or report it separately, because the emissions would swing with positions held for days. PCAF’s Part B, which covers facilitated emissions from underwriting (see C01), does not cover secondary market-making. The honest statement in Meridian’s sustainability report is that its trading book holds and hedges carbon-intensive issuers’ bonds for clients, that these positions are excluded from financed emissions targets, and why.
Targets and sector policies. Meridian’s sector policies (for example on thermal coal or unconventional oil and gas) normally apply to lending and underwriting; whether they apply to market-making in existing bonds of excluded issuers is a policy choice. Many banks allow market-making in secondary as a client service even for issuers they no longer finance, and disclose it. A trader must know the bank’s policy before advertising axes in such bonds.
| Risk | What can go wrong | Control | Owner |
|---|---|---|---|
| Spread risk | Inventory loses value before it is sold | CS01, stress and ES limits; hedge policy | Trader, market risk |
| Jump and transition risk | Policy news hits one sector or issuer | Issuer JTD limits; sector stress scenarios | Market risk |
| Basis risk | Index or ESG-index hedge diverges from bonds | Basis limits; daily P&L explain | Trader, product control |
| Liquidity risk | Cannot exit at the mark in stress | Concentration limits by bond float; prudent valuation | Market risk, finance |
| Mis-marking | Illiquid green bonds marked at stale levels | Independent price verification; Level 3 review | Product control |
| Greenium mis-statement | Single-method greenium quoted to clients | Approved methodology; range and date required | Desk analyst, compliance |
| Label misdescription | Bond called green without standard or review | Golden-source label data; sales scripts | Sustainable finance, compliance |
| ESG product mis-selling | ESG index or TRS sold as a climate hedge | Product governance; target market; risk disclosure | Product approval committee |
| Market abuse | Trading ahead of a green issue or a downgrade | Information barriers; surveillance | Compliance |
| Policy breach | Axes in issuers excluded by the bank’s policy | Restricted list check in the order system | Compliance, sustainable finance |
Red flags for a credit trading desk.
What changed recently. Transparency (new deferral regime, tapes), capital (FRTB with temporary relief), central-bank collateral (climate factor) and fund rules (names guidelines, SFDR revision) all moved in 2025 and 2026. Each shifts secondary demand for labelled and carbon-intensive bonds at the margin.
Open debates. Whether a secondary greenium exists for corporates once liquidity and coupon effects are controlled; whether green bonds are more or less liquid in stress, given that dedicated holders may become forced sellers if green funds face outflows; whether ESG-screened CDS indices will ever be liquid enough to replace Main as a hedge; whether market-making inventory belongs in banks’ financed emissions; and whether greater post-trade transparency will reduce dealers’ willingness to hold inventory in less liquid bonds, including many corporate green bonds.
INTERACTIVE LAB — Credit trading and secondary green bond markets: check yourself
Test the mechanics of market-making, hedging, greenium measurement, liquidity, FRTB capital and reporting covered in the module.
Q1. A bond with a modified duration of 7 is quoted 4 bp wide in spread. What is the bid-offer in price, approximately?
Answer: B. Price bid-offer is about duration times spread bid-offer: 7 x 0.04% = 0.28 points per 100.
Q2. Why does a credit trader sell Bobl or Bund futures after buying a corporate bond?
Answer: C. A fixed-coupon bond’s yield has a rate and a spread component; futures remove the rate part.
Q3. iTraxx Main trades at 55 bp with a 100 bp fixed coupon. The buyer of protection:
Answer: A. The spread is below the coupon, so the buyer is compensated up front for paying the higher fixed coupon.
Q4. Interpolating between an issuer’s 6-year (124 bp) and 10-year (156 bp) conventional bonds, what is the greenium of its 8-year green bond at 137 bp?
Answer: D. The synthetic 8-year conventional spread is 140 bp; 137 − 140 = −3 bp.
Q5. Why did the linear curve fit give a positive greenium in the case?
Answer: B. Credit curves are steep at the front and flat at the back; a line underestimates mid-curve spreads, so the green bond looks wide.
Q6. In the case, hedging with the iTraxx MSCI ESG Screened index instead of Main lost about EUR 87,000 more. Why?
Answer: C. The screened index moved 27 bp against 30 bp for Main; the protection gained less while the bonds lost the same.
Q7. Which statement about green bond liquidity is best supported by the studies cited?
Answer: A. ECMI 2023 found tighter spreads mainly for sovereigns and certified corporates; Dulak and Wolff 2026 found no systematic disadvantage.
Q8. Under the FRTB SBM, a Helios bond is in bucket 7 with a 1.5% risk weight and a Vantoria bond in bucket 4 with 3%. What does this imply?
Answer: B. Weighted sensitivity is risk weight times sensitivity; buckets are set by sector and rating, not by label.
Q9. What is the hedge benefit ratio in the DRC?
Answer: C. HBR = net longs divided by net longs plus absolute net shorts; it limits the offset between different obligors.
Q10. How are the trading book’s green bonds treated in Meridian’s Green Asset Ratio in 2026?
Answer: D. Trading-book exposures are outside the GAR; Delegated Regulation (EU) 2026/73 postponed the trading-book KPI.
Q11. Since 2 March 2026, what governs EU post-trade deferrals for corporate bonds?
Answer: B. The MiFIR review replaced national deferral regimes with harmonised categories in the amended Delegated Regulation 2017/583.
Q12. In the case RAROC, what is the effect of holding green rather than conventional bonds?
Answer: A. Carry falls by about EUR 18,700 a year because of the greenium; capital is unchanged.
Q13. Which is the main lever to raise the RAROC of a flow credit book at constant client volume?
Answer: C. Flow revenue depends on client volume, capital on average inventory: less inventory for the same flow raises RAROC.
An issuer’s conventional bonds trade at 98 bp (5 years) and 128 bp (10 years) over mid-swaps; its 7-year green bond trades at 112 bp, price 101.0, modified duration 6.2. (a) Estimate the greenium by linear interpolation. (b) What is it worth on EUR 20m? (c) The quoted bid-offer is 4 bp: what is it in price?
Solution. (a) Synthetic 7-year conventional spread: bp. Greenium bp: the green bond trades 2 bp cheap to the curve, which happens and should be reported as such. (b) CS01 EUR 12,524 per bp; 2 bp are worth about EUR 25,000. (c) points per 100. Before telling a client the bond is cheap, check its coupon and age against the comparison bonds, and repeat with a curve fit.
A book has a CS01 of EUR 40,000 on one issuer with a spread beta of 1.2 to iTraxx Main. Index protection has a CS01 of EUR 470 per EUR 1m. (a) What notional hedges 75% of the beta-weighted CS01? (b) Over a week the index widens by 20 bp and the issuer by 30 bp. Compute the P&L of the bonds, the hedge (linear approximation) and the net, and split the net loss into its two causes.
Solution. (a) Beta-weighted CS01 ; notional EUR 76.6m. (b) Bonds: EUR 1.20m. Hedge: EUR 0.72m. Net: −EUR 0.48m. Split: the beta predicted bp; the extra 6 bp are idiosyncratic and cost EUR 0.24m; the unhedged 25% of the beta-predicted move costs EUR 0.24m. The two add to the EUR 0.48m net loss.
A desk is long EUR 30m of a BBB issuer’s bond at a price of 98 and short EUR 10m of another bond of the same issuer at 102. CSR delta capital is EUR 1.5m. Annual revenues EUR 900k, cost-income 60%, PD 0.20%, market LGD 60% on the net long nominal, tax 25%, capital 13% of RWA. (a) Compute the DRC. (b) Compute the RAROC.
Solution. (a) Long JTD EUR 21.9m. Short JTD EUR 7.7m. Same obligor and seniority, so they net: EUR 14.2m. DRC EUR 0.852m. (b) Own funds EUR 2.352m; RWA EUR 29.4m; capital EUR 3.822m. Pre-tax EUR 336k; net EUR 252k; RAROC .
1. How does a credit market-maker make money? Mostly by capturing part of the bid-offer on client trades, turning inventory before spreads move. Second, by earning carry on inventory, net of hedges and funding. It loses through adverse selection, through spread moves on inventory and through hedges that do not track the bonds. The business works on volume and short holding periods, which is why turnover matters more than position size.
2. How would you measure the greenium of a green bond in secondary? I would compare its spread with a synthetic conventional bond of the same issuer and maturity, interpolated from the bonds that bracket it, and cross-check with a fitted curve. I would adjust for coupon, age and liquidity where I can and use mid levels on the same date. I would report a range, not a point, because the method alone can move the result by several basis points. For a sovereign with twin bonds, the direct comparison is much cleaner.
3. Are green bonds less liquid than conventional bonds? Not systematically, according to recent evidence: studies find similar or slightly tighter bid-offers, mostly for sovereigns and certified issues, and higher trading volumes. Individual corporate green bonds can be thin because a large share is held by buy-and-hold funds, which also makes them harder to borrow. In stress the stable holder base can help, but if green funds face outflows they become sellers. I would judge each bond by its float, borrowability and recent trading.
4. You are long a high-carbon issuer’s bonds and hedged with iTraxx Main. What risk remains? The idiosyncratic risk of the issuer and its sector, including transition-risk events such as a change in carbon pricing or allowances, which the index barely moves on. The unhedged share of the beta-weighted risk. Basis risk between bonds and CDS. Hedging with an ESG-screened index would make the first problem worse, because it excludes some of the most exposed names.
5. What changes for your desk with FRTB in January 2027? Capital is computed from sensitivities by bucket, with correlation scenarios, plus a default risk charge and a residual risk add-on under the standardised approach, or from expected shortfall with liquidity horizons under internal models, which require desk-level approval and P&L attribution tests. For a credit book the standardised approach is costly. The EU’s temporary multiplier until 2029 softens the impact. Hedges must be designed with the capital formulas in mind, since an over-hedge in an index bucket creates its own capital.
6. Does a green bond consume less capital? No. The FRTB and the CRR do not distinguish green from conventional bonds of the same issuer, and the EBA concluded in 2023 that there was no evidence to justify a green supporting factor. The only differences are indirect: a slightly lower spread, and possibly a larger prudent valuation adjustment if the position is concentrated in a thin bond.
7. A client asks you to sell them an “ESG hedge” for their credit portfolio. What do you say? I would explain what the ESG-screened index contains and excludes, how liquid it is, and that its basis to the portfolio may be large, especially in transition-risk events. I would check the client’s target market and the product approval for the instrument. I would avoid any claim that the hedge reduces the portfolio’s climate impact; it changes market exposure, not emissions.
8. Why would a bank make markets in bonds of an issuer it no longer finances? Because investors need a counterparty to exit or adjust positions, and refusing to trade can trap clients in positions they want to sell, including those divesting for climate reasons. Market-making is a service to investors, not new financing of the issuer. The bank should have a written policy, apply it consistently and disclose it, so that the trading activity does not contradict its public commitments.
Regulation and supervisory texts
Market infrastructure and indices
Market data and commentary
Academic