Credit Desk Product Control

CHAPTER 1

What Does a Credit Desk Do?

Before looking at how Product Control checks a Credit book, it helps to understand what the desk you are supporting is actually there to do — what it trades, who it trades with, how it makes money, and how it is typically organised. Every product in this manual exists because it solves a real problem for a real client or for the bank's own balance sheet.

1.1 Role of the Desk

A Credit desk trades instruments whose value depends on the creditworthiness of a company, sovereign, or portfolio of names — corporate bonds, credit default swaps (CDS), credit indices (CDX, iTraxx), and index tranches. Within an investment bank it sits inside the Fixed Income, Currencies and Commodities (FICC) division, alongside Rates, FX, Commodities and Securitised Products. Credit is usually split into Investment Grade (IG) and High Yield (HY) trading, and further into cash bonds, single-name CDS, index/flow trading, and structured credit (tranches, bespoke portfolios).

The desk exists to serve three overlapping purposes:

• Client market-making: quoting prices to clients (asset managers, hedge funds, insurers, corporates) who want to buy or sell credit risk, or hedge existing exposure, earning the bid-offer spread on that flow.
• Risk warehousing and hedging: the desk ends up holding residual credit risk from client flow (e.g. an unsold piece of a bond it bought from a client) and manages that risk using CDS, indices, or other bonds.
• Relative value / flow trading: taking positions where the desk sees value — e.g. a single name trading cheap to its sector, or the CDS-bond basis being wider than historical norms — within its risk mandate.

1.2 Who the Desk Trades With

1.3 How the Desk Is Typically Organised

• Cash bonds (IG / HY): trading corporate bonds, both flow market-making and new-issue distribution.
• Single-name CDS: buying/selling protection on individual reference entities, used for hedging and relative value.
• Index / flow credit: trading CDX (North America) and iTraxx (Europe) indices — the most liquid, most frequently used hedging instruments on the desk.
• Structured credit: index tranches, bespoke correlation products — more complex, less common outside specialist desks (introduced briefly in Chapter 8).
• Credit derivatives trading / flow desk: often manages the overall index and single-name CDS risk that flows in from client business across cash and derivatives.

1.4 How the Desk Makes (and Loses) Money

Understanding the desk's P&L sources is exactly why Chapters 11–15 of this manual exist — Product Control's core job is separating out each of the following into its own explainable bucket, every single day:

CHAPTER 2

A Day in the Life of a Credit Product Controller

Before any product-specific detail, it helps to see the whole job in one place: what actually happens from the moment a trader executes a Credit trade to the moment it appears, fully explained, in a month-end pack. Every chapter that follows plugs into one of the stages below.

2.1 The Trade Lifecycle Through a Controller's Eyes

2.2 The Question to Ask for Every New Position

Whenever a trader takes on a new Credit position — a bond, a single-name CDS, an index trade — a controller should be able to answer five questions before moving on. These five questions are the spine of every product chapter in this manual:

1. What exactly was traded? (Reference entity/issuer, direction, notional, spread/price, restructuring clause, maturity — verified against the confirmation, not assumed from the blotter.)
2. How is it valued, and off which curve? (Which credit curve, which recovery assumption, which discount curve — see Chapters 5 and 15.)
3. What is its risk, and does that risk match expectation? (CS01, jump-to-default — Chapter 11 — sanity-checked against the trader's stated intent for the trade.)
4. What cash flows will it generate, and when? (Coupons, premium payments, upfront exchange — so a break can be anticipated rather than discovered.)
5. What could go wrong, and how would I know? (The specific FOBO break patterns for this product — covered at the end of every product chapter and consolidated in Chapter 17.)

2.3 Who You Work With

2.4 Independent Price Verification (IPV) — How to Perform It

IPV is not simply a month-end comparison of the trader mark to one external number. A robust Credit IPV control fixes the instrument, valuation timestamp, market-data definition and tolerance before comparing values, and preserves enough evidence for another reviewer to reproduce the conclusion.

• Independent sources: use approved external pricing services, CCP settlement prices where relevant, observable market composites, broker/dealer quotes or executable levels under the firm's valuation policy. A trader mark should never be the only evidence for an illiquid position.
• Valuation timestamp: compare like with like. If the official valuation cut-off is 4:00pm New York, do not compare that mark to a midday quote and call the difference an IPV exception.
• Tolerance: set tolerances by product and liquidity, and assess both the market-data difference (bp / price points) and the resulting valuation impact. There is no single sensible bank-wide bp tolerance for an on-the-run index and an illiquid distressed bond.
• Illiquid names: corroborate with at least two genuinely independent inputs where available, or use a documented proxy / curve-interpolation methodology. Where uncertainty remains, carry an appropriately governed valuation adjustment rather than accepting an unsupported trader mark.
• Evidence and escalation: retain the source, timestamp, raw quote, tolerance, valuation impact, exception owner and approval. Repeated small IPV differences in the same direction are a control signal even when each one is individually below materiality.

2.5 Valuation Reserves and Prudent Valuation

Credit books often contain positions that cannot be exited at the theoretical mid value shown by the front-office model. Product Control should understand the purpose of each valuation adjustment and ensure it is independently governed rather than used as a plug to force a P&L or reconciliation to tie.

• Bid-offer / close-out adjustment: reflects the cost of exiting or hedging a position away from mid, particularly where the desk is large relative to normal market depth.
• Liquidity / concentration adjustment: reflects uncertainty or exit cost where a position is illiquid, concentrated, off-the-run or difficult to hedge quickly.
• Market-price uncertainty adjustment: captures uncertainty in observable inputs when independent prices are sparse or dispersed.
• Model-risk adjustment: captures material uncertainty from model choice, calibration, correlation / recovery assumptions or other non-observable inputs.
• Governance distinction: accounting fair-value adjustments and regulatory prudent-valuation / AVA calculations are related but not identical concepts. Product Control should know which reserve is in P&L, which is a capital adjustment, who owns it, and what evidence supports it.

2.6 Regulatory Context — FRTB and Credit Risk

Under the Fundamental Review of the Trading Book (FRTB), credit positions feed market-risk capital through credit-spread risk and a separate default-risk framework. Product Control does not normally own the capital calculation, but the quality and consistency of the desk P&L, sensitivities and default exposures are important inputs and control evidence.

• Sensitivities-based measures: credit spread risk is captured by prescribed spread sensitivities and buckets under the standardised framework.
• Default Risk Charge (DRC): captures jump-to-default risk that ordinary spread shocks do not capture — reinforcing why CS01 and JTD must both be visible on a Credit desk.
• P&L / risk alignment: where an internal-model framework applies, clean and risk-theoretical P&L used for model testing must be explainable against the same underlying risk factors and valuation conventions.
• Product Control role: reconcile material differences between valuation P&L and risk-system sensitivities, maintain data-quality discipline, and escalate breaks that could contaminate regulatory risk or capital reporting.

2.7 Systems, Sources of Truth and Cut-Offs

A Credit controller rarely works in one system. The practical control problem is understanding what each platform is authoritative for and where timing differences arise between them.

• Front-office trading / risk system: source for intended trade economics, position risk and trader valuation; not automatically the accounting book of record.
• Sub-ledger / accounting system: official books-and-records source for accounting positions, accruals and ledger posting; differences versus front office are the subject of FOBO control, not evidence that one side is always wrong.
• Market-data / valuation engines: source of curves, recovery assumptions, bond prices and independent valuation inputs. Version and timestamp matter as much as the number itself.
• CCP / collateral platforms: authoritative for cleared positions, variation margin, initial margin and clearing cash statements, subject to reconciliation back to internal books.
• Interfaces and cut-offs: record expected file arrival times and valuation cut-offs. A partial feed, late batch or stale snapshot should be treated as a controlled data incident with a documented fallback and re-run, not silently absorbed into unexplained P&L.

CHAPTER 3

Bond Pricing and Credit Spreads

A corporate bond's price reflects both the risk-free rate environment (Chapter 3 of the Rates manual) and the additional yield investors demand for taking on the issuer's credit risk — the credit spread. For a Credit controller, the spread is usually the more important number to understand day to day, because it's the spread that moves for credit-specific reasons.

3.1 Spread Measures

3.2 Spread Duration and Price Impact

3.3 Bond Accrued Interest — Clean Price vs Dirty Price

Corporate bonds are normally quoted on a clean-price basis, excluding coupon accrued since the previous payment date, but they settle at the dirty price: clean price plus accrued interest. This distinction is one of the most common reasons a bond position appears to reconcile on price but not on cash or daily carry.

3.4 New Issues and Primary-Market Effects

A new bond issue can create P&L on an existing Credit book even when the issuer's fundamental credit view has not changed. The new supply, underwriting process and new-issue concession can temporarily move both the cash bond curve and the issuer's CDS-bond basis.

• Separate underwriting / new-issue inventory and associated fees or concession from ordinary secondary-market P&L where the desk participates in distribution.
• Expect existing bonds to cheapen temporarily when a large new issue comes at a concession; this can widen the bond side of a CDS-bond basis trade without any CDS move.
• Track allocation, re-offer price, accrued interest, settlement date and any unsold inventory retained by the desk — these are common sources of temporary position and cash breaks.
• In the daily commentary, distinguish a supply-driven technical move from an idiosyncratic deterioration in the issuer's credit quality.

CHAPTER 4

Credit Default Swaps — Structure and Mechanics

A credit default swap (CDS) is a contract where the protection buyer pays a periodic premium to the protection seller, in exchange for a payment if a specified credit event (default, bankruptcy, or restructuring, depending on documentation) occurs on the reference entity. It is, in effect, insurance against a company's default — and it is the single most important instrument on a Credit desk, both for hedging and for expressing views.

4.1 Key Terms

• Protection buyer: pays the premium, receives a payout if a credit event occurs — economically “short” the credit (benefits if the credit worsens).
• Protection seller: receives the premium, pays out on a credit event — economically “long” the credit (similar exposure to owning the bond).
• Reference entity: the company or sovereign the CDS references — not necessarily the specific bond issuer if the entity has multiple issuing subsidiaries.
• Standard coupon: fixed-coupon / points-upfront trading became the market standard during the 2009 CDS standardisation reforms. The 2009 Big Bang introduced the Standard North American Corporate 100bp/500bp convention; fixed-coupon conventions were also adopted across other major regions during the same standardisation wave. The 2009 Small Bang dealt primarily with restructuring auction hardwiring. The 2014 ISDA Credit Derivatives Definitions updated the legal definitions and documentation framework; they did not create the 100bp/500bp coupon convention.

4.2 Bilateral vs Centrally Cleared CDS

Standardised index CDS and many eligible single-name CDS can be cleared through central counterparties (CCPs), while other trades remain bilateral under an ISDA Master Agreement and, where applicable, a Credit Support Annex (CSA). A controller must know the clearing status because the operational cash flows and reconciliation evidence differ materially.

• Cleared trade: after clearing, the CCP becomes the counterparty. Product Control should reconcile trade population and notional to CCP statements, daily settlement / variation margin, initial margin and clearing fees.
• Uncleared bilateral trade: valuation and collateral are governed by the bilateral documentation and CSA. Margin calls and collateral interest create additional cash movements that do not themselves represent trading P&L.
• Do not assume every liquid single-name CDS is centrally cleared. Eligibility and actual clearing usage vary by product, reference entity, jurisdiction and counterparty; verify the trade's clearing status from the booking and clearing records.
• A cleared-versus-uncleared mismatch in static data can create persistent valuation, cash and counterparty breaks even when the trade economics are otherwise identical.

CHAPTER 5

CDS Pricing and the Credit Curve

A CDS is priced by comparing the present value of the premium payments (the “premium leg”) against the present value of the expected credit-event payout (the “protection leg”), using a survival curve derived from market-observed CDS spreads for that reference entity.

5.1 The Building Blocks

5.2 Hazard Rate

The hazard rate (or default intensity) is the instantaneous probability of default per unit time, conditional on survival to that point. A flat CDS spread curve implies a roughly constant hazard rate; an upward-sloping curve (higher spreads at longer tenors) implies the market sees default risk increasing further into the future, which is the normal shape for most credits.

CHAPTER 6

Upfront Points, Running Spread and CDS Accrued Premium

Since CDS contracts trade with a standardised running coupon, the actual market price information is expressed through an upfront payment exchanged at trade inception, plus an accrued premium adjustment for the part of the current coupon period that has already elapsed — conceptually very similar to clean vs dirty bond pricing (Rates manual, Chapter 4), but for a very different reason.

6.1 Accrued Premium

Like a bond's accrued interest, CDS premium accrues daily on Actual/360 from the last premium payment date (or trade date). If a credit event occurs, the protection seller is typically entitled to the accrued premium up to the event determination date. On a normal (non-defaulted) trade, accrued premium affects the cash amount exchanged at any point the position is unwound or novated.

6.2 Premium Dates, Accrual and Settlement Timing

Standard CDS premium periods are aligned to quarterly dates (20 March, 20 June, 20 September and 20 December for the standard quarterly cycle), with Actual/360 accrual under standard conventions. The economic accrual date, cash payment date and upfront settlement date can therefore be different dates and can be represented differently across systems.

• Premium accrual: standard contracts generally accrue through the quarterly cycle even when the trade is entered between payment dates; the trade economics include an accrued-premium adjustment so that buyer and seller are economically aligned.
• Upfront settlement: standard single-name CDS upfront cash commonly settles on a short post-trade cycle (often T+3 under standard bilateral convention), but clearing venue, confirmation and local market rules should be treated as authoritative.
• Netting: a CCP or settlement platform may show one net cash movement for many trades while the front-office system shows gross trade-level cash flows. Reconcile from gross expected cash to net settlement rather than treating the net amount as a missing trade.
• Break pattern: one-day differences in accrual start, holiday calendar, cut-off or settlement status can create a systematic cash break across an entire population.

CHAPTER 7

CDS-Bond Basis

The CDS-bond basis is the difference between a reference entity's CDS spread and its bond spread (typically the Z-spread or asset-swap spread of a comparable-maturity bond). It is one of the most commonly traded relative-value strategies on a Credit desk, and one of the trickiest P&L stories for a controller to explain cleanly.

7.1 Deliverability and Recovery Basis

A CDS-bond basis trade is not a perfect hedge merely because the reference entity and maturity look similar. The specific bond held may not be an eligible deliverable obligation into the CDS, and its realised recovery can differ from the market-wide auction recovery used to settle the CDS.

• Confirm seniority, currency, maturity and deliverability characteristics of the cash bond against the CDS documentation and the eligible-obligation rules.
• The CDS auction final price reflects the deliverable-obligation set / cheapest-to-deliver dynamics, not necessarily the specific bond held by the desk.
• A persistent basis can therefore be structural — reflecting funding, repo, liquidity, deliverability and recovery differences — rather than a simple mispricing that must converge to zero.
• When a name becomes distressed, reassess the hedge relationship explicitly; basis behaviour can be dominated by recovery and deliverability rather than ordinary spread convergence.

CHAPTER 8

Credit Indices (CDX / iTraxx) and Index-Tranche Basics

Credit indices are standardised baskets of single-name CDS, and are the most liquid instruments in the credit derivatives market — used far more heavily than single names for hedging and expressing broad market views. CDX covers North American (and some emerging market) reference entities; iTraxx covers European and Asian entities.

8.1 Key Index Families

8.2 Index Roll Mechanics — What the Controller Checks

CDX and iTraxx families create a new on-the-run series roughly every six months, generally in March and September. The exact roll date is family-specific, so Product Control should use the published index calendar rather than hard-code one date for every index. The old series remains tradeable but normally becomes less liquid.

• To maintain the same direction of broad credit exposure, the desk typically closes or offsets the old-series position and enters the same protection direction in the new on-the-run series.
• The old and new series are not economically identical: constituents, fixed coupons, maturity, index factor and market spread / upfront can differ. Treat the roll as two separately controlled trades before looking at the net package.
• The economic roll difference is the difference in fair PV / upfront between the old and new series, adjusted for direction, notional and accrued premium. It is not automatically Day-1 trading P&L; Day-1 P&L is execution price versus the independent fair value of each leg.
• After constituent credit events, reconcile the published index factor / effective notional and the separate credit-event settlement. Do not simply overwrite the original trade notional without understanding the index convention.

8.3 A Brief Note on Tranches

An index tranche isolates a specific slice of the loss distribution of the underlying index basket — e.g. a 0–3% “equity” tranche absorbs the first 3% of losses across the basket, while a 3–7% “mezzanine” tranche absorbs losses only after the equity tranche is wiped out. Tranches introduce correlation risk (how defaults are correlated across the basket) as a new, additional risk dimension beyond spread and jump-to-default risk. Tranche trading is a specialist area; most Credit Product Controllers will encounter tranches far less often than bonds, single-name CDS and index CDS, but should recognise that a tranche book requires additional correlation-sensitivity monitoring beyond what this manual covers in detail.

CHAPTER 9

Credit Events and Settlement

A credit event is a contractually defined trigger — bankruptcy, failure to pay, or restructuring — that activates the CDS protection payout. Handling a credit event correctly, quickly, and with the right settlement mechanism is one of the highest-stakes, lowest-frequency tasks a Credit controller performs.

9.1 Types of Credit Event

• Bankruptcy: the reference entity files for bankruptcy protection or is otherwise legally insolvent.
• Failure to Pay: the reference entity fails to make a payment (subject to a grace period and materiality threshold) on any of its obligations.
• Restructuring: the terms of the entity's debt are changed in a way adverse to creditors (maturity extension, coupon reduction, subordination) — the most contractually complex trigger, and the one where restructuring clause variants matter most.

9.2 Restructuring Clause Variants

9.3 Settlement Mechanism

Since the 2009 “big bang” protocol, most CDS credit events are settled via a standardised ISDA-administered auction process, producing a single, market-wide final price for the defaulted obligation. This determines the payout for cash-settled CDS across the entire market simultaneously, replacing the older, more operationally complex process of physical settlement (delivering an eligible bond in exchange for full par payment).

CHAPTER 10

Recovery Rate and Loss Given Default

The recovery rate is the assumed (or, after a credit event, realised) percentage of face value a creditor recovers in a default. It is a critical input to CDS pricing (Chapter 5) and to any distressed-name valuation, and one of the most frequently mis-communicated numbers between desks and control functions.

10.1 Standard vs Bespoke Recovery Assumptions

Standard CDS pricing conventions typically assume a flat 40% recovery rate for senior unsecured corporate CDS (25% for subordinated debt) purely as a market quoting convention — this is NOT a genuine estimate of what the entity would actually recover in default, and should not be treated as one. For a genuinely distressed name, trading desks and IPV will typically override this with a bespoke, name-specific recovery estimate based on the entity's actual capital structure and expected recovery.

CHAPTER 11

CS01 (Credit DV01) and Jump-to-Default Risk

CS01 (credit spread 01, sometimes called credit DV01) measures the change in a position's value for a 1 basis point parallel move in credit spread — the direct credit-market analogue of DV01 in Rates (Rates manual, Chapter 11). Jump-to-default (JTD) risk is a separate, complementary risk measure unique to credit: the loss (or gain) if the reference entity defaults immediately, right now, rather than its spread simply moving gradually.

11.1 Distressed and Pre-Default Positions — When CS01 Stops Being Enough

As a name moves into distressed territory, the relationship between spread and price becomes highly non-linear and recovery assumptions become increasingly important. A simple CS01 × spread move may still be shown by the risk system, but it should no longer be the controller's primary reasonableness test.

• Very wide spreads: once spreads are several hundred basis points — and especially around 1,000bp or more — small-spread linear approximations become increasingly unreliable. Focus on full revaluation and scenario P&L.
• Default-implied view: compare the current valuation to explicit default scenarios using plausible recovery rates. Ask, “If the name defaults tonight at 20%, 30% or 40% recovery, what is the desk P&L?”
• Recovery sensitivity: the closer the name is to default, the more a recovery assumption can dominate valuation. A change in recovery can create a large P&L even with little change in quoted spread.
• Defaulted bonds: after a credit event, cash bonds may continue trading as recovery instruments. Price / expected recovery becomes more meaningful than a conventional spread measure.
• Realised versus unrealised: distinguish P&L crystallised through auction / settlement from remaining mark-to-market on residual bonds or unsettled claims. Do not leave post-event P&L inside ordinary “spread” explain.

CHAPTER 12

Carry and Roll-Down in Credit

As with Rates positions (Rates manual, Chapter 10), carry and roll-down explain the P&L a credit position earns purely from the passage of time, assuming the credit curve doesn't move. This is an important P&L bucket to separate cleanly, particularly because credit carry (bond coupon or CDS premium income) is often large relative to day-to-day spread P&L.

12.1 Carry

For a bond, carry is the coupon income earned less funding cost. For a CDS position, carry is the premium received (protection seller) or paid (protection buyer) over the period, adjusted for the standardised coupon versus market spread difference captured at inception via the upfront (Chapter 6).

12.2 Roll-Down

Roll-down is the P&L from a position moving to a shorter point on the credit curve as time passes. If the credit curve is upward sloping (longer tenors trade at wider spreads — the normal shape for most credits), a position rolling down to a shorter remaining maturity will, if the curve shape is unchanged, reprice at a tighter spread, generating a roll-down gain for a protection seller / bond holder.

CHAPTER 13

Credit P&L Explain — Spread, Basis, Curve and Jump-to-Default

A full P&L explain for a Credit book breaks total daily P&L into named, understood risk-factor buckets, in the same spirit as the Rates P&L explain (Rates manual, Chapter 13) but with buckets specific to credit risk.

13.1 How the Daily P&L Attribution Is Actually Calculated

The attribution should use close-to-close changes between the same official valuation snapshots used to produce the desk P&L. Intraday market moves may be useful commentary, but they should not be mixed into a close-to-close P&L calculation unless the methodology explicitly requires it.

• Single-name spread P&L: Directional CS01 × close-to-close spread change, followed by full-revaluation checks for large moves or non-linear names.
• Curve P&L: use tenor / key-rate CS01 sensitivities against the change at each point of the credit curve. A single parallel CS01 cannot explain steepening or flattening.
• Rates / discount P&L: for cash bonds and CDS discounting, separate risk-free-curve movement from credit-spread movement so a rates sell-off is not reported as a credit deterioration.
• FX P&L: translate non-base-currency position and cash-flow exposures using the approved close-to-close FX move, and show the hedge separately if the desk uses FX forwards or cross-currency swaps.
• Carry / roll-down: calculate coupon or premium accrual and time / curve roll separately from market movement.
• Day-1 P&L: trade price versus independent fair value at initial recognition. Positive Day-1 can be genuine bid-offer capture; negative Day-1 can be an intentional client concession or a valuation concern.
• Cross-effects / non-linearity: spread convexity, recovery sensitivity, rate-spread interaction and other second-order effects should be calculated or isolated before a residual is accepted.
• Residual: Total reported P&L minus all explained components. A residual is an investigation queue, not a balancing plug.

13.2 FX Risk in Credit Positions

Credit desks often hold bonds, CDS cash flows and hedges in USD, EUR, GBP and other currencies while reporting P&L in a single desk or legal-entity currency. FX can therefore create material P&L even when the underlying credit spread is unchanged.

13.3 Day-1 P&L Is Not Always Bid-Offer Capture

Bid-offer capture is one common source of Day-1 P&L, but the two terms are not interchangeable. Day-1 P&L is the difference between the executed trade price and the independently determined fair value at initial recognition. It can be positive because the desk earned spread, negative because the desk gave a client concession, or unusual because the valuation input or model is uncertain.

CHAPTER 14

Coupon, Premium and Cash Breaks

Coupon and premium breaks are the most operationally frequent (though usually individually small) breaks a Credit controller deals with day to day — they arise whenever an expected cash flow (a bond coupon, a CDS premium payment, an upfront exchange) does not match what actually settled.

14.1 Typical Root Causes

• Day-count convention mismatch — CDS premium is Actual/360 unadjusted; bond coupons follow whatever convention is specified in the prospectus (30/360, Actual/Actual) — comparing the two without adjusting is a frequent false break.
• Incorrect business-day convention shifting a CDS or bond payment date across a weekend/holiday.
• Upfront payment booked with the wrong sign or on the wrong settlement date (CDS upfronts typically settle T+3 business days from trade date).
• Wrong standard coupon applied (100bp vs 500bp, or a legacy non-standardised coupon on an older trade) for the reference entity's rating category.
• Restructuring clause or ISDA definitions vintage mismatch between the trade capture system and the confirmation, causing valuation or eligible-obligation disputes down the line.

14.2 Investigation Sequence

1. Confirm the expected cash flow amount independently (recompute from trade economics, not from either system's output).
2. Compare day-count and business-day conventions used by each system for this specific trade.
3. Confirm the standard coupon and upfront calculation basis match the ISDA standard model conventions for that reference entity's documentation.
4. If a credit event is involved, confirm the auction timeline and payout calculation independently before assuming a booking error (Chapter 9).

14.3 Settlement, Netting and Payment Timing

Many Credit cash breaks are not valuation errors at all — they are differences in when systems expect a cash flow, whether it has settled, and whether the operational platform presents gross or net amounts.

• CDS premiums are typically quarterly on the standard March / June / September / December cycle; bonds can be semi-annual, annual or follow issuer-specific schedules.
• The CDS accrual period and the actual adjusted payment date are distinct concepts. Weekend / holiday adjustments can move cash settlement without changing the contractual accrual end date.
• CCPs and payment platforms may net multiple premium, upfront and margin flows into one currency-level cash amount. Build a gross-to-net reconciliation before raising a missing-cash break.
• A front-office system may recognise accrued income daily while the accounting cash ledger moves only on settlement date. Reconcile accrued receivable / payable separately from cash movement.
• When a system cut-off is missed, identify whether the trade is genuinely absent or simply queued for the next batch. Age and track the timing break until the official books-and-records are corrected.

CHAPTER 15

Discounting, Curve Construction and the ISDA Standard Model

CDS valuation across the market is standardised around the ISDA CDS Standard Model, which fixes the conventions for discounting, day-count, and the bootstrapping methodology used to build the survival probability curve from market spread quotes. Product Control should understand the key inputs well enough to identify when a valuation difference is a genuine market data issue rather than a model or convention mismatch.

15.1 Collateral Support Annex (CSA) and Margining

For bilateral derivatives, the CSA governs how collateral is posted against mark-to-market exposure. The collateral terms matter to both cash reconciliation and valuation because the economically appropriate discounting framework depends on the collateral arrangement.

• Variation margin (VM): moves frequently — often daily — as MTM changes. VM is a collateral cash movement, not a separate trading P&L source; reconcile it to the underlying exposure and collateral statement.
• Initial margin (IM): protects against potential future exposure during close-out. IM is generally not Day-1 trading P&L, but it creates funding / liquidity cost and must reconcile to margin systems and statements.
• Collateral currency: the discount curve should be consistent with the collateral currency and agreement. Modern OIS examples include SOFR for USD, €STR for EUR and SONIA for GBP.
• Multi-currency / optional collateral: if a CSA permits multiple eligible collateral currencies, the valuation methodology may need to capture the economics of the collateral option. Treat persistent cross-currency valuation differences as a model / CSA issue, not a random trade break.
• Collateral interest and settlement: interest on posted collateral and timing of margin settlement can create separate cash and accrual breaks that need their own reconciliation.

15.2 CCP Clearing — Daily Product Control Checks

Major CCPs clear CDS indices and a range of eligible single-name CDS. Clearing reduces bilateral counterparty exposure but introduces a new set of daily controls around CCP positions, settlement prices and margin.

• Position reconciliation: internal cleared trade population and effective notional must agree to the CCP account / clearing broker statement.
• Daily settlement / variation margin: reconcile the CCP settlement-price move to the cleared portfolio and the resulting cash movement. A cash mismatch can be a price, position, timing or netting issue.
• Initial margin: reconcile changes in IM separately from P&L. IM can move materially when portfolio risk, concentration or stress parameters change even if daily trading P&L is small.
• Clearing fees and account structure: confirm house versus client account, currency, clearing broker and fees are mapped to the correct desk / legal entity.
• Uncleared positions: do not force a CCP reconciliation onto bilateral trades. Maintain a clear population split between cleared and uncleared CDS.

CHAPTER 16

Documentation Basis and Restructuring Clause Mismatches

This chapter consolidates a risk that is unique to credit derivatives and doesn't have a direct Rates equivalent: two positions that look economically identical on the surface — same reference entity, same notional, same maturity — can behave completely differently in a credit event if their underlying documentation differs.

16.1 What to Check

• Restructuring clause (CR / MR / MM / XR — Chapter 9.2): determines whether restructuring triggers the CDS at all, and which obligations are deliverable/reference for auction purposes.
• ISDA Definitions vintage (e.g. 2003 vs 2014 Definitions): older and newer trades on the same entity can have materially different credit event definitions and settlement mechanics.
• Seniority (senior unsecured vs subordinated): a CDS on senior debt and a CDS on subordinated debt for the same issuer are entirely different instruments with different standard recovery assumptions and different payout in default.
• Reference obligation vs reference entity: some trades specify a particular reference obligation; most standard contracts simply reference the entity, with a broader set of eligible obligations available at settlement.

CHAPTER 17

Typical Credit FOBO Breaks

This chapter consolidates the recurring Front-Office-to-Back-Office (FOBO) reconciliation break patterns seen across a Credit book, building on the individual issues raised in earlier chapters. Use this as a quick BAU reference when a new break lands on your desk.

CHAPTER 18

Sample Daily P&L Explain — Credit Book

This worked example brings every concept in this manual together into the single deliverable a Credit Product Controller actually produces each day: a fully attributed P&L explain, tying total P&L to named, understood risk factors.

18.1 Scenario

Book: USD Credit Trading desk. Total P&L reported by the front-office system for the day: +USD 554,200. Portfolio: a mix of IG and HY single-name CDS, two CDS-bond basis positions, an index CDS hedge (CDX.NA.IG), and a small book of cash bonds.

18.2 Market Moves Observed

• CDX.NA.IG index: tightened 3bp — broad risk-on sentiment in credit markets.
• One single-name reference entity (a retail-sector credit on which the desk holds bought protection) widened sharply (+40bp) following a weak earnings release — an idiosyncratic move unrelated to the broader market.
• CDS-bond basis on two relative-value positions: broadly stable, minor net convergence.
• Two new client trades booked during the day, net Day-1 P&L of +USD 28,000.

18.3 Attribution

The desk is net short credit risk overall (it has bought more protection than it has sold), expressed partly through the single name and partly through the CDX.NA.IG index used as a broad hedge. Today the two legs moved in opposite directions:

Sum of attributed components: 512,600 − 186,400 + 38,900 + 142,700 + 0 + 28,000 + 18,400 = 554,200 — ties exactly to reported total P&L.

18.3.1 How the Main P&L Components Were Calculated

The CS01 examples above use signed directional sensitivity. The same calculation must be performed from the official prior-close to current-close market snapshots; otherwise the risk-based explain and the reported P&L are not measuring the same period.

18.3.2 Bridge to Accounting Accruals and Cash

Daily P&L does not equal daily cash. Spread, basis and Day-1 components are predominantly mark-to-market; coupon / premium income may be partly settled and partly accrued. A controller should be able to bridge the same USD 554,200 total into accounting-style components without forcing cash to equal P&L.

18.3.3 What to Do When the P&L Does Not Tie

• Reconfirm the total: make sure front-office total P&L, sub-ledger P&L and the attribution all use the same book scope, currency and valuation date.
• Check new trades / lifecycle events: late bookings, amendments, cancellations, novations, index rolls and credit events are the first place to look for a one-day residual.
• Check cash and accruals: recompute bond coupon, CDS premium and upfront cash independently; separate clean / dirty price and settled / accrued amounts.
• Check market-data timing: compare prior-close and current-close curves, spreads, recovery assumptions and FX rates; one stale snapshot can create a large false residual.
• Check non-linearity: for large spread moves or distressed names, run full revaluation / scenario explain rather than relying on CS01 alone.
• Check system / mapping changes: interface failures, book mapping, currency translation, curve mapping and CCP / collateral feeds often create portfolio-wide residuals.
• Only after these checks should a residual remain. Assign an owner, document the hypothesis, set an SLA and carry it forward visibly until resolved — never bury it by changing another P&L bucket.

18.4 Sign-Off Commentary (example wording)

CHAPTER 19

Glossary and Quick Reference

Closing Note

The concepts in this manual are the same ones tested in Product Control interviews, and the same ones a controller draws on every single day to move from “there is a break” to “here is exactly why, and here is the number that proves it.” Revisit Chapter 18 whenever you need a template for how a complete, defensible Credit P&L explain should read — and revisit Chapter 16 every time a book is described to you as “flat” or “fully hedged.”