← Glossary

Devancore Post-Trade Glossary

Credit Curve Reference Data

Credit curve reference data is the governed fixed income data used to support spread curves, valuation controls, P&L explain, collateral marks, accounting inputs, and review evidence.

Definition

Credit curve reference data is the governed fixed income data used to support valuation, P&L explain, collateral marks, accounting inputs, performance attribution, reporting, and review evidence. The operational question is not whether a curve exists. The question is whether the firm can prove which curve state was used, why it applied to the instrument, how the method worked, and who approved any exception or override.

Fixed income positions often need valuation support when direct market quotes are unavailable, stale, thin, or inconsistent across sources. Credit curve data helps estimate the spread component of valuation by issuer, sector, rating, seniority, maturity, and instrument type. That estimate then flows into operating records used by portfolio teams, risk, accounting, fund administration, collateral, performance, and control functions.

Credit curve data components

Credit curve data components

The operating record should identify the curve, the method, the source, and the downstream control that consumed it.

Component What it contains Control question
Benchmark curve Treasury, sovereign, swap, or other base curve used as the valuation anchor Was the correct benchmark curve used for the instrument and currency?
Spread curve Issuer, sector, rating, seniority, instrument type, and tenor-specific spread points Does the selected curve match the bond's risk and reference data?
Tenor points Observed or evaluated points such as 1Y, 2Y, 5Y, 10Y, and 30Y Are key points current, complete, and sourced?
Methodology Interpolation, extrapolation, OAS, Z-spread, G-spread, proxy curve, and fallback rules Can the valuation method be reconstructed?
Governance Source version, timestamp, approval, override, exception reason, and distribution status Who accepted the curve state and why?
Downstream use IBOR, ABOR, PBOR, collateral, risk, reporting, and reconciliation Which records consumed the curve state?

The benchmark curve is the base rate input. Depending on instrument, currency, and market convention, the workflow may use a Treasury, sovereign, swap, or other base curve. The control should show the selected benchmark, its source, its timestamp, and the reason it applies to the instrument.

The spread curve is the credit component. The curve may be issuer-specific where enough observable data exists. If issuer data is thin, the workflow may use a sector, rating, seniority, or peer curve. That mapping should be explicit. A private placement, illiquid corporate bond, municipal bond, or structured instrument should not inherit a curve accidentally because a field was missing.

Tenor points determine the shape of the curve. A curve built from 1-year, 2-year, 5-year, 10-year, and 30-year points may still need to value a 7-year bond. The interpolation method matters because small differences in tenor logic can affect evaluated price, spread attribution, and collateral marks.

Methodology determines how the curve becomes a valuation input. G-spread, Z-spread, and option-adjusted spread answer different questions. Interpolation and extrapolation carry different risk. Proxy curves require stronger documentation than direct issuer curves. Manual overrides require clear evidence, approval, and expiry.

Governance turns curve data into an operating record. Source, timestamp, version, tolerance, stale-data checks, override reason, approval, and downstream distribution should be retained. Without that record, teams may know a valuation changed but cannot show whether the cause was rates, spread, liquidity, security master mapping, or a manual judgment.

How it works

Credit curve reference data works as a controlled input into valuation and post-trade records. It starts with market and reference data, then moves through mapping, methodology, review, approval, and downstream distribution.

Credit curve control workflow

Credit curve control workflow

The workflow governs the curve as a reference-data input before it becomes a valuation, risk, or performance record.

Step Data checked Evidence produced
Select curve Issuer, sector, rating, seniority, currency, maturity, and instrument type Curve selection rule and mapped instrument population
Validate source Provider, timestamp, tenor completeness, contributor depth, liquidity indicator, and version Accepted source record or stale-data exception
Apply method G-spread, Z-spread, OAS, interpolation, extrapolation, proxy, and fallback rules Methodology record tied to the valuation run
Compare marks Manager mark, evaluated price, secondary source, tolerance, and threshold Independent price verification result or review item
Approve override Override reason, supporting quote, owner, approval, expiry, and affected instruments Controlled override with audit trail
Distribute state IBOR, ABOR, PBOR, collateral, risk, reporting, and reconciliation consumers Versioned downstream delivery record

Curve selection starts with the instrument master. The workflow uses CUSIP, ISIN, issuer, LEI, currency, country, sector, rating, seniority, maturity, call features, security type, and liquidity context to identify the appropriate benchmark and credit curve family.

Source validation checks whether the curve state is usable. The system should preserve provider, timestamp, version, tenor completeness, liquidity indicator, source depth, and any missing values. A stale curve should become a data exception before it becomes a valuation result.

Method application records how the input was transformed. If a bond falls between tenor points, the interpolation logic should be visible. If an instrument falls beyond the available curve, the extrapolation or fallback rule should be documented. If a callable instrument uses OAS, the optionality method should be tied to the valuation run.

Price validation compares marks against controlled references. The manager mark, evaluated price, secondary source, tolerance, curve-derived value, and supporting quote should be compared with clear thresholds. A difference inside tolerance may be accepted. A difference outside tolerance should create a review item with owner, reason, and resolution.

Override workflow is a control event. A portfolio manager, valuation committee, or operations team may have evidence that the curve-derived value is wrong for a specific instrument. The override should carry source support, reason, approver, affected instrument population, effective date, expiry date, and downstream impact.

Distribution turns the accepted curve state into a consumed record. IBOR may use it for intraday valuation and exposure. ABOR may use it for accounting support. PBOR may use it for performance attribution. Collateral workflows may use it for mark and haircut support. Reporting and reconciliation workflows may use it to explain why a value changed.

In Devancore™

Devancore supports credit curve reference data by maintaining controlled records around curve source, instrument mapping, valuation method, review state, override evidence, and downstream consumption. The platform should be framed as a post-trade record and workflow layer, not as a pricing vendor, investment adviser, valuation agent, accountant, fund administrator, broker, custodian, or compliance authority.

In a Devancore-style workflow, a fixed income instrument is mapped to issuer, LEI, sector, rating, seniority, maturity, currency, settlement context, and security master terms. Curve inputs arrive with timestamp, source version, tenor state, methodology, and coverage flags. The record shows which curve was selected, which rule selected it, and which workflows consumed it.

This gives teams a way to identify stale tenor points, unsupported proxy mappings, curve source differences, tolerance breaches, manual overrides, missing approvals, and downstream records that used a prior curve state.

The same controlled record can support IBOR valuation state, ABOR accounting handoff, PBOR attribution, collateral review, independent price verification, regulatory reporting support, reconciliation, and audit trail. The practical value is that a valuation movement can be traced to a curve state and method rather than reconstructed from spreadsheets, vendor files, and comments.

Conversational finance becomes useful when curve data is structured. A user may ask which bonds used proxy curves, which issuer curves moved more than a threshold, which positions have stale curve inputs, which valuation overrides expire today, which P&L movement came from credit spread widening, or which PBOR attribution record consumed a specific curve version. The answer should resolve to the underlying instruments, curve state, method, owner, and evidence.